Heider's Five Levels of Causality and Assignment of Responsibility by Actors and Observers.

Transcription

1 Louisiana State University LSU Digital Commons LSU Historical Dissertations and Theses Graduate School 1977 Heider's Five Levels of Causality and Assignment of Responsibility by Actors and Observers. David Conner Blouin Louisiana State University and Agricultural & Mechanical College Follow this and additional works at: Recommended Citation Blouin, David Conner, "Heider's Five Levels of Causality and Assignment of Responsibility by Actors and Observers." (1977). LSU Historical Dissertations and Theses This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion in LSU Historical Dissertations and Theses by an authorized administrator of LSU Digital Commons. For more information, please contact

2 INFORMATION TO USERS This material was produced from a microfilm copy of the original document. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the original submitted. The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction. 1.The sign or "target" for pages apparently lacking from the document photographed is "Missing Page(s)". If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting thru an image and duplicating adjacent pages to insure you complete continuity. 2. When an image on the film is obliterated with a large round black mark, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image. You will find a good image of the page in the adjacent frame. 3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning" the material. It is customary to begin photoing at the upper left hand corner of a large sheet and to continue photoing from left to right in equal sections with a small overlap. If necessary, sectioning is continued again beginning below the first row and continuing on until complete. 4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints of "photographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, title, author and specific pages you wish reproduced. 5. PLEASE NOTE: Some pages may have indistinct print. Filmed as received. University Microfilms International 300 North Z eeb Road Ann Arbor, M ichigan USA St. John s Road, Tyler's Green High Wycombe, Bucks. England HP10 8HR

3 I I BLOUIN, David Conner, HEIDER'S FIVE LEVELS OF CAUSALITY AND ASSIGNMENT OF RESPONSIBILITY BY ACTORS AND OBSERVERS. The Louisiana State University and Agricultural and Mechanical College, Ph.D., 1977 Psychology, social University Microfilms International, Ann Arbor, M ichigan 48106

4 HEIDER'S FIVE LEVELS OF CAUSALITY AND ASSIGNMENT OF RESPONSIBILITY BY ACTORS AND OBSERVERS A D is s e r t a t io n Subm itted to th e Graduate F a c u lty o f the L o u isia n a S ta te U n iv e r s ity and A g r ic u ltu r a l and M echanical C o lle g e in p a r t i a l f u l f i ll m e n t o f th e req uirem en ts fo r th e degree o f D octor o f P h ilosop h y in The Department o f P sych ology ty David C. B lou in B.S., B aylor U n iv e r s ity, 1970 M.A., L o u isia n a S ta te U n iv e r s ity, 1973 M.A p.s ta t., L o u isia n a S ta te U n iv e r s ity, 197^ December 1977

5 ACKNOWLEDGEMENTS The au th or ex ten d s h i s thanks to th e members o f h i s com m ittee, d o c to r s Nathan G o ttfr ie d, Robert Coon, I r v in Lane, and P r e n t is s S c h i l l i n g, f o r t h e i r a s s is t a n c e in th e w r itin g and e v a lu a tio n o f t h i s stu d y. The au th or e s p e c i a l l y a p p r e c ia te s th e guid ance and work o f th e com m ittee chairm an, Dr. P erry H. P r e s th o ld t, from th e i n i t i a l c o n c e p tu a liz a t io n o f th e p r o j e c t to i t s c o m p letio n.

6 TABLE OF CONTENTS Page A. TITLE... i B. ACKNOWLEDGEMENTS... i i C.. TABLE OF CONTENTS... i i i D. LIST OF TABLES... v E. LIST OF FIGURES... v i F. ABSTRACT... i x G. INTRODUCTION O verview o f A ttr ib u t io n P r o c e ss O verview o f A t t r ib u t io n R esearch L ite r a tu r e Review ii-. Developm ent o f R e so lu tio n S tatem en t o f O b je c tiv e s... 6? 6. D e r iv a tio n o f H ypotheses H. METHOD S u b je c ts C o n str u c tio n o f A c tio n S it u a tio n s P rocedure... 8ifij-. D e sig n I. RESULTS AND DISCUSSION J. GENERAL DISCUSSION K. REFERENCES i i i

7 Table o f C on ten ts (c o n t'd ) Page L. APPENDIX A M. APPENDIX B N. APPENDIX C APPENDIX D P. APPENDIX E... 16k Q. VITA

8 LIST OF TABLES Table Page 1 A n a ly sis o f v a ria n ce fo r o b s e r v e r s assign m en t o f r e s p o n s ib ilit y to a c to r s Mean amount o f r e s p o n s ib i li t y a ssig n ed by a c to r s, o b se r v e r s, and a c to r s w ith th e s e lf-a w a r e n e ss m an ip u lation (SA a c to r s) fo r p o s i t i v e and n e g a tiv e outcom es a t each l e v e l o f c a u s a lit y and o v e r a ll: a s s o c ia t io n (L e v el I ), com m ission (L evel I I ), f o r e s e e a b i li t y (L evel I I I ), j u s t i f i c a t i o n (L evel IV) and i n t e n t i o n a l i t y (L evel V) Mean amount o f r e s p o n s ib i li t y a s s i g n e d by a c to r s, o b se r v e r s, and a c to r s w ith th e se lf-a w a r e n e s s m an ip u lation (SA a c to r s) f o r low or h ig h p o s i t i v e or n e g a tiv e outcom es a t each l e v e l o f c a u s a lit y : a s s o c ia t io n (L ev el I ), com m ission (L evel I I ), f o r e s e e a b i li t y (L evel I I I ), j u s t i f i c a t i o n (L e v el IV) and i n t e n t i o n a l i t y (L ev el V) k A n a ly sis o f v a ria n ce fo r th e amount of r e s p o n s ib i li t y (AR) a ssig n e d by a c to r s, o b se r v e r s, and a c to r s w ith th e s e l f - aw areness m an ip u lation v

9 LIST OF FIGURES F igure Page 1 Mean p e r ce n t f o r e s e e a b i li t y by l e v e l fo r q u e stio n, "To what degree co u ld X fo r e s e e th a t h is a c tio n s would le a d to Y?" Mean p ercen t i n t e n t io n a lit y by l e v e l fo r q u e s tio n, "To what degree was Y X 's in te n tio n? " Mean p ercen t c a u s a lit y by l e v e l f o r q u e stio n "To what degree did X 's a c tio n s cau se Y?" Ur Mean p ercen t j u s t i f i c a t i o n by l e v e l fo r q u e stio n, "To what degree were X ' s a c tio n s j u s t i f i e d? " Mean amount o f r e s p o n s ib ilit y (AR) a ssig n e d by o b ser v er s f o r p o s i t i v e and n e g a tiv e outcomes a t each l e v e l o f c a u sa lity * a s s o c ia t io n ( I ), com m ission ( I I ), f o r e s e e a b i li t y ( I I I ), j u s t i f i c a t i o n (IV ), and i n t e n t i o n a l i t y (V ) Mean amount o f r e s p o n s ib i li t y (AR) a ssig n ed by o b serv ers a t th e l e v e l o f a s s o c ia t io n depending upon outcome v a len ce ( p o s it i v e, n e g a tiv e ) and outcome i n t e n s it y (lo w, h ig h ) Mean amount o f r e s p o n s ib ilit y (AR) a ssig n ed by o b serv ers a t th e l e v e l o f com m ission depending upon outcome v a len ce ( p o s it i v e, n e g a tiv e ) and outcome in t e n s it y (lo w, h ig h ) Mean amount o f r e s p o n s ib i li t y (AR) a ssig n e d by o b serv ers a t th e l e v e l o f f o r e s e e a b i li t y depending upon outcome v a len ce ( p o s it i v e, n e g a tiv e ) and outcome in t e n s i t y (low, h ig h ) Mean amount o f r e s p o n s ib ilit y (AR) a ssig n e d by o b serv ers a t th e l e v e l o f j u s t i f i c a t i o n depending upon outcome v a len ce ( p o s it i v e, n e g a tiv e ) and outcome i n t e n s it y (low, h ig h ) 117

10 L i s t o f F ig u r e s (c o n t'd ) F igu re Page 10 Mean amount o f r e s p o n s i b i l i t y (AR) a ssig n e d by o b ser v er s a t th e l e v e l o f i n t e n t i o n a l i t y depending upon outcome v a len ce ( p o s i t i v e, n e g a tiv e ) and outcome i n t e n s i t y (lo w, h ig h ) Mean amount o f r e s p o n s i b i l i t y (AR) a ssig n e d by a c to r s, o b s e r v e r s, and a c to r s w ith the s e lf-a w a r e n e s s m a n ip u la tio n (SA a c to r s ) fo r p o s i t i v e and n e g a tiv e outcom es a t each l e v e l o f c a u s a lit y : a s s o c i a t i o n ( I ), com m ission ( I I ), f o r e s e e a b i l i t y ( I I I ), j u s t i f i c a t i o n (IV ), and i n t e n t i o n a l i t y (V ) Mean amount o f r e s p o n s i b i l i t y (AR) a ssig n e d by a c to r s, o b s e r v e r s, and a c to r s w ith the s e lf-a w a r e n e s s m a n ip u la tio n (SA a c to r s ) at th e l e v e l o f a s s o c i a t i o n depending upon outcome v a le n c e ( p o s i t i v e, n e g a tiv e ) and outcome i n t e n s i t y (lo w, h ig h ) Mean amount o f r e s p o n s ib i li t y (AR) a ssig n e d by a c to r s, o b s e r v e r s, and a c to r s w ith the s e lf-a w a r e n e s s m a n ip u la tio n (SA a c to r s ) at th e l e v e l o f com m ission depending upon outcome v a le n c e ( p o s i t i v e, n e g a tiv e ) and outcome i n t e n s i t y (lo w, h ig h ) Mean amount o f r e s p o n s i b i l i t y (AR) a ssig n e d by a c t o r s, o b s e r v e r s, and a c to r s w ith the s e lf-a w a r e n e s s m a n ip u la tio n (SA a c to r s ) a t th e l e v e l o f f o r e s e e a b i l i t y d ep en d in g upon outcome v a le n c e ( p o s i t i v e, n e g a tiv e ) and outcome i n t e n s i t y (lo w, h ig h ) Mean amount o f r e s p o n s i b i l i t y (AR) a ssig n e d by a c to r s, o b s e r v e r s, and a c to r s w ith the s e lf-a w a r e n e s s m a n ip u la tio n (SA a c to r s ) at th e l e v e l o f j u s t i f i c a t i o n dep en d in g upon outcome v a le n c e ( p o s i t i v e, n e g a tiv e ) and outcome i n t e n s i t y (lo w, h ig h ) v i i

11 L is t o f F ig u r e s (c o n t'd ) F igu re Page 16 Mean amount o f r e s p o n s i b i l i t y (AR) a s s ig n e d by a c t o r s, o b s e r v e r s, and a c to r s w ith th e s e lf-a w a r e n e s s m a n ip u la tio n (SA a c to r s ) a t th e l e v e l o f i n t e n t i o n a l i t y depending upon outcome v a le n c e ( p o s i t i v e, n e g a tiv e ) and outcome i n t e n s i t y (lo w, h i g h ) v i i i

12 ABSTRACT R esearch f in d in g s on o b s e r v e r s' r e s p o n s i b i l i t y assign m en t depending upon outcome v a len ce ( p o s i t i v e, n e g a tiv e ) and in t e n s i t y (lo w, h igh ) have been in c o n s is t e n t. However, r e se a r c h on H e id e r 's f i v e l e v e l s o f c a u s a l i t y s u g g e s ts th a t th e e f f e c t o f outcome c h a r a c t e r i s t i c s depends upon th e n atu re o f th e i n t e r a c t io n betw een a c to r and environm ent a s o b j e c t iv e ly d e p ic te d in a c t io n s i t u a t i o n s a t each l e v e l ( a s s o c ia t io n, com m ission, f o r e s e e a b i l i t y, j u s t i f i c a t i o n, i n t e n t i o n a l i t y ). P o s s ib ly r e se a r c h e r s have found in c o n s is t e n t e f f e c t s due to outcome c h a r a c t e r i s t i c s and i n c o n s is t e n t su pp ort f o r d e fe n s iv e a t t r i b u t io n because th e y u sed s i t u a t io n s r e p r e s e n tin g d if f e r e n t l e v e l s o f c a u s a l i t y. S im ila r ly, r e se a r c h on a c to r s and o b ser v er s has found p a tte r n s o f r e s p o n s i b i l i t y assign m en t in c o n s is t e n t w ith th e d i f f e r e n t i a l s e l f - i n t e r e s t m o tiv es o f a c to r s and o b ser v er s im p lie d by d e fe n s iv e a t t r i b u t i o n. P o s s ib ly p a tte r n s o f r e s p o n s i b i l i t y assign m en t r e f l e c t i n g th e se m o tiv a tio n a l d if f e r e n c e s depend upon th e n atu re o f th e i n t e r a c t io n among c a u s a l a g e n ts a t each l e v e l o f c a u s a l i t y. Hence, a major o b j e c t iv e was to examine a t w hich l e v e l s ob served p a tte r n s o f r e s p o n s i b i l i t y assign m en t by a c to r s and o b se r v e r s would in d ic a t e th e o p e r a tio n o f d e fe n s iv e a t t r ib u t io n and s e l f - i n t e r e s t. F in a lly, based on th e th e o r y o f o b j e c t iv e s e lf-a w a r e n e s s, a second o b j e c t iv e was to determ ine i f in c r e a s in g a c t o r s s e l f - aw areness (SA a c to r s ) would a f f e c t t h e ir r e s p o n s ib i li t y i x

13 assign m en t r e l a t i v e t o a c to r s w ith o u t in c r e a s e d s e lf-a w a r e n e s s. S ix t y male s tu d e n ts were o b s e r v e r s, a c t o r s, or SA a c to r s who w ere exposed to t h e i r v is u a l image v ia m ir r o r s. For each o f 20 h y p o th e tic a l s it u a t io n s r e p r e s e n tin g a 5 x 2 x 2 f a c t o r i a l arrangem ent o f l e v e l, v a le n c e and i n t e n s i t y, Ss in d ic a te d how much th ey ( f o r a c to r s ) or th e c e n tr a l c h a r a c te r ( f o r o b se r v e r s) were r e s p o n s ib le f o r th e outcom e. R e s u lts showed th a t the e f f e c t o f outcome c h a r a c t e r is t ic s on o b s e r v e r s ' r e s p o n s i b i l i t y assign m en t depended upon l e v e l. At th e p o la r l e v e l s o f a s s o c ia t io n and i n t e n t i o n a l i t y, th e e f f e c t o f outcome c h a r a c t e r i s t i c s was r a th e r m inim al, th e re b y not conform ing to d e fe n s iv e a t t r ib u t io n p r e d ic t io n s. However, a t th e in te r m e d ia te l e v e l s (com m ission, f o r e s e e a b i l i t y, j u s t i f i c a t i o n ), outcome c h a r a c t e r i s t i c s had an e f f e c t. The p a tte r n a t th e f o r e s e e a b i l i t y l e v e l conform ed r a th e r w e ll w ith d e fe n s iv e a t t r ib u t io n p r e d ic t io n s but o n ly somewhat a t th e com m ission l e v e l. The p a tt e r n a t the j u s t i f i c a t i o n l e v e l was o p p o site t o d e f e n s iv e a t t r ib u t io n p r e d ic t io n s. S im ila r p a t t e r n s were e x h ib ite d ir r e s p e c t iv e o f r o l e. As su c h, the o p e r a tio n o f d i f f e r e n t i a l p e r c e p tu a l an d /or m o tiv a tio n a l b ia s e s im p lied by d e fe n s iv e a t t r ib u t io n, s e l f - i n t e r e s t, and s e lf-a w a r e n e s s was n ot in d ic a t e d. The r e s u l t s were in te r p r e te d a s in d ic a t in g t h a t ir r e s p e c t iv e o f r o l e, Ss were e v a lu a tin g th e c a u sa l i n t e r a c t io n a t each l e v e l and a s s ig n in g r e s p o n s i b i l i t y from an o b j e c tiv e detached p e r s p e c t iv e r a th e r than from d i f f e r e n t i a l s u b j e c t iv e p e r s p e c t iv e s.

14 INTRODUCTION The a t t r ib u t io n p r o c e s s r e f e r s to th e ten d en cy o f an in d iv id u a l to make in f e r e n c e s about th e p e r so n a l d i s p o s i t i o n s o f p e o p le whose b e h a v io r th e y have ob served. In fe r e n c e s or " a ttr ib u tio n s " about a p e r s o n 's a t t i t u d e s, a b i l i t i e s, m o tiv e s, and o th e r p e r so n a l d is p o s it io n s serv e to e x p la in th e c a u s a l a n te c e d e n ts o f th e p e r s o n 's b e h a v io r. For exam ple, a p r o f e s s o r may a t t r ib u t e a s t u d e n t s s u c c e s s (o r f a i l u r e ) on an exam to h is own e f f e c t i v e (o r i n e f f e c t i v e ) te a c h in g or to the s t u d e n t 's a b i l i t y (o r la c k o f a b i l i t y ) i n th e c o u r s e. Over r e c e n t y e a r s, a t t r i b u t i o n r e s e a r c h has d e a l t w ith a wide v a r ie t y o f t o p ic s r a n g in g from th e a t t r i b u t i o n o f a t t i t u d e s and a b i l i t i e s to th e a t t r i b u t i o n o f o p in io n s and em otion s. A g e n e r a l d e s c r ip t io n o f th e a t t r ib u t io n p r o c e s s a s view ed by th e p r e s e n t auth or f o llo w s. The overa l l purpose o f th e d e s c r ip t io n i s to fo c u s on one p a r t ic u la r area o f a t t r i b u t i o n r e s e a r c h, nam ely r e se a r c h t h a t has d e a lt w ith th e " a ttr ib u tio n o f c a u s a lit y " and "assignm ent o f r e s p o n s i b i l i t y." R esearch on th e a t t r i b u t i o n o f c a u s a l i t y and a s s ig n ment o f r e s p o n s i b i l i t y has r e v o lv e d around two g e n e r a l i s s u e s. The i s s u e s may be s t a t e d a s f o llo w s. G iven a sequence o f e v e n ts th a t in v o lv e s one o r more p e o p le whose b eh a v io r le a d s to a p a r t ic u la r outcom e, (1) "What f a c t o r s in flu e n c e an in d iv i d u a l 's ' a t t r i b u t io n s ' about th e

15 2 c a u s e -a n d -e ffe c t r e la tio n s h ip s which serve to e x p la in a p erso n s a c tio n s and th e r e s u lta n t outcome?" and (2 ) "What f a c to r s in flu e n c e th e degree to which an in d iv id u a l h old s th e person r e sp o n sib le ' fo r th e outcome?" R esearch ers have a ssig n ed a ra th er g en era l meanin g to th e concept o f " r e sp o n sib ility " in th e m a jo rity o f p a s t resea r ch. As viewed in p a s t resea r ch, r e s p o n s ib ilit y h as not t y p ic a lly been o p e r a tio n a lly or s p e c if i c a lly d efin ed as e ith e r "moral" or "legal" r e s p o n s ib ilit y. R e s p o n s ib ility has a more s u b je c tiv e, judgm ental, or phenome n o lo g ic a l m eaning. Sim ply p u t, i f a p erson i s h eld r e s p o n s ib le, then th e person m ight be thanked fo r som ething "good" th a t happens o r blamed fo r som ething "bad" th a t happens. T h erefo re, in th e most g en eral se n se, " assign ment o f r e s p o n s ib ility " has refe rred to what e x te n t a person i s p henom en alogically c r e d ite d w ith th e p rod u ction o f an outcom e. T h e o r e tic a lly, a t tr ib u t io n s o f c a u s a lity and a ssig n ment o f r e s p o n s ib ilit y are r e la t e d. The u n d erly in g n o tio n in a l l co n cep tu a l approaches to th e a ttr ib u tio n p ro c e ss i s th a t an in d iv id u a l's assignm ent o f r e s p o n s ib ilit y to a person depends on th e in d iv id u a l's a ttr ib u tio n s about th e c a u sa l a n teced en ts o f the p e r s o n 's a c tio n s and r e s u lta n t outcome. As m ight be ex p ected, th o se fa c to r s which in f l u ence an in d iv id u a l's a ttr ib u tio n s o f c a u s a lity a ls o in f l u ence th e in d iv id u a l s assignm ent o f r e s p o n s ib ilit y.

16 3 For example, "both a p r o fe s s o r and a stu d en t are norm ally in v o lv ed in produ cin g an outcome o f stu dent s u c c e s s (or f a i lu r e ) on an exam. The p r o fe s s o r may a ttr ib u t e th e s t u d e n t s s u c c e s s (o r f a ilu r e ) p a r t i a l l y to h is own e f f e c t i v e (o r i n e f f e c t i v e ) te a c h in g a b i l i t y and p a r t i a ll y to the s tu d e n t's a b i l i t y (or la c k o f a b i l i t y ), and in tu r n, hold h im s e lf and th e stu d en t p a r t i a l l y r e sp o n sib le f o r th e s tu d e n t's s u c c e s s (o r f a ilu r e ) on th e exam. O verview o f th e A ttr ib u tio n P r o c e s s The fo llo w in g d is c u s s io n o f th e a t t r ib u t io n p r o c e ss has been d esign ed to accom p lish two g o a ls. The f i r s t g oal i s to in tro d u ce th e g e n e ra l c o n c e p ts, b a sic p r o p o s itio n s, and exp erim en tal f a c t o r s t h a t have rec eiv ed m ajor a t t e n t io n in co n cep tu a l and exp erim en tal approaches to th e " a ttr ib u tio n o f c a u s a lity " and "assignm ent o f r e s p o n s ib ilit y." The second g oal i s to d e lim it th e area o f a t t r ib u t io n r e se a r c h th a t i s o f major i n t e r e s t in the p r e se n t study. Three fundam ental p r o p o s itio n s are in v o lv e d in a l l c o n cep tu a l approaches to th e a t t r ib u t io n p r o c e s s. Each p r o p o s itio n w i l l be d is c u sse d i n tu rn. To i l l u s t r a t e th e se p r o p o s itio n s, th e fo llo w in g e v e n ts w i l l be d is c u s s e d and expanded: ( l ) a p erson buys a fr ie n d a b eer, (2 ) a person sa v e s a boy from drowning, (3 ) a person f a i l s an exam, and (*}-) a p erson k i l l s a man.

17 These ev en ts d i f f e r in two im portant ways th a t are r e le vant to th e b a sic p r o p o sitio n s o f the a ttr ib u tio n p r o c e s s. F ir s t, th e consequences or outcome o f th e p erso n s a c tio n s d if f e r in terms o f p o s i t i v i t y and n e g a tiv ity ( e.g. " savin g a boy from drowning" versu s " k illin g a man"). Secon d ly, an in d iv id u a l may be th e person who a c ts in each s it u a t io n or may m erely be a n o n -a ctin g w itn e ss o f another p e r s o n 's a c tio n s in each s it u a t io n. For example, th e in d iv id u a l h im self may "buy a fr ie n d a beer" or may m erely w itn e ss another person "buy a frie n d a b eer." These d iffe r e n c e s have receiv ed m ajor a tte n tio n from resea r ch e r s who have attem pted to d e lin e a te th e in flu e n c e o f outcome c h a r a c te r i s t i c s ( e.g. d egree o f p o s i t i v i t y or n e g a tiv ity ) and th e p e r sp e c tiv e or r o le o f the in d iv id u a l ( e. g. a c tiv e p a r t i c i pant o r n o n -a ctin g w itn e ss) on a ttr ib u tio n s o f c a u s a lity and assignm ent o f r e s p o n s ib ilit y. The p r o p o sitio n most fundam ental to th e a ttr ib u tio n p r o c e ss i s th a t when an in d iv id u a l o b serv es h is own a c tio n s or th e a c tio n s o f o th e r p eop le in a p a r tic u la r s it u a t io n, th e in d iv id u a l i s o fte n m otivated to e s t a b lis h s u f f ic ie n t rea so n s fo r th e a c tio n s th a t were taken and the r e s u lta n t outcom e. Now, a s th e exam ples i l l u s t r a t e, an in d iv id u a l may be e ith e r an a c to r or an ob server o f an even t. As su ch, th e in d iv id u a l who i s an a cto r may be m otivated to e x p la in h is own a c tio n s and r e s u lta n t outcome w hile th e in d iv id u a l who i s an ob server may be m otivated to e x p la in

18 a n oth er p e r s o n 's a c tio n s and r e s u lta n t outcom e. F u rth erm ore, th e n a tu re o f th e produced outcome in term s o f p o s i t i v i t y or n e g a t iv it y may d i f f e r e n t i a l l y a f f e c t th e ty p e s o f c a u sa l a t t r ib u t io n s and th e amount o f r e s p o n s ib ilit y a ssig n e d by a c t o r s and o b se r v e r s. The second fundam ental p r o p o s itio n i s th a t inform a t i o n p r o c e ssin g i s b a sic to th e a t tr ib u t io n p r o c e s s. The in d iv id u a l, a s an a c to r or o b se r v e r, accum ulates and e v a lu a te s a v a ila b le in fo r m a tio n about e v e n ts in o rd er to d eterm in e th e m ost p la u s ib le and rea so n a b le ex p la n a tio n f o r th e observed a c tio n s and outcom e. The in d iv id u a l i s assumed to p r o c e s s a v a ila b le in fo rm a tio n about th e a c tu a l sequence o f e v e n ts, th e b eh a v io r o f th e p eo p le who were in v o lv e d, th e environm ental s e t t in g, and th e e x ten u a tin g circu m sta n ces th a t f a c i l i t a t e d or in h ib ite d th e observed a c t io n s and outcom e. Based on th e p r o c e s s in g o f t h i s in fo r m a tio n, th e in d iv id u a l a r r iv e s a t a d e c is io n a s to w hich o f s e v e r a l a lt e r n a t iv e e x p la n a tio n s fo r th e a c tio n s ta k en and th e outcome produced i s th e m ost p la u s ib le or r e a so n a b le. T his second p r o p o s itio n im p lie s th a t an in d iv id u a l p r o c e s s e s a g r e a t d ea l o f in fo rm a tio n in making c a u s a l in fe r e n c e s and in a s s ig n in g r e s p o n s i b ilit y. However, in th e exam ples o f a c tio n seq u en ces d is c u s s e d above, th e a v a ila b le in fo rm a tio n i s q u ite lim it e d. For exam ple, n o t ic e th a t in th o se a c tio n sequences o n ly one p erso n a c ts

19 and th e outcome i s a d ir e c t r e s u lt o f h is a c t s. Thus, based on th e a v a ila b le in fo rm a tio n in each c a s e, a lte r n a t i v e c a u sa l e x p la n a tio n s fo r th e outcome are v ir t u a lly n o n - e x is te n t. A lso n o tic e th a t numerous in fe r e n c e s co u ld be made < about th e c a u s a l a n teced en ts o f th e p e r so n 's a c t io n s, but v ir t u a lly no in fo rm a tio n i s g iv en to support o r su ggest th e p l a u s i b i l i t y o f such in fe r e n c e s. For exam ple, "Was th e person who bought a fr ie n d a b eer r e c ip r o c a tin g a p rev io u s favor?" o r "Was th e p erso n who saved th e boy from drowning th e o n ly o th er p erso n a t the beach?" or "Did th e p erso n who f a ile d th e exam study?" o r "Did the p erson who k i l l e d th e man do so in te n tio n a lly? " Thus, i f th e on ly in fo rm a tio n a v a ila b le to th e a ttr ib u to r i s a sim ple a c tio n sequence such a s, "a person buys a fr ie n d a b e e r," th en a l l ev id en ce p o in t s to the p erso n a s the s o le c a u sa l agen t in producing th e outcome and th e person would l i k e l y be h eld t o t a l l y r e s p o n s ib le fo r th e outcom e. However, an a ttr ib u to r u s u a lly has a g r e a t d eal more in fo rm a tio n about h is own a c tio n s o r another p e r so n 's a c tio n s and th e circu m sta n ces surrounding th o s e a c tio n s. T h erefo re, each s it u a t io n may be expanded to in clu d e more in fo rm a tio n and th e im pact o f t h i s a d d itio n a l in form ation may be examined r e la t iv e to a t tr ib u t io n s o f c a u s a lit y and assignm ent o f r e s p o n s ib ilit y. A d d itio n a l in fo rm a tio n ab ou t each a c tio n sequence

20 7 might bei ( l ) the p e r so n s fr ie n d had no money and asked th e person to buy him a b eer, (2 ) th e person was the o n ly p erson around who c o u ld swim and sa v e th e boy from drownin g, (3) th e person sta y ed up a l l n ig h t stu d y in g fo r the exam and was not a le r t when ta k in g th e exam, or (*0 the person was d riv in g and ran over a man who step ped out from beh ind a parked c a r. In th e s e expanded s it u a t io n s, more in form ation i s now a v a ila b le not o n ly about th e p erso n s a c tio n s but a ls o about environm ental fo r c e s which in flu en ced th e a c tio n s tak en and th e outcome produced. There now e x i s t s a somewhat g lo b a l dichotomy o f c a u sa l a g e n ts, namely "the person" and "the environm ental fo rces" which a f f e c t th e p erson s a c tio n s and the outcom e. The a v a ila b ilit y o f t h is additio n a l in form ation im p lie s th a t th e a ttr ib u to r must d ea l w ith th e nature o f th e in te r a c tio n between th e person and environm ental fo r c e s in a r r iv in g a t p la u s ib le ex p la n a tio n s fo r th e a c tio n s and th e outcome, in a ttr ib u tin g c a u s a lity, and in a ssig n in g r e s p o n s ib ilit y. The r e s u ltin g in form ation p ro cessin g ta sk o f the a ttr ib u to r might be co n cep tu a lized as fo llo w s. F ir s t, th e a ttr ib u to r p ro cesse s th e a v a ila b le in form ation to a s s e s s the r e la t iv e c o n tr ib u tio n o f th e person and th e environm ental fo r c e s in c a u sin g the p e r so n 's a c tio n s and the outcome. Second, to th e ex ten t th a t e x te r n a l environm ental fo r c e s are ruled ou t as a p o ten t c a u sa l agent in producing

21 th e p e r s o n 's a c tio n s a n d /o r outcom e, th e a t t r ib u t o r makes c a u sa l a ttr ib u t io n s about th e in te r n a l p e r so n a l m o tiv es or d is p o s it io n s o f th e p erso n which serv e to e x p la in h is a c tio n s. F in a lly, based on th e s e c a u s a l a t t r ib u t io n s con cern in g th e p e r s o n 's a c tio n s and th e outcom e, th e a t t r ib u to r a s s ig n s r e la t iv e amounts o f r e s p o n s i b ilit y to th e c a u sa l a g e n ts. For exam ple, an in d iv id u a l may a s s ig n l e s s r e s p o n s ib ilit y to a p erso n when th e outcome i s a ttr ib u te d p u rely to environm ental fa c t o r s such a s "chance" than when th e outcome i s a ttr ib u te d to in te r n a l d is p o s it io n s such a s th e " c a r e le ss" or "prem editated" a c tio n s o f th e p erso n. As m ight be ex p ecte d, th e way in w hich th e a t t r ib u to r d e a ls w ith th e a v a ila b le in fo rm a tio n i s a f fe c t e d by th e p e r s p e c tiv e or r o le o f th e a ttr ib u to r and th e n atu re o f th e outcom e. Thus, th e th ir d fundam ental p r o p o s itio n in co n cep tu a l approaches to th e a t t r ib u t io n p r o c e ss i s th a t th e d egree to which an in d iv id u a l h o ld s h im s e lf or an oth er person r e s p o n s ib le f o r th e occu rren ce o f an outcome may n o t be sim ply th e r e s u lt o f an o b je c t iv e a n a ly s is o f th e c a u se - a n d -e ffe c t r e la t io n s h ip s in an e v e n t. D epending upon h is r o le o r p e r s p e c tiv e i n an even t (a c to r or o b serv er) and upon outcome c h a r a c t e r is t ic s ( p o s i t i v i t y o r n e g a t i v i t y ), th e a t tr ib u t o r l i k e l y has p e r c e p tu a l an d /or m o tiv a tio n a l b ia s e s to lo c a te c a u s a lit y and a s s ig n r e s p o n s ib ilit y to d if f e r e n t c a u sa l a g e n ts. These b ia s e s le a d to m o d ific a t io n s o r d is t o r t io n s t h a t would n o t be p r e sen t i f

22 9 a ttr ib u tio n s o f c a u s a lity were th e r e s u lt o f a p u r ely o b je c tiv e a n a ly s is o f c a u s a lit y. These m o d ific a tio n s or d is t o r t io n s o f c a u sa l a ttr ib u tio n s a ls o lea d to a c o r r e sponding m o d ific a tio n or d is t o r t io n in assignm ent o f r e s p o n s ib ilit y. For exam ple, suppose a person i s r o l lin g a n a ir o f d ic e in a gam bling s it u a t io n. O b je c tiv e ly, given th e d ic e are f a i r, a w inning t o s s or a lo s in g t o s s i s due p u rely to chance. However, an in d iv id u a l who throws th e d ic e may a ttr ib u te c a u s a lity to h im self ra th er than to o th e r environm ental fo r c e s ( e.g. chance) and take more r e s p o n s ib ilit y fo r a p o s it iv e outcome ( e.g. a w inning to s s ) than fo r a n e g a tiv e outcome ( e.g. a lo s in g t o s s ). On the o th er hand, an in d iv id u a l who m erely w itn e s s e s t h is a c tio n sequence may be more o b je c tiv e and may tend to a ttr ib u te c a u s a lity to chance and a ssig n l i t t l e r e s p o n s ib ilit y to th e person r e g a r d le ss o f th e outcome. In summary, the g en eral co n cep tu a l framework in a t t r ib u tio n resea rch has been to p o s tu la te a p ro cess which media te s an in d iv id u a l's resp o n ses to a s e t o f s tim u li. The s tim u li i s the in form ation th a t i s a v a ila b le about th e sequence o f ev en ts th a t le d to th e outcom e. The resp on ses are a ttr ib u tio n s o f c a u s a lity and assignm ent o f r e sp o n sib i l i t y. The p ro cess b a s ic a lly in v o lv e s th e p r o c e ssin g o f a v a ila b le in fo rm a tio n. The p r o c e ssin g o f in form ation le a d s to d e c is io n s as to th e c a u s e -a n d -e ffe c t r e la tio n s h ip s which

23 10 se rv e to ad eq u a tely e x p la in th e a c tio n s th a t were taken and th e outcome th a t was produced. However, th e p r o c e ss may n ot sim p ly in v o lv e an o b je c tiv e a n a ly s is o f th e c a u se - a n d -e ffe c t r e la t io n s h ip s. Depending upon h is r o le in an ev en t and upon outcome c h a r a c t e r is t ic s, an in d iv id u a l may have p e r c e p tu a l an d/or m o tiv a tio n a l b ia s e s. These b ia s e s o p era te to m odify or d i s t o r t c a u sa l a t t r ib u t io n s and r e s p o n s ib ilit y assignm en t which would r e s u lt from a p u r ely o b je c tiv e e v a lu a tio n o f th e c a u sa l r o le shared by a person and environm ental fo r c e s in producing a p a r tic u la r outcom e. Overview o f A ttr ib u tio n R esearch R esearch ers o f th e a t tr ib u t io n p r o c e s s have attem pted to d e lin e a te th e in flu e n c e o f s e v e r a l f a c t o r s on th e a t t r ib u tio n o f c a u s a lit y and assignm en t o f r e s p o n s ib ilit y in " a ctio n s it u a t io n s." M ost resea r c h h a s been d ir e c te d toward th e in flu e n c e o f outcome v a le n c e ( p o s i t iv e o r n e g a t iv e ), outcome in t e n s it y (low o r h ig h ), and s u b je c t r o le v i s - a - v i s th e a c tio n s it u a t io n ( a c t o r o r o b s e r v e r ). A review o f t h i s r e se a r c h w i l l dem onstrate th a t s tu d ie s on outcome c h a r a c t e r is t ic s and s u b je c t r o le have le d to in c o n s is t e n t fin d in g s and an e x te n s iv e but p e r p le x in g body o f l it e r a t u r e. The o v e r a ll g o a l o f th e p resen t study i s to propose and t e s t a t h e o r e t ic a l r e s o lu tio n fo r th e se in con s i s t e n t f in d in g s. B efore p roceed in g w ith a review o f p a s t lit e r a t u r e, r e le v a n t c o n c e p ts must be e x p l i c i t l y d efin ed to a ssu re a c le a r u n d erstan d in g o f p r io r resea r ch and th e

24 11 v a r io u s co n cep tu a l approaches to th e a t tr ib u t io n p r o c e s s. In th e t y p ic a l a t tr ib u t io n experim ent! s u b je c ts are exposed to an " a ctio n s i t u a t i o n, " An a c tio n s it u a t io n may be d efin ed a s any sequence o f e v e n ts th a t in v o lv e s one or more p a r tic ip a n ts whose a c tio n s r e s u lt in a o a r t ic u la r o u t com e. An example o f an a c tio n s it u a t io n m ight b e, "Person A d iv e s o f f a d iv in g board and j u s t as he d iv e s o f f, Person B swims up from under w ater and P erson A la n d s on top o f P erson B whose arm i s b ro k en." D if fe r in g in e x p lic it n e s s and d ep th, a c tio n s it u a t io n s p r e s e n t o b je c tiv e in form ation about th e b eh avior o f th e p a r t ic ip a n t s, th e environm ental s e t t i n g, and th e e x te n u a tin g circu m sta n ces surrounding the sequence o f e v e n ts. T h erefo re, an a c tio n s it u a t io n rep res e n ts th e s t im u li, i. e. th e in fo rm a tio n a v a ila b le to an in d iv id u a l on which to base a t t r ib u t io n s o f c a u s a lit y and assignm ent o f r e s p o n s i b ilit y. In th e t y p ic a l paradigm o f a t tr ib u t io n r e se a r c h, a c tio n s it u a t io n s may a c t u a lly be l i v e b eh a v io ra l ev e n ts or may be p r e se n ted as a w r itte n d e s c r ip tio n o f a b e h a v io r a l sequence o f e v e n ts. However, r e g a r d le s s o f th e p r e s e n ta tio n mode, s u b je c ts are ex p erim en ta lly induced to be e it h e r a c to r s who p a r tic ip a te in th e a c tio n s it u a t io n or o b servers who m erely w itn e ss th e a c tio n s it u a t io n. For exam ple, i f a w r itte n d e s c r ip tio n o f an a c tio n s it u a t io n i s g iv e n, then a c to r -s u b je c ts may be asked to im agine o r to c o n sid e r thems e lv e s as p a r t ic ip a n ts in th e s it u a t io n w h ile o b s e r v e r -s u b je c ts

25 12 may be asked to c o n sid e r th em selv es a s w itn e s s e s to th e a c tio n s it u a t io n. F o llo w in g exposure to an a c tio n s it u a t io n, th e a c to r - and o b s e r v e r -s u b je c ts are t y p ic a lly asked a v a r ie ty o f q u e stio n s co n cern in g th e a c tio n s it u a t io n. In p a r tic u la r, a su b je c t may be asked to s p e c ify th e degree to which a p a r tic ip a n t, e ith e r h im s e lf as an a c to r or another person a s an o b se r v e r, i s r e s p o n s ib le fo r th e outcom e. As an exam ple, a c to r - and o b s e r v e r -s u b je c ts may be p resen ted a w r itte n d e s c r ip tio n o f th e fo llo w in g s it u a t io n and asked to a s s ig n r e s p o n s ib ilit y! "You (P erry) dove o f f a d iv in g board and j u s t a s you (P erry) dove o f f, someone swam up from under w ater and you (P erry) landed on top o f him and broke h is arm. A scrib e a p e r c e n ta g e a t f a u lt ' to y o u r s e lf (P erry) and to th e o th e r p erso n." The example was adapted from a stu d y th a t was conducted by D uval and Wicklund ( ). To be more s p e c i f i c about th e n a tu re o f a c tio n s it u a t io n s u sed in r e se a r c h, ev ery a c tio n s it u a t io n can be c o n s id ered to c o n ta in th r e e b a sic components: a "primary a c to r," an " a ctiv e environm ent," and a " fin a l outcom e." The d is t in c t io n s among th e com ponents are fundam ental to the r e sea rch and con cep tu a l approaches to a t tr ib u t io n o f c a u s a lit y and assignm ent o f r e s p o n s ib ilit y. The term "primary actor" r e fe r s to th e c e n tr a l charact e r in an a c tio n s it u a t io n and th e p erso n to whom r e sp o n sib i l i t y i s to be a ssig n e d. Thus, th e t y p ic a l dependent

26 13 measure i s th e amount o f r e s p o n s ib ilit y th a t i s a ssig n ed to the prim ary a c to r. E xp erim en tally, an a c to r -su b je c t t y p ic a lly s e r v e s as th e prim ary a cto r in an a c tio n s it u a t io n w hile an o b server-su b je c t t y p ic a lly serv es a s th e w itn ess o f th e behavior o f th e primary a c to r. Even though more than one person may p a r tic ip a te in an a c tio n s it u a t io n, only one p a r tic ip a n t i s co n sid ered to be th e primary a c to r. In th e above exam ple, Terry" i s the prim ary a cto r fo r an o b se r v e r -su b je c t. The term " active environment" r e f e r s to a l l a sp e c ts o f th e a c tio n s it u a tio n o th e r than the prim ary a c to r. The a c tiv e environm ent in c lu d e s a l l fa c to r s or fo r c e s which in flu e n c e o r surround th e behavior o f th e primary a cto r and/or the p rod uction o f th e outcome. In any p a r tic u la r s itu a tio n, any number o f fa c to r s may op era te as th e a c tiv e environm ent. The a c tiv e environment may in clu d e o th e r p eop le who may o r may not a c tu a lly a c t in producing th e outcome, chance f a c t o r s, f a c i l i t a t i v e fo r c e s, in h ib itin g f o r c e s, or ex ten u a tin g circu m sta n ces. The im p lic a tio n i s th a t the primary a c to r and the a c tiv e environm ent may in te r a c t in any number o f ways to c r e a te a sequence o f ev en ts le a d in g to an outcom e. For exam ple, the prim ary actor may or may n ot be co erced (a f a c i l i t a t i v e fo r c e ) in to perform ing a s o c ia lly u n accep tab le a c t (an in h ib it in g fo r c e ). F in a lly, the th ir d component o f any a ctio n s it u a tio n i s the " fin a l outcome." The f in a l outcome may be p o s it iv e

27 (fa v o r a b le or d e s ir a b le ) or n e g a tiv e (u n fa v o rab le or u n d e sir e a b le ) and may vary in i n t e n s i t y, t h a t i s, th e d egree o f p o s i t i v i t y or n e g a t iv it y (lo w to h ig h ). Although th e f i n a l outcome may n o t be th e o n ly outcome produced by th e in te r a c tio n betw een th e prim ary a c to r and th e a c tiv e environm ent, th e f i n a l outcome i s t y p ic a lly th e outcome fo r w hich th e s u b je c t i s to in d ic a t e th e prim ary a c t o r 's re s p o n s ib ili t y. To rep e a t, r e se a r c h e r s o f th e a t t r ib u t io n p ro cess have sought to d e lin e a te th e e f f e c t o f su b je c t r o le (a c to r o r o b serv er) and outcom e c h a r a c t e r is t ic s (v a le n c e and in t e n s i t y ) on a t tr ib u t io n s o f c a u s a lit y and assignm en t o f resp o n s i b i l i t y. In r e v ie w in g s tu d ie s w hich examined a c to r -o b se r v e r d iffe r e n c e s in th e a t tr ib u t io n o f c a u s a lit y and assignm en t o f r e s p o n s ib ilit y, s tu d ie s can be grouped in to two g e n e r a l c a t e g o r ie s. The f i r s t ca te g o r y in c lu d e s th o s e s tu d ie s th a t have examined a c to r -o b se r v e r d iffe r e n c e s w ith o u t manipul a t in g outcome c h a r a c t e r is t ic s. The second ca teg o ry in c lu d e s th o se s tu d ie s th a t have examined a c to r -o b se r v e r d iffe r e n c e s a s a fu n c tio n o f outcome c h a r a c t e r is t ic s. In g en era l, th e stu d ie s i n th e f i r s t ca teg o r y have y ie ld e d a f a i r l y c o n s is t e n t f in d in g. The g e n e r a lly c o n s is t en t fin d in g has b een th a t a c to r s and o b se r v e r s tend to make d if f e r e n t ty p es o f a t t r ib u t io n s. A ctors have been shown to u s u a lly fo cu s on e x te r n a l, en vironm en tal e x p la n a tio n s fo r t h e ir own a c t io n s w h ile o b se r v e r s have been shovm to

28 15 u s u a lly fo cu s on th e in te r n a l p erso n a l d is p o s it io n s o f th e a cto r in a ccountin g fo r th e a c t o r s "behavior. C oncep tu ally, what th e se fin d in g s su ggested was th a t a c to r s and ob serv ers have d iffe r e n t hut s ta b le p ercep tu a l b ia s e s to a tten d to d iffe r e n t c a u sa l a g en ts in a ccountin g fo r a p e r s o n 's a c tio n s. Simply by im p lic a tio n th en, one m ight ex p ect a c to r s to p la ce r e s p o n s ib ilit y p rim a rily on e x te r n a l environm ental fo r c e s and o b servers to p la c e r e s p o n s ib ilit y p rim a rily on th e a c to r. However, when tu rn in g to resea rch on a cto r-o b server d iffe r e n c e s where outcome c h a r a c t e r is t ic s were a c e n tr a l is s u e, th e fin d in g s have in d ic a te d th a t outcome ch a ra c te r i s t i c s d if f e r e n t i a l l y in flu e n c e th e degree o f r e s p o n s ib ilit y a ssig n ed by a c to r s and o b se r v e r s. Furthermore, th e fin d in g s have n ot been c o n s is t e n t. For exam ple, some resea rch has shown a c to r s to tak e more r e s p o n s ib ilit y fo r p o s it iv e o u t comes than fo r n e g a tiv e outcomes w h ile o th e r resea rch has shown th e o p p o site. In a d d itio n, some resea rch has shown o b serv ers to a s s ig n more r e s p o n s ib ilit y to a c to r s a s negat iv e outcomes became more severe w h ile o th e r resea rch has shown th e o p p o s ite. C o n cep tu a lly, th e se fin d in g s le d resea r ch ers to su g g est th a t a c to r s and o b servers have d i f f e r en t m o tiv a tio n a l b ia s e s, n ot s ta b le p ercep tu a l b ia s e s, to a ttr ib u te c a u s a lity and a s s ig n r e s p o n s ib ilit y to d iffe r e n t c a u sa l a g en ts depending upon the nature o f the f in a l o u t come The p o in t i s th a t resea rch on a cto r-o b serv er d iffe r e n c e s

29 16 h a s r e s u lte d n o t o n ly i n in c o n s is t e n t fin d in g s "but a l s o in a v a r ie t y o f t h e o r e t i c a l e x p la n a tio n s to account f o r th e f in d in g s. The s t u d ie s to be review ed i n th e next s e c t i o n have b een chosen to p o in t ou t th e s e i n c o n s is t e n c ie s and to e x p lic a t e th e v a r io u s t h e o r e t i c a l e x p la n a tio n s t h a t have been g en era ted to accou n t f o r th e s e f in d in g s. F o llo w in g t h i s r e v ie w, a d d it io n a l r e se a r c h and t h e o r e t i c a l p o s i t i o n s w i l l be exam ined to d e v e lo p a p o s s i b l e r e s o lu t io n o f the in c o n s is t e n t f in d in g s. T h is a d d it io n a l r e se a r c h w i l l a lso p r o v id e th e e m p ir ic a l b a s e s f o r th e m ethodology to be used in th e p r e sen t stu d y. L ite r a tu r e Review The m ajor purpose o f th e f o llo w in g b r i e f r ev iew i s to examine th e g e n e r a lly c o n s is t e n t f in d in g th a t a c to r s and o b se r v e r s ten d to make d i f f e r e n t ty p e s o f a t t r i b u t i o n s. G en er a lly, a c t o r s ten d t o a t t r ib u t e c a u s a l i t y to e x te r n a l s i t u a t i o n a l f a c t o r s w h ile o b ser v er s te n d to a t t r ib u t e c a u s a l i t y to th e in t e r n a l p e r s o n a l d i s p o s i t i o n s o f th e a c t o r h im s e lf. N is b e t t, C aputo, L eg a n t, and M aracek (1973) a sk ed s u b j e c t s, a s a c t o r s, to g iv e th e r e a s o n s fo r c h o o sin g t h e ir m ajor f i e l d o f stu dy and f o r l i k i n g th e g i r l s th a t w ere dated m ost f r e q u e n tly. S im ila r ly, th e same s u b je c ts, a s o b s e r v e r s, were asked to g iv e th e r e a so n s th a t t h e i r b e st f r i e n d s had fo r c h o o sin g a m ajor and f o r l i k i n g th e g i r l s th a t were d a te d. R easons were c a te g o r iz e d a s " stim u lu s a ttr ib u tio n s " (e.g. "Chemistry i s a h ig h p a y in g f ie ld " o r "She's a v e r y warm

30 17 person") or a s "person a ttr ib u tio n s " (e.g. "X want to make a l o t o f money" or "I lik e warm g i r l s " ). The fin d in g s showed th a t the s u b je c ts, as a c t o r s, gave p r im a r ily "stim ulus a ttr ib u tio n s " or e x te r n a l reason s when answ ering f o r thems e l v e s. On th e other hand, th e s u b je c ts, a s o b se r v e r s, gave p r im a r ily "person a ttr ib u tio n s " or in te r n a l d is p o s it io n a l r ea so n s when answ ering f o r t h e ir b e s t f r ie n d s. Jones, Rock, Shaver, G oeth als, and Ward (1968) showed th a t s u b je c ts who observed th e perform ance o f accom p lices (a c to r s) on a rig g ed "IQ t e s t " a ttr ib u te d perform ance m ainly to the a b i l i t y o f a c to r s, an in t e r n a l d is p o s it io n a l e x p la n a tio n. In c o n tr a s t, su b je c ts who were a c to r s a ttr ib u te d t h e i r own perform ance on the "IQ t e s t " m ainly to th e l e v e l o f item d i f f i c u l t y, an e x te rn a l environm ental e x p la n a tio n. Other r e se a r c h has a ls o in d ic a te d th a t a c to r s tend to a ttr ib u te t h e i r own b e h a v io r to s it u a t io n a l determ inan ts w h ile o b serv ers tend to a t t r ib u t e th e b eh avior o f a c to r s p r im a r ily to th e p erso n a l d is p o s it io n s o f th e a c to r s (McArthur, 1970} 1972). A t h e o r e t ic a l account o f t h i s g e n e ra l fin d in g was form u lated by J on es and N is b e tt (1971)* The fo rm u la tio n was c a l l e d th e "discrepancy h yp oth esis" and s tr e s s e d th e p ercep tu a l b ia s e s o f a c to r s and o b ser v er s. The b a s ic prem ise o f th e "discrepancy h yp oth esis" i s th a t in d iv id u a ls have a r a th e r s ta b le tendency to fo c u s on th e most s a l i e n t fe a tu r e s o f an a c tio n sequence when making

31 18 a t t r ib u t io n s. However, th e fe a tu r e s o f th e a c tio n s it u a t io n th a t are most s a l i e n t fo r an a c to r and an o b serv er are d if f e r e n t. The m ost s a lie n t fe a tu r e fo r an o b server i s th e b eh avior o f th e a c to r h im s e lf (e.g. what he i s d o in g, what he i s sa y in g, and what a c tio n s he i s s p e c i f i c a l l y t a k in g ). The most s a l i e n t fe a tu r e s f o r an a c to r are s it u a t io n a l f a c t o r s, be th e y o th e r p a r t ic ip a n ts an d/or ex ten u a tin g circu m sta n ces surrounding th e ev en t (i.e. th e a c tiv e en vironm en t). Thus, th e d iscrep a n cy h y p o th e sis argu es th a t a c to r s have a p e r v a siv e and s ta b le ten d en cy to fo c u s t h e ir a tte n tio n on s it u a t io n a l cu es and a s such to d ir e c t a t tr ib u t io n s o f c a u s a lit y toward th e environm ental circu m sta n ces and away from th e m se lv e s. In c o n tr a s t, o b serv ers have a p e r v a siv e and s ta b le tendency to fo c u s t h e i r a t te n t io n on th e b eh avior o f th e a c to r and as such to make a t tr ib u t io n s about th e in te r n a l d is p o s it io n s, m o tiv a tio n s, and p erso n a l c h a r a c te r i s t i c s o f th e a c to r. The d iscrep a n cy h y p o th e sis a s form ulated by J o n es and N is b e tt (1971) d oes n ot c l e a r l y and s p e c i f i c a l l y d ea l w ith a c to r -o b se r v e r d iffe r e n c e s in assignm ent o f r e s p o n s i b i l i t y a s a fu n c tio n o f outcome c h a r a c t e r is t ic s. However, by com bining a fundam ental p r o p o s itio n o f th e a t tr ib u t io n p r o c e ss and th e d iscrep a n cy h y p o th e s is, a rea so n a b le s u g g e stio n co u ld be d eriv ed a s to a c to r -o b se r v e r d iffe r e n c e s in r e s p o n s ib ilit y a ssign m en t.

32 19 The r e le v a n t p r o p o s itio n s t a t e s th a t assignm ent o f r e s p o n s ib ilit y depends upon o n e 's a ttr ib u t io n s o f c a u s a lit y. I t would fo llo w th a t i f one a t t r ib u t e s c a u s a lity p r im a r ily to e x te r n a l environm ental f o r c e s, th en one should a s s ig n r e s p o n s ib ilit y p r im a r ily to th e s e e x te r n a l fo r c e s r a th e r than to th e a c to r h im s e lf, and v ic e v e r s a. In a d d itio n, th e d iscrep a n cy h y p o th e sis s t ip u la t e s th a t a c to r s have a s ta b le tendency to a t tr ib u t e c a u s a lit y p rim a rily to e x te r n a l s it u a t io n a l d eterm in ants w h ile o b serv ers have a s t a b le ten dency to a t tr ib u t e c a u s a lit y p r im a r ily to th e a c to r. T h erefo re, a rea so n a b le su g g e stio n i s th a t a c to r s should dem onstrate a ra th er s ta b le tendency to a ssig n r e s p o n s ib ilit y p r im a r ily to e x te r n a l f o r c e s, n ot to h im s e lf, w h ile o b servers should dem onstrate a r a th e r s ta b le ten d en cy to a s s ig n r e s p o n s ib ilit y p r im a r ily to th e a c to r, n ot to e x te r n a l environm ental f o r c e s. However, resea rch which i s review ed below has in d ic a te d th a t th e d iscrep a n cy betw een a c to r s and o b serv ers i s not s t a b le, but v a r ie s depending upon th e n atu re o f th e f in a l outcom e. For exam ple, rese a r ch has su g g ested th a t fo r n e g a tiv e outcom es, o b se r v ers tend to h old a c to r s more r e sp o n sib le a s n e g a tiv e outcom es become more sev ere w h ile a c to r s tend to hold th em selv es l e s s r e s p o n s ib le a s n e g a tiv e outcom es become more s e v e r e. As fo r p o s it iv e outcom es, rese a r ch has su g g ested th a t th e te n d e n c ie s are r e v e r se d.

33 20 Based on th e s e fin d in g s and o th e r s, resea rch ers have attem pted to e x p la in the in flu e n c e th a t outcome ch a ra c te r i s t i c s have on a c to r s ' and o b serv ers' a ttr ib u tio n s o f c a u s a lit y and assignm ent o f r e s p o n s ib ilit y. G enerally sp ea k in g, r e sea r ch ers have p o stu la te d th a t a cto rs and o b serv ers have a m o tiv a tio n a l b ia s to o p era te in term s o f t h e ir own s e l f - i n t e r e s t s. The u n d erlyin g n o tio n i s th a t a c to r s and o b serv ers atten d t o, a t tr ib u t e, and a ssig n r e s p o n s ib ilit y to th o se c a u sa l fo r c e s w hich are in accord ance w ith a m o tiv a tio n a l b ia s to enhance o n e 's own d e lf - p e r c e p tio n or s e lf-c o n c e p t. A thorough d is c u s s io n o f th e t h e o r e t ic a l ex p la n a tio n s which d ea l w ith outcome c h a r a c t e r is t ic s w i l l be given a f t e r review in g f iv e s tu d ie s th a t have provided em p irical support fo r th e n o tio n th a t a c to r s and ob servers op erate in term s o f t h e ir own s e l f - i n t e r e s t s. The a ttr ib u tio n r o le o f s u b je c ts d iffe r e d from stu d y to stu d y. In two o f th e s tu d ie s, th e s u b je c ts were o n ly ob servers (W alster, 1966; Shaw and S k oln ick, 1971) w h ile in two oth er s t u d ie s, the s u b je c ts were o n ly a cto rs (Johnson, Feigenbaum, & Weiby, 1969; H a rris and Harvey, 1975) Only one study (Harvey, H a r ris, & Barnes, 1975) d e a lt w ith b oth r o le s. The f i r s t m ajor study t h a t d e a lt w ith outcome v a le n c e and in t e n s it y was conducted by W alster ( 1966). The amount o f r e s p o n s ib ilit y th a t o b serv ers a ssig n ed to a " resp on sib le person" (Lennie) fo r an autom obile a c c id e n t was exam ined.

34 21 E s s e n t ia lly, th e a ccid en t in volved Lennie parking h is car a t the top o f a h i l l and s e t t in g the handbrake, but w h ile he was gone, the car r o lle d down th e h i l l. L a ter, p o lic e found th a t the brake c a b le s were ru sted and had broken. The outcome o f the a ccid en t was e ith e r m ild or se v e r e. Mild outcomes in v o lv ed e ith e r a tin y dent in the bumper or a tin y dent in th e bumper, but oth er p eo p le cou ld have been h u rt. Severe outcomes in volved e ith e r a dem olished car or o th er people who were a c tu a lly h u rt. O v e ra ll, th e r e s u lt s showed th a t o b serv ers a ssig n ed more r e s p o n s ib ilit y to Lennie when the outcom es were s e v e r e than when the outcom es were m ild. While W alster was concerned o n ly w ith the e f f e c t o f s e v e r ity o f n eg a tiv e outcom es, Shaw and Skolnick (1971) stu d ied th e e f f e c t o f outcome in t e n s it y fo r both p o s it iv e and n eg a tiv e outcom es. O bservers were given a d e s c r ip tio n o f a Chem istry stu d en t (Jim) who wanted to com plete a lab assignm ent so th a t he could lea v e e a r ly fo r C hristm as v a ca tio n. The outcome o f "Jim m ixing two ch em icals w hile tr y in g to fo llo w th e lab in s tr u c tio n s " was e ith e r an o ffe n s iv e odor (m ild -n e g a tiv e ), an ex p lo sio n (s e v e r e -n e g a tiv e ), a p le a sa n t odor ( m ild - p o s it iv e ), or a major d isco v ery ( s e v e r e - p o s it iv e ). R e su lts showed th a t outcome in te n s ity had a d iffe r e n t e f f e c t fo r p o s it iv e and n eg a tiv e outcom es. For p o s it iv e outcomes evid en ce in d ica ted th a t ob servers a ssig n ed le s s r e s p o n s ib ilit y to the a c to r as p o s it iv e o u t comes became more fa v o r a b le. In c o n tr a s t, fo r n eg a tiv e

35 22 outcomes evidence in d ic a te d th a t o b serv ers a ssig n ed more r e s p o n s ib ilit y to th e a cto r as n e g a tiv e outcomes became more se v e r e. R esearchers have a lso examined th e e f f e c t s o f outcome c h a r a c t e r is t ic s on th e degree to which a c to r s hold thems e lv e s r e sp o n sib le fo r the consequences o f t h e ir own a c tio n s. Johnson, e t. a l. (1969) examined th e e f f e c t o f outcome v a le n c e. A c to r -su b je c ts taught a rith m etic to two stu d en ts who were f i c t i t i o u s l y lo c a te d in an ad jacen t room. The a c to r -s u b je c ts tau gh t th e stu d en ts v ia a one-way speaker system. One stu d en t continued to improve (a p o s it iv e ou t come) and th e second student con tin u ed to do p o o rly (a negat i v e outcom e). The a c to r s tended to take p erso n a l c r e d it fo r th e stu d en t who continued to improve and tended to blame the stu d en t who continued to do p o o rly. H arris and Harvey (1975) examined the e f f e c t o f outcome v a len ce on th e degree to which a c to r s h eld th em selves r e sp o n sib le fo r an outcome. A ctors were asked to s e le c t one o f two le a r n in g ta sk s th a t would be performed la t e r by another s u b je c t, th e "learn er." A fter each a c to r s e le c te d a ta s k, an experim enter to ld each a cto r th a t th e s e le c te d ta sk would e ith e r be p le a sa n t or u n p leasant fo r th e le a r n e r. To measure assignm ent o f r e s p o n s ib ilit y, each a cto r was asked to in d ic a te how much c h o ic e he had in s e le c t in g the ta sk which th e le a r n e r would perform. The fin d in g s showed th a t p erceiv ed c h o ic e was h igher when th e ta sk was to be

36 2 3 p le a s a n t (a p o s it iv e outcome) th a n when th e ta sk was to he u n p lea sa n t (a n e g a tiv e ou tcom e). H arvey, e t. a l. (1975) compared a cto rs* and o b s e r v e r s assign m en t o f r e s p o n s ib ilit y a s a fu n c tio n o f th e s e v e r it y o f n e g a tiv e outcom es. A c to r -s u b je c ts were asked to p resen t a le a r n in g ta sk to a p u p il ( a c t u a lly a co n fed era te) and to a d m in iste r a shock f o r each in c o r r e c t resp o n se made by th e p u p il. The p u p il's d is t r e s s r e a c tio n to th e shocks was e i t h e r m oderate o r se v e r e. O b se r v e r -su b je c ts m erely watched th e p r o c e e d in g s. A ll s u b je c ts, a c to r s and o b se r v e r s, were asked to s p e c ify th e degree to which th e a c to r, th e p u p il, and th e experim enter were r e s p o n s ib le fo r th e p u p il s d i s t r e s s. The fin d in g s r e v e a le d th a t ( l ) o b servers h eld th e a c to r s more r e s p o n s ib le fo r th e n e g a tiv e outcom es th an a c to r s h eld th e m se lv e s, (2 ) o b servers h e ld th e a c to r s more r e sp o n sib le a s n e g a tiv e outcom es became more s e v e r e, but (3 ) a c to r s h e ld th em selv es l e s s r e s p o n s ib le as n e g a tiv e outcom es became more s e v e r e. The "theory o f d e fe n siv e a ttr ib u tio n " was form ulated to a ccou n t fo r th e e f f e c t s o f outcom e v a le n c e and in t e n s it y on a t t r ib u t io n s o f c a u s a lit y and assignm ent o f r e s p o n s ib ilit y. The th eo ry was i n i t i a l l y form u lated to d e a l o n ly w ith o b se r v e r s ' a t t r ib u t io n s o f c a u s a lit y and assignm ent o f r e s p o n s ib ilit y (W a lster, 1$ 66 ; Shaw and S k o ln ic k, ). O ther authors l a t e r extended th e b a sic co n cep ts o f th e th eory

37 ZK to account f o r a c t o r s a ttr ib u tio n s o f c a u s a lity and a ssig n ment o f r e s p o n s ib ilit y (Johnson, e t. a l., 1969; H arris and Harvey, 1975; Harvey, e t. a l., 1975)* The u n d erly in g p r o p o sitio n o f th e theory i s than an ob server o p era te s as i f he might be in v o lv ed in an a c tio n s it u a tio n s im ila r to th e one he j u s t w itn essed and i f so, he could be a p o te n t ia l cau se o f a s im ila r outcom e. I f he c o n sid e r s h im self to be a p o te n t ia l cause o f th e f in a l o u t come in a s im ila r s e t o f circu m sta n ces, then th e ob server w i l l a t tr ib u t e c a u s a lity and a s s ig n r e s p o n s ib ilit y in accordance w ith h is own s e l f - i n t e r e s t s. The ob serv er o p era tes in accordance w ith h is own s e l f - in t e r e s t s in two r e la te d ways. F i r s t, the th eory argues th a t as p o s it iv e and n e g a tiv e outcom es become more extrem e, th e o b serv er s c r u tin iz e s more c l o s e ly th e sequence o f b eh a v io ra l ev en ts in an attem pt to e s ta b lis h th e p r e d ic t a b i l i t y of th e outcome. B e lie v in g th a t th e c a u s e -a n d -e ffe c t r e la tio n s h ip s were p r e d ic ta b le, th e ob server co n v in ces hims e l f th a t he would have caused th e p o s it iv e outcome or would have avoided th e n e g a tiv e outcome in a sim ila r s e t o f circum s ta n c e s. Second, the th eory argues th a t as p o s it iv e and n eg a tiv e outcomes become more extrem e, th e ob server i s m otivated to make d if f e r e n t ty p es o f in fe r e n c e s about the c a u sa l r o le o f th e a c to r. As n eg a tiv e outcomes become more se v e r e, th e ob server i s in c r e a s in g ly m otivated to in fe r th a t the n e g a tiv e

38 25 outcom es co u ld have "been caused o n ly by th a t s p e c if ic a c to r. As such, th e ob serv er i d e n t i f i e s or a s s o c ia t e s th e a c to r as th e unigue sin g u la r ca u se o f th e u n d e sir a b le n e g a tiv e o u t com es. In c o n tr a s t, a s p o s it iv e outcom es become more fa v o r a b le, th e o b server i s in c r e a s in g ly m o tiv a ted to avoid th e in fe r e n c e th a t th e p o s it iv e outcom es cou ld have been caused o n ly by th a t s p e c if ic a c to r. As such, th e o b serv er d is s o c i a t e s th e a c to r as th e unigue sin g u la r ca u se o f th e d e s ir a b le p o s it iv e outcom es. Through th e se two r e la te d p r o c e s s e s, th e o b server opera t e s in accordance w ith h is own s e l f - i n t e r e s t s. By co n v in c in g h im s e lf th a t he would have avoided th e n e g a tiv e outcomes and by a s s o c ia t in g th e a c to r a s th e u nigu e cause o f th o se n e g a tiv e outcom es, th e o b server defends o r in c r e a s e s th e p o s s i b i l i t y th a t o n ly th e a c to r cou ld have caused th e u n d e s ir a b le outcom es. On th e o th e r hand, by c o n v in c in g h im s e lf th a t he a lso would have caused th e p o s it iv e outcom es and by d is s o c ia t in g th e a c to r a s th e u nigu e ca u se o f th o se p o s it iv e outcom es, th e o b server d efen ds a g a in st o r d ecrea se s th e p o s s i b i l i t y th a t o n ly th e a c to r cou ld have cau sed th e b e n e f ic ia l outcom es. Based on t h i s argument a s form ulated by W alster ( 1966) and Shaw and S k oln ick ( ), th e th eo ry makes d if f e r e n t p red ic t io n s about o b se r v e r s' a t tr ib u t io n s o f c a u s a lity and assignm ent o f r e s p o n s ib ilit y fo r p o s it iv e and n e g a tiv e o u t com es. As p o s it iv e outcom es become more fa v o r a b le, th e

39 26 o b s e r v e r 's b ia s i s to co n v in ce h im self th a t he a lso would have caused th e p o s it iv e outcome and to d is s o c ia t e th e a c t o r a s th e u n igu e cau se o f th a t outcom e. The endp ro d u ct i s th a t a s p o s it iv e outcom es become more fa v o r a b le, th e o b server a llo c a t e s l e s s c a u s a lit y and l e s s r e sp o n sib i l i t y to th e a c t o r. As n e g a tiv e outcom es become more s e v e r e, th e o b s e r v e r 's b ia s i s to co n v in ce h im self th a t he would have avoided the n e g a tiv e outcome and to a s s o c ia t e th e a c to r a s th e unigue ca u se o f th a t outcom e. The end- p rod u ct i s th a t a s n e g a tiv e outcom es become more s e v e r e, th e o b serv er a l l o c a t e s more c a u s a lit y and more r e sp o n sib i l i t y to th e a c to r. However, f o r sake o f co m p leten ess and as a v ia b le a lt e r n a t iv e to th e above argum ent, W alster (1967) proposed a somewhat d if f e r e n t and more parsim onious argument to d e fe n s iv e a t t r ib u t io n. W a lster* s argument (196?) d id n o t le a d to p r e d ic tio n s o f o p p o s ite a t t r ib u t io n a l te n d e n c ie s f o r p o s it iv e and n e g a tiv e outcom es. Her argument d e a lt w ith th e f i r s t p r o c e ss,- a s p r e v io u s ly d is c u s s e d, by w hich o b se r v e r s o p era te in accordance w ith t h e ir own s e l f - i n t e r e s t s. W a lster argued f o r a d ir e c t r e la tio n s h ip betw een outcome in t e n s it y and r e s p o n s ib ilit y assignm en t. Simply p u t, W alster proposed th a t a s p o s it iv e o r negat i v e outcomes become more extrem e, o b serve rs tend to con v in c e th em selv es th a t th ey would have p r e d ic te d th e outcom e. H aving e s ta b lis h e d th e p r e d ic t a b ilit y o f th e outcome,

40 ob servers h old a cto rs resp o n sib le fo r p o s it iv e outcomes th ey would have caused and h old a cto rs r e sp o n sib le fo r n eg a tiv e outcomes th ey would have avoided. The end-product i s th a t a s p o s it iv e and n eg a tiv e outcomes become more extrem e, o b servers w i l l hold a c to r s more r e sp o n sib le fo r th e f in a l outcom e. The p rev io u s d is c u s s io n o f d e fe n siv e a ttr ib u tio n d e a lt o n ly w ith o b serv ers' a ttr ib u tio n s o f c a u s a lity and a ssig n ment o f r e s p o n s ib ilit y. However, e x te n sio n o f th e b asic con cep ts to a c to r s ' a ttr ib u tio n s o f c a u s a lity and a ssig n ment o f r e s p o n s ib ilit y f o r th e ir own a c tio n s an d/or produced outcome i s f a i r l y d ir e c t. To m aintain s e lf-e s te e m or to make a ttr ib u tio n s th a t are in accordance w ith o n e 's own s e l f - i n t e r e s t s, a cto rs have a m o tiv a tio n a l b ia s to d is s o c i a te th em selves from th e production o f a n eg a tiv e outcome and as such, to a ttr ib u te l e s s c a u s a lity and to a ssig n l e s s r e s p o n s ib ilit y to th em selv es as n e g a tiv e outcomes become more se v e r e. S im ila r ly, a s p o s it iv e outcomes become more fa v o ra b le, a c to r s have a m o tiv a tio n a l b ia s to a s s o c ia te th em selves w ith the p ro d u ctio n o f a p o s it iv e outcome and as such, to a ttr ib u te more c a u s a lity and to a ssig n more r e s p o n s ib ilit y to th em selv es. The g en e r a l c o n c lu sio n to be drawn from th e p revious d is c u s s io n s o f d e fe n siv e a ttr ib u tio n and the d iscrep ancy h y p o th esis i s as fo llo w s. As s tip u la te d by the d iscrep ancy h y p o th e sis, a c to r s may have a rath er stable tendency to focu

41 28 on s it u a tio n a l determ inants o f h is own behavior and/or outcome w h ile o b servers may have a ra th er s ta b le tendency to fo cu s on in te r n a l d is p o s it io n a l determ inants o f th e a c t o r 's b eh avior. However, th e d if f e r e n t i a l ten d en cies o f a c to r s and o b serv ers may be superseded by m o tiv a tio n a l b ia s e s to a ttr ib u te c a u s a lity and a s s ig n r e s p o n s ib ilit y in accordance w ith o n e 's own s e l f - i n t e r e s t s as a fu n c tio n o f th e p o s i t i v i t y or n e g a tiv ity o f th e f i n a l outcome. As in d ica ted e a r lie r, se v e r a l s tu d ie s have y ie ld e d fin d in g s th a t have f a ile d to support th e gen eral p ro p o sit io n s d erived from d e fe n siv e a ttr ib u tio n. Four s tu d ie s w i l l be review ed th a t have co n tra d ic te d th e d e fe n siv e a ttr ib u tio n approach to r e s p o n s ib ilit y assignm ent. F ollow in g t h is review, a p o s s ib le r e s o lu tio n o f the problem s w i l l be developed. Shaver (1970) conducted a study sim ila r to W alster ( 1966) in which th e e f f e c t o f s e v e r ity o f n eg a tiv e outcomes on o b se r v e r s assignm ent o f r e s p o n s ib ilit y to a c to r s was examined. O b serv er-su b jects were given th e d e s c r ip tio n o f th e autom obile a c c id e n t in v o lv in g L ennie. The outcome was e ith e r m ild or se v e r e. The m ild outcome in v o lv ed a t in y dent in the bumper, but o th er p eop le could have been h u rt. The severe outcome in v o lv ed o th er p eop le who were a c tu a lly h u rt. U n lik e th e study by W alster ( ), the a ccid en t account was p resen ted as a courtroom case which was to be used in a la t e r study on jury functioning.

42 29 W alster (1966) p resen ted th e a ccid en t account as stim u lu s m a te r ia ls which were to be used in a l a t e r study on s o ld ie r s ' a b i l i t y to a c c u r a te ly r e c a ll e v e n ts. C ontrary to p r e d ic tio n s from d e fe n siv e a ttr ib u tio n, Shaver found th a t s e v e r ity o f n e g a tiv e outcomes did not have a s ig n if ic a n t e f f e c t on o b serv ers' assignm ent o f r e s p o n s ib ilit y. W alster (196?) conducted two experim ents to a s s e s s o b serv ers' assignm ent o f r e s p o n s ib ilit y a s a fu n ctio n o f th e in t e n s it y o f p o s it iv e and n e g a tiv e outcom es. R ecall th a t W a lste r 's form u lation o f d e fe n siv e a ttr ib u tio n p red icted th a t a s p o s it iv e and n e g a tiv e outcomes become more extrem e, o b servers should a s s ig n more r e s p o n s ib ilit y to the a c to r. For both exp erim en ts, o b se r v e r -su b je c ts were given a d e s c r ip tio n o f an a cto r who decided to buy a house, but had gained or lo s t d if f e r e n t amounts o f money due to environm ental circu m sta n ces. The f i r s t experim ent in volved "Mrs. W allace who had decided to buy a house a f t e r much d e lib e r a t io n o f th e p o s s ib le r is k s and ad van tages. The environm ental circu m stan ces th a t le d to a g a in was a la n d s lid e th a t d ep o sited a v a lu a b le m ineral on th e p rop erty. The circu m stan ces th a t led to a l o s s was a la n d s lid e th a t d estroyed th e p ro p erty. The second experim ent in volved "Alex K endler who had d ecid ed to buy a house and knew th a t th e value o f th e house would r is e o r f a l l depending upon government renew al o f resea rch c o n tr a c ts in th e a rea. The fin d in g s o f both experim ents were sim ila r. O bservers

43 30 tended to hold th e a c t o r s l e s s r e s p o n s ib le a s both p o s it iv e and n e g a tiv e outcom es became more extrem e. Thus, w ith r e sp e c t to s e v e r ity o f n e g a tiv e outcom es, th e fin d in g s were o p p o site to p r e d ic tio n s b ased on d e fe n s iv e a t tr ib u t io n. W ith r e s p e c t to th e f a v o r a b ilit y o f p o s it iv e outcom es, th e f in d in g s were o p p o s ite to p r e d ic tio n s o f d e fe n siv e a t t r ib u t io n a s form u lated by W a lster (1967)* However, the fin d in g s r e la te d to p o s it iv e outcomes d id conform to p r e d ic tio n s o f d e fe n s iv e a t t r ib u t io n as form u lated by W alster (1966) and Shaw and S k o ln ick ( ). R esearch on a c t o r s ' assignm en t o f r e s p o n s ib ilit y have a lso y ie ld e d fin d in g s in c o n s is t e n t w ith p r e d ic tio n s based on d e fe n s iv e a t t r ib u t io n and th e co rresp o n d in g con cep t o f s e l f - i n t e r e s t. R oss, B ierb ra u er, and P o lly (197*0 conducted a stu d y s im ila r to t h a t by Johnson, e t. a l. (1969) in exam ining th e e f f e c t o f p o s i t iv e and n e g a tiv e outcom es on a c to r s ' assignm ent o f r e s p o n s ib ilit y to th e m se lv e s. In a d d itio n, R oss, e t. a l. compared th e assignm en t o f r e s p o n s ib ilit y by a c to r s and o b se r v e r s. A c to r -s u b je c ts, p r o f e s s io n a l and n o n -p r o fe ssio n a l in s t r u c t o r s, ta u g h t s p e llin g to a boy who was f i c t i t i o u s l y lo c a te d behind a one-way m irror. O b serv er-su b je cts were p a ired w ith n o n -p r o fe ssio n a l in s t r u c to r s o n ly. The fin d in g s r e v e a le d th a t o b serv ers saw th e nonp r o fe s s io n a l in s t r u c t o r s as a more im portant determ inant o f stu d en t f a ilu r e th a n o f stu d e n t s u c c e s s. N o n -p r o fessio n a l

44 31 in s tr u c to r s saw th em selves as a l e s s im portant determ inant fo r stu d en t f a ilu r e than fo r studen t s u c c e s s. Even though th e se fin d in g s supported p r e d ic tio n s based on d efen siv e a ttr ib u tio n, o th er fin d in g s were not in lin e w ith d efen s iv e a ttr ib u tio n. F ir s t, p r o fe s s io n a l in s tr u c to r s saw th em selves a s a more im portant determ inant fo r student f a ilu r e than fo r stu d en t s u c c e s s. Second, when stu d en ts f a ile d, both groups o f in s tr u c to r s did n ot a c tu a lly blame the stu d en ts and when stu d en ts succeeded, both groups o f in s tr u c to r s did not a c tu a lly take c r e d it. F in a lly, Wortman (1975) conducted a study which in d i ca ted th a t a c to r s hold th em selves more resp o n sib le fo r n eg a tiv e outcomes than fo r p o s it iv e outcomes under c e r ta in c o n d itio n s. A c to r -su b je c ts were given a chance to r e c e iv e e ith e r an a t tr a c t iv e item (p o s itiv e outcome) or an u n a ttr a c tiv e item (n e g a tiv e outcome) by a chance drawing o f a blue or red m arble. Some su b je c ts were to ld b efo rehand- which marble corresponded to th e a t tr a c t iv e and u n a ttr a c tiv e item w h ile o th er su b je c ts had no such fo r e knowledge. The fin d in g s in d ica ted th a t a c to r s who had foreknowledge held th em selves more r e sp o n sib le when the outcome was n eg a tiv e than when th e outcome was p o s it iv e. A ctors who d id n o t have foreknowledge tended to hold them selves more r e sp o n sib le fo r p o s it iv e outcomes than fo r n eg a tiv e outcom es. Wortman su ggested th a t the degree to which a person h o ld s h im self resp o n sib le defends upon

45 32 having foreknow ledge about th e p o s s ib le consequences o f h i s b eh avior. Developm ent o f R eso lu tio n As th e p rev io u s rev iew in d ic a t e s, p a st resea rch th a t has d e a lt s p e c i f i c a l l y w ith a c to r s ' and o b se r v e r s' assignm ent o f r e s p o n s ib ilit y a s a fu n c tio n o f outcome c h a r a c t e r is t ic s has y ie ld e d in c o n s is te n t f in d in g s. For exam ple, some resea rch found o b serv ers a ssig n ed more r e s p o n s ib ilit y to a c to r s a s n eg a tiv e outcom es became more sev ere ( e.g. W a lster, 1966) w h ile o th e r resea r ch found th e o p p o site ( e.g. W a lster, 1967) and s t i l l o th er rese a r ch found no s u b s ta n tia l e f f e c t due to th e s e v e r ity o f n e g a tiv e outcom es ( e. g. Shaver, ). Furtherm ore, some resea r ch found a c to r s took more r e s p o n s i b i l i t y f o r p o s it iv e outcom es than fo r n e g a tiv e outcomes ( e.g. Johnson, e t. a l., 19&9) w h ile o th er resea r ch found th e o p p o site ( e.g. R oss, e t. a l., 197*0* T h erefo re, n e it h e r th e d iscrep a n c y h y p o th e sis nor d e fe n s iv e a t tr ib u t io n have been c o n s is t e n t ly supported in a cco u n tin g f o r a c to r -o b se r v e r d iffe r e n c e s in r e s p o n s ib ilit y assignm en t. M oreover, a major gap e x is te d in the lit e r a t u r e co n cern in g a c to r -o b se r v e r d iffe r e n c e s a s a fu n c tio n o f outcome c h a r a c t e r is t ic s. P a st rese a r ch h a s not examined a c to r- o b serv er d iffe r e n c e s in r e s p o n s ib ilit y assignm en t as a fu n c tio n o f outcome c h a r a c t e r is t ic s in th e same exp erim en tal s it u a t io n. Hence, an o v e r a ll o b je c tiv e o f th e p resen t stu d y i s to f i l l t h is void in e x is t in g l it e r a t u r e. But more

46 33 im p ortan tly, t h is stu d y w ill a lso propose and t e s t a p o s s i b le r e s o lu tio n o f th e in c o n s is te n t fin d in g s, a su b ject to which we now tu rn. An ex p la n a tio n o f th e co n fu sio n and in c o n s is te n c ie s in a ttr ib u tio n resea r ch may be p o s s ib le i f co n sid e r a tio n i s given to some o f th e major p r o p o sitio n s issu e d in con cep tu al approaches to th e a ttr ib u tio n p r o c e s s. To b egin th e d evelop ment o f a p o s s ib le r e s o lu tio n o f th e problem s in p a st resea r ch, r e c a ll th e gen eral co n cep tu a l framework o f a t t r ib u tio n rese a r c h. The o v e r a ll approach has been to p o stu la te a p ro c e ss which m ed ia tes an a t t r ib u t o r s resp o n ses to a s e t o f s t im u li. The resp onses are th e a ttr ib u tio n s o f c a u s a lity and assignm ent o f r e s p o n s ib ilit y by a c to r s and o b serv ers. The s tim u li are th e a c tio n s it u a t io n s, i. e. th e a v a ila b le in form ation about th e in te r a c tio n between the "primary actor" and th e " active environment" in producing the " fin a l outcome." The m ed iating p r o cess does d i f f e r depending upon the s p e c if ic approaches o ffe r e d by th e d iscrep ancy h y p o th esis and by d e fe n siv e a t tr ib u t io n. However, both approaches r e s t upon the same fundamental p r o p o sitio n o f the a ttr ib u tio n p r o c e s s. Soth th e o r ie s assume th a t a c to r s and ob serv ers base t h e ir a ttr ib u tio n s o f c a u s a lity and assignm ent o f r e s p o n s ib ilit y on th e p r o cessin g o f a v a ila b le inform ation about th e in te r a c tio n between th e primary a cto r and the a c tiv e environm ent in producing th e f in a l outcome.

47 E m p irica lly though, resea r ch ers have focu sed t h e ir a tte n tio n on ly on th e e f f e c t s o f outcome c h a r a c te r is tic s on th e resp onses o f a c to r s and o b serve rs, i. e. on ly th e dim ensions o f one component o f a c tio n s it u a tio n s have receiv ed major a t te n t io n, namely th e valen ce and in t e n s it y o f th e f in a l outcome. Thus, l i t t l e em p irica l a tte n tio n has been given to id e n t if ia b le dim ensions o f th e rem aining two components o f a c tio n s it u a t io n s, namely th e,p rim ary a c to r and the a c tiv e environm ent. But both t h e o r e t ic a l approaches have su ggested th a t d iffe r e n t dim ensions o f th e se components are im portant determ inants o f c a u sa l a ttr ib u tio n s and r e s p o n s ib ilit y assign m en t. As w i l l be proposed and illu s t r a t e d s h o r tly, s tu d ie s have in a d v e r te n tly varied th e se stim u lu s dim ensions and as such have produced in co n s is t e n t fin d in g s. The d iscrep ancy h y p o th e sis and d e fe n siv e a ttr ib u tio n have each su ggested a d iffe r e n t stim u lu s dim ension to be an im portant determ inant o f a c to r s ' and o b se r v e r s resp o n se s. The d iscrep ancy h y p o th e sis su g g e sts th a t th e most im portant dim ension o f a c tio n s it u a t io n s in d eterm ining ca u sa l a t t r ib u tio n s and r e s p o n s ib ilit y assignm ent i s th e " sa lien ce o f in form ation." The th eory argues th a t an a c to r and an ob server tend to d ir e c t a ttr ib u tio n s o f c a u s a lity and a s s ig n r e s p o n s ib ility to th e most s a lie n t components o f an a c tio n s itu a tio n. Thus, th e most im portant stim u lu s dim ension in

48 35 * d eterm in in g th e a ttr ib u to r * s resp o n ses i s th e p ercep tu a l s a lie n c e o f th e prim ary a c to r or th e a c tiv e environment in producing th e f i n a l outcom e. The th eo ry su g g e sts then th a t i f th e b eh avior o f th e primary a c to r i s p e r c e p tu a lly most s a lie n t to th e a ttr ib u to r, then th e a ttr ib u to r w i l l tend to d ir e c t c a u sa l a ttr ib u tio n s and a ssig n r e sp o n sib i l i t y to the prim ary a c to r. On th e o th er hand, i f th e a c t iv e environm ent i s p e r c e p tu a lly m ost s a lie n t, then th e a ttr ib u to r w i l l d ir e c t c a u sa l a ttr ib u tio n s and a s sig n r e s p o n s ib ilit y to th e a c t iv e environm ent. In c o n tr a st to th e d iscrep an cy h y p o th e sis, th e d e fe n siv e a ttr ib u tio n approach su g g e sts th a t th e most im portant dimens io n o f a c tio n s it u a t io n s in d eterm in in g ca u sa l a ttr ib u t io n s and r e s p o n s ib ilit y assignm en t, o th er than th e nature o f th e f in a l outcom e, i s th e "nature o f th e in te r a c tio n between the prim ary a c to r and th e a c t iv e environm ent." T his l a t t e r stim ulu s dim ension r e fe r s to th e nature o f th e r o le s o f th e r e s p e c tiv e com ponents, th a t i s, th e primary a cto r and a sp e c ts o f th e a c tiv e environm ent in a p a r tic u la r s it u a t io n. T h is dim ension r e la t e s to q u e stio n s such a s, "Does th e prim ary a cto r in t e n t io n a lly produce th e f in a l outcome?" and "Does th e a c tiv e environm ent f a c i l i t a t e or in h ib it th e a c t o r 's behavior?" The th eory o f d e fe n siv e a ttr ib u tio n su g g ests thn.t th e o p era tio n o f s e l f - i n t e r e s t r e s t s upon the p o ssib ility th a t th e ca u sa l r o le shared by th e primary a cto r and th e a c tiv e

49 36 environm ent i s a t l e a s t su b je c t to d i f f e r e n t i a l in te r p r e ta t io n s on th e p a rt o f a c to r s v e r su s o b se r v e r s. For exam ple, a b a sic p o s tu a lte o f d e fe n s iv e a t t r ib u t io n i s th a t an o b server a ttem p ts to e s t a b lis h th e p r e d ic t a b ilit y o f th e f i n a l outcome so as to con vin ce h im s e lf th a t he would have had c o n tr o l in p rod ucing th e p o s it iv e outcome or in a v o id in g th e n e g a tiv e outcome in a s e t o f circu m sta n ces s im ila r to th a t w itn e sse d. However, i f a l l ev id en ce o r in fo rm a tio n about th e s it u a t io n d ic t a t e s th a t th e prim ary a c to r had a b s o lu te ly no c o n tr o l over th e a c tiv e environm ent in p rod u cin g th e f i n a l outcom e, th en th e o b serv e r, and th e a c to r fo r th a t m a tte r, would have d i f f i c u l t y i n in te r p r e tin g th e sequence o f e v e n ts so as to d i s count th e a c t iv e environm ent a s c a u sin g o r a s b e in g r e s p o n s ib le fo r th e p ro d u ctio n o f th e f i n a l outcom e. On th e o th e r hand, i f a l l in fo r m a tio n p o in ts to th e prim ary a c to r a s h avin g had t o t a l and com plete c o n tr o l over th e p ro d u ctio n o f th e f i n a l outcom e, th en b o th th e o b serv er and th e a c to r would have d i f f i c u l t y in in te r p r e tin g th e sequence o f ev e n ts so a s to c r e d it th e a c t iv e environm ent w ith th e p ro d u ctio n o f th e f i n a l outcom e. However, i f th e in fo r m a tio n or ev id en ce i s r a th e r ambiguous or e q u iv o ca l in d ic t a t in g th e r e la t iv e c o n tr ib u tio n o f th e prim ary a c to r and th e a c t iv e environm ent in p rod u cin g th e f i n a l outcom e, th en th e o b server and th e a c to r a re more a b le to in te r p r e t th e sequence o f e v e n ts so a s to a t tr ib u t e c a u s a lit y and a s s i g n.r e s p o n s ib i lit y in accordance w ith t h e ir own s e l f - in t e r e s t s.

50 37 To r e it e r a t e, th e su g g estio n b ein g made i s th a t fo r a l l p r a c t ic a l purposes, resea r ch ers have e m p ir ic a lly s tr e s s e d on ly the dim ensions o f th e f in a l outcome and have t h e o r e t ic a lly s tr e s s e d on ly the p r o c e ss which i s assumed to m ediate an a t t r ib u t o r 's assignm ent o f r e s p o n s ib ilit y fo r a low to high p o s it iv e or n eg a tiv e outcom e. I t appears th a t th ese resea r ch ers were s t r e s s in g the study of on ly th o se fa c to r s ( i. e. outcome c h a r a c te r is tic s ) which were o f primary concern in t h e ir p r e d ic tio n s regard in g a c to r -o b se r v e r d iffe r e n c e s in r e s p o n s ib ilit y assign m en t. E m p irically though, resea rch ers have e s s e n t i a lly ign ored the two dim ensions o f a c tio n s it u a tio n s th a t are su g g ested to be o f major im portance in determin in g assignm ent o f r e s p o n s ib ilit y. These two dim ensions are the p e rcep tu a l s a lie n c e o f th e r o le o f the prim ary a c to r and th e a c tiv e environment in producing the f i n a l outcome and the nature o f th e in te r a c tio n among th e se ca u sa l a g en ts. As a r e s u lt, p r io r s tu d ie s have s y s te m a tic a lly v a ried th e r o le o f su b je c ts and th e c h a r a c te r is tic s o f th e f in a l outcome, but have seemed to in a d v e r te n tly and u n sy ste m a tic a lly vary the nature o f th e a c tio n s it u a t io n s in terms o f th e s a lie n c e o f and th e nature o f th e in te r a c tio n between th e primary a cto r and th e a c tiv e environm ent. P o s s ib ly, by f a i l i n g to system a t i c a l l y take in to account th e se two im portant stim u lu s dimens io n s, p a s t research has r e s u lte d in in c o n s is te n t and co n fu sin g f in d in g s. However, th ese in c o n s is t e n c ie s might be reso lv ed by a c lo s e exam ination o f th e experim ental s it u a t io n s r e la t iv e

51 38 to the " sa lie n c e o f in fo rm a tio n about the primary a cto r and th e a c tiv e environm ent and th e "nature o f th e in te r a c tio n between th e prim ary a c to r and th e a c tiv e environm ent." To i l l u s t r a t e th e v i a b i l i t y o f th e above argument in r e s o lv in g th e problem s o f p a st resea r ch on a c to r -o b se r v e r d iffe r e n c e s in r e s p o n s ib ilit y assignm en t, two p a ir s o f s tu d ie s th a t were p r e v io u sly review ed can be compared: (1) W alster (1966) and W alster (196?) and (2) Johnson, e t. a l. (1969) and R oss, e t. a l. (197^)* The com parison between W alster (1966) and W alster (196?) d e a ls p rim a rily w ith th e n atu re o f th e in te r a c tio n between th e primary a c to r and th e a c t iv e environm ent. W alster*s s tu d ie s examined o b serv ers' assignm ent o f r e s p o n s ib ilit y to a c to r s. On the oth er hand, the com parison between Johnson, e t. a l. (1969) and R oss, e t. a l, (197*0 d e a ls p rim a rily w ith th e s a lie n c e o f in form ation about the prim ary a c to r and th e a c tiv e environm ent. These two s tu d ie s examined a c to r s ' assignm ent o f r e s p o n s ib ilit y to th e m se lv es. W alster ( 1966; 1967) examined o b serv ers' r e s p o n s ib ilit y assignm ent to a cto rs a s a fu n c tio n o f th e s e v e r ity o f n eg a tiv e outcom es. W alster (1966) found th a t o b servers a ssig n ed more r e s p o n s ib ilit y to th e a c to r (L ennie) as th e n eg a tiv e outcome (autom obile a ccid en t) became more se v e r e. R e su lts supported th e d e fe n siv e a t tr ib u t io n p r e d ic tio n th a t a s n eg a tiv e outcomes become more se v ere, o b servers w i l l hold a c to r s more r e s p o n s ib le. Then in 1967, A s im ila r study by W alster produced o p p o site r e s u lt s. W alster (1967) found th a t ob servers a ssig n ed l e s s r e s p o n s ib ilit y to th e a c to r ( e.g. Mrs. W allace) a s th e n e g a tiv e outcome ( e. g. amount o f money l o s t due to a la n d s lid e ) became

52 39 more se v e r e. W alster ( 1967) did n o t o f f e r an ex p la n a tio n fo r th e s e fin d in g s. The a c tio n s it u a t io n s in the two s tu d ie s d iff e r e d not o n ly in term s o f th e nature o f th e a c t o r 's r o le in p rod ucing the outcome but a ls o in term s o f the in fo rm a tio n th a t was emphasized or made s a lie n t to the o b serv ers. P araphrasing, W alster (1966) d escrib ed L ennie as h avin g ju s t bought a 6 year o ld ca r w ith r u ste d brake c a b le s and a s having s e t th e handbrake a f te r parkin g a t th e top o f the h i l l. In c o n tr a s t, W alster (196?) d escrib ed M rs. W allace a s having q u ite a b it o f p r io r in form ation on which to base a d e c is io n fo r b uyin g the h ouse. Mrs. W allace was d escrib ed a s having d e lib e r a te d e x te n s iv e ly the r is k s and advantages in v o lv ed in b uyin g th e h ouse, but as h avin g no c o n tr o l over th e environm ental f a c t o r s (la n d s lid e ) which would determ ine th e amount o f money l o s t. Furthermore, she knew th a t the l i k e l i hood o f a la n d s lid e r e s u lt in g in a severe lo s s was q u ite low a s in d ic a te d in th e fo llo w in g statem ent: "each year one fam ily o f th e 4-00 in th e h i l l s had a home s e v e r e ly damaged by mud from a la n d s lid e, w h ile two o r th ree out o f th e 4-00 uncovered z e a lr it e and made a fo r tu n e." A rea so n a b le e x p la n a tio n fo r th e in c o n s is te n t fin d in g s may be a s fo llo w s. In her 1967 study, W alster s tr e s s e d th e a c t o r 's e x te n siv e d e lib e r a tio n in th e d e c is io n to buy th e house and th e r o le o f "chance" in p rod u cin g the outcom e. The o b serv ers sim ply made a ttr ib u tio n s and a ssig n e d r e s p o n s ib ilit y c o n s is te n t w ith th e ca u sa l f a c t o r s th a t were em phasized. By knowing o f Mrs. W a lla ce's e x te n siv e d e lib e r a tio n in th e d e c is io n to buy th e h ouse, th e ob servers h e ld her r e sp o n sib le ra th er th an not

53 4o r e s p o n s ib le r e g a r d le s s o f th e s e v e r it y o f th e n e g a tiv e outcom e. A lso, by knowing th a t th e lik e lih o o d o f a se v e r e lo s s was l e s s than th e lik e lih o o d o f a sm all l o s s, th e o b se r v e r s tended to a t tr ib u t e c a u s a lit y more to chance a s th e n e g a tiv e outcome became more se v e r e. However, in W alster ( 1966) n e ith e r th e a c t o r 's b eh a v io r nor th e r o le o f chance was em phasized. Thus, a rea so n a b le su g g e stio n i s th a t th e o b serv ers were more a b le to in t e r p ret th e s it u a t io n a s one in w hich th e a c to r had a more i n f l u e n t ia l r o le in p rod ucing th e outcome th a n chance f a c t o r s and a s su ch, o b serv ers h eld L ennie more r e s p o n s ib le a s th e n e g a tiv e outcome became more s e v e r e. The im p lic a tio n i s th a t th e two a c tio n s it u a t io n s a re a c tu a lly d if f e r e n t in th e n atu re o f th e c a u s a l r o le sh ared by the a c to r and th e a c tiv e environm ent in p rod ucing th e f i n a l outcom e. F urth er support f o r th e im portance o f d if f e r e n t i a t in g between th e two stim u lu s d im ensions in a cco u n tin g fo r in c o n s is t e n t fin d in g s can be based on a com parison betw een th e s t u d ie s by Johnson, e t. a l. ( 1969) and R o ss, e t. a l. (197^ ). Both s tu d ie s examined a c t o r s ' r e s p o n s ib ilit y a s s ig n ment f o r th e su c c e ss or f a ilu r e o f stu d en ts whom the a c to r s ta u g h t. Johnson, e t. a l. found a c to r s took c r e d it fo r stu d en t s u c c e s s and blamed th e stu d e n t fo r f a i l u r e. R oss, e t. a l. found a c to r s did n o t a c tu a lly ta k e c r e d it fo r stu d en t s u c c e s s nor d id a c to r s a c t u a lly blame th e stu d en t fo r f a i l u r e. However, J o h n so n 's a c to r s tau gh t stu d e n ts

54 v ia a one-way speaker system w h ile R o ss's a c to r s taught stu d en ts v ia a one-way m irror. R oss, e t. a l. p oin ted out th a t th e a c to r s a c tu a lly saw t h e ir own beh avior as r e fle c t e d by th e m irror and th e a c to r s knew th ey were b ein g ev a lu a ted. R oss, e t. a l, su g g ested th a t th e v is u a l p e r sp e c tiv e provided by th e m irror and the e v a lu a tio n a sp ect l i k e l y in flu e n c e d th e s a lie n c e o f a c to r s ' behavior and in tu rn, a c to r s ' a ttr ib u tio n s o f c a u s a lity and a ssig n ment o f r e s p o n s ib ilit y. In summary, th e key to r e s o lv in g th e in c o n s is te n c ie s in research may be th e r e c o g n itio n th a t c a u sa l a ttr ib u tio n s and r e s p o n s ib ilit y assignm ent are a fu n c tio n not on ly o f the a t tr ib u t o r s r o le and th e nature o f th e f in a l ou t come but a ls o two dim ensions o f a c tio n s it u a t io n s. Again, th e se dim ensions are ( 1 ) th e o b je c tiv e in form ation about th e nature o f th e in te r a c tio n between the prim ary a cto r and the a c tiv e environm ent and ( 2) th e o b je c tiv e inform a tio n which i s most s a lie n t. These two dim ensions o f a c tio n s it u a tio n s have been e m p ir ic a lly and t h e o r e t ic a lly addressed in the resea rch which i s d iscu sse d below. Each o f th ese dim ensions i s d iscu ssed a t le n g th, e s p e c ia lly In terms o f th e ir relev a n ce to th e p resen t stu d y. The dim ension r e la t in g to th e s a lie n c e o f Inform ation i s d iscu sse d f i r s t. Research has dem onstrated th a t th e s a lie n c e o f inform a tio n in flu e n c e s p e r c e p tio n s o f c a u s a lity and assignm ent o f r e s p o n s ib ilit y fo r both a c to r s and o b serv ers, storm s ( 1973)

55 examined th e e f f e c t o f m anip u lating th e v is u a l p e r s p e c tiv e s o f a c to r s and o b servers on a ttr ib u tio n s o f c a u s a lit y. A lte r in g the v is u a l p e r s p e c tiv e o f a p e r c e iv e r i s funct io n a l ly eq u iv a len t to m anip u lating th e s a lie n c e o f in form ation to which an in d iv id u a l i s exposed. The b r ie f, u n stru ctu red "get acquainted" co n v ersa tio n s between acto r- s u b je c ts w ith in a p a ir were vid eo taped. Each co n v ersa tio n was w itn essed by a p a ir o f o b s e r v e r -s u b je c ts. Each observersu b je c t was matched w ith one a cto r and was asked to watch the behavior o f th a t a c to r. A fter th e c o n v ersa tio n ended, the s u b je c ts (a c to r s and o b serv ers) e it h e r did or did not see a vid eo tape rep la y o f th e co n v ersa tio n. The o r ie n ta tio n o f the rep la y s were m anipulated. For some s u b je c ts, th e rep la y rep eated th e o r ig in a l o r ie n ta tio n o f th e s u b je c t, th a t i s, a c to r s viewed th e b eh avior o f the o th er a c to r and o b servers viewed the beh avior o f th e matched a c to r. For o th er s u b je c ts, th e rep la y reversed the o r ig in a l o r ie n ta tio n, th a t i s, a c to r s viewed h is own behavior and o b serv ers viewed the behavior o f th e unmatched a c to r. F ollow in g th e rep la y, each su b ject was asked to rate the r e la t iv e im portance o f "personal c h a r a c te r is tic s " and " s itu a tio n a l c h a r a c te r is tic s " in ca u sin g h is own behavior, a s an a c to r, or in ca u sin g the behavior o f the matched a c to r, as an o b serv e r. The fin d in g s showed th a t (1) when su b je c ts saw no re p la y or a rep la y th a t repeated th e

56 43 o r ig in a l o r ie n ta tio n, a c to r s rated " s itu a tio n a l ch a ra cteri s t i c s " as more im portant than "personal c h a r a c te r is tic s " in ca u sin g t h e ir own behavior w h ile ob servers rated "personal c h a r a c te r is tic s " as more im portant in ca u sin g th e behavior o f th e matched a cto r and ( 2) when su b je c ts saw a rep lay th a t reversed the o r ig in a l o r ie n ta tio n, a c to r s rated "personal c h a r a c te r is tic s " as more im portant in cau sin g t h e ir own behavior and o b serv ers rated " s itu a tio n a l c h a r a c te r is tic s " a s more im portant in ca u sin g th e behavior o f th e matched a c to r. Another study which m anipulated the s a lie n c e o f a c to r s ' behavior was conducted by Duval and Wicklund (1973)* These authors examined the e f f e c t o f m anip u lating a c to r s ' v is u a l p e r s p e c tiv e s on th e degree to which a cto rs h eld them selves r e sp o n sib le fo r p o s it iv e and n e g a tiv e outcom es. The study by Duval and Wicklund i s p a r tic u la r ly im portant r e la t iv e to the m ethodology to be employed in the p resen t study. A c to r-su b jects were asked to im agine th em selves in a v a r ie ty o f s it u a tio n s th a t r e s u lte d in e ith e r p o s it iv e or n eg a tiv e outcom es. The s it u a tio n s were p resen ted in a q u e stio n n a ir e. For exam ple, "Imagine th a t you have s e le c te d and purchased a race h o rse. You e n te r the h orse in a major race and h ir e a good jockey to r id e him. The horse wins f i r s t p la c e. To what degree did your a c tio n s cause the v ic to r y and to what degree did th e a c tio n s o f the jockey cau se the victory?"

57 A m irror was p la ced in th e exp erim en tal room and was turned such th a t th e m irror e it h e r did o r d id not fa c e th e a c to r -s u b je c t. The fin d in g s rev ea le d th a t (1) o v e r a ll, a c to r s h eld th em selv es about e q u a lly r e sp o n sib le fo r p o s it iv e and n e g a tiv e outcomes and (2) a c to r s a scrib ed a la r g e r d egree o f r e s p o n s ib ilit y to th em selv es when th e m irror was turned toward them than when th e m irror was turned away from them. The above fin d in g s le d to an e x te n sio n o f th e d is c r e p ancy h y p o th e sis c a lle d th e "theory o f o b je c tiv e s e l f - awareness" (W icklund and D uval, 1971* Duval and W icklund, 1973) The d iscrep a n cy h y p o th e sis proposed th a t a c to r s tend to fo cu s on e x te r n a l s it u a t io n a l d eterm in ants in a cco u n tin g fo r t h e ir own a c tio n s and th e occurrence o f an outcome w h ile o b se r v ers tend to fo cu s on th e a c t o r s b eh a v io r. The reaso n g iv en was th a t environm ental fa c t o r s are p e r c e p tu a lly m ost s a lie n t fo r a c to r s w h ile th e a c to r h im self and h is b eh avior are p e r c e p tu a lly most s a lie n t fo r o b se r v e r s. As a r e s u l t, a c to r s and o b servers are p r o c e ssin g d if f e r e n t in fo rm a tio n and co n seq u en tly have d iv erg en t p erc e p tio n s o f c a u s a lit y. Based on th e above r e a so n in g, Duval and Wicklund p ro posed th a t th e d iscrep a n cy betw een a c to r s ' and o b se r v ers' a ttr ib u t io n s o f c a u s a lit y can be e f f e c t i v e l y reduced by d ir e c tin g an a c t o r 's fo cu s o f a tte n tio n to h im self and h is b eh avior. By d ir e c t in g an a c t o r 's a t te n t io n to h im s e lf,

58 4-5 h is own b eh avior w i l l be p e r c e p tu a lly more s a lie n t than e x te r n a l s it u a t io n a l d eterm in ants and as su ch, th e a c to r and th e o b server w i l l be more s im ila r r e la t iv e to th e in fo rm a tio n w hich th e y a tte n d to and p r o c e s s. Thus, i f an a c to r i s induced to a tte n d to h is own b eh a v io r, th en th e d iscrep a n cy betw een a c to r s ' and o b se r v e r s' a ttr ib u t io n s o f c a u s a lit y should be reduced sin c e s im ila r ty p e s o f in fo rm a tio n w i l l be most s a lie n t fo r both a cto rs and o b se r v e r s. A consequence o f th e argument based on th e d is c r e p ancy h y p o th e sis and th e th eory o f o b je c tiv e se lf-a w a r e n e ss r e la t e s to ran k ing th r e e groups o f s u b je c ts r e la t iv e to th e degree to which a t t r ib u t io n s o f c a u s a lit y are d ir e c te d toward th e a c to r. The th r e e r e le v a n t groups are o b se r v e r s, a c to r s whose fo c u s o f a t te n t io n i s r e d ir e c te d, and a c to r s whose fo cu s o f a t t e n t io n i s not r e d ir e c te d. The g e n e r a l e x p e c ta tio n i s th a t a c to r s whose fo c u s o f a tte n tio n i s n ot r e d ir e c te d should a t tr ib u t e l e s s c a u s a lit y to thems e lv e s than a c to r s whose fo c u s o f a t te n t io n i s r e d ir e c te d who in turn should a t t r ib u t e l e s s c a u s a lit y to th em selv es than o b serv ers should a t tr ib u t e to a c to r s. The em p iric a l fin d in g s and t h e o r e t ic a l p o s it io n o f Duval and Wicklund in d ic a te th a t d ir e c t in g an a c t o r 's fo c u s o f a t te n t io n upon h im self n o t on ly a f f e c t s h is a t t r ib u tio n s o f c a u s a lit y but s im ila r ly a f f e c t s h is r e s p o n s i b i l i t y assignm en t. That i s, an a c to r whose fo cu s o f

59 46 a t t e n t io n i s r e d ir e c te d n ot o n ly a t t r ib u t e s more c a u s a lit y to h im self than an a cto r whose fo c u s o f a tte n tio n i s not r e d ir e c te d but a ls o ta k es more r e s p o n s ib ilit y fo r p o s it iv e and n eg a tiv e ou tcom es. However, in ex ten d in g th e p o s it io n o f Duval and W icklund to compare th e th r ee groups o f s u b je c ts in term s o f r e s p o n s ib ilit y assign m en t, d e fe n s iv e a t tr ib u t io n d ic t a t e s th a t th e d iscrep a n cy between a c t o r s and o b s e r v e r s resp o n s i b i l i t y assignm ent must be examined as a fu n c tio n o f o u t come c h a r a c t e r is t ic s. Based on d e fe n s iv e a ttr ib u tio n (W a lster, 1966; Shaw and S k o ln ic k, 1971)* as n e g a tiv e o u t comes become more s e v e r e, a c to r s tend to take l e s s resp onsi b i l i t y w hile o b se r v e r s tend to a s s ig n more r e s p o n s ib ilit y. Thus, a s n e g a tiv e outcom es become more s e v e r e, th e d isc r e p ancy between a c t o r s ' and o b se r v e r s' r e s p o n s ib ilit y a s s ig n ment i s exp ected to in c r e a s e. Based on th e th eory o f o b je c tiv e s e lf-a w a r e n e s s, a c to r s whose fo c u s o f a t te n t io n i s r e d ir e c te d should take more r e s p o n s ib ilit y fo r n e g a tiv e outcom es than a c to r s whose fo c u s o f a t te n t io n i s n ot r e d ir e c te d. Thus, th e d is c r e p ancy between a c t o r s and o b s e r v e r s r e s p o n s ib ilit y a s s ig n ment would be ex p ected to be reduced by r e d ir e c tin g an a c t o r 's focu s o f a t t e n t io n. Comparing th e th ree groups o f s u b je c ts in term s o f r e s p o n s ib ilit y assignm ent fo r n e g a tiv e outcom es, i t i s ex p ected th a t a c to r s whose fo cu s o f a t te n t io n i s n o t r e d ir e c te d should take l e s s r e s p o n s ib ilit y than a c to r s

60 4-7 whose fo cu s o f a t t e n t io n i s r e d ir e c te d who in turn should ta k e l e s s r e s p o n s i b ilit y than o b servers should a s s ig n to a c t o r s. But in th e c a se o f p o s it iv e outcom es, th e e x p e c ta tio n d eriv ed from d e fe n s iv e a t tr ib u t io n and th e th eory o f o b je c tiv e s e lf-a w a r e n e s s would be th e r e v e r se o f th a t fo r n e g a tiv e outcom es. Based on d e fe n s iv e a t tr ib u t io n, as p o s it iv e outcom es become more fa v o r a b le, a c to r s tend to ta k e more r e s p o n s ib ilit y w h ile o b serve rs ten d to a s s ig n l e s s r e s p o n s ib ilit y. Thus, th e d iscrep a n cy between a c to r s ' and o b se r v e r s' r e s p o n s ib ilit y assignm ent i s exp ected to in c r e a se a s p o s it iv e outcom es become more fa v o r a b le. However, based on th e th eo ry o f o b je c tiv e s e l f - aw areness, a c to r s whose fo cu s o f a t t e n t io n i s r e d ir e c te d should take more r e s p o n s ib ilit y fo r p o s it iv e outcom es than a c to r s whose fo cu s o f a tte n tio n i s n ot r e d ir e c te d. Thus, u n lik e fo r n e g a tiv e outcom es, th e d iscrep a n cy between a c to r s ' and o b s e r v e r s r e s p o n s i b ilit y assignm en t would be expected to in c r e a se by r e d ir e c tin g an a c t o r 's fo cu s o f a t t e n t io n. T h erefo re, r e la t iv e to th e rank order o f the th r e e groups o f s u b je c ts in r e s p o n s ib ilit y assignm ent fo r p o s it iv e outcom es, i t i s exp ected th a t a c to r s whose fo cu s o f a tte n tio n i s r e d ir e c te d should take more r e s p o n s ib ilit y than a c to r s whose fo c u s o f a t te n t io n i s n ot r e d ir e c te d who in turn should take more r e s p o n s ib ilit y than o b serv ers should a s s ig n to a c t o r s.

61 ^8 As su g g ested p r e v io u s ly, p a st resea r ch has in ad v e r te n t ly v a ried two im portant stim u lu s d im en sion s. R esearch ers have v a ried not o n ly th e in fo rm a tio n th a t was m ost s a lie n t to a c to r s and o b se r v e r s, as j u s t d is c u s s e d, b ut a ls o th e o b je c t iv e in fo rm a tio n about th e n atu re o f th e in te r a c tio n betw een th e prim ary a c to r and th e a c tiv e environm ent. T h is l a t t e r dim ension i s d isc u sse d a t le n g th b elow. The r e se a r c h and t h e o r e t ic a l p o s it io n s r e la te d to t h i s dim ension have m ajor im p lic a tio n s fo r th e p r e sen t in v e s t ig a t io n. F ish b e in and A jzen (1973) have a ls o n oted th e im portance o f t h i s l a t t e r stim u lu s d im en sion, nam ely th e n atu re o f th e in t e r a c t io n among th e c a u s a l a g en ts in an a c tio n s it u a t io n. They d is c u s s e d v a r io u s ty p e s o f in te r a c tio n s betw een th e prim ary a c to r and th e a c t iv e environm ent. They d is tin g u is h e d betw een f iv e d if f e r e n t l e v e l s o f c a u s a lit y th a t were d evelop ed by H eider (1958)* " a s s o c ia tio n," "com m ission," " f o r e s e e a b ilit y," " j u s t if ic a t io n," and " i n t e n t i o n a lit y." Each l e v e l n o t o n ly r e p r e s e n ts a d i s t i n c t c a u sa l in t e r a c tio n betw een th e prim ary a c to r and th e a c t iv e environm ent a s o b j e c t iv e ly d e p ic te d in d if f e r e n t ty p e s o f a c tio n s it u a t io n s but a ls o r e p r e s e n ts a d if f e r e n t d egree to which th e prim ary a c to r and th e a c t iv e environm ent share a c a u sa l r o le in p rod u cin g th e f i n a l outcom e. At each s u c c e s s iv e l e v e l, th e prim ary a c to r i s co n sid ered to have a more

62 ^9 in f lu e n t ia l r o le in producing th e f in a l outcome than th e a c tiv e environm ent. C orrespondingly, the degree to which th e primary a c to r i s held r e sp o n sib le by o b serv ers fo r p o s it iv e and n e g a tiv e outcomes i s a fu n c tio n o f the l e v e l o f c a u s a lity. As such, F ish b e in and Ajzen argued th a t th e d iscrep a n t fin d in g s from p a st research on r e s p o n s ib ilit y assignm ent were due to s u b je c ts ev a lu a tin g th e a c to r 's r o le in producing th e f i n a l outcome a t d iffe r e n t le v e ls o f ca u s a l i t y as o b j e c t iv e ly d ep icted in a ctio n s it u a t io n s. Shaw and h is a s s o c ia te s a ls o conducted research th a t provided a d d itio n a l e lu c id a tio n and em p iric a l support fo r H eid er's l e v e l s o f c a u s a lity (Shaw and R eita n, 1969; Shaw and S u lzer, 196*J-; S u lz e r, 196*0. The exam ples used to d is c u s s and i l l u s t r a t e each l e v e l were adapted from q u e stio n n a ires th a t were d eveloped by th e s e r e se a r ch ers. For each s it u a t io n, "Perry" i s d esign ated a s the "primary actor" and th e outcome could be p o s itiv e o r n eg a tiv e and vary in in t e n s it y. In t h e ir resea rch, o b se r v e r -su b je c ts are asked to s p e c ify the d egree to which "Perry i s resp o n sib le fo r th e outcom e. "A ssociation" i s the l e v e l a t which th e stim ulu s person i s d ep icted as b ein g only in d ir e c t ly r e la te d to the product io n o f the f i n a l outcome. The stim ulus p erson does n ot a c tu a lly a ct in producing th e outcome. The stim ulus person i s m erely r e la te d to the produced outcome by v ir tu e o f ownership o f th e instrum ent used or by v ir tu e o f a p erso n a l

63 50 u n it r e la tio n s h ip w ith th e p eo p le who do a c t u a lly a c t in p rod u cin g th e outcom e. For exam ple, "One day s e v e r a l o f P e r r y 's fr ie n d s were f is h in g a t th e la k e. They found a f is h in g rod in th e hushes and broke i t. I s Perry resp o n s i b l e fo r th e f is h in g rod b ein g broken? The f in a l outcome ( f i s h i n g rod b ein g broken) i s n e g a tiv e. The le v e l o f c a u s a lit y i s a s s o c ia t io n b ecau se even though P erry did n o t a c t or p a r t ic ip a t e in b reak in g th e f is h in g rod, Perry was r e la te d to th e produced outcome by v ir tu e o f fr ie n d sh ip w ith th e p eop le who d id produce the f i n a l outcome. "Commission" i s th e l e v e l a t which th e stim u lu s person d o es a c t in th e s it u a t io n, but ( 1) he a c t s to produce an outcome th a t i s d if f e r e n t from th e outcome th a t a c tu a lly r e s u l t s and ( 2 ) th e f i n a l outcome th a t d o es r e s u lt could n o t have p o s s ib ly been a n tic ip a te d or fo r e s e e n by th e stim u lu s p erso n. The outcome i s not p art o f th e stim u lu s p e r s o n 's in te n tio n s or g o a ls, but h is a c tio n s m erely s e t in m otion a sequence o f e v e n ts th a t le a d to an u n fo r e se e a b le outcome beyond h is c o n t r o l. For exam ple, "Perry was making some b u sin e ss telep h o n e c a l l s. When th e phone ran g in one home he c a lle d, i t awakened a man who was s le e p in g near a broken gas h e a te r. I f he had n ot awakened, th e le a k in g gas would have k ille d him. I s Perry r e sp o n sib le fo r th e man w aking up in tim e to escape death?" The f in a l outcome (man e sca p in g d eath ) i s p o s i t i v e. The l e v e l o f c a u s a lit y i s com m ission because P e r r y 's a c tio n s s e t in

64 m otion an u n fo r e se e a b le and u n c o n tr o lla b le sequence o f e v e n ts th a t d id r e s u lt in an outcome th a t was n ot p art o f P e r r y 's in t e n t io n s. " F o r ese ea b ility " i s th e l e v e l a t which th e stim u lu s p erso n does a c t in th e s it u a t io n, but ( 1) he a c t s to p roduce an outcome th a t i s d if f e r e n t from th e outcome th a t a c t u a lly r e s u lt s and ( 2 ) th e f i n a l outcome th a t does r e s u lt co u ld have been a n tic ip a te d or p r e d ic te d by th e stim u lu s p erso n. The f i n a l outcome i s n ot p a rt o f th e stim u lu s p e r s o n 's in te n t io n s or g o a ls, but th e consequences o f h is a c t io n cou ld have been a n tic ip a te d, p r e d ic te d, or c o n tr o lle d by th e stim u lu s p erson i f he had taken in to account or had g iv e n enough thought to th e circu m sta n ces surrounding h is a c t io n s. to s c h o o l. For exam ple, "Perry was ta k in g h is l i t t l e s i s t e r She s ta r te d to step in to a busy s t r e e t but P erry wanted to look in a s to r e window, so he p u lled her back. T h is k ep t h is s i s t e r from b ein g h it by a sp eed in g c a r. I s P erry r e s p o n s ib le fo r sa v in g h is s i s t e r ' s lif e? " The f i n a l outcome (sa v in g s i s t e r ' s l i f e ) i s p o s i t i v e. The l e v e l o f c a u s a lit y i s f o r e s e e a b ilit y because P e r r y 's in te n t i o n was n ot to save h is s i s t e r ' s l i f e and th e con sequen ces o f h is a c tio n s cou ld have been a n tic ip a te d. " J u s tific a tio n " i s th e l e v e l a t which th e stim u lu s p erso n does a c t in th e s it u a t io n, but u n lik e th e f o r e s e e a b i l i t y l e v e l, he d oes fo r e s e e th e con sequen ces o f h is a c t io n s. However, h is a c tio n s are warranted by the

65 52 circu m stan ces an d/or san ction ed by s o c ie t a l norms. The f i n a l outcome i s a d ir e c t r e s u lt o f h is a ctio n s* but circum sta n c e s or environm ental f o r c e s, such as c o e r c iv e fo r c e s, f a c i l i t a t e the b eh avior o f th e stim ulu s p erso n. The stim u lu s person behaves and a c ts a s most o th er p eop le would have acted in th e same s e t o f circu m sta n ces. For exam ple, "A man tr ie d to k i l l Perry w ith a la rg e k n ife. Perry grabbed the k n ife and stabbed th e man to keep from b ein g k ille d h im self. I s Perry r e sp o n sib le fo r th e man b ein g stabbed?" The f i n a l outcome (man b ein g stab b ed) i s negat i v e. The le v e l o f c a u s a lity i s j u s t if ic a t io n because P erry was coerced in to sta b b in g the man. Under c o n d itio n s o f s e lf-d e f e n s e, P e r r y s a c tio n s were warranted by the s e t o f circu m stan ces and most oth er p eo p le would defend th em selv es. " I n te n tio n a lity ' i s th e le v e l a t which the stim u lu s p erson a c ts to produce th e outcom e, bub u n lik e the j u s t i f i c a tio n le v e l, h is a c tio n s a re n ot warranted by the circum sta n c e s and/or n o t san ction ed by s o c ie t a l norms. The stim u lu s person d oes not a c t and behave as most other p eo p le would have acted in th e same s e t o f circu m stan ces. The f in a l outcome i s a d ir e c t r e s u lt o f h is a c tio n s and th e person a c ts in s p ite o f circu m stan ces or environm ental fo r c e s which in h ib it h is a c t io n s. For exam ple, "Perry was f is h in g when he saw a boy drowning in th e r iv e r. Parry co u ld not swim, but he fough t h is way out to the boy and

66 53 p u lle d him ou*fc. I s P erry r e sp o n sib le fo r savin g th e boy s lif e? " The f i n a l outcome (sa v in g b o y 's l i f e ) i s p o s it iv e. The l e v e l o f c a u s a lity i s in t e n t io n a lit y because P erry d e lib e r a te ly saved th e boy s l i f e in s p it e o f the f a c t th a t he cou ld not swim. In summary, a t each le v e l o f c a u s a lity, the nature o f th e in te r a c tio n between th e primary a c to r and th e a c tiv e environm ent i s q u ite d if f e r e n t. At th e le v e l o f a s s o c ia tio n, th e stim u lu s person i s on ly in d ir e c t ly r e la te d to th e p rod u ction o f th e outcome which i s caused e x c lu s iv e ly by th e a c tiv e environm ent. The p e r so n s r o le i s m inim al, i. e. he i s sim ply a sso c ia te d w ith c a u sa l fo r c e s in th e a c tiv e environm ent. At th e le v e l o f com m ission, the stim u lu s person i s more in v o lv ed by v ir tu e o f th e f a c t th a t he does a c t, but h is a c tio n s m erely s e t in m otion a sequence o f even ts beyond h is c o n tr o l. At th e l e v e l o f f o r e s e e a b ilit y, the stim u lu s p e r so n 's a c tio n s s e t in m otion a sequence o f ev en ts th a t i s not beyond h is c o n tr o l, but cou ld have been cont r o lle d i f he had taken in to account environm ental c o n tin g e n c ie s. At th e l e v e l o f j u s t i f i c a t i o n, th e outcome i s a d ir e c t r e s u lt o f th e p e r so n 's a c tio n s, but th e person I s coerced in to perform ing an a c t by v ir tu e o f f a c i l i t a t i n g e x ten u a tin g circu m stan ces th a t warrant h is b eh avior. At th e le v e l o f in t e n t io n a lit y, th e stim u lu s person d e lib e r a te ly a c t s to produce th e outcome even though th ere are environm ental fo r c e s and c o n tin g e n c ie s a c tin g to in h ib it h is

67 54 behavior. T h e o r e tic a lly, a s u b j e c t s 's in te r p r e ta tio n o f c a u s a lit y and assignm ent o f r e s p o n s ib ilit y depends upon the l e v e l o f c a u s a lit y as o b j e c t iv e ly d e p ic ted in an a c tio n s it u a t io n. The in form ation p resen ted in a s it u a tio n d e fin e s the f iv e " contextu al le v e l s o f c a u sa lity " and a llo w s th e su b ject to answer q u e stio n s, such as (1 ) "Was the person on ly in d ir e c t ly r e la te d to th e p rod u ction o f th e outcome by v ir tu e o f h is a s s o c ia tio n w ith o th er ca u sa l agents?" (a s s o c ia tio n l e v e l ), (2 ) "Did th e p e r so n 's a c tio n s m erely s e t in motion the a c tiv e environm ent which y ie ld e d an u n c o n tr o lla b le and u n fo reseea b le outcome?" (com m ission l e v e l ), (3) "Was th e in d iv id u a l ca rel e s s by not ta k in g in to account in form ation su p p lied by the a c tiv e environment?" ( f o r e s e e a b ilit y l e v e l ), (4) "Did the environm ental circu m stan ces warrant the in d iv id u a l's a ctio n s? " ( j u s t i f i c a t i o n l e v e l ), and (5) "Was th e behavior o f the in d iv id u a l prem editated and unwarranted w ith in th e c o n tex t o f th e situ a tio n? " ( in t e n t io n a lit y l e v e l ). I f th e l e v e l s are ordered from a s s o c ia tio n to com m ission to f o r e s e e a b i li t y to j u s t i f i c a t i o n to i n t e n t io n a lit y, then th e assum ption i s th a t a t th e s u c c e s s iv e l e v e l s, the a c tiv e environm ent i s view ed a s a l e s s i n f l u e n t i a l agent in c a u sin g th e behavior o f th e a c to r and in producing th e outcome. Thus, the p r e d ic tio n i s th a t ob serv ers should a s s ig n in c r e a sin g amounts o f r e s p o n s ib ilit y to th e prim ary a c to r a t each s u c c e s s iv e l e v e l. However, as w i l l be d isc u sse d s h o r tly,

68 55 some argument e x i s t s fo r o rd erin g th e l e v e l o f j u s t i f ic a t i o n p r io r to th e l e v e l o f f o r e s e e a b ilit y. M oreover, em p irical evid en ce has shown th a t a complex in te r a c tio n e x i s t s between th e l e v e l s o f c a u s a lity and outcome c h a r a c t e r is t ic s in determin in g th e degree to which an a c to r i s h eld r e sp o n sib le fo r th e f in a l outcom e. F ive experim ents have been conducted to examine o b serv ers' assignm ent o f r e s p o n s ib ilit y to an a c to r as a fu n c tio n o f th e f iv e l e v e l s o f c a u s a lity (Shaw and R eitan, 1969; Shaw and S u lz e r, 1964; S u lz e r, 1964). The r e s u lt s o f the experim ents w i l l be summarized to g e th e r. A ll experim ents used q u e stio n n a ir e s to p resen t a v a r ie ty o f a c tio n s it u a tio n s to ad u lt o b s e r v e r -s u b je c ts. Two experim ents e x p l i c i t l y d iffe r e n tia te d th e outcome o f each s it u a t io n as p o s it iv e or n e g a tiv e, but d id not m anipulate th e in t e n s it y o f th e outcome (Shaw and S u lz e r, ). Three experim ents e x p l i c i t l y d if f e r e n tia te d th e outcome fo r each s it u a t io n as a low to high p o s it iv e or n eg a tiv e outcome (Shaw and R eitan, 1969; S u lz er, 1964). The o b se r v e r -su b je c ts were asked to s p e c ify th e degree to which th e prim ary a cto r should be h eld r e sp o n sib le fo r th e outcom e. To ex p ed ite th e d is c u s s io n o f the fin d in g s, "AR" w i l l r e fe r to o b se r v e r s assignm ent o f r e s p o n s ib ilit y to th e prim ary a c to r. The fin d in g s g e n e r a lly supported th e expected ord erin g o f th e l e v e l s. O v e ra ll, o b se r v e r s r e s p o n s ib ilit y a s s ig n ment in crea sed from a s s o c ia tio n to com m ission to

69 56 f o r e s e e a b ilit y to j u s t i f i c a t i o n to in t e n t io n a lit y. However, th e order o f f o r e s e e a b i li t y and j u s t i f i c a t i o n was u s u a lly rev ersed f o r p o s i t i v e and n e g a tiv e outcom es. For p o s i t i v e outcomes AR in c r e a se d from a s s o c ia t io n to commission to f o r e s e e a b i li t y to j u s t i f i c a t i o n to in t e n t io n a lit y. For n e g a tiv e outcomes AR in crea sed from a s s o c ia t io n to comm ission to j u s t i f i c a t i o n to f o r e s e e a b i li t y to i n t e n t i o n a l i t y. W ithin the c o n te x t o f o rd erin g th e l e v e l s, an oth er fin d in g i s worth n o tin g reg a rd in g o b se r v e r s' AR at th e p o la r l e v e l s o f a s s o c ia t io n and in t e n t io n a lit y. At th e le v e l o f a s s o c ia tio n AR tended to be minimal w h ile at the l e v e l o f in t e n t io n a lit y AR tended to be maximal ir r e s p e c tiv e o f outcome c h a r a c t e r i s t i c s, i. e. AR was r e l a t i v e l y u n a ffe c te d by outcome v a len ce and in t e n s i t y a t th e p o la r l e v e l s o f a s s o c ia t io n and i n t e n t i o n a l i t y. However, a t th e th ree in term ed ia te l e v e l s o f c a u s a lit y (com m ission, f o r e s e e a b i li t y, j u s t i f i c a t i o n ), o b ser v er s' AR was d i f f e r e n t i a l l y a ffe c te d by outcome v a le n c e and in t e n s it y. F ir s t, a t th e l e v e l o f com m ission the ev id en ce r e v e a le d th at AR tended to be h ig h er fo r p o s it iv e outcom es than f o r negat i v e outcom es. M oreover, AR tended to in c r e a se as p o s it iv e outcomes became more fa v o r a b le, but ten ded to d ecrea se as n e g a tiv e outcom es became more se v er e. Second, at th e le v e l o f f o r e s e e a b ilit y AR tended to be h igh er fo r n e g a tiv e o u t comes than fo r p o s i t i v e outcom es. A lso, AR tended to in c r e a se a s p o s i t i v e outcomes became more favorab le (some

70 57 evidence in d ic a te d th e r e v e r se) and tended to in c r e a se as n e g a tiv e outcom es became more se v e r e. F in a lly, a t th e l e v e l o f j u s t i f i c a t i o n AR tended to be h igher fo r p o s it iv e outcomes than fo r n e g a tiv e outcom es, but some evid en ce in d i c a te d the r e v e r s e. Furthermore, AR tended to in c r e a se as both p o s it iv e and n eg a tiv e outcomes became more extrem e. C oncep tu ally, th e se fin d in g s su g g est then th a t o b servers' AR i s not sim ply a fu n c tio n o f the r e la t iv e contr ib u tio n o f th e primary a cto r and th e a c tiv e environm ent as o b je c t iv e ly d ep icted in a c tio n s it u a t io n s at each l e v e l. Wot.only did th e o rd erin g o f th e l e v e l s change depending upon outcome valen ce but a ls o th e p a tte r n o f AR as a funct io n o f outcome valen ce and in t e n s it y was d iffe r e n t w ith in each le v e l. Moreover, th e evidence did not always p ortray th e same p a tte r n o f AR w ith in l e v e l s, e s p e c ia lly w ith in the l e v e l s o f f o r e s e e a b ilit y and j u s t i f i c a t i o n. In g en era l, the o rd erin g o f th e l e v e l s conformed to th e n o tio n th a t in a c tio n s itu a tio n s a t each s u c c e s s iv e l e v e l, the prim ary a c to r i s o b je c t iv e ly d ep icted a s having a more in f lu e n t ia l c a u s a l r o le in producing the f i n a l ou t come than th e a c tiv e environm ent. However, the f a c t th a t th e p a ttern o f AR changed a cro ss l e v e l s su g g ests th a t i t i s th e dynamics or nature o f th e in te r a c tio n among ca u sa l a g en ts which determ ine o b servers' assignm ent o f resp onsi b i l i t y as a fu n c tio n o f outcome c h a r a c t e r is t ic s. For example, a t th e p o la r l e v e l s o f c a u s a lity { a s s o c ia tio n and

71 58 i n t e n t i o n a l i t y ), th e r e la t iv e amount o f AR by o b serv ers was q u ite d if f e r e n t but th e p a tte r n o f AR was q u ite s im ila r. At th e l e v e l o f a s s o c ia t io n AR tended to be m inimal w h ile a t th e l e v e l o f in t e n t io n a lit y AR tended to be maximal. However, a t both l e v e l s th e p a tte r n o f AR was r e l a t iv e ly u n a ffe c te d by outcome c h a r a c t e r is t ic s. In c o n tr a s t, a t th e in term ed ia te l e v e l s o f c a u s a lit y (com m ission, f o r e s e e a b i l i t y, j u s t i f i c a t i o n ) n ot o n ly was th e r e la t iv e amount o f AR d if f e r e n t but a ls o th e p a tte r n o f AR was d if f e r e n t f o r each l e v e l. An ex p la n a tio n f o r th e se g en era l fin d in g s b eg in s w ith th e o b serv a tio n th a t th e l e v e l s o f a s s o c ia t io n and in te n t i o n a l i t y rep r e se n t o p p o s ite p o le s o f a c a u s a lit y continuum. At the l e v e l o f a s s o c ia t io n th e prim ary a c to r has v ir t u a lly no in flu e n c e in p rod u cin g th e f i n a l outcome w h ile a t th e l e v e l o f in t e n t io n a lit y th e prim ary a c to r has e s s e n t i a l l y t o t a l c o n tr o l and in flu e n c e in p rod u cin g th e f i n a l outcom e. Thus, th e c a u s a l r o l e s o f th e prim ary a c to r and the a c t iv e environm ent are c l e a r l y d if f e r e n t ia t e d and unambiguous and are th e r e fo r e, m in im ally su b je c t to d i f f e r e n t i a l in te r p r e ta t io n by o b se r v e r s. The r e s u lt i s th a t th e prim ary a c to r i s n o t h eld r e s p o n s ib le a t th e l e v e l o f a s s o c ia t io n, but i s h eld t o t a l l y r e s p o n s ib le a t th e l e v e l o f in t e n t io n a lit y ir r e s p e c t iv e o f th e n atu re o f th e f i n a l outcom e. However, a t th e in term ed ia te l e v e l s, n ot o n ly does th e prim ary a cto r and th e a c tiv e environm ent more c le a r ly sh are

72 59 a ca u sa l r o le in producing the f i n a l outcome than at th e p olar l e v e l s o f c a u s a lit y but a ls o th e nature o f the c a u s a l r o le shared by the prim ary a cto r and the a c t iv e environment changes from le v e l to l e v e l. F ir s t, in a c tio n s itu a tio n s a t the l e v e l o f com m ission, the prim ary a c to r a c ts as most o th e r peop le would have a cted in order to accom plish a p a r tic u la r outcome, i. e. h is s p e c if ic a c tio n s are j u s t i fie d r e l a t i v e to h is intended g o a l. Furthermore, h is s p e c ific a c tio n s m erely s e t in m otion the a c t iv e environment w hich a c tu a lly produces a f i n a l outcome which i s d e f in it e ly u n fo r e se e a b le, u n c o n tr o lla b le, and unintended. Second, in a c tio n s it u a tio n s a t the l e v e l o f fo r e s e e a b ilit y, th e primary a c to r could have fo r e seen the f i n a l outcome i f he had taken in to account a l l o f th e circu m stan ces surrounding h is b eh a v io r. However, the a v a ila b le inform a tio n a t t h i s le v e l i s q u ite ambiguous as to whether or n o t the a c to r a c tu a lly foresaw the f i n a l outcome and whether or not h is s p e c if ic a c tio n s were j u s t i f i e d. F in a lly, in a c tio n s it u a tio n s a t the l e v e l of j u s t i f i c a tio n, th e primary a c to r does a c tu a lly fo r e s e e the f i n a l outcome. In a d d itio n, h is s p e c if ic a c tio n s are e x p l i c i t l y j u s t if ie d by ex ten u a tin g circu m stan ces and co e r c iv e fo r c e s such a s o rd ers, t h r e a t s, or ultim atum s, i. e. th e c o e r c iv e a c tiv e environm ent h as a share in producing th e a c tio n s that le d to the f i n a l outcome. Thus, the a c tio n s itu a tio n s a t the in term ed iate l e v e l s

73 60 d i f f e r p rim a rily in terras o f th e "dynamics o f f o r e s e e a b ilit y and j u s t i f i c a t i o n. " At th e le v e l o f com m ission, th e outcome i s u n fo reseea b le and th e s p e c if ic a c tio n s o f th e primary a c to r are j u s t i f i e d. At th e le v e l o f f o r e s e e a b ilit y, the outcome i s fo r e s e e a b le, but the j u s t i f i a b i l i t y o f th e a c to r 's b eh avior i s ambiguous. At th e l e v e l o f j u s t if ic a t io n, the outcome i s fo r e s e e a b le, but th e s p e c if ic a c tio n s o f the a c to r are e x p l i c i t l y j u s t i f i e d. Thus, the a c tio n s it u a tio n s a t th e l e v e l s o f com m ission and j u s t i f i c a t i o n are s im ila r, y e t d is t in c t from th o se a t th e l e v e l o f f o r e s e e a b ilit y, in th a t th e a c t o r 's s p e c if ic a c tio n s are j u s t i f i e d. In c o n tr a s t, th e a c tio n s it u a tio n s a t th e l e v e l s o f f o r e s e e a b ilit y and j u s t i f i c a t i o n are s im ila r, y e t d i s t i n c t from th o se a t the l e v e l o f com m ission, in th a t th e outcomes are a t le a s t p o t e n t ia lly fo r e s e e a b le. The im portance o f th e se s i m ila r it ie s and d iffe r e n c e s in e x p la in in g th e d if f e r e n t p a tte r n s o f AR w i l l be examined in more d e t a il s h o r tly. However, a t t h i s p o in t, th e im portant n o tio n i s th a t th e r e la t iv e c o n tr ib u tio n s o f th e primary a c to r and the a c tiv e environment are not as c le a r ly d i f f e r e n tia te d and unambiguous a s was th e ca se w ith the p o la r l e v e l s o f a s s o c ia tio n and in t e n t io n a lit y. The r e s u lt i s th a t the o b je c tiv e in form ation about th e dynamics o f th e in te r a c tio n among c a u sa l agen ts i s more conducive to s u b je c t iv e in te r p r e ta tio n on th e p art o f o b serv ers. As such, o b serv ers' assignm ent o f r e s p o n s ib ilit y i s more l ik e l y to

74 61 be a f f e c t e d by outcome c h a r a c t e r is t ic s a t the in term ed iate l e v e l s o f c a u s a lit y than a t th e p o la r l e v e l s o f c a u s a lit y. As in d ic a te d, th e d is t i n c t i o n s among the th r ee in t e r m ediate l e v e l s in term s o f th e "dynamics o f f o r e s e e a b ilit y and j u s t if ic a t io n " may e x p la in the d if f e r e n t p a tte r n s o f AR th a t were found a t th e se l e v e l s. The fa c t th a t the p a tte r n s o f AR changed a c r o ss l e v e l s has major im p lic a tio n s fo r r e s o lv in g th e in c o n s is t e n c ie s from p a st r e se a r c h. The f a ilu r e in p a s t resea rch to fin d th e same p a tte r n o f AR r a is e s th e p o s s i b i l i t y th a t r ese a rc h e r s have in a d v e r te n tly employed a c tio n s it u a t io n s which rep resen ted d if f e r e n t l e v e l s o f c a u s a lit y. As a r e s u lt, r ese a rc h e r s have f a ile d to fin d o v e r a ll support fo r any p a r t ic u la r t h e o r e t ic a l approach to o b ser v er s' r e s p o n s ib ilit y assign m en t, in p a r tic u la r, d e fe n siv e a t t r ib u t io n. However, i f th e p a tte r n s o f AR found by Shaw and h is a s s o c ia t e s are examined c l o s e l y, th en support fo r d e fen siv e a t t r ib u t io n can be found. But th e degree to which d e fen siv e a t t r ib u t io n i s supported v a r ie s from l e v e l to l e v e l. R eca ll the p r e d ic tio n s o f d e fe n siv e a t t r ib u t io n. D efen sive a t t r ib u tio n p r e d ic ts th a t o b ser v er s hold a c to r s more r esp o n sib le fo r n e g a tiv e outcom es than fo r p o s i t i v e outcomes and th at r e s p o n s ib ilit y in c r e a se s as n e g a tiv e outcomes become more s e v e r e. When the p a tte r n s o f AR a t th e th ree in term ed iate l e v e l s are examined, th e m ajority o f th e evid en ce seems to

75 62 in d ic a te t h a t s u b s ta n tia l support f o r d e fe n siv e a t t r ib u t io n i s found i n a c tio n s i t u a t i o n in v o lv in g fo r e se e a b le outcom es. However, su pp ort fo r d e f e n s iv e a t t r ib u t io n appears to weaken i n a c tio n s it u a t io n s in v o lv in g j u s t i f i a b l e a c t io n s. For exam ple, a t th e l e v e l o f f o r e s e e a b i l i t y where outcom es are a t l e a s t p o t e n t ia lly f o r e s e e a b le, d e fe n s iv e a t t r ib u t io n i s s u b s ta n tia te d. As p r e d ic te d by d e f e n s iv e a t t r ib u t io n, o b se r v e r s h e ld a c to r s more r e sp o n s ib le fo r u n d e sir a b le negat i v e outcom es than fo r d e s ir a b le p o s i t i v e outcom es and more r e s p o n s ib le as th ese u n d e sir a b le outcom es became more se v er e. However, when we tu r n to the l e v e l o f j u s t i f i c a t i o n which in c lu d e s not o n ly fo r e s e e a b le outcom es but a ls o j u s t i f i a b l e a c t io n s, support f o r d e fe n siv e a t t r ib u t io n d im in ish e s. Even though some e v id en ce rev ea led th e same p a tt e r n us was found a t th e l e v e l o f f o r e s e e a b i l i t y, th e m a jo r ity o f the ev id en ce c o n tr a d ic te d d e fe n s iv e a t t r ib u t io n in ter m s of the e f f e c t o f outcome v a le n c e. O bservers h eld a c to r s more r e s p o n s ib le fo r d e s ir a b le outcomes th a n fo r u n d e sir a b le o u t com es. I t appears th e n t h a t j u s t i f i a b i l i t y, e s p e c i a l l y th e e x is t e n c e o f c o e r c iv e f a c i l i t a t i v e fo r c e s, d e t r a c t s s u b s t a n t ia lly from th e r e s p o n s i b i l i t y o f the a c t o r fo r negat i v e outcom es but not f o r p o s it iv e outcom es. T hus, ob servers seem to h o ld a c to r s r e s p o n s ib le f o r p o s it iv e outcom es which are th e r e s u l t o f a c t io n s c o n s is t e n t w ith c o e r c iv e fo rc es d ir e c te d toward the p r o d u c tio n o f b e n e f i c i a l ou tcom es. However, o b serv ers do n o t seem to h o ld a cto rs r e s p o n s ib le f o r

76 63 n e g a tiv e outcom es which are th e r e s u l t o f a c tio n s produced under c o e r c io n. T h is l a t t e r c o n te n tio n i s fu r th e r supported by th e fin d in g s w hich r e l a t e to o r d e r in g th e l e v e l s o f fo r e s e e a b i l i t y and j u s t i f i c a t i o n fo r p o s i t i v e and n e g a tiv e o u t com es. R o c a ll t h a t f o r p o s i t i v e outcom es AR g e n e r a lly in crea sed from f o r e s e e a b i l i t y to j u s t i f i c a t i o n w h ile fo r n e g a tiv e outcom es AR g e n e r a lly d ecrea sed from f o r e s e e a b i li t y to j u s t i f i c a t i o n. The d i s t i n c t i v e fe a tu r e betw een th e se two l e v e l s i s th a t th e a c tio n s o f th e prim ary a c to r are j u s t i f i a b l e a t th e l e v e l o f j u s t i f i c a t i o n but n ot a t th e l e v e l o f f o r e s e e a b i l i t y. Thus, i t appears th a t o b serv ers h o ld -a c to r s more r e s p o n s ib le fo r p o s i t i v e outcom es which are fo r e s e e a b le and c o n s is t e n t w ith c o e r c iv e fo r c e s than fo r p o s i t i v e outcom es which are fo r e s e e a b le but la ck e x p l i c i t j u s t i f i c a t i o n. In c o n t r a s t, i t app ears th a t o b serv ers hold a c to r s l e s s r e s p o n s ib le fo r fo r e s e e a b le outcom es produced under c o e r c io n th an fo r n e g a tiv e outcom es which could have been a v o id ed. Thus, a t t h i s p o in t, a c tio n s i t u a t i o n s w ith fo r e s e e a b le outcom es appear to be con d u cive to th e o p e r a tio n o f d e fe n s iv e a t t r ib u t io n. At th e l e v e l o f f o r e s e e a b i l i t y AR I s h ig h er fo r n e g a tiv e outcom es than f o r p o s i t i v e outcom es and AR in c r e a s e s a s n e g a tiv e outcom es become more se v e r e. L ik e w ise, a t th e l e v e l o f j u s t i f i c a t i o n which in v o lv e s fo r e s e e a b le outcom es AR in c r e a s e s a s n e g a tiv e outcom es become more s e v e r e.

77 64 However, th e in c lu s io n o f j u s t i f i a b l e a c tio n s through c o er c io n does n ot appear to be conducive to th e o p e r a tio n o f d e fe n siv e a tt r ib u tio n w ith r e sp e c t to outcome v a le n c e. U n lik e a t th e l e v e l o f f o r e s e e a b ilit y, th e m ajority o f th e evid en ce in d ic a te s th a t ob serv ers hold a c to r s more r e sp o n sib le fo r b e n e f ic ia l outcom es than fo r u n d esira b le n e g a tiv e outcom es. M oreover, AR d ecrea ses fo r u n d esira b le outcomes from the l e v e l o f f o r e s e e a b ilit y to the l e v e l o f j u s t i f i c a t i o n. But when we tu rn to th e l e v e l o f com m ission which in c lu d e s not only j u s t i f i a b l e a c tio n s but a ls o u n fo r e se e ab le outcom es, support fo r d e fe n siv e a ttr ib u tio n e s s e n t ia lly d isa p p ea r s. At t h i s l e v e l o b serv ers held a c to r s more r e sp o n sib le fo r p o s i t i v e outcom es than fo r n e g a tiv e outcomes and a lso l e s s r e sp o n s ib le as th o se n e g a tiv e outcom es became more se v e r e. Thus, th e com bination o f j u s t i f i a b l e a c tio n s and u n fo re se ea b le outcom es le a d s to a c o n tr a d ic tio n o f d e fe n siv e a t t r ib u t io n. Up to t h i s p o in t, the d is c u s s io n has on ly concerned the e f f e c t s o f outcome va len ce and s e v e r ity o f n e g a tiv e outcom es. But when the fin d in g s w ith r e sp e c t to the f a v o r s b iiit y o f p o s it iv e outcomes are examined, one must 0 e a l w ith the two a lte r n a t iv e fo rm u la tio n s o f d e fe n siv e a ttr ib u tio n d iscu sse d e a r lie r in the l it e r a t u r e rev iew. R e ca ll th a t th o se two a lte r n a tiv e form u lation s made o p p o site p r e d ic tio n s o n ly w ith r e so e c t to in t e n s it y o f p o s it iv e outcom es. The fo.emulation

78 65 "by Shaw and Sk olnick (1971) p r e d ic te d th a t a s p o s it iv e ou t comes become more fa v o ra b le AH should d ecrease w h ile the form u lation by W alster (196?) p r e d ic te d e x a c tly the opposi t e. When th e fin d in g s are examined at a l l th ree in term ed iate l e v e l s, th e approach o f W alster (196?) seems to r e c e iv e the most su pp ort. At a l l th r e e in term ed ia te l e v e l s, evidence r ev ea led t h a t as p o s i t i v e outcomes became more favorab le AR tended to in c r e a se. Thus, th e s e fin d in g s are c o n s is te n t w ith V Jalster's p o s itio n th a t o b serv ers operate as i f they could p o s s ib ly be th e cau se o f th e observed outcomes in a sim ila r s e t o f circu m sta n ces. As such, th ey con vin ce th em selves th a t th ey would have a n tic ip a te d the outcomes and appear to be w i l li n g to hold a c to r s r e sp o n sib le fo r the d e sir a b le outcomes t h a t they a ls o would have caused had they been in th a t s it u a t io n. However, th ere are se v e r a l im portant o b se r v a tio n s worth n o tin g h e r e. F ir s t, th e fin d in g s a t the l e v e l o f fo r e s e e a b i l i t y were somewhat eq u iv o ca l w ith r e sp e c t to the favor- a b i l i t y o f p o s it iv e outcom es. Some evid en ce in d ic a te d th a t AR d ecreased, not in c r e a se d, as p o s it iv e outcomes became more fa v o r a b le. Thus, both fo rm u la tio n s o f d e fe n siv e a ttr ib u tio n r eceiv ed some support a t t h i s l e v e l. Even though th e re was c o n f l ic t in g ev id en ce, th e im portant p o in t i s th a t th e basic p r e d ic tio n s from e it h e r v e rsio n o f d e fe n siv e a t tr ib u tio n fin d s o v e r a ll suppoi t a t th is l e v e l. Observer*:-; held a c to r s

79 66 more r e sp o n sib le fo r u n d e sir a b le outcomes than fo r d e sir a b le outcom es and more r e sp o n s ib le a s u n d e sir a b le outcomes became more s e v e r e. Second, to some e x te n t, support fo r d e fe n s iv e a t tr ib u t io n i s found a t th e l e v e l o f j u s t i f i c a t i o n a s w e ll. S im i la r to the l e v e l o f f o r e s e e a b i li t y, the l e v e l o f j u s t i f i c a t io n in c lu d e s fo r e s e e a b le outcom es and the evid en ce again r e v e a le d th at AR in c re a se d as both p o s it iv e and n e g a tiv e outcom es became more extrem e. However, w ith th e in trod u c t io n o f j u s t i f i a b l e a c tio n s, d e fe n siv e a t t r ib u t io n p r e d ic t io n s fo r the e f f e c t o f outcome v a len ce are n ot supported. In f a c t, most evid en ce in d ic a te d AR was h igh er for* b e n e fic i a l outcomes than fo r u n d e sir a b le n eg a tiv e outcom es. F in a lly, support fo r th e d e fe n siv e a t t r ib u t io n th eory form ulated by H a lste r ( 1967) d im in ish es even fu r th e r a t th e l e v e l o f com m ission, bhcn th e outcom es were u n fo re se ea b le and u n c o n tr o lla b le, o b serv ers h eld a c to r s more r e sp o n sib le fo r d e sir a b le outcom es than fo r u n d e sir a b le outcom es. M oreover, o b serv ers held a c to r s l e s s r e sp o n sib le as th e se u n d e sir a b le outcom es became more severe but more r e sp o n sib le as th e d e sir a b le outcomes became more fa v o r a b le. This p a tte r n o f AR i s p r e c is e ly o p p o site to th a t p r e d icte d by Shaw and Skolnicic ( ). In c o n c lu sio n, the fin d in g s dem onstrate th a t p a tte rn s o f Al'i a s a fu n c tio n o f outcome c h a r a c t e r is t ic s and in tu rn, support fo r d e fe n siv e a t t r ib u t io n, depends upon the l e v e l o f

80 6? c a u s a lit y as o b j e c t iv e ly d e p ic ted in d if f e r e n t ty p e s o f a c tio n s it u a t io n s. T h erefore, th e se fin d in g s support the lik e lih o o d th a t one major problem in p a s t resea rch has been th a t d if f e r e n t r ese a r c h e r s have employed a c tio n s it u a t io n s which rep resen ted d if f e r e n t l e v e l s o f c a u s a lit y. As such, dem onstrated p a tte r n s o f AR have v a r ie d from study to study because o b serv ers were in te r p r e tin g the nature o f th e a c t o r 's r o le a t d if f e r e n t l e v e l s o f c a u s a lit y. C on sequ en tly, r ese a r c h e r s have f a i le d to c o n s is t e n t ly fin d o v e r a ll support fo r any p a r tic u la r approach to o b se r v e r s' assignm ent of r e s p o n s ib ilit y as a fu n c tio n o f outcome c h a r a c t e r is t ic s. In f a c t, th e o v e r a ll im p lic a tio n fo r resea rch on l e v e l s o f c a u s a lit y i s th a t a l l co n cep tu a l approaches to c a u sa l a t t r i b u tio n s and r e s p o n s ib ilit y assignm ent may be c o r r e c t. However, support fo r any p a r tic u la r approach l i k e l y depends upon th e l e v e l o f c a u s a lit y a t which a c to r s and ob serv ers arc o p era t in g. Statem ent o f O b je c tiv e s R esearchers have examined th e e f f e c t s o f outcome valen ce ( p o s it iv e or n e g a tiv e ) and outcome in t e n s i t y (low or high) on a c t o r s and o b se r v e r s' assignm ent o f r e s p o n s ib ilit y. S tu d ie s have in d ic a te d th a t o b serv ers hold a c to r s more r e sp o n s ib le fo r n e g a tiv e outcom es than fo r p o s it iv e outcomes w h ile a c to r s hold th em selv es more r e sp o n s ib le fo r p o s it iv e outcomes than fo r n e g a tiv e outcom es. Moreover, s tu d ie s have in d ic a te d th a t th e d iscrep an cy between p o s it iv e anrj n eg a tiv e

81 68 outcomes in crea sed fo r both a c to r s and o b serv ers as outcomes became more extrem e. The th eory o f d e fe n siv e a ttr ib u tio n and th e correspon d ing con cept o f s e l f - i n t e r e s t were d ev eloped to account fo r th e se r e s u lt s. However, o th er resea rch th a t has examined a cto r-o b serv er d iffe r e n c e s as a fu n c tio n o f outcome c h a r a c te r is tic s have y ie ld e d fin d in g s th a t were in c o n s is te n t and sometimes o p p o site o f th e p r e d ic tio n s based on the d e fe n siv e a ttr ib u tio n approach. One p o s s ib le problem w ith p a st resea rch has been th a t resea rch ers have employed a d iv e r se s e t o f a c tio n s i t u a t io n s. These resea r ch ers have te s te d th e same hypotheses but have employed a c tio n s it u a tio n s which have varied alon g two d e fin a b le d im ensions. 3 in ce a c tio n s it u a tio n s rep resen t s tim u li fo r making a ttr ib u tio n s o f c a u s a lity and a ssig n in g r e s p o n s ib ilit y, the f a c t th a t s tu d ie s have v a ried the two dim ensions may account fo r th e in c o n s is te n c ie s from p a st resea rch. The f i r s t dim ension r e la t e s to th e o b je c tiv e in form ation th a t i s made a v a ila b le to s u b je c ts about th e nature o f th e in te r a c tio n between th e a cto r and th e a c tiv e environment in producing the f i n a l outcome, 3haw and h is a s s o c ia te s have dem onstrated th a t th e degree to which o b servers hold a c to r s r e sp o n sib le fo r a low to high p o s it iv e or n e g a tiv e outcome depended upon the nature o f th e in te r a c tio n between the a c to r and environm ental fo r c e s as o b je c t iv e ly d ep icted in a c tio n s it u a tio n s.

82 69 The r esea rch o f Shaw and h is a s s o c ia t e s was "based on f iv e " con tex tu a l l e v e l s o f c a u s a lity " which d i f f e r e n t i a l l y rep resen ted th e nature o f th e a c t o r s r o le and th e r e l a t iv e i n f l u ence th a t th e a c to r had in producing th e f i n a l outcom e. The f iv e l e v e l s were a s s o c ia t io n, com m ission, f o r e s e e a b ilit y, j u s t i f i c a t i o n, and i n t e n t i o n a l i t y. S im ila r ly, F ish b ein and A jzen (1973) proposed th a t th e in c o n s is t e n c ie s in p a st resea rch may be due to th e f a c t th a t c o n f l ic t in g s t u d ie s employed a c tio n s it u a t io n s a t d if f e r e n t l e v e l s o f c a u s a lit y. However, p a st r e se a r c h has n ot s y s t e m a tic a lly examined both a c to r s ' and o b ser v er s' r e s p o n s ib ilit y assignm ent a s a fu n c tio n o f the f iv e l e v e l s o f c a u s a lit y. Hence, a major o b je c tiv e o f th e p resen t study i s to examine b oth a c to r s ' and o b serv ers' assignm ent o f r e s p o n s ib ilit y a s a fu n c tio n o f outcome c h a r a c t e r is t ic s a c ro ss the f i v e l e v e l s o f c a u s a lit y. The second dim ension r e l a t e s to th e s a lie n c e o f the in fo rm a tio n a v a ila b le to a c to r s and o b se r v e r s. R esearchers have in d ic a te d th a t a c to r s tend to a t t r ib u t e c a u s a lit y and a s s ig n r e s p o n s ib ilit y to s it u a t io n a l determ inan ts o f h is own b eh avior w h ile o b ser v er s tend to a tt r ib u te c a u s a lit y and a s s ig n r e s p o n s ib ilit y to th e a c to r. The d iscrep an cy hypothes i s was form ulated to accou nt fo r th e s e fin d in g s. The hypot h e s i s su g g ested th a t a c to r s tend to fo c u s t h e ir a t t e n t io n on e x te r n a l s it u a t io n a l d eterm in an ts o f h is own b eh avior and th e f i n a l outcome w h ile ob serv ers ten d to fo c u s th e ir a t t e n t io n on the b eh avior o f th e a c to r in acco u n tin g fo r th e f i n a l outcome. T h erefore, environm ental fo r c e s are

83 70 p e r c e p tu a lly most s a l i e n t fo r a c to r s v/hilo tho behavior o f the a c to r i s p e r c e p tu a lly most s a l i e n t fo r o b se r v e r s. However, the d iscrep an cy h y p o th e sis did n ot d eal s p e c i f i c a l l y w ith the e f f e c t o f outcome c h a r a c t e r is t ic s. Duval and tlic k lu n d s "theory o f o b je c tiv e self-a w a ren ess" i s an e x te n sio n o f the d iscrep an cy h y p o th e sis a p p lied to a c to r 's assignm ent o f r e s p o n s ib ilit y. The research generated by t h i s theory in d ic a te d th a t i f an a c t o r 's a tte n tio n i s red ir ec ted to h is own behavior, th en the a c to r holds h im se lf more r e sp o n s ib le fo r both p o s it iv e and n e g a tiv e outcom es than an actor whose a tte n tio n i s not r e d ir e c te d. The su p p o sitio n i s th a t fo c u sin g an a c t o r s a tte n tio n upon h im se lf in c r e a se s h is se lf-a w a r e n e ss and i s fu n c tio n a lly e q u iv a len t to in c r e a s ing th e s a lie n c e o f th e a c to r 's behavior and h is r o le in producing the f i n a l outcome. Hence, the second o b je c tiv e o f the p r e se n t study i s to examine th e impact o f in c r e a sin g an a c to r 's self-sv.'sren ess on h is assignm ent o f r e s p o n s ib ilit y as a fu n c tio n o f outcome c h a r a c t e r is t ic s. O p e r a tio n a lly, Duval and b ic k lu n d s study dem onstrated th a t an a c to r 's se lf-a w a r e n e ss could be e f f e c t iv e ly in creased by ex p osin g th e a c to r to h is own v isu a l image w ith the use o f m irrors. Hence, the m ethodology to be employed in the p r e se n t study fo r th e fo cu s c f a tte n tio n m anipulation w i l l bo p attern ed a f t e r th a t o f Duval and D icklund.

84 71 D e r iv a tio n o f H ypotheses In lin e w ith th e sta te d o b je c t iv e s, th e p r e se n t stud y was d esig n ed to t e s t s e v e r a l h yp oth eses th a t were d eriv ed from th e t h e o r e t ic a l p o s it io n s and resea r ch r e la te d to H e id e r 's f iv e l e v e l s o f c a u s a lit y, th e th eory o f d e fe n siv e a t t r ib u t io n, the d iscrep a n cy h y p o th e sis, and th e th eory o f o b je c tiv e s e lf-a w a r e n e s s. The work o f Shaw and h is a s s o c ia t e s on H e id e r 's f i v e l e v e l s o f c a u s a lit y in d ic a te d th a t th e f i v e l e v e l s should be ordered d if f e r e n t l y fo r p o s it iv e and n e g a tiv e outcom es. Based on th e r e se a r c h o f Shaw and h is a s s o c ia t e s, i t was ex p ected th a t in th e ca se o f p o s it iv e outcom es, o b serv ers would hold a c to r s more r e s p o n s ib le from a s s o c ia t io n to com m ission to f o r e s e e a b ilit y to j u s t i f i c a t i o n to in te n tio n - a l i t y ( H yp oth esis l ). However, in th e c a se o f n e g a tiv e outcom es, t h e ir r e se a r c h su g g ested th a t th e o rd erin g o f f o r e s e e a b ilit y and j u s t i f i c a t i o n should be th e r e v erse o f th e o rd erin g found f o r p o s it iv e outcom es. T h erefo re, th e e x p e c ta tio n was t h a t fo r n e g a tiv e outcom es, o b servers would h old a c to r s more r e s p o n s ib le from a s s o c ia t io n to com m ission to j u s t i f i c a t i o n to f o r e s e e a b ilit y to in t e n t io n a lit y ( H y p o th esis 2). The h y p o th esized r e v e r s a l o f th e f o r e s e e a b i l i t y and j u s t i f i c a t i o n l e v e l s was based on th e c o n te n tio n th a t th e c o e r c io n w hich i s p r e s e n t in a c tio n s it u a t io n s a t th e l e v e l o f j u s t i f i c a t i o n d e tr a c ts s u b s t a n t ia lly from th e r e s p o n s ib ilit y o f th e a cto r f o r n e g a tiv e outcom es but n ot fo r p o s it iv e outcom es.

85 72 A d d ition al h ypotheses p e r ta in in g to o b serv ers' assignm ent o f r e s p o n s ib ilit y were d erived based on th e work o f Shaw and h is a s s o c ia te s w hich in d ic a te d th a t the e f f e c t of outcome v a le n c e and in t e n s it y depended upon the l e v e l o f c a u s a lit y. F i r s t, a t the l e v e l o f a s s o c ia t io n, o b serv ers' AR was expected to be r e la t iv e ly u n a ffected by outcome c h a r a c t e r is t ic s (H ypoth esis 3 ). T h is h y p o th e sis was based on the fin d in g s o f Shaw and h is a s s o c a te s th a t AR tended to be minimal ir r e s p e c t iv e o f outcome c h a r a c t e r is t ic s. The h y p o th e sis was a d d itio n a lly b a sed on the c o n te n tio n th a t th e o b je c tiv e cau sal r o le o f the a c to r i s minimal and c le a r ly d iffe r e n tia te d from th e ca u sa l r o le o f the a c tiv e environment and as such m inim ally su b je c t to in te r p r e ta tio n on th e p a rt of o b servers. Second, a t th e le v e l o f com m ission, o b serv ers' AR was exp ected to be h ig h er fo r p o s it iv e outcomes than fo r n eg a tiv e outcom es. In a d d itio n, AR was expected to in crea se a s p o s it iv e outcom es became more fa v o ra b le but d ecrea se a s n eg a tiv e outcomes became more se v e r e (H ypothesis *0. T his h y p o th esis was o p p o site to p r e d ic tio n s b ased on d e fe n s iv e a ttr ib u tio n as form ulated by W alster ( 1966) and Shaw and S kolnick ( )- The h y p o th esis was based on th e co n ten tio n th a t a t the com m ission l e v e l, the a c t o r 's behavior i s j u s t if ia b le and th e outcomes are u n fo reseea b le and u n c o n tr o lla b le. As such, a c tio n s itu a tio n s a t t h is l e v e l do not r e p r ese n t c o n d itio n s h y p o th esized to produce d e fe n siv e a t tr ib u t io n. M oreover, the fin d in g s o f Shaw and h is a s s o c ia t e s p rovid ed support fo r t h is p r e d ic te d p a tte r n o f AR by o b s e r v e r s.

86 73 Third, a t th e le v e l o f f o r e s e e a b ilit y, AR by ob servers was expected to be h igh er fo r n eg a tiv e outcomes than fo r p o s it iv e outcomes and was expected to in c r ea se a s both p o s it iv e and n eg a tiv e outcomes became more extreme (H ypothesis 5). T h is h y p o th e sis was based on th e co n ten tio n th a t th e o b je c tiv e in form ation a t t h i s le v e l regard in g the f o r e s e e a b ilit y o f the outcome and th e j u s t i f i a b i l i t y o f the a c t o r 's b eh avior i s q u ite ambiguous. As such, a c tio n s it u a t io n s a t t h i s le v e l rep resen t c o n d itio n s which are. conducive to th e o p era tio n o f d e fe n siv e a t tr ib u t io n. The p r e d ic tio n conform s to th e p o s it io n o f W alster (1967) which receiv ed s u b s ta n tia l support in the resea rch o f Shaw and h is a s s o c ia t e s. However, i f th e p resen t study fin d s th a t AR d ecrea ses as p o s it iv e outcomes become more fa v o r a b le, i t would support the d e fe n siv e a t tr ib u t io n approach o f W alster ( 1966) and Shaw and S k oln ick (1971)* Fourth, a t the l e v e l o f j u s t i f i c a t i o n, o b serv ers' AR was expected to be h igher fo r p o s it iv e outcomes th an fo r n eg a tiv e outcomes and was exp ected to in c r e a se a s p o s it iv e and n eg a tiv e outcomes became more extreme ( H yp oth esis 6), T h is h y p o th esis was based on th e fin d in g s su p p ortin g th e co n ten tio n th a t the c o e r c iv e a c t iv e environm ent d e tr a c ts s u b s ta n tia lly from th e r e s p o n s ib ilit y o f th e a cto r fo r n e g a tiv e outcomes but not fo r p o s it iv e outcom es. However, th e f a c t th a t th e a cto r fo r e s e e s th e consequences o f h is a c tio n s le a d s ob servers to hold a c to r s r e sp o n sib le commensurate w ith the in t e n s it y o f the f in a l outcome whether th a t outcome i s p o s it iv e or n e g a tiv e. The

87 ?b h y p o th esis was a ls o supported by th e m a jo rity o f th e fin d in g s o f Shaw and h is a s s o c ia t e s. F in a lly, a t th e l e v e l o f in t e n t i o n a lit y, o b serv ers' AR was exp ected to be r e l a t i v e l y u n a ffe c te d by outcome charact e r i s t i c s (H ypothesis 7 ) T his h y p o th e sis was based on th e fin d in g s o f Shaw and h is a s s o c ia t e s th a t AR tended to be maximal ir r e s p e c tiv e o f outcome c h a r a c t e r is t ic s. The hypot h e s is was fu r th er based on th e c o n te n tio n th a t th e o b je c tiv e c a u sa l r o le o f th e a cto r i s maximal and c le a r ly d iffe r e n tia te d from th e ca u sa l r o le o f th e a c tiv e environm ent and as such m inim ally su b je c t to in te r p r e ta tio n on th e p a rt o f o b serv ers. H ypotheses 1 and 2 which concerned o b serv ers' assignm ent o f r e s p o n s ib ilit y fo r p o s it iv e and n e g a tiv e outcomes a c r o ss le v e l s can be used as a r e fe r e n c e p o in t fo r exam ining a c to r - o b serv er d iffe r e n c e s. What fo llo w s i s th e d e r iv a tio n o f h ypoth eses p e r ta in in g to a c to r -o b se r v e r d iffe r e n c e s in r e s p o n s ib ilit y assignm ent a s a fu n c tio n o f outcome charact e r i s t i c s, as a fu n c tio n o f th e f iv e l e v e l s o f c a u s a lity, and a s a fu n c tio n o f r e d ir e c tin g an a c t o r 's a tte n tio n to h im self so a s to in c r e a se h is s e lf-a w a r e n e ss. H ypotheses 1 and 2 sta te d th a t th e o rd erin g o f the l e v e l s in term s o f o b serv ers' assignm ent o f r e s p o n s ib ilit y would be d iffe r e n t fo r p o s it iv e and n e g a tiv e outcom es. H ypothesis 1 s ta te d th a t fo r p o s it iv e outcom es, o b serv ers' AR would in crea se from a s s o c ia tio n to com m ission to f o r e s e e a b ilit y to j u s t i f i c a tio n to in t e n t io n a lit y. H yp oth esis 2 s ta te d th a t fo r negat iv e outcom es, o b serv ers' AR would in c r e a se from a s s o c ia t io n to

88 75 com m ission to f o r e s e e a b ilit y follow ed, by a drop a t th e l e v e l o f j u s t i f i c a t i o n and then a n o th er s u b s ta n tia l in c r e a se a t th e l e v e l o f i n t e n t io n a lit y. But r e se a r c h on d e fe n siv e a t t r ib u t io n has in d ic a te d t h a t th e assignm ent o f r e s p o n s ib ilit y by a c to r s and o b servers d i f f e r s a s a fu n c tio n o f th e p o s i t i v i t y or n e g a t iv it y o f th e f i n a l outcom e. R esearch fin d in g s r e la te d to d e fe n siv e a t t r ib u t io n have in d ic a te d th a t compared to o b se r v e r s, a c to r s ten d to h old th em selv es more r e s p o n s ib le fo r p o s it iv e outcomes but ten d to h old th em selv es l e s s r e s p o n s ib le fo r n e g a tiv e outcom es. Thus, th e two i n i t i a l h y p o th eses can be expanded to in clu d e ex p ected d is c r e p a n c ie s betw een a c to r s ' and o b serve rs' r e s p o n s ib ilit y assignm en t a s a fu n c tio n o f outcome c h a r a c te r i s t i c s and th e f i v e l e v e ls o f c a u s a lit y. In th e ca se o f p o s i t i v e outcom es, a c t o r s should a s s ig n g r e a te r r e s p o n s i b ilit y to th e m se lv es than o b se r v ers sh o u ld a s s ig n to a c to r s a c r o ss a l l f i v e l e v e l s o f c a u s a lit y (H y p o th esis 8 ). In th e ca se o f n eg a t i v e outcom es, a c t o r s should a s s ig n l e s s r e s p o n s ib ilit y to th e m se lv es than o b servers sh o u ld a s sig n to a c to r s a c r o ss a l l f i v e l e v e l s o f c a u s a lit y ( H y p o th esis 9?. N ext, r e se a r c h on the th e o r y o f o b je c tiv e s e lf-a w a r e n e s s has in d ic a te d t h a t fo r p o s it iv e and n e g a tiv e outcom es, a c t o r s whose s e lf-a w a r e n e s s i s in c r e a s e d tend to h o ld th em selv es more r e s p o n s ib le than a c to r s whose s e lf-a w a r e n e s s i s not in c r e a s e d. These fin d in g s a lo n g w ith th e resea r c h on th e d iscrep a n cy h y p o th e sis su g g e st a rank o rd er fo r th ree groups o f s u b je c ts t

89 76 ( 1 ) a c to r s whose s e lf-a w a r e n e s s i s in c r e a se d, ( 2 ) a c to r s whose s e lf-a w a r e n e ss i s n ot in c r e a se d, and ( 3 ) o b se r v e r s. However, th e ran k ing o f th e th r ee groups o f s u b je c ts was ex p ected to be d if f e r e n t fo r p o s it iv e and n e g a tiv e outcom es. For p o s it iv e outcom es, AR by a c to r s whose s e lf-a w a r e n e ss i s in c r e a se d should be h ig h er than AR by a c to r s whose s e l f - aw areness i s n ot in c r e a se d which in tu rn should be h ig h er than AR by o b se r v e r s. Furtherm ore, th e ranking o f th e th r e e groups should be th e same a c r o ss a l l f i v e l e v e l s o f c a u s a lit y ( H yp oth esis 10) In c o n tr a s t, fo r n e g a tiv e outcom es, AR by a c to r s whose s e lf-a w a r e n e s s i s n ot in c r e a se d should be l e s s th a n AR by a c to r s whose s e lf-a w a r e n e ss i s in c r ea se d which in tu rn should be l e s s th an AR by o b se r v e r s. M oreover, th e ran k in g o f th e th r e e groups should be th e same a c r o ss a l l f i v e l e v e l s o f c a u s a lit y ( H yp oth esis 1 1 ). F in a lly, r e la t iv e to a c to r s ' assignm en t o f r e s p o n s ib ilit y w ith in each l e v e l o f c a u s a lit y r a th er than a c r o ss l e v e l s, two h y p o th eses may be d eriv ed based on d e fe n s iv e a t tr ib u t io n and th e corresp o n d in g n o tio n o f s e l f - i n t e r e s t. A ctors should tak e more r e s p o n s ib ilit y f o r p o s it iv e outcom es than fo r n e g a tiv e outcom es a c r o ss a l l f i v e l e v e l s o f c a u s a lit y (H y p o th esis 12). M oreover, a c to r s sh ould ta k e more r e s p o n s ib ilit y as p o s it iv e outcom es become more fa v o r a b le but sh ould take l e s s resp o n s i b i l i t y a s n e g a tiv e outcom es become more severe a c r o s s a l l f i v e le v e l s o f c a u s a lit y ( H yp oth esis 13).

90 77 METHOD S u b jects The Ss were 60 m ale undergraduate stu d e n ts e n r o lle d in p sy ch o lo g y co u r ses a t L o u isia n a S ta te U n iv e r s ity. The Ss were v o lu n te e r s, p a r tic ip a te d in d iv id u a lly in th e experim ent, and r e c e iv e d e x tr a cou rse c r e d it fo r p a r t ic - ip a t in g. C o n str u c tio n o f A c tio n S it u a t io n s A p o o l o f 20 a c t io n s i t u a t i o n s was u sed (Appendix A ). The 20 a c t io n s i t u a t i o n s corresp ond ed to a 5 x 2 x 2 f a c t o r i a l arrangem ent o f th e f i v e l e v e l s o f c a u s a lit y ( a s s o c ia t io n, com m ission, f o r e s e e a b i l i t y, j u s t i f i c a t i o n, i n t e n t i o n a l i t y ), th e two l e v e l s o f outcome v a le n c e ( p o s i t i v e, n e g a t iv e ), and th e two l e v e l s o f outcome in t e n s it y (lo w, h i g h ). Through th e a p p ro p ria te u se o f names and pronouns i n r e f e r r in g to th e prim ary a c to r in each s it u a t io n, th e w ording o f th e s i t u a t i o n s was c o n s is t e n t w ith th e r o le o f th e Ss a s a c to r s or o b se r v e r s. The p o o l o f 20 a c t io n s it u a t io n s was d evelop ed and v a lid a te d i n th r e e s t a g e s. s it u a t io n s was o b ta in e d. F i r s t, a la r g e p o o l o f a c tio n Each s i t u a t i o n was e it h e r o r ig in a l or was adap ted from q u e s tio n n a ir e s d ev elo p ed by Shaw and h i s a s s o c i a t e s. The s i t u a t i o n s were w r itt e n to s a t i s f y o p e r a tio n a l d e f i n i t i o n s o f th r ee d im en sion s (outcom e v a le n c e, outcome i n t e n s i t y, l e v e l o f c a u s a lit y ) and some s p e c i f ic c r i t e r i a w hich are g iv e n b elow.

91 78 The f i r s t two dim ensions (outcome v a le n c e and in t e n s it y ) r e fe r to th e nature o f the f i n a l outcome a s sta ted in th e "assignment o f r e s p o n s ib ility " q u estio n w hich fo llo w s each s it u a tio n (se e Appendix A). The f in a l outcome i s p o s it iv e (fa v o r a b le, d e s ir a b le ) or n eg a tiv e (unfavorable or u n d esir ab le) and v a r ie s in in t e n s it y, i. e. the degree of p o s i t i v i t y or n e g a tiv ity (low, h ig h ). The le v e l o f c a u s a lity dim ension r e f e r s to the nature o f the primary a c t o r 's ro le in producing th e f in a l outcome I. A sso c ia tio n l e v e l. The a c tio n s o f th e primary a c to r are u n rela ted to th e f in a l outcom e, but th e primary a c to r i s in d ir e c tly r e la te d to the f i n a l outcome by v ir tu e o f a p erson al u n it r e la tio n s h ip w ith oth er p eo p le whose a c tio n s do lead to th e f in a l outcome. I I. Commission l e v e l. The a c tio n s o f th e primary a c to r lead to th e f in a l outcome, but th e f in a l outcome i s not intended and the primary a cto r could not have known or a n tic ip a te d the consequences o f h is a c tio n s. I I I. F o r e s e e a b ility l e v e l. The a c tio n s o f the prim ary a cto r r e s u lt in th e f in a l outcome and th e f in a l outcome i s not in ten d ed, but th e primary a cto r cou ld have known or a n tic ip a te d the consequences o f h is a c t io n s. IV. J u s t i f ic a t i o n l e v e l. The primary a cto r fo r e s e e s the consequences o f h is a c tio n s, but ex ten u a tin g circum sta n ces and/or e x te r n a l c o e r c iv e fo r c e s op erate to warrant or f a c i l i t a t e h is a c tio n s.

92 79 V. In ten tio n a lity l e v e l. The primary actor foresees the consequences of h is action s and he a cts to produce the fin a l outcome in sp ite o f circum stances and environmental forces which operate to in h ib it h is behavior. The action situ a tio n s were also w ritten to s a tis fy four sp ec ific c r ite r ia : ( i) at the le v e l o f a sso cia tio n, the primary actor i s always in d irectly related to the fin a l outcome by v irtu e of a s ib lin g rela tio n sh ip with h is brother, (2) the name of the primary actor is d iffer e n t for each situ a tio n, (3) "the primary actor i s never the actual recip ien t of the fin a l outcome, i. e. someone or something e ls e is always the d irect object of harm or b en efit, and (40 the grammatical structure o f the assignment o f responsi b i l i t y question i s constant in that the recip ien t is always the subject of a verb phrase which sta te s the fin a l outcome (e.g. To what degree i s "the actor" responsible for "the recipient" being k ille d? ). The second stage in developing the 20 action situ a tio n s involved a s e r ie s o f v a lid a tio n sessio n s with -psychology graduate students (n=12). The students were requested to categorize each action situ a tio n according to the operational d efin itio n s o f the three dimensions,,'dach dimension was discussed and illu s tr a te d in d e ta il p rior to categorizin g the situ a tio n s. The wording o f the situ a tio n s always corresponded to obsorver-ss. I f disagreement occurred in categorisin g an action situ a tio n, then the situ a tio n was

93 80 re-w ritten or deleted based on d iscu ssion with the graduate students concerning the source of the d if f ic u lt y. disagreement occurred, then the situ a tio n was kept. I f no The fin a l pool of 20 action situ a tio n s for which to ta l agreement was obtained was selected for use in a second prelim inary study with undergraduate students enrolled in an introductory psychology course. This second preliminary study con stitu ted the third and f in a l stage in v erify in g that the fiv e le v e ls o f c a u sa lity were v a lid ly represented by the 20 action situ a tio n s. The purpose was to determine i f ob jective d istin c tio n s among the fiv e le v e ls o f c a u sa lity were v a lid ly reflected in the 20 action s itu a tio n s. Based on the operational d efin itio n s of the fiv e le v e ls, the fiv e le v e ls were expected to be d istin gu ish ab le along four dimensions: fo r e se e a b ility, in te n tio n a lity, ca u sa lity, and ju s tific a tio n. Each dimension re fe rs to the nature o f the a c to r 's role in producing the fin a l outcome. Each dimension i s defined below and the expected d is tin c tio n s among the le v e ls along each dimension are stip u la ted. To determine i f ob jective d istin c tio n s among the le v e ls were v a lid ly r e flected in the action situ a tio n s, undergraduate students were asked to evaluate the role of the primary actor along one of the four dimensions. The 3s evaluated the role of the actor along only one dimension for a l l 20 action situ a tio n s. The Ss (n=55) were asked to

94 81 read each s itu a tio n and to in d ica te the degree to which they honestly thought th at the outcome was foreseeab le (n = ll), that the outcome was intended (n=15)» that the action s o f the actor caused the outcome (n^ l^ ), or that the action s o f the actor were j u s t if ie d (n=15)» Each of the four dimensions were defined as fo llo w s. F o reseea b ility referred to the degree to which the actor could have a n ticip a ted that h is a ctio n s would r e s u lt in the f in a l outcome. In te n tio n a lity referred to the degree to which the f in a l outcome was part o f the goals o f the primary a cto r. C au sality referred to the degree to which the fin a l outcome was a r e s u lt o f the a ctio n s o f the actor. J u s tific a tio n referred to the degree to which the action s of the actor were warranted. For each s itu a tio n the 3s were asked to in d ica te what they thought by checking a 21-point percentage sca le which ranged from to 100;i in u n its o f 5'^. The wording of the situ a tio n s always corresponded to observer-3s and answer sh eets were provided with the appropriate question and the percentage sc a le fo r each s itu a tio n (e.g. see Appendix D). The general form o f the q uestions fo r each situ a tio n were as follow s: (1) To what degree could "the actor" foresee "the fin a l outcome?", (2) To what degree was "the f in a l outcome" "the a cto r's" in ten tio n?, (3) To what degree did "the actor's" actio n s cause "the fin a l outcome?", or (*! ) To what degree were the a ction s o f "the a c to r ju s tifie d?

95 82 The mean percentage-response of the Ss was determined fo r the four situ a tio n s at each le v e l o f c a u s a lity. The mean fo r e s e e a b ility, in te n tio n a lity, c a u s a lity, and j u s t if ic a t io n are gra p h ica lly presented in Figures 1 through r e s p e c tiv e ly. The general p a ttern o f the means for each dimension was expected to be as fo llo w s based on the op eration al d e fin itio n s o f th e le v e ls and was supported o v er a ll. F ir s t, the o v er a ll exp ectation was that f o r e s e e a b ility (Figure l), in te n tio n a lity (Figure 2), and c a u sa lity (Figure 3) would in crease from a sso c ia tio n to commission to fo r e s e e a b ility to j u s t if ic a t io n to in te n tio n a lit y. However, j u s t if ic a t io n (Figure *0 was expected to generally decrease from a sso c ia tio n to commission to j u s t if ic a t io n to fo r e s e e a b ility to in te n tio n a lity. Second, some s p e c ific exp ectation s fo r each dimension were as fo llo w s and were gen erally supported: ( l) fo re see a b ilit y (Figure 1) was expected to be r e la t iv e ly high at the le v e ls o f j u s t if ic a t io n and in te n tio n a lity, r e la t iv e ly low at th e le v e ls o f a sso c ia tio n and commission, but in te r mediate a t the le v e l o f fo r e s e e a b ility ; (2) in te n tio n a lity (Figure 2) was expected to be r e la t iv e ly high at the le v e ls o f j u s t if ic a t io n and in te n tio n a lity but r e la t iv e ly low at the le v e ls o f a sso c ia tio n, commission, and fo r e s e e a b ility ; (3) c a u s a lity (Figure 3) was expected to be r e la tiv e ly high at the le v e ls o f fo r e s e e a b ility, j u s t if ic a t io n, and in te n tio n a lity but r e la t iv e ly low at the le v e ls of

96 Figure 1. H H J o - II III IV Mean percent fo r e s e e a b ility by le v e l for question, "To what degree could X foresee that h is actions would lead to Y?" Ju stification {%) m Intentionality (js) c o I II lli IV Mean percent in te n tio n a lity by le v e l fo r question, "To what degree was Y X's intention?" Jb- V I II I I I IV V I II III IV V Figure 3* Mean percent ca u sa lity by le v e l for question, "To what degree did X's action s cause Y?" Figure 4-. Mean percent ju s tific a tio n by le v e l for question, "To what degree were X's actions ju s tifie d?

97 association and commission; (i)0 ju s t if ic a t io n (Figure *0 was expected to "be r e la tiv e ly high a t the le v e ls of asso cia tio n, commission, and ju s tific a tio n but r e la tiv e ly low at the le v e ls of fo r e s e e a b ility and in te n tio n a lity. Since th ese general expectations were supported overall the fiv e le v e ls of c a u sa lity were s a tis fa c to r ily represented in the 20 action situ a tio n s used in the present study. Procedure The 60 Ss were assigned at random to the three "subject role - focus o f attention" con d itions with 20 Ss per group (actors w ith focus, a cto rs without focus, and observers). When each S arrived a t the experimental room, the S was greeted by the E. Each S was then led into the experimental room and asked to s i t a t a ta b le. At th is p o in t, the procedure fo r each S in the "no focus" and "focus" condition s was d iffe r e n t. Actor-Ss and observer-ss in the "no fo cu s condition only p articip ated in the "assignment of resp o n sib ility " task Each S was handed a w ritten se t o f in stru ctio n s which corresponded to h is ro le as an actor or as an observer (Appendix B ). An answer sheet was provided along with these in stru ctio n s (Appendix C). As the E handed the m aterials to the S, the E told each S, "Here are some in stru ctio n s which w i l l explain what w e 'll be doing. Would you please take your time and read these in stru ctio n s thoroughly very c a r e fu lly. When you're fin ish e d, I w ill

98 85 answer any questions you might have," In th ese in stru ctio n s, each S was asked to read each situ a tio n s and to in d ica te the degree to which he, as the actor, or the central character, as an observer, was responsible for the f in a l outcome in each situ a tio n. The 20 situ a tio n s were presented in random order to each S by means o f a slid e p ro jecto r. Each s lid e contained one action s itu a tio n along w ith the assignment of responsi b i l i t y question which was appropriately phrased fo r h is role as an actor or as an observer. Each action situ a tio n was numbered on the s lid e and the order in which the situ a tio n s were presented to each S was given on the S's answer s h e e t. Each S was asked to in d icate what he honestly thought was th e appropriate degree of responsi b i l i t y by u sin g the percentage sc a le which was provided at the top o f the answer sheet (Appendix C). The scale was a 21-point percentage scale ranging from 0?2 to 100%in u n its of 5?. A fter each S fin ish e d reading the in stru ctio n s and a l l questions were answered, the S was shown two p ra ctice s lid e s to fa m ilia rize the S w ith the procedure, to make sure the projector was focused properly, and to make sure the S had no fu rth er q uestions. The p r a ctic e s lid e s contained the fo llo w in g two a c tio n situ a tio n s and the S was asked to in dicate h is answer to the assignment of r e sp o n sib ility question a t the end o f the in stru ction s: (1) One day

99 86 several of K arl s (your) fr a te r n ity brothers were fis h in g at the lak e. They found a fis h in g rod in the bushes and broke i t. To what degree i s Karl (are you) responsible for the fish in g rod being broken? and (2) Walter was (You were) making some business telephone c a lls. When the phone rang in one home Walter (you) ca lle d, i t awakened a man who was sleep in g near a broken gas heater. I f he had not awakened, the leaking gas would have k ille d him. To what degree i s Walter (are you) responsible for the man's l i f e being saved? A fter the two p ractice s lid e s had been shown and a l l questions were answered, the E to ld the S, "Just to repeat before we sta r t, remember th at there are no right or wrong answers. The purpose of what we're doing i s to find out what people honestly think i s the appropriate degree of r e sp o n sib ility fo r things that happen. Again, please read each story very c a r e fu lly and tr y to imagine that you are a ctu a lly a w itness ( i f the S was an observer) or a ctu a lly involved ( i f the S was an actor) before you indicate what you h onestly think is the appropriate degree of responsi b i l i t y in each ca se. I f you don't have any questions, then we w ill s t a r t." The procedure for each actor-ss in the "focus" conditio n was as fo llo w s. Each S was greeted by the E as was done w ith the Ss in the "no focus" con d ition. The S was led in to the experimental room and asked to s i t at a ta b le.

100 87 The ta b le had m irrors lo ca ted on the fron t and on two sid es o f the tab le lik e a lib r a r y study c a r r e l. The Ss were to ld th a t they were to perform two ta sk s, the f i r s t one being a sim ple c l e r ic a l task which required the use o f the m irrors. The ca rrel was constructed such th a t the acto r-s s under focu s of a tte n tio n were exposed to th e ir f u l l fr o n ta l and p r o f ile image throughout th e ir p a r tic ip a tio n in the a ssig n ment of r e s p o n s ib ility ta sk. The s lid e s were projected onto the w a ll in fron t o f the S such th at the S had to look up over the fron t m irror to read the s itu a tio n s. As the S looked up and down in reading the s itu a tio n s and answering the assignment o f r e s p o n s ib ility q u estio n s, he was co n sta n tly exposed to h is own image. Patterned a fte r the study by Duval and Wicklund (1973)* t h is m anipulation was to in crease the self-aw aren ess of the acto r-s s throughout h is p a r tic ip a tio n in the assignment of r e s p o n s ib ility task. Before the S p a r ticip a ted in the assignment o f responsi b i l i t y ta sk, he was asked to perform a sim ple c l e r ic a l ta sk. The purpose o f the c l e r i c a l task was to provide a ra tio n a le fo r the presence of th e m irrors. The Ss were given a verbal s e t of in str u c tio n s fo r performing the c l e r i c a l task (Appendix D ). The c le r i c a l ta sk required each S to copy an IBM answer sh eet that was already marked onto a blank IBM s h e e t. The two IBM sh eets were placed between an opaque p a r titio n and the front m irror. Upon com pletion of the c l e r ic a l ta sk, the S was informed of h is performance. Then, the IBM sh eets and

101 88 p a r titio n were removed, but the m irrors remained fo r the "focus o f attention" m anipulation. The E then handed the S the m aterials fo r the assignment of r e s p o n s ib ility task. The remainder of the procedure was ex a c tly the same as that fo r actor- and observer-ss in the "no focus" con d ition. A ll s were debriefed a fter th e ir p a rticip a tio n and each was given a card signed by the E so that each S could n o tify h is teacher th at he had p a rticip a ted in the experiment for extra course c r e d it. Design A com pletely randomized design w ith a s p lit -p lo t arrangement o f fa cto rs was employed (Appendix E). The Between-Ss p a r titio n o f the an alysis o f variance involved the three subject ro le - focus of a tte n tio n groups (actors with focu s, a cto rs without focu s, and observers). The W ithin-ss p a r titio n involved the 20 action situ a tio n s which represented a 5 x 2 x 2 fa c to r ia l arrangement of le v e l of c a u sa lity, outcome valence, and outcome in te n s ity. The dependent variable was the amount o f r e sp o n sib ility assigned for the outcome in each a ctio n situ a tio n (AH).

102 89 RESULTS AND DISCUSSION The purpose of the present study was to compare actors' and ob servers' assignment of r e sp o n sib ility in action s itu a tio n s. The action situ a tio n s d iffer ed in two resp ects. The action situ a tio n s d iffered not only in terms o f fiv e le v e ls of ca u sa lity but also in terms o f the nature o f the fin a l outcome. The fiv e le v e ls of c a u sa lity were a sso cia tio n, commission, fo r e s e e a b ility, ju s tific a tio n, and in te n tio n a lity. The fin a l outcome in each action situ a tio n d iffer ed in both valence and in te n sity. That i s, the fin a l outcome in each action situ a tio n was eith er a low or high p o s itiv e or negativ e outcome. Thus, a l l research questions concerned actors' and observers' assignment of r e sp o n sib ility as a function of le v el of ca u sa lity and outcome c h a r a c te r istic s as objectiv e ly depicted in a ctio n situ a tio n s. Of th e 13 hypotheses tested in th is study, seven hypotheses (Hypotheses 1-7) dealt ex clu siv ely with observers' assignment o f r e sp o n sib ility. The findings rela ted to these seven hypotheses w ill be considered p rio r to the findings rela ted to the hypotheses dealing w ith actorobserver d ifferen ces, namely Hypotheses Hypotheses 1-7 concerned o b servers assignment of r e sp o n sib ility to a cto r s. To in v e stig a te these hypotheses, an a n a ly sis of variance was conducted using data obtained from observers only. The ob jective was to a sse ss the e f f e c t s of le v e ls o f ca u sa lity and outcome c h a r a c te r istic s on the

103 90 degree to which observers held a cto rs resp o n sib le. The dependent variab le was th e amount o f r e s p o n s ib ility (AR) assigned by observers fo r the f in a l outcome in each action situ a tio n. The r e s u lts o f t h is a n a ly sis o f variance are given in Table 1. The r e s u lts of t h i s a n a ly sis and the r e su lts o f orthogonal comparisons w i l l be d iscu ssed w ithin the context o f the re lev a n t hypotheses. Hypotheses 1 and 2 concerned observers' assignment of r e s p o n s ib ility for p o s itiv e and negative outcomes across the fiv e le v e ls of c a u s a lity. I t was expected that observers' assignment o f r e s p o n s ib ility would in crease across the f iv e le v e ls o f c a u sa lity, but the in crease across le v e ls ( i. e. the ordering o f the le v e ls ) in th e mean AR assigned by observers would depend upon outcome valence. The r e s u lts o f the a n a ly sis of variance supported th is expectation (Table 1 ). The h ighly s ig n ific a n t le v e l x valence in te r a c tio n, F (4,3 6 l) = 11.96, p <.0 1, indicated that the changes in o b ser v ers assignment of r e s p o n s ib ility across le v e l s was a ffe c te d by the valence o f the f in a l outcomes. The fin d in g s r e la te d to H ypothesis 1 and Hypothesis 2 w ill be presented sep a ra tely. The data relev a n t to these two hypotheses are contained in Table 2 and Figure 5* Table 2 con tain s the mean AR by observers fo r p o s itiv e and negative outcomes a t each le v e l o f c a u sa lity w hile Figure 5 presents th ese means g ra p h ica lly.

104 91 Table 1 A n a ly s is o f v a r ia n c e f o r o b ser v er s' assign m en t o f r e s p o n s ib i li t y to a c to r s Source df ss MS F S u b ject L ev el (L) if lifo 9if ^ ** V alen ce (V) ** I n t e n s it y ( I ) L x V if ^0 66?if ** L x I if ififlif V x I L x V x I if if. 62 ** erro r i f. l l t o t a l, * P <.0 5 ** p <.01

105 92 Table 2 Mean amount o f r e sp o n sib ility assigned by actors, observers, and actors w ith the self-aw areness manipulation (SA actors) for p o s itiv e and negative outcomes a t each le v e l of ca u sa lity and o v era ll! a sso c ia tio n (Level X), commission (Level II), fo r e s e e a b ility (Level III), ju s tific a tio n (Level IV) and in te n tio n a lity (Level V) Level Role Valence I II III IV V O verall Actor P o sitiv e Negative Observer P o sitiv e Negative SA Actor P o sitiv e * Negative

106 93 Mean AR P o sitiv e Negative ---- lo I ll III IV V Level Figure 5* Mean amount of r e sp o n sib ility (AR) assigned by observers for p o s itiv e and negative outcomes a t each le v e l of ca u sa lity : a sso c ia tion (I), commission (II), fo r e s e e a b ility (III), j u s t if ic a t io n (IV), and in ten tio n - a l i t y (V).

107 9 ^ Hypothesis 1 predicted th at for p o s itiv e outcomes, observers would hold the actor more responsible from asso c ia tio n to commission to fo r e s e e a b ility to ju s tific a tio n to in te n tio n a lity. This p red iction was based on the contention that r e la tiv e to the causal ro le of the a ctiv e environment, the potency of the a c to r 's cau sal role in action situ a tio n s o b je c tiv e ly in creases from a sso c ia tio n to commission to fo r e s e e a b ility to j u s t if ic a t io n to in te n tio n a lity. As such, a corresponding increase in r e sp o n sib ility assignment by observers was expected. The order of the means confirmed th is hypothesis (Table 2 and Figure 5) Observers held the actor more responsible for p o s itiv e outcomes from a sso c ia tio n (X = 6. 9?S) to commission (X = ^2.190 to fo r e s e e a b ility (X = S) to ju s tific a tio n (X = 7 2.^ ) to in te n tio n a lity (X = ). In contrast to Hypothesis 1, Hypothesis 2 d ealt with negative outcomes and stated a d iffe r en t ordering o f the fiv e le v e ls o f c a u sa lity. For negative outcomes, the expectation was th at r e sp o n sib ility assignment would increase from assoc ia tio n to commission to fo r e s e e a b ility but then would drop at the le v e l of j u s t if ic a t io n follow ed by a su b stan tial increase at the le v e l of in te n tio n a lity. The expected drop in r e sp o n sib ility assignment a t the le v e l of ju s tific a tio n was based on the contention th at the external coercive forces operating at th is le v e l detract su b sta n tia lly from the perceived contribution of the actor in the production o f

108 95 n egative outcomes. However, fo r p o sitiv e outcomes, these extern al coercive fo rce s were not expected to operate in a d etractin g manner. Thus, as stated in Hypothesis 1 for p o s itiv e outcomes, observers1 r e s p o n s ib ility assignment was expected to be lower at the le v e l of fo r e s e e a b ility than at the le v e l of ju s t if ic a t io n w hile for negative outcomes, as stated in Hypothesis 2, observers' r e sp o n sib ility assignment was expected to be greater a t the le v e l o f fo r e se e a b ility than a t the le v e l o f ju s tific a tio n. The order o f the means fo r negative outcomes supported H ypothesis 2 (Table 2 and Figure 5) Observers held the actor more resp onsible for negative outcomes from a sso cia tio n (Y =* 6.6%) to commission (X = 10.*$) to fo r e s e e a b ility (X = 78.*$) follow ed by the expected drop at the le v e l of ju s tific a tio n (X = 53*970 and then a su b sta n tia l increase toward a maximum a t the le v e l o f in te n tio n a lity (X = 9^*370* There were a d d itio n a l hypotheses p ertain in g to observers1 assignment of r e s p o n s ib ility. These hypotheses, namely Hypotheses 3 through 7, concerned the p attern of r e sp o n sib ility assignment as a fu n ction of outcome c h a r a c te r istic s w ithin each le v e l of ca u sa lity rather than across le v e ls of c a u sa lity. In general, Hypotheses 3-7 were based on the expectation that the e ffe c t of outcome ch a ra c te r istic s on observers' responsi b i l i t y assignment would change from le v e l to le v e l. That i s, observers' assignment of r e sp o n sib ility fo r low to high p o s itiv e outcomes and for low to high negative outcomes would

109 96 depend upon the le v e l of c a u sa lity as o b je ctiv ely depicted in the action situ a tio n s. The r e su lts of the an alysis of variance supported t h is expectation (Table 1). The r e s u lts revealed a highly s ig n ific a n t le v e l x valence in tera ctio n, F(^»36l) = I I. 96, p <. 01, and a highly sig n ific a n t le v e l x valence x in ten sity in te ra ctio n, F(^,36l) = 4-.62, p<.01. Therefore, observers' assignment o f r e sp o n sib ility for outcomes o f varying valence and in te n sity did depend upon the le v e l o f c a u sa lity. Before examining the r e su ltin g p atterns of r e sp o n sib ility assignment, the conceptual framework underlying th is examinatio n w ill be b r ie fly reviewed. B a sica lly, the present study was designed to determine which le v e ls of ca u sa lity would y ie ld a pattern of r e sp o n sib ility assignment that would conform to the p attern predicted from defensive a ttrib u tio n theory. Defensive a ttr ib u tio n theory p red icts that observers' assignment of r e sp o n sib ility for any action sequence depends upon both the valence and in te n s ity of the outcome. Thus, a review of the p attern o f r e sp o n sib ility assignment predicted from d efen sive a ttrib u tio n theory is e s se n tia l for understanding the d iscu ssio n o f the predicted and obtained r e s u lts. F irst, w ith respect to the e ffe c t of outcome valence, observers are expected to hold actors more responsible for negative outcomes than for p o s itiv e outcomes. Second, the expected e ffe c t of outcome in te n s ity on observers' assignment o f r e sp o n sib ility i s d ifferen t for p o sitiv e and negative outcomes.

110 97 For negative outcomes, observers are expected to hold a cto rs more resp onsible as th ese negative outcomes become more severe. That i s, o b serv ers assignment o f r e s p o n s ib ility i s expected to be greater fo r high negative outcomes than for low negative outcomes. In contrast to the predicted e ffe c t o f in te n s ity of negativ e outcomes, the p red icted e f f e c t of in te n s ity of p o s itiv e outcomes i s le s s c le a r. As p o s itiv e outcomes become more favorable, the r e s u lt could be e ith e r increased or decreased amounts of r e s p o n s ib ility assignm ent. E ither an increase or a decrease would support defen sive a ttr ib u tio n theory sin ce these con trad ictory p red ic tio n s correspond to th e a lte r n a tiv e form ulations of d efen sive a ttr ib u tio n theory proposed by W alster ( 1967) and Shaw and Skolnick (1971). W alster p o stu l ated that r e s p o n s ib ility assignment would in crease as p o s itiv e outcomes became more favorable w hile Shaw and Skolnick predicted the opposite trend, i.e. r e s p o n s ib ility assignment would decrease as p o s itiv e outcomes became more favorable. But the p attern o f r e s p o n s ib ility assignment p redicted by d efen sive a ttr ib u tio n theory was not expected a t a l l fiv e le v e ls o f c a u sa lity sin ce the nature o f the in te ra c tio n between the primary actor and the a ctiv e environment i s d is t in c t ly d iffe r e n t a t each le v e l. I t was p r e c ise ly the unique nature o f the in te r a c tio n among cau sal agents as depicted in a ctio n s itu a tio n s at each le v e l which was to determine the e ffe c t o f outcome c h a r a c te r is tic s on ob servers

111 98 r e s p o n s ib ility assignm ent at each le v e l. The nature o f t h is in te r a c tio n at each le v e l was a lso expected to determine whether or not the r e su ltin g p a ttern of observers' responsi b i l i t y assignment would conform to the p red ictio n s of d efen siv e a ttr ib u tio n theory. At some le v e ls the nature of the causal in te r a c tio n was expected to be in terp reted by observers in such a way that would y ie ld a p a ttern of r e s p o n s ib ility assignment th at would conform to th at p red icted by d efen sive a ttr ib u tio n. At other le v e ls though, the nature of the in te r a c tio n was not expected to be in terp reted in such a way th at would y ie ld a p attern o f r e s p o n s ib ility assignm ent conforming to th a t p red icted by d efen siv e a ttr ib u tio n. Therefore, in examining the fin d in g s at d iffe r e n t le v e ls o f c a u sa lity, the d iffe r in g nature of the in te r a c tio n among causal agents was expected to r e s u lt in d iffe r e n t observed p attern s o f r e s p o n s ib ility assignm ent. In tu rn, these p a ttern s were to be comparable to the p attern expected from d efen siv e a ttr ib u tio n theory. In re la tin g the d iffe r e n c e s in the nature of the in te r a c tio n among cau sal agents a t each le v e l to the observed p a tter n at each le v e l and in turn, comparing the observed p a ttern at each le v e l with the p attern p red icted from d efen siv e a ttr ib u tio n theory, two ca teg o ries o f data were examined -- means and frequency d is tr ib u tio n s. The la t t e r category of data (I.e. frequency d istr ib u tio n s) i s not p resen ted at every le v e l of c a u s a lity. These d istr ib u tio n s

112 99 supplemented th e inform ation represented by comparisons among the means and were presented when appropriate fo r e lu c id a tin g the fin d in g s. The f i r s t category o f data was the mean amount of r e s p o n s ib ility (AR) assign ed by ob servers. Three in d iv id u a l orthogonal comparisons were conducted a t each le v e l o f causa l i t y to determine the e f f e c t of outcome valence and in te n s ity. Each comparison corresponded to a separate p red ictio n of d efen sive a ttr ib u tio n th eo ry. The fo llo w in g comparisons were made at each le v e l o f ca u sa lity : ( l) the mean AR for p o s itiv e outcomes was compared to the mean AR fo r negative outcomes; (2) the mean AR fo r low p o s itiv e outcomes was compared to the mean AR fo r high p o s itiv e outcomes; (3) the mean AR for low negative outcomes was compared to the mean AR for high n egative outcomes. Thus, at each le v e l of c a u s a lity, comparison 1 represented th e e ffe c t o f outcome valence w hile comparisons 2 and 3 represented the e ffe c t of in te n s ity of p o s itiv e and negative outcomes. The second category o f data th at was examined at each le v e l was the frequency o f observers whose in d iv id u a l p a ttern o f r e s p o n s ib ility assignment matched the pattern r e fle c te d by the means. For example, i f the mean AR assigned by a l l 20 observers at a p a rticu la r le v e l was higher fo r p o s itiv e outcomes than fo r n egative outcomes, then the q uestion was how many of th e 20 observers showed the pattern r e fle c te d by the means. That is, "How many observers assigned more r e s p o n s ib ility for p o s it iv e outcomes than fo r n egative outcomes

113 100 a t th a t level?" S im ila rly, i f the mean AR assigned "by a l l 20 observers a t a p a r tic u la r le v e l was le s s fo r high p o s itiv e outcomes than fo r low p o s itiv e outcomes, then the q uestion was "How many observers assigned le s s r e s p o n s ib ility fo r high p o s itiv e outcomes than fo r low p o s itiv e outcomes at th at level?" L ikew ise, the same q u estion was asked regarding the in te n s ity o f negative outcomes. Based on these q u estio n s, frequency d istr ib u tio n s were constructed at each le v e l of c a u s a lity. F ir s t, for each in d iv id u a l observer at a p a rticu la r le v e l o f c a u sa lity, the mean AR fo r p o s itiv e outcomes (two p o s itiv e outcomes per le v e l) was compared to the mean AR fo r negative outcomes (two n egative outcomes per level). Each observer was then categorized as having a mean AR fo r p o s itiv e outcomes greater than, equal t o, or le s s than h is mean AR fo r n egative outcomes. F in a lly, a determ ination was made of the frequency o f observers whose mean AR fo r p o s itiv e outcomes was greater than, equal t o, or l e s s than that fo r n egative outcomes. The r e s u ltin g frequency d istr ib u tio n thus r e fle c te d the e f f e c t o f outcome valence and sim ila r frequency d is tr ib u tio n s were produced fo r each le v e l o f c a u sa lity. In con stru ctin g th ese d istr ib u tio n s and the d istr ib u tio n s described below, i f the d ifferen ce between means was le s s than 10#, then the amount o f r e s p o n s ib ility assigned for p o s itiv e and negative outcomes was said to be "equal." That i s, a d iffer e n ce of 10% or more was needed to declare "greater than" or " less than."

114 101 A sim ila r procedure was used in producing frequency d istr ib u tio n s to in v e stig a te the e ffe c t of in te n s ity of outcomes a t each l e v e l. For each in divid ual observer a t each le v e l, the amount o f r e s p o n s ib ility assigned fo r the high p o s itiv e outcome (one high p o s itiv e outcome per le v e l) was compared to the amount o f r e s p o n s ib ility assign ed for the low p o s itiv e outcome (one low p o s itiv e outcome per level). A fter c la s s if y in g each observer based on t h i s comparison, a determ ination was then made o f the frequency o f observers whose amount of r e s p o n s ib ility fo r the high p o s itiv e outcome was greater than, equal to, or le s s than th a t fo r the low p o s itiv e outcome based on a d ifferen ce of 10?2!. Thus, th ese frequency d istr ib u tio n s r e fle c te d the e f f e c t o f in te n s ity of p o s itiv e outcomes at each le v e l o f c a u sa lity. U sing the same method th a t was used fo r p o sitiv e outcomes, a frequency d istr ib u tio n r e f le c t in g the e f f e c t of in te n s ity of negative outcomes a t each le v e l was con stru cted. That i s, a determ ination was made o f the frequency of observers whose amount o f r e s p o n s ib ility fo r the high n egative outcome a t each le v e l was greater than, equal to, or l e s s than th at fo r the low negative outcome at th a t le v e l. H ypothesis 3 concerned the e ffe c t of outcome ch aracteri s t i c s a t the le v e l o f a s s o c ia tio n. At the le v e l of a s s o c ia tio n, observers* assignment o f r e s p o n s ib ility was expected to be minimal reg a rd less of th e nature o f the fin a l outcome. That i s, the exp ectation was th a t r e s p o n s ib ility assignment would not be a ffe c te d by outcome c h a r a c te r is tic s at t h is

115 102 polar le v e l of c a u sa lity. This expectation was based on the contention that in action situ a tio n s a t the le v e l o f a sso cia tio n, the primary actor has nothing d ir e c tly to do with the production o f the fin a l outcome. Since the a c to r 's causal r o le in producing the f in a l outcome i s minimal or e s s e n tia lly zero in a ctio n situ a tio n s at th is le v e l, the premise was th at observers would have d if f ic u lt y in p erceiv in g or evaluating the actor as being responsible regardless of type o f outcome. As such, there was no expectation that defensive a ttrib u tio n would operate at th is l e v e l. The findings a t the le v e l of a sso cia tio n conformed to expectation (Tables 2-3 and Figures 5-6). At the le v e l of a sso cia tio n, o b servers assignment of r e sp o n sib ility was near minimum irresp ectiv e of outcome c h a r a c te r is tic s. The overall mean AR a t th is le v e l was 6.?57&. Moreover, a l l comparisons among means -- p o s itiv e versus negative outcomes (Table 2 and Figure 5)» low versus high p o sitiv e outcomes (Table 3 and Figure 6), and low versus high negative outcomes (Table 3 and Figure 6) - - were not sig n ific a n t at the.05 le v e l. In contrast to the le v e l o f a sso cia tio n, outcome charact e r i s t i c s were expected to a f f e c t observers' assignment of r e sp o n sib ility at the le v el o f commission. Observers assignment o f r e sp o n sib ility was expected to be affected by outcome ch a ra c teristic s a t th is le v e l sin ce the actor and the active environment more c le a r ly share a causal ro le in producing the

116 103 T able 3 Mean amount of r e sp o n sib ility assigned by a ctors, observers, and actors with the self-aw areness manipulation (SA actors) for low or high p o sitiv e or negative outcomes at each le v e l of causality* a sso cia tio n (L evel I ), commission (Level XI), fo r e se e a b ility (Level III), ju s tific a tio n (Level XY) and in te n tio n a lity (Level V) Level Role Valence In te n sity I II III IV V Actor P o sitiv e Negative Low High Low High Observer P o sitiv e Negative Low High Low High SA Actor P o sitiv e Negative Low High Low High *

117 10^ fo O - fo~ So?o- Mean AR io - P ositive Negative so- JO- 2o- / Low High In ten sity Figure 6. Mean amount o f r e sp o n sib ility (AR) assigned by observers a t the le v e l o f a sso cia tio n depending upon outcome valence (positive, negative) and outcome in te n s ity (low, high).

118 105 f in a l outcome than a t the le v e l o f a sso c ia tio n. However, observers were not expected to in terp ret or evaluate the nature of the causal in te r a c tio n between a ctor and a c tiv e environment in such a way th at would lead observers to assig n r e s p o n s ib ility implying the operation o f d efen sive a ttr ib u tio n. In fact, the p attern o f r e s p o n s ib ility a ssig n ment was expected to be the opposite o f that predicted from d efensive a ttr ib u tio n as formulated by Shaw and Skolnick (1971). That i s, observers were expected to hold the acto r more resp onsible fo r p o s itiv e outcomes than for negative outcomes. Moreover, w hile observers were expected to hold the actor more resp on sib le as p o s it iv e outcomes became more favorable, observers were expected to hold the actor l e s s resp on sib le as negative outcomes became more severe (H yoothesis 4-). W ithin the le v e l o f commission, i t was p red icted th at observers would hold the actor more resp o n sib le for p o s itiv e outcomes than for n eg a tiv e outcomes. This was expected because th e inform ation in a ctio n s itu a tio n s a t th is le v e l not only in d ica tes th a t the actor's behavior i s j u s t if ie d r e la tiv e to h is intended goals but a lso in d ic a te s that the f in a l outcome i s the r e s u lt of an u n co n tro lla b le and unforeseeable sequence o f ev en ts. The nature of the a c to r s r o le i s such th a t the actor j u s t if ia b ly or innocently a cts in order to accomplish a p a rticu la r outcome. However, the outcome which a c tu a lly r e s u lts i s not intended by the a cto r since the a c to r s behavior merely s e t s in motion an unforeseeable and u n con trollab le sequence of events which ends

119 106 with the fin a l outcome. The contention of Hypothesis 4- was that t h is information regarding the j u s t i f i a b i li t y o f the a c to r 's behavior and the u n fo reseea b ility of the fin a l outcome would not be conducive to the operation of defensive a ttrib u tio n. Instead, when the a c to r 's behavior i s j u s tifie d, observers were expected to demonstrate a w illin g n ess to a sso cia te the actor with and give the actor more cred it fo r p o sitiv e outcomes than for negative outcomes, e s p e c ia lly sin ce these n egative outcomes are unavoidable. This was expected even' though the informatio n in d ica tes that the actor does not a ctu a lly intend eith er type of outcome. The fin d in gs supported expectations regarding the effect o f outcome valence at the le v e l o f commission (Table 2 and Figure 5)- Observers' assignment of r e sp o n sib ility for p o sitiv e outcomes was s ig n ific a n tly higher than that for negative outcomes, F(l,36l) = 3&.0, p<.01. The mean AR for p o sitiv e outcomes was $ while the mean AR for negative outcomes was 10.4$. Moreover, the frequency d istrib u tio n s at the le v e l of commission revealed that none o f the observers assigned more r e sp o n sib ility for negative outcomes than for p o sitiv e outcomes. Therefore, as expected, the e ffe c t of outcome valence was opposite to that predicted by defensive a ttrib u tio n theory. S im ilarly, with resp ect to the e ffe c t o f in te n sity of p o s itiv e and negative outcomes at the le v e l of commission, the information regarding the j u s t if i a b i l i t y of the a cto r s

120 107 behavior and the u n fo reseea b ility of the f in a l outcome was not expected to he conducive to the operation of defensive a ttrib u tio n as formulated by Shaw and Skolnick (1971)* Instead, observers were expected to hold the actor more responsible as p o sitiv e outcomes became more favorable but le s s responsible as n egative outcomes became more severe (Hypothesis 4-). With respect to p o s itiv e outcomes, the contention was that when information e x is t s in d ica tin g that an a cto r's behavior i s j u s tifie d, observers would be w illin g to assoc ia te the actor with and hold the actor more responsible commensurate with outcome fa v o r a b ility. This was expected in sp ite of inform ation in d icatin g th at the actor does not intend or foresee the consequences of h is behavior. However, contrary to expectation, observers did not hold the actor more resp onsible for the high p o sitiv e outcome than for the low p o sitiv e outcome. opposite occurred (Table 3 and Figure 7 ). In fa c t, the Observers held the actor le s s responsible as p o sitiv e outcomes became more favorable. The r e su lts of comparing the mean AR for the high p o sitiv e outcome (X = 29.8#) with the mean AR fo r the low p o s itiv e outcome (X = 5^*5#) revealed a highly sig n ific a n t decrease as p o sitiv e outcomes became more favorable, F( 1,361) = 10.9, p<.01. Moreover, most in d ivid u al observers exhibited the same p attern reflected by the means. The m ajority of the observers (70^) assigned le s s r e sp o n sib ility

121 108 P o ~ P o sitiv e Negative ---- Mean AR Low In te n sity High Figure 7- Mean amount of r e sp o n sib ility (AR) assigned by observers at the le v e l of commission depending upon outcome valence (positive, negative) and outcome in te n sity (low, high).

122 109 for the high p o sitiv e outcome than for the low p o sitiv e outcome while only 20# assigned more r e sp o n sib ility and 10# assigned an equal amount of r e sp o n sib ility as p o s itiv e outcomes became more favorable. While observers were expected to hold the actor more responsible as p o s itiv e outcomes became more favorable, observers were expected to hold the actor le s s responsible as negative outcomes became more severe (Hypothesis k). This expected pattern was based on the notion that observers would not be w illin g to a sso cia te the actor w ith and hold the actor responsible for negative outcomes which are unavoidable, esp e cia lly when these negative outcomes are severe. Again, r e s u lts did not conform to expectation (Table 3 and Figure 7) Contrary to expectation, there appeared to be a s lig h t tendency on the part o f observers to hold the actor more responsible as negative outcomes became more severe, ^he mean AR for the low negative outcome was 7-0# while the mean AR for the high negative outcome was 13*8#. However, the differen ce between these means was not sig n ific a n t a t the.05 le v e l, F ( l, 36l) = O.8 3. Moreover, contrary to expectation, the frequency d istrib u tio n a t th is le v e l revealed that only a few observers ( 10#) assigned le s s r e sp o n sib ility as negative outcomes became more severe. Of the remaining observers, 35fa assigned more r e sp o n sib ility and assigned an equal amount o f r e sp o n sib ility as negative outcomes became more severe. Thus, action situ a tio n s at the le v e l of commission appeared to be at le a s t p a r tia lly conducive to observers holding the

123 110 actor more resp onsible for the more severe n egative outcome even though the n egative outcome i s o b je c tiv e ly depicted as being unavoidable. However, even though the p attern in the means for low to high negative outcomes was contradictory to Hypothesis 4-, th is trend as a fu n ction of se v e r ity of n egative outcomes should not be taken too l i t e r a l l y. The mean AR for both the low negative outcome and the high n egative outcome tended to be rath er minimal and these means were not s ig n ific a n tly d iffe r e n t. Therefore, th ese r e s u lts were c o n siste n t w ith the p ro p o sitio n in H ypothesis 4 th at observers do not hold actors resp on sib le fo r n egative outcomes which cannot be avoided reg a rd less of the se v e r ity o f these negative outcomes. Even though an equivocal in terp re ta tio n was suggested regarding the e f f e c t o f se v e r ity o f negative outcomes, the obtained r e s u lts in d ica ted th at a t the le v e l o f commission, observers hold a cto rs le s s resp on sib le as p o s itiv e outcomes become more favorable but somewhat more resp onsible as negative outcomes become more severe. Thus, the o v era ll p attern as a function o f outcome in te n s ity which was observed at t h is le v e l approximated th at p red icted from defensive a ttr ib u tio n theory as formulated by Shaw and Skolnick (1971)* At the le v e l o f fo r e s e e a b ility, lik e a t the le v e l o f commission, outcome c h a r a c te r is tic s were expected to a f fe c t observers' assignment of r e s p o n s ib ility. However, unlike at the l e v e l of commission, the p attern of r e s p o n s ib ility a ssig n ment a t the le v e l o f fo r e s e e a b ility was expected to conform

124 I l l to the p attern based on defensive a ttrib u tio n theory. The pattern was expected to match the p red iction s from defensive a ttrib u tio n theory sin ce action situ a tio n s at th is le v e l were conceptualized as being h ighly ambiguous re la tiv e to the j u s t if ia b ilit y o f the a cto r's behavior and the fo resee a b ility o f the fin a l outcome. In action situ a tio n s at t h is le v e l, the information does not c le a r ly in d ic a te whether or not the actor a n tic ip a te s or fo resees the consequences o f h is behavior and whether or not the a cto r's behavior i s j u s t if ia b le. However, the inform ation does in d icate that the fin a l outcome is at le a s t under the p o te n tia l control o f the actor. Mainly because the fo r e s e e a b ility o f the outcome i s not e x p lic it in a ctio n situ a tio n s at th is l e v e l, the ambiguous nature of the a cto r's ro le was expected to be conducive to subjective in terp reta tio n on th e part of observers and, as such, the operation of d efen sive a ttr ib u tio n was lik e ly. Based on the contention th at defensive a ttrib u tio n would be operative a t t h is le v e l, i t was p red icted that observers would hold the actor more responsible for negative outcomes which could have been avoided than fo r p o sitiv e outcomes which may or may not have been expected or a n tic ipated by the primary actor. Moreover, sin ce the negative outcomes could be interpreted by observers a s p reventab le, observers were expected to a ttr ib u te greater r e sp o n sib ility to the actor as these avoidable negative outcomes became more severe (Hypothesis 5)

125 112 The find in gs revealed a pattern of r e sp o n sib ility assign ment which seemed to support these expectations. Observers held the actor more resp onsible for negative outcomes than for p o sitiv e outcomes (Table 2 and Figure 5)* The mean AR assigned by observers for negative outcomes (X = 7 8.^ ) was higher than th a t for p o s itiv e outcomes (X = 6?.6? ). However, the d ifferen ce was not sig n ific a n t at the,05 le v e l, F ( l, 36l) = 2.1. But as predicted, observers held the actor s ig n ific a n tly more responsible as negative outcomes became more severe (Table 3 and Figure 8). The mean AR for the high negative outcome was 83.^fo while the mean AR for the low negative outcome was 68. 5^. This d ifferen ce was highly s ig n ific a n t, F (l,3 6 l) = 7. 02, p <.0 1. Thus, even though the differen ce between the mean AR for negative outcomes and the mean AR fo r p o sitiv e outcomes was not s ig n ific a n t, the o v era ll p attern of r e sp o n sib ility assignment as a functio n o f outcome valence and se v erity o f negative outcomes conformed to the pattern expected on the b a sis o f defensive a ttrib u tio n theory. Regarding the e ffe c t of in te n s ity of p o sitiv e outcomes, the p red iction s derived from defensive a ttrib u tio n theory i s somewhat unclear. The confusion a r ise s because o f the two a lter n a tiv e form ulations of defensive a ttrib u tio n theory proposed by W alster (1967) and Shaw and Skolnick (1971)* W alster predicted an increase while Shaw and Skolnick predicted a decrease in r e sp o n sib ility assignment as p o sitiv e outcomes became more favorable. Moreover, the research of

126 113 Mean AR 3o - P o sitiv e Negative ---- Low In ten sity High Figure 8. Mean amount of r e sp o n sib ility (AR) assigned by observers at the le v e l of fo r e s e e a b ility depending upon outcome valence (p o sitiv e, negative) and outcome in te n s ity (low, h ig h ).

127 11^ Shaw and h is a sso c ia te s on le v e ls o f cau sality y ie ld e d r e su lts which supported both p red ictio n s. W alster's 19^7 form ulation o f defensive a ttr ib u tio n received su b stan tial support in the work of Shaw and h is a sso c ia te s and as such, H ypothesis 11 predicted that observers would be w illin g to give the actor more cred it as p o sitiv e outcomes became more favorab le. The mean amount o f r e s p o n s ib ility assigned by observers in the present study did not conform to p red iction s of eith er form ulation o f defensive a ttr ib u tio n. Observers' assignment o f r e sp o n sib ility did not appear to be affected by the in te n s ity of p o sitiv e outcomes (Table 3 and Figure 8), The mean AR for the low p o s itiv e outcome was 66.5% w h ile the mean AR for the high p o s itiv e outcome was 68.8%. The same lack of evidence for a systematic e f f e c t of in te n s ity of p o sitiv e outcomes was found with regard to the frequency o f observers whose in d iv id u a l pattern o f responsi b i l i t y assignment increased, decreased, or remained the same as a function of outcome fa v o r a b ility. The number of observers whose assignment of r e sp o n sib ility increased (35%) «decreased (bq%), or remained the same (25%) was rather eq u ally d is tr ib uted. The fa ct that some observers showed an in crea se, some showed a decrease, and others showed no change su ggests that the nature of the a c to r 's role in producing the p o s itiv e outcomes was interpreted in a v a r ie ty of ways. Although th is v a riety in the patterns was in keeping with the ambiguous nature of the information in a ctio n situ a tio n s at t h is le v e l, sim ilar evidence for a variable

128 115 in terp reta tio n on the p art of observers was not indicated fo r sev erity of negative outcomes. The m ajority of observers ( 65^) assigned more r e sp o n sib ility as negative outcomes became more severe while 25!^ assigned le s s and 15$ assigned the same amount o f r e sp o n sib ility as negative outcomes became more severe. Thus, i t appears that a more co n sisten t response in observers' assignment o f responsi b i l i t y occurs in a ctio n situ a tio n s at th is le v e l when negative outcomes increase in se v erity than when p o sitiv e outcomes increase in fa v o r a b ility. At the le v e l of j u s t if ic a t io n, observers' assignment o f r e sp o n sib ility was expected to be a ffected by outcome c h a r a c te r is tic s, both valence and in te n s ity. However, while the pattern o f r e sp o n sib ility assignment as a function of outcome valence was not expected to conform to expectations from defensive a ttr ib u tio n theory, the pattern o f re su lts for in te n s ity of p o s itiv e and negative outcomes was expected to conform to p red ictio n s of defensive a ttrib u tio n theory. With respect to the e ffe c t o f outcome valence, the pattern w ithin the le v e l of ju s tific a tio n was expected to be opposite to that p redicted by defensive a ttrib u tio n theory. Hypothesis 6 predicted th at observers would hold the actor more responsible for p o sitiv e outcomes than for negative outcomes because of the operation of external coercive fo rces in action situ a tio n s at t h is le v e l. This expectation was based on the proposition th at the impact of external coercion i s interpreted d iffe r e n tly for negative outcomes than for

129 116 p o s itiv e outcomes. Actors are le s s lik e ly to be held resp onsible for n egative outcomes r e s u ltin g from behavior produced under co n d itio n s o f extern al coercion than fo r p o s itiv e outcomes r e su ltin g from a ctio n s co n siste n t w ith extern al coercive fo rces d irected toward the production of b e n e fic ia l outcomes. The fin d in g s supported t h is exp ectation. Observers held the actor more resp on sib le fo r p o s itiv e outcomes (X = 72.*ff5) than fo r negative outcomes (X = 53*970 at the le v e l o f j u s t if ic a t io n (Table 2 and Figure 5 ) * T his d ifferen ce was s ig n ific a n t at the.01 le v e l, F (1,361) = 12.3* With resp ect to the e f f e c t of in te n s ity of p o s itiv e and negative outcomes, observers were expected to hold the actor more resp on sib le as both p o s itiv e and negative outcomes became more in ten se (H ypothesis 6). This p red ictio n was based on the fa c t th at in a ctio n s itu a tio n s at th is le v e l, the actor fo resees the consequences o f h is a ctio n s. Because o f the apparent fo r e s e e a b ility o f the consequences o f h is a c tio n s, observers were expected to hold the actor more resp onsible as outcomes became more extreme in sp ite o f the fa c t th a t extern al coercion f a c ilit a t e d the a c to r 's behavior. This p red ictio n was a lso c o n siste n t w ith the p attern predicted from d efen sive a ttr ib u tio n theory as formulated by W alster (1967). The fin d in g s supported the above exp ectation w ith respect to the in te n s ity o f p o s itiv e outcomes but not with resp ect to the in te n s ity o f negative outcomes (Table 3 an(l Figure 9)

130 Mean AH P o s i t i v e N e g a tiv e ---- /o - Figure 9- Low In te n s ity High Mean amount o f r e s p o n s ib ility (AR) assigned by observers at the l e v e l of j u s t if ic a t io n depending upon outcome valence (p o s itiv e, n egative) and outcome in te n s ity (low, high),

131 118 Observers held the actor s ig n ific a n tly more resp onsible for the high p o sitiv e outcome (X = 8 0.Ofo) than for the low p o sitiv e outcome (X = 6k-.8fo). This d ifferen ce was sig n ific a n t a t the.05 le v e l, F (1,361) = ^.14. Contrary to expectation regarding negative outcomes, observers held the actor le s s responsible as negative outcomes became more severe, but t h is difference was not s ig n ific a n t at the.05 le v e l, F (1, 361) = 1.2. The mean AR for the low negative outcome was 58*0# while the mean AR for the high negative outcome was h9.8%. Though not predicted, t h is dcreasing trend suggests that when external coercive forces e x is t, they operate in a d etractin g manner and as such, observers hold the actor le s s responsible as negative outcomes become more se v e r e. F in ally, Hypothesis 7 concerned the le v e l of in ten tio n - a l i t y. At the le v e l of in te n tio n a lity, observers' assign ment of r e sp o n sib ility was expected to be maximal regardless o f the nature of the f in a l outcome. In action situ a tio n s a t the le v e l o f in te n tio n a lity, the primary actor i s portrayed as the so le causal agent in the production of the f in a l outcome. Since the causal ro le of the actor i s maximal, the assignment of r e sp o n sib ility was expected to be maximal and unaffected by outcome c h a r a c te r istic s at t h is le v e l. Therefore, no evidence fo r defensive a ttr ib u tio n was expected. The findings revealed that observers' assignment of r e sp o n sib ility was equivalent and near maximum fo r a l l types

132 119 of outcomes except low p o s itiv e outcomes (Tables 2-3 and Figures 5 and 10). The mean AR for low and high negative outcomes was 9^ -8?S and 93*8^, re sp e ctiv e ly. The mean AR for the high p o sitiv e outcome was also near maximum (X = 93-0JS), but the mean AR fo r the low p o sitiv e outcome was le s s than for other types of outcomes (X = 79.0^ ). Thus, while re sp o n sib ility assignment tended to be maximal and not change as a function o f the se v erity of negative outcomes, responsi b i l i t y assignment did tend to increase as p o sitiv e outcomes became more favorable. The increase from low to high p o sitiv e outcomes was quite close to sig n ifica n ce a t the.05 le v e l, F(1,361) = 3.5. As a r e s u lt, r e sp o n sib ility assignment was higher, but not s ig n ific a n tly higher, fo r negative outcomes (X = 94.37S) than fo r p o s itiv e outcomes (X = 86.0fa). The d ifferen ce was not s ig n ific a n t at the.05 le v e l, F (1, 361) = 1.2. Even though the e ffe c t o f in te n s ity of p o sitiv e outcomes was not quite s ig n if ic a n t, the trend which showed an increase from low to high p o sitiv e outcomes was contrary to expectation. On the other hand, the fa c t that r e sp o n sib ility assignment was le s s for low p o sitiv e outcomes than other types of outcomes was sim ila r to r e s u lts found by Sulzer ( 196^). These fin d in gs suggest that since the actor must overcome in h ib itory forces in order to accomplish a b e n e fic ia l outcome, observers perceive the actor as more in flu e n tia l and more responsible for circumventing forces p ro h ib itin g the production of the h igh ly favorable outcome than for circumventing forces prohibit in g the production of the le s s favorable p o sitiv e outcome.

133 120 Mean AR P o sitiv e Negative ---- / o - Low Figure 10. In ten sity High Mean amount of r e s p o n s ib ility (AR) assigned by observers at the le v e l o f in te n tio n a lity depending upon outcome valence {p o sitiv e, negative) and outcome in te n s ity (low, h igh ).

134 121 The d iscu ssion of fin d in gs related to the hypotheses p ertain in g to ob servers assignment of r e sp o n sib ility has been concluded. The remainder of th is se ctio n deals with s ix additonal hypotheses (Hypotheses 8-13) which d ealt e x c lu siv ely with actor-observer d ifferen ces in r e sp o n sib ility assignment. Along with the observer su b jects, two groups of actors were employed actors without the self-aw areness (SA) manipulatio n and actors w ith the SA m anipulation. In th is d iscu ssion of the r e su lts, these three su bject groups w ill be referred to as observers, acto rs, and SA actors, r e sp ectiv e ly. Hypotheses 8-13 dealt w ith actor-observer d ifferen ces in r e s p o n s ib ility assignment as a function o f the fiv e le v e ls o f ca u sa lity as w ell as the valence and in te n s ity of the f in a l outcomes. The o b jectiv e was to compare the p attern s of r e sp o n sib ility assignment as a function of outcome c h a r a c te r istic s demonstrated by the three subject groups both across the f iv e le v e ls o f ca u sa lity and w ithin these le v e ls. The amount of r e sp o n sib ility assigned across and w ith in le v e ls as a function o f outcome valence and in te n s ity was expected to depend upon the role of subjects as actors or observers. Expected d iffer en ce s were based on the theory o f defensive a ttrib u tio n and the corresponding notion of s e lf - in t e r e s t as w e ll as the theory of ob jective s e lf - awareness. The o verall claim of Hypotheses 8-13 was that observers, a c to r s, and SA actors would demonstrate d iffe r in g patterns of r e sp o n sib ility assignment as a function of le v e l and

135 122 outcome c h a r a c te r is tic s. To in v estig a te these hypotheses, an a n a ly sis of variance was conducted using data obtained from observers, a cto rs, and SA a cto rs. The dependent variable was the amount of r e sp o n sib ility assigned by each actor or observer for the fin a l outcome in each action situ a tio n. The r e s u lts of th is a n a ly sis of variance are given in Table 4. The r e su lts o f the an alysis o f variance did not support the claim s o f Hypotheses 8 through 13* Since Hypotheses 8 through 13 predicted th at assignment of r e sp o n sib ility as a function o f le v e l and outcome c h a r a c te r istic s would depend upon su bject ro le, three sources o f v a r ia b ility were expected to be sig n ific a n t the role x valence in tera ctio n, the role x le v e l x valence in tera ctio n, and the ro le x le v e l x valence x in te n sity in tera ctio n. None of these in teraction s were sig n ific a n t (Table *0. Only the role x le v e l x valence in tera ctio n approached sig n ifica n ce at the.05 le v e l, P(8,1083) = 1.93* Therefore, there was very l i t t l e evidence that actors and observers d iffered in th eir assignment o f r e sp o n sib ility as a function of le v e l of c a u sa lity and outcome c h a r a c te r is tic s. Instead, the mean AR assigned for low to high p o s itiv e or negative outcomes indicated th a t both groups o f actors exhibited the same p atterns of r e sp o n sib ility assignment exhibited by observers, both across le v e ls and w ithin le v e ls (Tables 2-3 and Figures 11-16). Figure 11 shows the pattern

136 123 T a b le 4 A n a ly s is o f v a r ia n ce fo r th e amount o f r e s p o n s i b i l i t y (AR) a ssig n e d by a c t o r s, o b s e r v e r s, and a c to r s w ith th e s e lf-a w a r e n e s s m a n ip u la tio n Source d f SS MS F R ole (R) * e r r o r a L e v e l (L) ** V alen ce (V) ** I n t e n s it y (I) L x V ** L x I ** V x I L x V x I ** R x L * R x V R x I * R x L x V R x L x I R x V x I R x L x V x I e r r o r b IO t o t a l * p <.0 5 * * p <.0 1

137 o f r e s p o n s ib ilit y assignm ent by th e th r ee su b je c t groups fo r p o s i t i v e and n e g a tiv e outcom es a c ro ss th e f i v e l e v e l s o f c a u s a lit y. T able 2. The means p resen ted in F igu re 11 are g iv e n in F ig u re s show the p a tte r n o f r e s p o n s ib ilit y assignm ent fo r low to high p o s i t i v e and n e g a tiv e outcom es by each su b ject group at each l e v e l o f c a u s a lit y. The means p resen ted i n F igu res are g iv e n in Table 3.

138 125 o >c" Mean AR J o - Jo- P o sitiv e o Negative Actors ---- Observers SA Actors /o - I II I I I IV Level V.Figure 11. Mean amount of r e sp o n sib ility (AR) assigned by acto rs, observers, and actors with the self-aw areness manipulation (SA actors) for p o s itiv e and negative outcomes at each le v e l of ca u sa lity ; a sso c ia tio n ( I ), commission ( I I ), fo r e s e e a b ility ( I I I ), ju s tific a tio n (IV ), and in te n tio n a lity (V).

139 126 go Mean AR (a - P o sitiv e o Negative Actors Observers SA Actors m Low In ten sity High Figure 12. Mean amount of r e sp o n sib ility (AR) assigned by acto rs, observers, and actors with the. self-aw areness manipulation ( SA actors) at the le v e l of a sso cia tio n depending upon outcome valence (p o sitiv e, negative) and outcome in te n s ity (low, high).

140 12? 9o - P o sitiv e o Negative Actors Observers SA Actors Mean AR Ok. j«j - Low In te n s ity High Figure 13 Mean amount of r e s p o n s ib ility (AR) assigned by a cto rs, observers, and actors w ith the self-aw aren ess m anipulation (SA a cto rs) at the le v e l o f commission depending upon outcome valence (p o s itiv e, n eg a tiv e) and outcome in te n s ity (low, high).

141 to - Mean AR 4o P o s i t iv e o N eg a tiv e * A c to rs O bservers SA A cto rs _2o - Low I n t e n s it y High F igure lk. Mean amount o f r e s p o n s i b i l i t y (AR) a ssig n e d by a c to r s, o b s e r v e r s, and a c to r s w ith the s e lf-a w a r e n e s s m a n ip u la tio n (SA a c to r s ) a t th e l e v e l o f f o r e s e e a b i l i t y depending upon outcome v a le n c e ( p o s i t i v e, n e g a tiv e ) and outcome i n t e n s i t y (low, high).

142 129 Mean AR 70 - o~ /oo-?0- po- os'< t * /a P o sitiv e o Negative Actors Observers - SA Actors - /o , Low In te n sity High Figure 15. Mean amount of r e sp o n sib ility (AR) assigned by actors, observers, and actors with the self-aw areness manipulation (SA actors) at the le v e l of ju s tific a tio n depending upon outcome valence (p o sitiv e, negative) and outcome in te n s ity (low, h ig h ).

143 130 yo - 7o - Mean AR JO - P o s itiv e o N egative Actors Observers SA Actors - Low In te n sity High Figure 16. Mean amount o f r e s p o n s ib ility (AR) assigned by a cto rs, observers, and actors w ith the self-aw aren ess m anipulation (SA a cto rs) at the le v e l of in te n tio n a lity depending upon outcome valence (p o s itiv e, negative) and outcome in te n s ity (low, high).

144 131 GENERAL DISCUSSION The fin d in g s r e la t e d to o b serv ers' assignm ent o f resp o n s i b i l i t y dem onstrated th a t o b ser v er s' r e s p o n s ib ilit y a s s ig n ment a s a fu n c tio n o f outcome c h a r a c t e r is t ic s (v a len ce and in t e n s it y ) depended upon th e l e v e l o f c a u s a lit y a s o b j e c t iv e ly d ep icted in a c tio n s it u a t io n s. The observed p a tte r n o f r e s u lt s fo r o b serv ers' r e s p o n s ib ilit y assignm ent at each le v e l can be examined from two r e la te d p o in ts o f view. F ir s t, each l e v e l can be examined by comparing the r e s u lt s o b tain ed a t each l e v e l to p r e d ic tio n s o f d e fe n siv e a tt r ib u tio n th e o ry. Second, d is t i n c t i o n s betw een l e v e l s in term s o f th e nature o f th e in t e r a c t io n among c a u sa l a g e n ts (i.e. a c to r and a c tiv e environm ent) can be r e la te d to th e d if f e r e n t p a tte r n s th a t observed over l e v e l s. The d i s t i n c t io n s among th e l e v e l s were a c tu a lly th e b a s is fo r p r e d ic t io n s reg a rd in g o b ser v er s' p a tte r n s o f r e s p o n s ib ilit y assignm ent a s a fu n c tio n o f outcome c h a r a c t e r is t ic s a c r o ss a s w e ll a s w ith in l e v e l s. A lso, th e s e d i s t in c t io n s among th e l e v e l s d e fin e th e c o n d itio n s which determ ine where and to what degree d e fe n siv e a t t r ib u t io n i s l i k e l y to occu r. T h erefore, th e se two p o in ts o f view are r e la te d sin c e th ey both depend upon d e f i n i t io n a l d if f e r e n c e s among l e v e l s in th e nature o f th e in t e r a c t io n among c a u sa l a g e n ts. From th e f i r s t p o in t o f view, th a t i s, l e v e l s a t which d e fe n siv e a t t r ib u t io n th eory seems a p p lic a b le, th e observed p a tte r n s r e v e a le d th e fo llo w in g. At th e p o la r l e v e l s o f

145 132 c a u s a lit y ( a s s o c ia t io n and i n t e n t i o n a l i t y ), th e e f f e c t o f outcome c h a r a c t e r is t ic s tended to he m inim al. Since d i f f e r e n t i a l e f f e c t s due to outcome c h a r a c t e r is t ic s are n e c essa ry to support i t s p r e d ic tio n s, d e fe n siv e a t t r ib u t io n did not appear to be a p p lic a b le a t th e se l e v e l s. At th e l e v e l o f a s s o c ia t io n, th e re was no in d ic a t io n o f an e f f e c t fo r both outcome v a len ce and outcome i n t e n s i t y. At th e l e v e l o f i n t e n t i o n a l i t y, o b se r v e r s' assignm ent o f resp o n s i b i l i t y was near maximum fo r a l l ty p e s o f outcom es ex cep t low p o s it iv e outcom es. In c o n tr a s t to th e p o la r l e v e l s, outcome c h a r a c t e r is t ic s d id have an e f f e c t on o b se r v e r s' assign m en t o f r e s p o n s ib ilit y a t th e th ree in te rm e d ia te l e v e l s o f c a u s a l i t y (com m ission, f o r e s e e a b i li t y, j u s t i f i c a t i o n ). However, th e degree to which the p a tte r n matched p r e d ic tio n s from d e fe n siv e a t tr ib u t i o n th eory v a r ie d from l e v e l to l e v e l. E vidence fo r d e fe n siv e a t t r ib u t io n th eory was found a t th e l e v e l o f fo r e s e e a b i l i t y and to some e x te n t a t th e l e v e l o f com m ission. But evid en ce a g a in s t th e d e fe n siv e a t t r ib u t io n p r o c e ss was found a t th e l e v e l o f j u s t i f i c a t i o n. At the le v e l o f j u s t if ic a t io n, the pattern was opposite to that predicted from the d efen sive a ttr ib u tio n theory of Shaw and Skolnick (1971). Observers held the actor more responsible fo r p o s itiv e outcomes than for negative outcomes, more responsible as p o s itiv e outcomes became more favorable, and le s s responsible as negative outcomes became more severe.

146 133 At the le v e l of commission, the observed pattern only approximated the p red ictio n s of defensive a ttrib u tio n theory (Shaw and Skolnick, 1971) with respect to the e ffe c t of in te n s ity of p o s itiv e and negative outcomes. But with respect to the e ffe c t o f outcome valence, the r e su lts were opposite to p red ictio n s derived from defensive a ttrib u tio n theory. As to the e ffe c t o f outcome valence, observers held the actor more responsible for p o sitiv e outcomes than for negative outcomes, thereby con trad ictin g the a p p lic a b ility of defensive a ttr ib u tio n. For outcome in te n sity, observers held the actor le s s responsible as p o sitiv e outcomes became more favorable, but only s lig h tly more responsible as negative outcomes became more severe, thereby approaching the pattern predicted by Shaw and Skolnick. F in ally, at the le v e l of fo r e s e e a b ility, the observed pattern came c lo se to matching the pattern predicted by Shaw and Skolnick. Observers held the actor more responsible fo r negative outcomes than for p o sitiv e outcomes and more responsible as negative outcomes became more severe, However, sin ce observers' assignment of r e sp o n sib ility appeared to be unaffected by the fa v o ra b ility of p o sitiv e outcomes, at le a s t a portion of these r e s u lts fa ile d to support defensive a ttrib u tio n (W alster, 19&7, Shaw and Skolnick, 1971)* From the second p oint of view, namely the rela tio n sh ip s between the conceptual d istin c tio n s among le v e ls and the observed p attern s, the d istin c tio n s among le v e ls accounted rather w ell fo r the e ffe c t of outcome valence across and

147 13^ within le v e ls, "but l e s s w ell for the e ffe c t of outcome in te n s ity within le v e ls. With respect to the e ffe c t of outcome valence across le v e ls, ob servers assignment of r e sp o n sib ility for p o sitiv e outcomes was found to increase from a sso c ia tio n to commission to fo r e s e e a b ility to ju s tific a tio n to in te n tio n a lity. This increase in r e sp o n sib ility assignment across le v e ls was expected because the r e la tiv e in flu en ce of the actor in producing p o sitiv e outcomes o b je c tiv e ly increases from a sso cia tio n to commission to fo r e s e e a b ility to ju s tific a tio n to in te n tio n a lity. In contrast for negative outcomes, observers assignment of r e sp o n sib ility was found to increase from a sso c ia tio n to commission to fo r e s e e a b ility followed by a drop a t the le v e l o f ju s tific a tio n and then another su b stan tial increase a t the le v e l o f in te n tio n a lity. This deviation from the p a ttern obtained fo r p o sitiv e outcomes was predicted because o f the e ffe c t o f coercive forces at the le v e l o f ju s t if ic a t io n. External coercion operative at the le v e l o f ju s t if ic a t io n detracts from the perceived contribution of the a cto r when h is behavior r e s u lts in negative outcomes but not when h is behavior r e s u lts in p o sitiv e outcomes. With respect to the e ffe c t o f outcome valence w ithin le v e ls, outcome valence was not expected to have an e ffe c t at the p olar le v e ls o f ca u sa lity (a sso cia tio n, in te n tio n a lity ), but was expected to have an e ffe c t a t the three interm ediate le v e ls o f ca u sa lity (commission, fo r e s e e a b ility, j u s t if ic a t io n ). This was expected because at the p olar le v e ls, the actor and

148 135 the activ e environment do not share a causal role in producing the fin a l outcomes while a t the three interm ediate le v e ls, the actor and the active environment do share a cau sal r o le. As such, observers' a ttrib u tio n s o f c a u sa lity and assignment of r e sp o n sib ility to the actor are more subject to individual in terp retation and more lik e ly to be a ffected by the nature of the fin a l outcome at the interm ediate le v e ls. As expected, the r e su lts indicated that outcome valence did not have an e ffe c t at the polar le v e ls of a sso cia tio n and in te n tio n a lity, but did a ffe c t observers' assignment o f r e sp o n sib ility at the three interm ediate le v e ls. At the le v e ls of commission and ju s tific a tio n, observers held the actor more responsible for p o s itiv e outcomes than for negative outcomes. The explanation fo r these r e su lts was based on the fa ct that the a c to r 's behavior i s ju s tifie d at these le v e ls, even though for d iffer en t reasons. At the le v e l of commission, the actor ju s tifia b ly or innocently a cts in order to accomplish a p articu lar outcome. But the outcome which a ctu a lly does re su lt i s d iffe r e n t from the a cto r's intended outcome because o f an uncontrollable and unforeseeable sequence of events se t in motion by the a c to r 's behavior. In con trast, at the le v e l of ju s tific a tio n, the a c to r's behavior i s ju s tifie d due to the operation of external coercion directed toward the production of the f in a l outcome. When an a cto r s behavior is j u s tifie d as at the le v e ls o f commission and ju s tific a tio n, observers are not lik e ly to blame actors or hold actors responsible for undesirable negative outcomes.

149 136 In a d d itio n to th e apparent j u s t i f i c a t i o n o f th e a c t o r 's b eh a v io r a t th e s e l e v e l s, th e n e g a tiv e outcom es are u n avoid a b le. level. But th ey are u n avoid ab le f o r d if f e r e n t r ea so n s a t each At th e l e v e l o f com m ission, th e n e g a tiv e outcom es are beyond th e c o n tr o l o f th e a c to r s in c e th e n e g a tiv e outcom es r e s u l t from an u n fo r e se e a b le sequence o f e v e n ts. At th e l e v e l o f j u s t i f i c a t i o n, th e n e g a tiv e outcom es m ight be c o n c e p tu a liz e d a s u n avoid ab le due to th e fo r c e o f e x te r n a l c o e r c io n, i.e. e x te r n a l c o e r c io n im p lie s "no ch o ice" on th e p a r t o f th e a c to r. A p p aren tly, a s a r e s u l t o f th e in fo rm a tio n th a t th e a c t o r ' s b e h a v io r i s j u s t i f i e d and th a t th e n e g a tiv e outcom es are u n a v o id a b le, o b se r v e r s hold th e a c to r l e s s r e s p o n s ib le fo r n e g a tiv e outcom es than fo r p o s i t i v e outcom es. The in fo r m a tio n th a t th e a c to r cannot fo r e s e e th e con seq u en ces o f h is b eh a v io r a t th e l e v e l o f com m ission but does fo r e s e e th e con seq u en ces o f h i s b eh a v io r a t th e l e v e l o f j u s t i f i c a t i o n d oes n o t appear to have a b e a r in g on th e e f f e c t o f outcome v a le n c e. In c o n tr a s t to th e l e v e l s o f com m ission and j u s t i f i c a t i o n, a t th e l e v e l o f f o r e s e e a b i l i t y, th e p r e d i c t a b i l i t y o f th e f i n a l outcom es and th e in t e n t io n s o f th e a c to r are somewhat open to q u e s tio n b eca u se o f th e ambiguous in fo rm a tio n in a c tio n s it u a t io n s a t t h i s l e v e l. However, th e f i n a l outcome i s a t l e a s t under th e p o t e n t i a l c o n tr o l o f th e a c to r and a s su ch, o b ser v er s were exp ected to h old th e a c to r more r e s p o n s ib le fo r n e g a tiv e outcom es which co u ld have been avoid ed than fo r p o s i t i v e outcom es which may or may n o t be

150 137 perceived as foreseen or an ticip ated by the a ctor. Even though the d ifferen ce was not quite s ig n ific a n t, the r e su lts su ggest that on ly at the le v e l of fo r e s e e a b ility do observers hold the a c to r more resp o n sib le for negative outcomes than for p o s it iv e outcomes. With resp ect to th e e ffe c t of in te n s ity of outcomes at each le v e l, the in te n s ity of p o s it iv e outcomes must be considered separately from the e f f e c t o f in te n s ity of negative outcomes. Based on d iffe r e n c e s in the nature of the causal r o le shared by the actor and the a ctiv e environment, the e f f e c t of in t e n s it y of p o s it iv e outcomes was expected to vary from l e v e l to le v e l. Observers' a ssig n ment of r e s p o n s ib ility as a fun ction o f in te n s ity of p o sitiv e outcomes was expected to be (1) minimal fo r both low and h igh p o sitiv e outcomes a t the le v e l o f a sso c ia tio n since the a c to r 's ca u sa l role is minimal, (2) greater for high than fo r low p o s it iv e outcomes a t the le v e l o f commission sin ce the a c to r 's behavior i s apparently j u s t i fie d, (3) greater for high than fo r low p o s itiv e outcomes at the le v e l o f fo r e s e e a b ility sin ce observers were expected to convince them selves th at they would have caused the p o sitiv e outcomes and as such would hold the a ctor more responsible as p o s itiv e outcomes became more favorable (W alster, 196?), (4) greater for high than fo r low p o s itiv e outcomes a t the le v e l o f j u s t if ic a t io n since the a c to r 's behavior i s e x p lic it ly ju s tifie d and the actor fo resees the consequences of h is behavior, and (5) maximal fo r both low

151 138 and high p o sitiv e outcomes a t the le v e l of in te n tio n a lity sin ce the causal role o f the actor i s maximal. Reviewing the r e s u lts from le v e l to le v e l, i t was found that these expectations were supported only at the le v e ls of a sso cia tio n and ju s t if ic a t io n. As expected, outcome in te n sity had no e ffe c t on ob servers' assignment o f r e sp o n sib ility at the le v e l of a sso c ia tio n w hile at the le v e l of ju s tific a tio n, observers' assignment of r e sp o n sib ility was greater for high p o sitiv e than fo r low p o s itiv e outcomes. Expectations were contradicted at the other three le v e ls of commission, fo r e s e e a b ility, and in te n tio n a lity. At the le v e l o f commission, the r e su lts were opposite the p red iction, i.e. observers' assignment o f r e sp o n sib ility was greater fo r low than for high p o sitiv e outcomes. Then a t the le v e l o f fo r e s e e a b ility, in te n sity of p o sitiv e outcomes had no e ffe c t on observers' assignment of r e s p o n s ib ility. F in a lly, a t the le v e l of intentio n a lity, outcome in te n s ity had an e ffe c t where no e ffe c t was expected, i.e. observers' r e s p o n s ib ility assignment was greater for high than for low p o sitiv e outcomes. The unexpected fin d in gs at the le v e ls of commssion, fo r e s e e a b ility, and in te n tio n a lity r e fle c t the inadequacy of present explanations. An a ltern a tiv e rela tio n sh ip between the nature of the a c to r 's ro le at the f iv e le v e ls and the observed e ffe c t o f in te n s ity of p o sitiv e outcomes might be proposed. This a ltern a tiv e i s based prim arily on the dynamics of fo r e s e e a b ility and in te n tio n a lity. The basic

152 139 prem ise i s th a t in form ation con cern in g f o r e s e e a b ilit y and in t e n t io n a lit y are in tim a te ly r e la te d. For p o s it iv e outcom es, when an a c to r can fo r e s e e th e consequences o f h is b eh avior, i t i s l i k e l y th a t th e a c to r w i l l be p erceiv ed a s in ten d in g to produce th e f in a l outcomes and a s such, w i l l be h eld more r e sp o n sib le a s p o s it iv e outcomes become more fa v o r a b le. U sing t h i s prem ise in e v a lu a tin g th e nature o f th e a c t o r 's r o le a t the f i v e l e v e l s o f c a u s a lity, th e observed e f f e c t o f in t e n s it y o f p o s it iv e outcomes a t each l e v e l m ight be accounted fo r as fo llo w s. At th e l e v e l o f a s s o c ia t io n, the a c t o r 's ca u sa l r o le i s minimal and as su ch, f o r e s e e a b ilit y and in t e n t io n a lit y on the p art o f the a c to r are n ot r e a lly an is s u e to be reso lv ed by th e o b serv er. T h erefo re, o b servers' assignm ent o f r e s p o n s ib ilit y i s n ot a ffe c te d by in t e n s it y o f p o s it iv e outcom es. At th e le v e l o f com m ission, th e p o s it iv e outcomes are not fo r e seen and not in ten d ed by th e a cto r and as such, ob servers h old the a c to r l e s s r e sp o n sib le a s th e se u n fo reseea b le and unintended p o s it iv e outcomes become more fa v o r a b le. At th e le v e l o f f o r e s e e a b ilit y, th e f o r e s e e a b ilit y o f th e f in a l outcome i s ambiguous and a s such th e in te n tio n s o f the a c to r are open to q u e stio n. As a r e s u lt, ob servers a ssig n ed r e s p o n s ib ilit y in a l l p o s s ib le ways a s a fu n c tio n o f in t e n s it y o f p o s it iv e outcom es. That i s, some ob servers

153 iko a ssig n e d more r e s p o n s ib ilit y as p o s it iv e outcomes became more fa v o r a b le, some a ssig n ed l e s s r e s p o n s ib ilit y as p o s it iv e outcomes became more fa v o r a b le, and some a ssig n e d an equal amount o f r e s p o n s ib ilit y as p o s it iv e outcom es became more fa v o r a b le. The r e s u lt was no sy stem a tic e f f e c t due to the in t e n s it y o f p o s it iv e outcom es. I t appeared then th a t the in form ation about t h e -a c to r s ca u sa l r o le was h ig h ly ambiguous and open to d iffe r e n t in te r p r e ta tio n s. T his su g g ested in t e r p r e t a b ilit y con cern in g th e f o r e s e e a b ilit y o f th e f in a l outcome and th e in te n tio n s o f th e a c to r was n ot ev id en t a t th e l e v e l s o f com m ission, j u s t i f i c a t i o n, or in t e n t io n a lit y where th e m a jo rity o f th e o b serv ers ex h ib ited th e same p a tte r n as r e f le c t e d by th e means. At th e l e v e l s o f j u s t i f i c a t i o n and in t e n t io n a lit y, th e a cto r can fo r e s e e th e consequences o f h is b eh avior and o b serv ers conclude th a t he in te n d s to produce th e p o s itiv e f i n a l outcom es. As such, o b servers tend to h old th e actor more r e sp o n sib le commensurate w ith outcome f a v o r a b ilit y. In summary then, i t appears th a t when an a c to r does not fo r e s e e or in ten d p o s it iv e outcomes a s a t th e l e v e l o f com m ission, o b serv ers hold th e a cto r l e s s r e s p o n s ib le as p o s it iv e outcomes become more fa v o r a b le. On th e other hand, when the outcome i s fo r e s e e n or in ten d ed as a t th e le v e ls o f j u s t i f i c a t io n and in t e n t io n a lit y, o b serv ers h o ld the a cto r more r e sp o n sib le as p o s it iv e outcomes become more fa v o ra b le. F in a lly, when th ere i s doubt as to th e f o r e s e e a b ilit y of the outcome and in te n tio n s o f th e a cto r a s a t th e l e v e l of

154 144 f o r e s e e a b ilit y, o b serv ers vary in t h e ir in te r p r e ta tio n o f c a u s a lity and. r e s p o n s ib ilit y. As such, no e f f e c t fo r in t e n s it y o f p o s it iv e outcomes i s in d ic a te d. With r e sp ect to th e e f f e c t o f s e v e r ity o f n eg a tiv e outcom es, th e gu id in g p r in c ip le may be th e " a v o id a b ility " o f th e f in a l outcomes ra th er than th e dynamics o f fo r e s e e a b ilit y and in t e n t io n a lit y p o stu la te d fo r p o s it iv e outcom es. The b a sic prem ise proposed here i s th a t i f n e g a tiv e outcomes are p erceiv ed as a v o id a b le, then ob servers w i l l hold th e a cto r more r e sp o n sib le a s th e se n eg a tiv e outcomes become more se v e r e. At th e le v e ls o f a s s o c ia tio n and in te n t io n a lit y, avoidance o f th e f i n a l outcome i s n ot r e a lly an is s u e. At th e le v e l o f a s s o c ia t io n, th e f in a l outcome i s com p letely beyond th e c o n tr o l o f th e a cto r s in c e h is a c tio n s are t o t a l l y u n rela ted to th e occurrence o f the n eg a tiv e outcom es. As such, ob servers were not exp ected and were not found to h old th e a c to r r e sp o n sib le fo r e ith e r m ild or severe negat i v e outcom es. In c o n tr a s t, a t th e le v e l o f in t e n t io n a lit y, th e occurrence o f th e n eg a tiv e outcomes i s co m p letely under th e c o n tr o l o f the a c to r and as such, o b serv ers were expected and were found to h old th e a cto r m axim ally resp o n sib le fo r b oth m ild and severe n eg a tiv e outcom es. At th e le v e l o f com m ission, th e a v o id a b ility o f the n e g a tiv e outcomes may be an is s u e. There was a d iscrep ancy between e x p e c ta tio n and the observed p a ttern a t the le v e l

155 14-2 o f com m ission. The r e s u lt s su g g e ste d, a t l e a s t, a s lig h t tendency on the p a r t o f o b servers to hold th e a c to r more r e sp o n sib le as n e g a tiv e outcom es became more s e v e r e. I t was exp ected th a t o b se r v ers would h old th e a c to r l e s s r e sp o n sib le a s n e g a tiv e outcomes became more sev ere s in c e th e s e outcomes appeared to be u n a v o id a b le. However, th e f a c t th a t o b se r v e r s' assignm ent o f resp o n s i b i l i t y d id tend to be ra th er m inim al r e g a r d le s s o f outcome s e v e r ity reduced th e s e r io u s n e s s o f t h is d iscrep a n cy. T his m inimal r e s p o n s ib ilit y assignm ent sim ply r e f le c t e d th e n o tio n th a t a c to r s a re not a ssig n e d much r e s p o n s ib ilit y fo r u n avoid ab le n e g a tiv e outcom es. N e v e r th e le s s, to fin d even a s l i g h t tendency f o r r e s p o n s ib ilit y assignm ent to in c r e a se w ith outcome s e v e r it y was s u r p r is in g. T his r e s u lt su g g e sts th a t a s lo n g as th e outcome i s in any way c o n tin g e n t upon the a c t o r ' s b eh a v io r, o b serv ers m ight p e r c e iv e th e outcome a s a v o id a b le and a s s ig n r e s p o n s ib ilit y commensurate w ith outcome s e v e r it y. The in c r e a se in r e s p o n s i b ilit y assignm ent a s n e g a tiv e outcomes became more severe was found a t th e l e v e l o f fo r e s e e a b i l i t y. Such a r e s u lt would be exp ected s in c e fo r a c tio n s it u a t io n s a t t h i s l e v e l, i t i s f a i r l y c le a r th a t th e outcom es cou ld have been avoid ed i f th e a c to r had g iv e n enough thought to h is a c tio n s. I n c o n tr a s t, a t th e l e v e l o f j u s t i f i c a t i o n, even though the n e g a tiv e outcom es are a r e s u lt o f th e a c t o r 's b eh a v io r, o b serv ers h eld th e a c to r l e s s r e s p o n s ib le a s n e g a tiv e outcomes

156 14-3 became more se v e r e. T his p a tte r n a t th e le v e l o f j r.s t if ic a - tio n p o s s ib ly r e f l e c t s th e o b se r v e r s p ercep tio n th a t the outcome i s more co n tin g en t upon th e e x te r n a l c o e r c iv e fo r c e s than upon th e a cto r and h is b eh avior. As n eg a tiv e outcomes become more se v e r e, th e e x te r n a l c o e r c iv e fo r c e s are p erceiv ed as having a more p o ten t in flu e n c e in producing the f i n a l outcome. When th e a c to r knowingly behaves in a manner which r e s u lt s in a severe n eg a tiv e outcome, o b servers p e r c e iv e th a t e x te r n a l fo r c e s are p ow erful and the a cto r had l i t t l e c h o ic e. Thus, ob servers p e r c e iv e l e s s a v o id a b ility fo r severe n eg a tiv e outcomes than m ild n eg a tiv e outcomes and as such hold th e a c to r l e s s r e sp o n sib le as th e se n eg a tiv e outcomes become more se v e r e. As d iscu sse d above, the e f f e c t o f outcome c h a r a c te r is tic s on o b serv ers' assignm ent o f r e s p o n s ib ilit y depended upon the l e v e l o f c a u s a lity. However, a s d isc u sse d below, when the data fo r a c to r s and o b servers were examined, th e o v e r a ll s im ila r it y among a c to r s, o b serv ers, and SA a c to r s was (juite unexpected. D iffe r e n c e s were exp ected based on th e theory o f d e fe n siv e a t tr ib u t io n, th e n o tio n o f s e l f - i n t e r e s t, and th e th eory o f o b je c tiv e se lf-a w a r e n e ss. Based on d e fe n siv e a ttr ib u tio n th eory and the n o tio n th a t a c to r s are l i k e l y to a ttr ib u te r e s p o n s ib ilit y in terms o f t h e ir own s e l f - i n t e r e s t s, d iffe r e n c e s were expected betw een a c to r s and o b serv ers. A ctors were expected to hold th em selves more r e sp o n sib le fo r p o s it iv e outcom es, but l e s s

157 iwt resp o n sib le fo r n eg a tiv e outcomes than ob servers were expected to hold a c to r s. Thus, as s tip u la te d in H ypothesis 8, r e s p o n s ib ilit y assignm ent fo r p o s it iv e outcomes was expected to be higher fo r a c to r s than fo r o b serv ers a cross a l l f iv e l e v e l s o f c a u s a lit y. In c o n tr a s t, H ypothesis 9 sta te d th a t r e s p o n s ib ilit y assignm ent fo r n eg a tiv e outcomes was expected to be lower fo r a cto rs than fo r o b servers a cro ss a l l f iv e l e v e l s o f c a u s a lit y. N eith er o f th e se h ypotheses were supported. A ctors did not a ssig n more r e s p o n s ib ilit y than o b serv ers fo r p o s it iv e outcom es nor l e s s r e s p o n s ib ilit y fo r n e g a tiv e outcom es a cro ss a l l f iv e l e v e l s o f c a u s a lity. Based on th e th eory o f o b je c tiv e se lf-a w a r e n e ss, d iffe r e n c e s a c r o ss l e v e l s were expected between th e two groups o f a c to r s as w e ll. The su p p o sitio n was th a t the SA m anip u lation would enhance th e a c to r s ' awareness o f thems e lv e s in a c tio n s it u a tio n s which in tu rn would in c r ea se t h e ir acceptance o f r e s p o n s ib ilit y. Thus, a c to r s w ith the SA m anipulation should tak e more r e s p o n s ib ilit y fo r both p o s it iv e and n eg a tiv e outcom es than a c to r s w ithou t th e SA m anip u lation. T his e f f e c t was expected a cro ss a l l f iv e l e v e l s o f c a u s a lit y. Thus, as s tip u la te d in H ypothesis 10, r e s p o n s ib ilit y assignm ent fo r p o s it iv e outcomes was expected to be higher fo r SA a c to r s than fo r a c to r s a cro ss th e f iv e l e v e l s o f c a u s a lit y. S im ila r ly, H yp oth esis 11 sta te d th a t r e s p o n s ib ilit y assignm ent fo r n eg a tiv e outcomes was expected to be higher fo r SA a c to r s than fo r a c to r s a cro ss th e f iv e

158 1^5 l e v e l s o f c a u s a lit y. A gain, th e se h ypotheses were not confirm ed by th e d ata. A ctors and SA a c to r s were g e n e r a lly s im ila r in t h e ir assignm ent o f r e s p o n s ib ilit y fo r b oth p o s it iv e and n e g a tiv e outcomes a c r o ss th e f iv e le v e ls o f c a u s a lit y. The f in a l two h ypoth eses concerned th e p a tte r n o f resp o n s i b i l i t y assignm ent by a c to r s w ith in le v e l s o f c a u s a lit y. Based on d e fe n siv e a t tr ib u t io n th eo ry and th e n o tio n th a t a c to r s a s s ig n r e s p o n s ib ilit y to th em selv es in accordance w ith t h e ir own s e l f - i n t e r e s t s, th e a c to r s ' p a tte r n o f r e s p o n s ib ilit y assignm ent w ith in any p a r tic u la r l e v e l o f c a u s a lity was exp ected to be a ffe c te d by outcome c h a r a c t e r is t ic s. A ctors were expected to hold th em selv es more r e sp o n sib le fo r p o s i t iv e outcomes than fo r n e g a tiv e outcomes (H ypoth esis 1 2 ). Moreover, a c to r s were exp ected to hold th em selv es more resp o n sib le a s p o s it iv e outcomes became more fa v o ra b le, but l e s s r e sp o n sib le a s n e g a tiv e outcomes became more severe (H ypoth esis 13) Only a t th e le v e l o f j u s t i f i c a t i o n was t h is p a tte r n found fo r a c to r s. Contrary to a l l e x p e c ta tio n s o f H ypotheses 8-13 was th e fin d in g o f an o v e r a ll s im ila r it y between o b servers and both groups o f a c to r s in t h e ir r e s p o n s ib ilit y assignm ent as a fu n c tio n o f outcome c h a r a c t e r is t ic s a t each l e v e l o f c a u s a lit y. As in d ic a te d by th e n o n -s ig n ific a n t r o le x v a le n c e, r o le x l e v e l x v a le n c e, and r o le x l e v e l x v a len ce x in t e n s it y in te r a c tio n s, not o n ly did o b servers and both groups o f a c to r s a s s ig n roughly th e same amount o f r e s p o n s ib ilit y fo r

159 1>6 p o s i t i v e and n e g a tiv e outcomes a c r o s s l e v e l s i n gen eral but a ls o th e y dem onstrated th e same p a tte r n o f r e s p o n s i b i l i t y assignm en t w ith in each le v e l a s a fu n c tio n o f outcome c h a r a c t e r i s t ic s. Based on th e o v e r a ll s im ila r it y among s u b je c t groups, l i t t l e su p p ortive ev id e n c e was found fo r d e fe n s iv e a t t r ib u t io n, s e l f - i n t e r e s t, or o b je c tiv e s e lf-a w a r e n e s s. However, a t th e p o la r l e v e l s o f c a u s a lit y, namely a s s o c ia t io n and in t e n t io n a lit y, th e la c k o f d iffe r e n c e s b etw een su b ject groups was a c t u a lly n ot s u r p r is in g. The c a u s a l r o le sh ared by th e a cto r and th e a c tiv e environm ent i s so c le a r ly d if f e r e n t ia t e d i n a c t io n s it u a t io n s at th e se l e v e l s t h a t c a u s a l a t tr ib u t io n s and r e s p o n s ib ilit y assignm en t would be m in im ally s u b je c t to d i f f e r e n t i a l in te r p r e ta tio n on th e p a r t o f a cto r s and o b se r v ers. In c o n tr a st to th ese p o la r l e v e l s, th e a c to r and th e a c t iv e environm ent more c l e a r l y share a c a u s a l r o le in p rod u cin g the f i n a l outcomes a t th e in te rm ed ia te l e v e l s. As su ch, ca u sa l a t t r ib u t io n s and r e s p o n s ib ilit y assignm ent would be su b ject to d i f f e r e n t i a l in te r p r e ta t io n based on p e r c e p tu a l and m o tiv a tio n a l d iffe r e n c e s b etw een a c to r s and o b s e r v e r s. These d i f f e r e n t i a l I n te r p r e ta tio n s le a d in g to d if f e r in g p a tte r n s o f r e s p o n s ib ilit y assignm en t depending upon su b ject r o le were ex p ected p a r t ic u la r ly a t the l e v e l o f f o r e s e e a b ilit y where th e a c t io n s it u a t io n s are h ig h ly am biguous.

160 1^7 E xpected d iffe r e n c e s were n ot found however to any s u b s ta n tia l degree a t the th r e e in te rm e d ia te l e v e l s. Thus, th e r e was minor support fo r th e m o tiv a tio n a l d iffe r e n c e s p r e d ic te d by d e fe n s iv e a t t r ib u t io n th eo ry and the c o r r e sponding n o tio n o f s e l f - i n t e r e s t. In f a c t, a t th e l e v e l o f com m ission, th e data s u g g e s ts th a t a c to r s op erate somewhat a g a in st t h e ir s e l f - i n t e r e s t. R e la tiv e to what o b se r v e r s a ssig n e d, a c to r s to o k l e s s r e s p o n s ib ilit y fo r p o s i t i v e outcomes and more r e s p o n s i b ilit y fo r n e g a tiv e outcom es, e s p e c ia lly sev ere n e g a tiv e outcom es. Then a t th e l e v e l s o f f o r e s e e a b ilit y and j u s t i f i c a t i o n, th e o n ly in d ic a t io n fo r th e o p era tio n o f a s e l f - i n t e r e s t m otive was t h a t a c to r s to o k s l i g h t l y more r e s p o n s ib ilit y fo r p o s it iv e outcomes than o b se r v e r s a ssig n e d to a c to r s. Moreover, th e r e s u lt s p ro v id e l i t t l e in d ic a tio n th a t s e lf-a w a r e n e ss was o p e r a tiv e. Only a t th e le v e l o f j u s t i f i c a t i o n was th e re an in d ic a t io n th a t th e SA m anipulat io n m ight be o p e r a tiv e to a s u b s ta n tia l d eg ree. At t h i s l e v e l th e d iffe r e n c e betw een a c to r s and SA a c to r s conformed to e x p e c ta tio n. SA a c to r s to o k more r e s p o n s i b ilit y fo r b oth p o s it iv e and n e g a tiv e outcom es th a n a c to r s. However, a t th e l e v e l o f f o r e s e e a b i l i t y, th ere was no d iffe r e n c e betw een th e two groups o f a c t o r s. F in a lly, a t th e l e v e l o f com m ission, even though SA a c to r s took more r e s p o n s ib ilit y fo r p o s it iv e outcom es than a c t o r s, SA a c to r s took l e s s r e s p o n s ib ilit y f o r n e g a tiv e outcom es th an a c to r s which was c o n tr a r y to th e th eory o f o b je c tiv e s e lf-a w a r e n e s s.

161 148 In an attem pt to ex p la in th e gen eral s im ila r it y among th e th r ee su b ject groups as r e v ea le d by th e la c k o f s ig n if ic a n t e f f e c t s due to r o le in the a n a ly s is o f v a ria n c e, th e fundamental con cep tu al b a s is fo r ex p ectin g d iffe r e n c e s in th e f i r s t p la ce needs to be review ed. As s ta te d p r e v io u sly, th e a ttr ib u tio n p ro cess was co n cep tu a lized as an o b je c tiv e assessm en t o f th e c a u s e -a n d -e ffe c t r e la tio n s h ip s among ca u sa l fo r c e s (a c to r and a c tiv e environm ent) which serve to ex p la in th e occurrence o f a p a r tic u la r outcome. These c a u s e -a n d -e ffe c t r e la tio n s h ip s between a cto r and a c tiv e environm ent are o b je c t iv e ly d ep icted in a c tio n s it u a t io n s, but th e se r e la tio n s h ip s are d if f e r e n t depending upon th e le v e l o f c a u s a lit y. I t was p r e c is e ly the nature o f th e se in te r -c a u s a l r e la tio n s h ip s which were exp ected to determ ine th e p a tte r n o f r e s p o n s ib ilit y assignm ent a s a fu n ctio n o f outcome c h a r a c t e r is t ic s. However, th e in fe r e n c e s or d e c is io n s about c a u s a lity and r e s p o n s ib ilit y which are based on an o b je c tiv e ev a lu a t io n o f ca u sa l in t e r - r e la t io n s h ip s d ep icted in a c tio n s it u a tio n s were p o stu la te d to be m odified by p ercep tu a l and/or m o tiv a tio n a l b ia s e s. These b ia s e s were p o stu la te d to be a fu n c tio n o f th e a t t r ib u t o r 's r o le a s e ith e r an a cto r or as an o b serv er. In turn, th e o p era tio n o f th e se b ia s e s were expected to be r e f le c t e d in d if f e r in g p a tte r n s o f r e s p o n s ib ilit y assignm ent by a c to r s and o b se r v ers. The fa c t th a t a l l th ree su b je c t groups showed the same p a tte r n o f r e s p o n s ib ilit y assignm ent a cro ss and w ith in

162 l e v e l s su g g e sts th en th a t th e se b ia s e s were n ot o p e r a tiv e, th a t i s, a l l s u b je c ts were a tte n d in g to and p r o c e ssin g th e same in form ation and drawing th e same c o n clu sio n s regard in g c a u s a lity and r e s p o n s ib ilit y ir r e s p e c t iv e o f t h e ir r o le as an a c to r or as an o b serv er. T h is seems to im ply th a t a l l s u b je c ts, not ju s t o b servers, were in te r p r e tin g and e s ta b lis h in g in te r -c a u s a l r e la tio n s h ip s from an o b je c tiv e, detached p o in t o f view and th a t th e p ersonal su b je c tiv e involvem ent expected o f a c to r s and ob servers was not o ccu rrin g. Thus, a l l su b ject groups e x h ib ite d sim ila r p a tte r n s o f r e s p o n s ib ilit y assign m en t. T his ex p la n a tio n does n ot n e c e s s a r ily su g g est th a t a c to r s and o b servers do not have b ia s e s depending upon outcome c h a r a c t e r is t ic s and m an ip u lation o f se lf-a w a r e n e ss. Other re sea rch ers have alread y dem onstrated d iffe r e n c e s as a fu n c tio n o f su b je c t r o le and o b je c tiv e se lf-a w a r e n e ss (Duval & Wicklund, 1973; H arris and Harvey, 1975; Harvey, e t. al., 1975; Johnson, e t. al., 19&9; R oss, al., 197^)- The most p oten t ex p la n a tio n fo r th e r e s u lt s o f t h is study may be found in i t s m ethodology. S p e c if ic a lly, i t su g g e sts th a t th e m ethodology employed in t h i s study was not e f f e c t iv e or conducive to th e op era tio n o f th e su b je c tiv e b ia s e s p o stu la te d by d e fe n s iv e a ttr ib u tio n and p ercep tu a l b ia s e s p o stu la te d by th e th eo ry o f o b je c tiv e se lf-a w a r e n e ss. As a r e s u lt th e se b ia s e s were not o p era tin g to su p ersed e, as ex p ected, an o b je c tiv e assessm en t o f c a u s a lity and

163 150 assignm ent o f r e s p o n s ib ilit y. The q u e stio n then becomes why did th e m ethodology not e l i c i t th ese b ia s e s. Two answers come im m ediately to mind regard in g th e c o n d itio n s which m ight be n ecessa ry to e l i c i t p ercep tu a l a n d /o r m o tiv a tio n a l b ia s e s. F ir s t, i t i s p o s s ib le th a t th e in s tr u c tio n s g iv e n in th e p resen t stu d y were too p o ten t in d e fin in g th e ta sk of th e su b je c ts a s an "honest" e v a lu a tio n o f r e s p o n s ib ilit y. That i s, th e in s tr u c tio n s may have p reclu d ed th e o p era tio n o f su b je c tiv e b ia s e s by s t r e s s in g th e ta sk o f the su b je c ts a s "honestly in d ic a tin g th e ap p rop riate degree o f r e s p o n s i b i l i t y. The in s tr u c tio n s th en may have demanded an o b je c tiv e e v a lu a tio n o f th e in form ation and were not nebulous enough to a llo w th e in tr u s io n o f p erso n a l b ia s e s. T h erefore, b oth a c to r s and o b serv ers were o b je c t iv e ly e v a lu a tin g th e s it u a t io n s and a s s ig n in g r e s p o n s ib ilit y as d ic ta te d by th e in s tr u c tio n s ra th er than as d ic ta te d by th e p erso n a l b ia s e s and p e r s p e c tiv e s o f t h e ir r o le. Secondly, r e la t iv e to th e m ethodology employed in th e p resen t stu d y, l i v e b eh a v io ra l e v e n ts may be a b asic requirem ent fo r the e l i c i t a t i o n, o p era tio n, and demonstrat io n of p ercep tu a l an d /or m o tiv a tio n a l b ia s e s. I t i s l i k e l y th a t in l i v e b eh a v io ra l e v e n ts, a c to r s and o b serv ers w i l l d if f e r In t h e ir e v a lu a tio n o f th e ca u sa l lin k between observed b eh avior and observed outcom e. E x p erim en ta lly, p a st resea rch ers have examined and found a cto r-o b serv e r

164 151 d iffe r e n c e s in l i v e b eh a v io ra l ev e n ts (H arris and Harvey, 1975; Harvey, e t. al., Johnson, e t. al., 1969; R oss, e t. al., 197*0 T h e o r e tic a lly, th ese d iffe r e n c e s are lik e l y sin c e the a c to r a c tu a lly behaves and s e e s th e consequences o f h is a c tio n w hile th e ob server a c tu a lly w itn e sse s th e a c t o r 's behavior and the r e s u lta n t outcom e. Whether or not th e a c to r -o b se r v e r d iffe r e n c e s p o stu la te d by th e d iscrep ancy h y p o th e sis, the th eory o f o b je c tiv e s e l f - aw areness, or d e fe n siv e a ttr ib u tio n w i l l depend upon th e l e v e l o f c a u s a lity i s s t i l l open to q u e stio n. Hence, fu tu re resea rch should extend th e in co rp o r a tio n o f the f iv e l e v e l s o f c a u s a lit y in to l i v e b eh a v io ra l e v e n ts and examine th e e f f e c t o f outcome c h a r a c t e r is t ic s in lig h t o f the fin d in g s o f th e p resen t study a s w e ll as th e fin d in g s o f Shaw and h is a s s o c ia t e s (Shaw and R eita n, 1969; Shaw and S u lz e r, 196*)-; S u lzer, 196*1-; 1971).

165 152 REFERENCES D uval, S., & W icklund, R.A. E f f e c t s o f o b je c tiv e se lf-a w a r e n e ss on a t tr ib u t io n o f c a u s a lit y. Journal o f E xperim ental S o cia l P sy ch o lo g y, 1973, 17-31* F ish b ein, M., & A jzen, I. A ttr ib u tio n o f r e s p o n s ib ilit y : A t h e o r e t ic a l n o te. Journal o f E xperim ental S o c ia l P sy ch o lo g y. 1973, 2, H a rris, B., & Harvey, J.H. S e lf -a t t r ib u t e d ch o ice as a fu n c tio n o f the consequence o f a d e c is io n. Journal o f P e r s o n a lity and S o c ia l P sy ch o lo g y, 1975, Harvey, J.H., H a rris, B., & Barnes, R.D. A cto r-o b serv er d i f f e r ences in th e p e r c e p tio n s o f r e s p o n s ib ilit y and freedom. Journal o f P e r s o n a lity and S o c ia l P sy ch o lo g y, 1975, 3 2, H eider, F. The P sych ology o f In terp erso n a l R e la tio n s. New York: W iley, 1958* Johnson, T.J., Feigenbaum, R., & Weiby, M. Some determ in ants and consequences o f the te a c h e r 's p e r c e p tio n o f c a u sa tio n. Journal o f E d u cational P sy ch o lo g y, 1964, jjji, J o n es, E.E., & N is b e tt, R.E. The Actor and th e Observer: D ivergent P ercep tio n s o f th e C auses o f B ehavior. New York: General Learning P r e s s, 1971* J o n es, E.E., Rock, L., Shaver, K.G., G oeth als, G.R., & Ward, L.M. P a ttern o f perform ance and a b i l it y a ttr ib u tio n : An unexpected primacy effect. 1968, 10, Journal o f P e r s o n a lity and S o c ia l P sych ology,

166 153 McArthur, L.Z. The how and. what o f why: Some d eterm in ants and con sequen ces o f c a u sa l a t t r ib u t io n s. U npublished Ph.D. d is s e r t a t io n, Y ale U n iv e r s ity, 1970* McArthur, L.Z. The how and what o f why: Some d eterm in ants and consequences o f c a u sa l a t t r ib u t io n s. Jou rn al o f P e r s o n a lity and S o c ia l P sy c h o lo g y. 1972, 22, * N is b e tt, R.E., C aputo, G.C., L egant, P., & M arecek, J. Behavior a s seen by th e a c to r and a s seen by th e o b server. Journal o f P e r s o n a lity and S o c ia l P sy ch o lo g y, 1973> 27, 15^-16^. R oss, L., B ierb rau er, G., & P o lly, S. A ttr ib u tio n o f ed u ca tio n a l outcom es by p r o f e s s io n a l and n o n p r o fe ssio n a l in s t r u c t o r s. Jou rn al o f P e r s o n a lity and S o c ia l P sy ch o lo g y. 197^, 2 9, Shaver, K.G. D e fe n siv e a ttr ib u tio n : E f f e c t s o f s e v e r it y and relev a n ce on th e r e s p o n s ib ilit y a ssig n e d fo r an a c c id e n t. Jou rn al o f P e r s o n a lity and S o c ia l P sy c h o lo g y. 1970, 1^, Shaw, J.I., & S k o ln ic k, P. A ttr ib u tio n o f r e s p o n s i b ilit y fo r a happy a c c id e n t. Jou rn al o f P e r s o n a lity and S o c ia l P sy ch o lo g y, 1971, 18, Shaw, M.E., & R e ita n, H.T. A ttr ib u tio n o f r e s p o n s ib ilit y as a b a s is fo r s a n c tio n in g b eh a v io r. B r it is h Journal o f S o c ia l and C lin ic a l P sy c h o lo g y, 1969, 8, Shaw, M.E., & S u lz e r, J.L. An em p irical t e s t o f H e id e r s l e v e l s in a t t r ib u t io n o f r e s p o n s ib ilit y. J o u rn a l o f Abnormal and S o c ia l P sy c h o lo g y, 1964-, 6 f 39-^6.

167 15^ Storm s, M.D. V ideotape and th e a ttr ib u tio n p ro cess: R eversing a c t o r s and o b serv ers' p o in ts o f view. Journal o f P e r s o n a lity and S o c ia l P sy ch o lo g y, 1973* 2» * S u lz er, J.L. A ttr ib u tio n o f r e s p o n s ib ilit y a s a fu n c tio n o f th e str u c tu r e, q u a lity, and in t e n s it y o f th e ev e n t. U npublished d o cto ra l d is s e r t a tio n, U n iv e r s ity o f F lo r id a, S u lzer, J.L. H e id e r's " le v e ls model" o f r e s p o n s ib ilit y a t t r ib u tio n. Paper p resen ted a t th e Symposium on A ttr ib u tio n o f R e s p o n s ib ility R esearch, C o lleg e o f W illiam and Mary, W illiam sburg, Va., Ju ly * W alster, E. Assignm ent o f r e s p o n s ib ilit y fo r an a c c id e n t. Journal o f P e r so n a lity and S o c ia l P sy ch o lo g y, 1966, 73-79* W alster, E. "Second guessing" im portant e v e n ts. Human R e la tio n s, 1967, 20, W icklund, R.A., & D uval, S. O pinion change and perform ance f a c i l i t a t i o n a s a r e s u lt o f o b je c tiv e se lf-a w a r e n e ss. Journal o f Experim ental S o c ia l P sy ch o lo g y, 1971» Z> 319-3^2. Wortman, C.B. Some determ inants o f p erceiv e d c o n tr o l. Journal o f P e r so n a lity and S o c ia l P sy ch o lo g y, 1975» 21*' ^*

168 A-l 155 APPENDIX A P o o l o f 20 a c tio n s it u a t io n s w ith in d e n t if ic a t io n o f l e v e l o f c a u s a lit y, outcome v a le n c e, and outcome in t e n s it y g iv en in key fo llo w in g th e a c tio n s it u a tio n s 1. Adam was a t th e "beach w ith h is g i r l. A woman and her h usband a l l o f a sudden came running over to Adam. Both o f them were h y s t e r ic a l. They were y e l lin g a t Adam t e l l i n g him th a t t h e ir l i t t l e boy had g o tte n c a r r ie d out by th e cu rren t and was drowning. The b o y 's fa th e r sa id he c o u ld n 't swim and would k i l l Adam i f he d id n t go out and save the boy. Adam sp o tted a sm all boat on th e beach. Adam ran over to th e b o a t, paddled o u t, and brought the boy to s a f e t y. To what degree i s Adam r e sp o n sib le fo r th e b o y 's l i f e b ein g saved? 2. One morning, w h ile Sam was a t sch o o l stu d y in g fo r an exam, one o f Sam' s fr ie n d s asked Sam' s b rother to d riv e him to th e auto r e p a ir shop to p ick up h is car th a t aftern o o n. Sam s brother sa id OK, but th a t a ftern oon Sam's brother fo r g o t and Sam's fr ie n d had to pay fo r a t a x i to go g e t h is c a r. To what degree i s Sam r e sp o n sib le fo r h is fr ie n d having to pay fo r a ta x i to go g e t h is car? 3* Roger was c le a n in g out h is garage and found some old sh o es. Roger put th e o ld sh o es in th e tr a sh pile. A hobo who happened to p a ss by la t e r in th e day found th e old sh oes and kept them fo r h im s e lf. To what degree i s Roger r e sp o n sib le fo r th e hobo g e ttin g some old shoes? k. Steve and h is g i r l were s i t t i n g next to another man a t a bar one n ig h t. The man got up to le a v e and l e f t a quarter on the bar as a t ip fo r th e b arten d er. Steve reached to p ick up the q u arter, but h is g i r l to ld him not to take th e t i p. S teve took th e b a rten d e r's t ip anyway and used i t to p la y th e p in b a ll m achine. To what degree i s Steve resp o n sib le fo r th e bartender not. g e t t in g h is tip? 5. K eith and h is fr ie n d went to th e Chem istry la b one ev en in g. While h ea tin g two ch em ica ls K eith had mixed to g e th e r, th e beaker broke and a f i r e broke o u t. I n s t in c t iv e ly, K eith sta r te d to fig h t th e f i r e. F earing an e x p lo sio n and w anting to save th e b u ild in g, K eith p u lle d th e f i r e alarm and c lo s e d an e x p lo sio n p roof door. The f i r e alarm went o f f ju s t in tim e fo r K e ith 's fr ie n d to escape through a back door j u s t a s th e la b exploded. To what degree i s K eith r e sp o n sib le fo r h is frien d esca p in g the ex p lo sio n? 6. One day, Ed c a lle d a fr ie n d to come over and see h is new stereo equipm ent. On th e way, Ed's fr ie n d had an autom obile a ccid en t and was k i l l e d. To what degree i s Ed r e sp o n sib le fo r h is fr ie n d b ein g k ille d?

169 A One d ay, w h ile P h il was p la y in g g o lf, P h i l s "brother c a l le d up a f r ie n d to go p la y some t e n n is. They p la y ed 3 s e t s o f te n n is and P h il's "brother won 2 out o f th e 3 s e t s. To what degree i s P h il r e s p o n s ib le fo r h is b ro th er w inning 2 o u t o f th e 3 s e t s o f te n n is? 8. One d ay, J a ck s g ir lf r ie n d c a l le d him from sc h o o l and asked Jack t o b rin g a c a r to n o f m ilk over to h er apartm ent. When Jack g o t to h is g i r l ' s apartm ent, he gave h i s g i r l ' s c a t a l i t t l e m ilk and p u t th e c a r to n o f m ilk he had opened on th e k itc h e n co u n ter. The c a t f in is h e d th e m ilk q u ick ly and s ta r te d to purr f o r some more. Jack l e f t and w h ile no one was t h e r e, the c a t knocked o v er th e open c a r to n o f m ilk w hich h is g i r l had to c le a n up when she got home. To what d egree i s Jack r e sp o n sib le fo r h is g i r l h aving to c le a n up a f t e r th e ca t when she got home? 9. One m orning, B ert was sta n d in g around w ith some p eo p le w atch in g an apartm ent b u ild in g th a t was b urnin g down. W hile w a tch in g, a man appeared a t a th ir d sto r y window and s ta r te d to y e l l f o r h elp. Everyone th e r e thought th a t th ere was so much f i r e th a t no one should go in to th e b u ild in g and b e s id e s th e f i r e departm ent was on th e way. When B ert s ta r te d to run in to th e b u ild in g, a cou p le o f men t r ie d t o hold him back, but Bert broke away, ra n in, and p u lle d th e man to s a f e t y. To what d eg ree i s B ert r e s p o n s ib le fo r th e m an's l i f e b e in g saved? 10. George was at a dance h a ll one n ig h t s h o o tin g p ool w ith h is fr ie n d i n a back room. Three men came in t o th e back room and fo r no apparent r e a so n grabbed G eorge's fr ie n d and s ta r te d to h it him. A b ig f i g h t s ta r te d, but George and h is fr ie n d were u nab le to esca p e. During th e f i g h t, one o f th e men p u lle d a k n ife and came a t George to k i l l him. George grabbed th e k n ife and stabbed th e man to d ea th. To what degree i s George r e sp o n sib le f o r th e m an's death? 11. One n ig h t, w hile D avid was e a t in g a ste a k f o r supper, D a v id 's hungry dog was b e g g in g fo r some o f th e m eat. David had not fed t h e dog y e t. When David fin is h e d e a t in g, he l e f t i n a hurry w ith ou t f e e d in g h is dog and went to a nearby s to r e to buy a pack o f c i g a r e t t e s. W hile David was gone, h is hungry dog a t e some o f th e t a s t y meat le f t o v e r s. To what d egree i s David r e sp o n sib le f o r h is dog g e t t in g some ta s ty meat l e f t o v ers to eat? 12. One day Ken was t a lk in g to a fr ie n d on campus and asked him to come over l a t e r f o r a b e e r. On th e way to K en's a p a r t ment, h i s f r ie n d 's c a r s t a l l e d and h is fr ie n d had to walk to a nearby gas s t a t i o n to g e t some h e lp. To what d egree i s Ken r e s p o n s ib le f o r h is f r ie n d 's ca r s t a l l i n g and h a v in g to walk to a gas s t a t i o n fo r some help?

170 A One n ig h t a f t e r a f o o t b a ll game, w h ile Fred was b rin g in g h is d ate home, F red 's b roth er g o t in to a f ig h t and pushed a man in to th e s t r e e t. The man was run over by a sp eedin g autom obile and was k i l l e d. To what degree i s Fred resp onsi b le fo r th e man b ein g k ille d? 1*4-. Wes was b ein g i n it ia t e d in to a f r a t e r n it y. The fr a t e r n it y members t o ld Wes he had to spray some black p a in t on th e door o f a r iv a l f r a t e r n it y house. When Wes s a ild no, th e f r a t e r n it y members gathered around Wes and t o ld him th ey would shave h is head and spray b la ck p a in t a l l over h is c a r. Wes sprayed b lack p a in t on th e door which had to be r e p a in ted. To what degree i s Wes r e sp o n sib le fo r the door n eed in g to be rep ain ted? 15 J e r r y s la n d lo rd le n t J erry a fan and to ld J erry not to u se th e a ir -c o n d itio n e r because i t needed to be se r v ic e d and i f i t broke down i t would be ex p en siv e to fix. That even in g when i t was not very hot o u ts id e, a fr ie n d came o v e r. In stea d o f tu r n in g on th e fa n, Jerry turned on th e a ir -c o n d itio n e r which made h is fr ie n d com fortable during th e v i s i t. To what degree i s Jerry r e sp o n sib le fo r h is fr ie n d b ein g com fortable during th e v i s i t? 16. One weekend when S c o tt was a t home v i s i t i n g h is p a ren ts, S c o t t 's b roth er was on campus one n ig h t and saw two men attem p tin g to rape a yound c o -e d. S c o t t 's b roth er rushed o v er, chased th e men away, and saved the co -ed from b ein g raped. To what degree i s S c o tt r e sp o n sib le fo r th e young co -ed not b ein g raped? 17. Chuck had an o ld ca r which had very poor b rak es. He was d r iv in g h is g ir lf r ie n d to th e shopping c e n te r one morning when th e t r a f f i c was p r e tty heavy. When he came to a busy in t e r s e c t io n he cou ld not stop th e ca r and ran in to th e p ath o f a tru ck. Chuck's g i r l was k ille d in th e c o l l i s i o n. To what degree i s Chuck r e sp o n sib le fo r h is g i r l b ein g k ille d in th e c o l lis i o n? 18. One weekend, Jim wanted to borrow h is g i r l ' s ca r so th a t he and some o f h is fr ie n d s cou ld d riv e around fo r a b ig n ig h t on th e town. Jim's g ir lf r ie n d t o ld Jim he co u ld u se her ca r i f he changed th e o i l because she c o u ld n 't r e a lly a ffo r d to pay someone to do i t fo r h er. Jim changed th e o i l which saved h is g i r l a l i t t l e money. To what degree i s Jim r e sp o n sib le fo r h is g i r l sa v in g a l i t t l e money? 19* Bob was working a t a la r g e apartm ent complex a s a p a rt tim e c u sto d ia n. Bob se n t a fe llo w worker to f i n i s h c le a n in g th e swimming p o o l. On th e way, th e worker sp o tted a l i t t l e boy f a l l in to th e p o o l and h i t h is head on th e s id e o f th e p o o l. The worker rushed o v er and saved th e l i t t l e boy from drownin g. To what degree i s Bob r e sp o n sib le fo r th e l i t t l e boy b e in g saved from drowning?