CORRELATION. two variables may be related. SAT scores, GPA hours in therapy, self-esteem grade on homeworks, grade on exams

Similar documents
CORRELATION. two variables may be related. SAT scores, GPA hours in therapy, self-esteem grade on homeworks, grade on exams

HYPOTHESIS TESTING. four steps. 1. State the hypothesis and the criterion. 2. Compute the test statistic. 3. Compute the p-value. 4.

HYPOTHESIS TESTING. four steps. 1. State the hypothesis. 2. Set the criterion for rejecting. 3. Compute the test statistics. 4. Interpret the results.

Chapter 12: Velocity, acceleration, and forces

Equations of motion for constant acceleration

Phys 221 Fall Chapter 2. Motion in One Dimension. 2014, 2005 A. Dzyubenko Brooks/Cole

One-Dimensional Kinematics

Kinematics: Motion in One Dimension

Comparing Means: t-tests for One Sample & Two Related Samples

Unit 1 Test Review Physics Basics, Movement, and Vectors Chapters 1-3

Diebold, Chapter 7. Francis X. Diebold, Elements of Forecasting, 4th Edition (Mason, Ohio: Cengage Learning, 2006). Chapter 7. Characterizing Cycles

first-order circuit Complete response can be regarded as the superposition of zero-input response and zero-state response.

Physics Notes - Ch. 2 Motion in One Dimension

HYPOTHESIS TESTING. four steps

Analyze patterns and relationships. 3. Generate two numerical patterns using AC

1. (16 points) Answer the following derivative-related questions. dx tan sec x. dx tan u = du d. dx du tan u. du tan u d v.

INSTANTANEOUS VELOCITY

Topic 1: Linear motion and forces

Econ107 Applied Econometrics Topic 7: Multicollinearity (Studenmund, Chapter 8)

CSE-4303/CSE-5365 Computer Graphics Fall 1996 Take home Test

Q2.4 Average velocity equals instantaneous velocity when the speed is constant and motion is in a straight line.

LAB # 2 - Equilibrium (static)

Physics 101: Lecture 03 Kinematics Today s lecture will cover Textbook Sections (and some Ch. 4)

NEWTON S SECOND LAW OF MOTION

Kinematics in two dimensions

CSE 5365 Computer Graphics. Take Home Test #1

Radical Expressions. Terminology: A radical will have the following; a radical sign, a radicand, and an index.

Biol. 356 Lab 8. Mortality, Recruitment, and Migration Rates

Random Processes 1/24

DESIGN OF TENSION MEMBERS

Institutional Assessment Report Texas Southern University College of Pharmacy and Health Sciences "P1-Aggregate Analyses of 6 cohorts ( )

1. Diagnostic (Misspeci cation) Tests: Testing the Assumptions

Math 2214 Solution Test 1 B Spring 2016

Solutions to Odd Number Exercises in Chapter 6

THE UNIVERSITY OF TEXAS AT AUSTIN McCombs School of Business

Physics for Scientists and Engineers. Chapter 2 Kinematics in One Dimension

KINEMATICS IN ONE DIMENSION

Bias in Conditional and Unconditional Fixed Effects Logit Estimation: a Correction * Tom Coupé

Math 116 Practice for Exam 2

Atmospheric Dynamics 11:670:324. Class Time: Tuesdays and Fridays 9:15-10:35

ACE 562 Fall Lecture 5: The Simple Linear Regression Model: Sampling Properties of the Least Squares Estimators. by Professor Scott H.

Vehicle Arrival Models : Headway

Linear Time-invariant systems, Convolution, and Cross-correlation

Advanced Organic Chemistry

Kriging Models Predicting Atrazine Concentrations in Surface Water Draining Agricultural Watersheds

Help students learn physics by doing physics

Amit Mehra. Indian School of Business, Hyderabad, INDIA Vijay Mookerjee

s in boxe wers ans Put

Srednicki Chapter 20

1. The graph below shows the variation with time t of the acceleration a of an object from t = 0 to t = T. a

(a) Set up the least squares estimation procedure for this problem, which will consist in minimizing the sum of squared residuals. 2 t.

3.1.3 INTRODUCTION TO DYNAMIC OPTIMIZATION: DISCRETE TIME PROBLEMS. A. The Hamiltonian and First-Order Conditions in a Finite Time Horizon

T L. t=1. Proof of Lemma 1. Using the marginal cost accounting in Equation(4) and standard arguments. t )+Π RB. t )+K 1(Q RB

OBJECTIVES OF TIME SERIES ANALYSIS

RECTILINEAR MOTION. Contents. Theory Exercise Exercise Exercise Exercise Answer Key

Derivatives of Inverse Trig Functions

Kinematics and kinematic functions

Conservation of Momentum. The purpose of this experiment is to verify the conservation of momentum in two dimensions.

arxiv:physics/ v1 [physics.ed-ph] 6 Feb 2004

Introduction to AC Power, RMS RMS. ECE 2210 AC Power p1. Use RMS in power calculations. AC Power P =? DC Power P =. V I = R =. I 2 R. V p.

Method of Moment Area Equations

PHYSICS 151 Notes for Online Lecture #4

Chapter 3 Kinematics in Two Dimensions

Vectorautoregressive Model and Cointegration Analysis. Time Series Analysis Dr. Sevtap Kestel 1

Non-uniform circular motion *

The average rate of change between two points on a function is d t

Traveling Waves. Chapter Introduction

The Effect of Nonzero Autocorrelation Coefficients on the Distributions of Durbin-Watson Test Estimator: Three Autoregressive Models

20. Applications of the Genetic-Drift Model

Localization and Map Making

Article from. Predictive Analytics and Futurism. July 2016 Issue 13

Zürich. ETH Master Course: L Autonomous Mobile Robots Localization II

Main Ideas in Class Today

The Simple Linear Regression Model: Reporting the Results and Choosing the Functional Form

Echocardiography Project and Finite Fourier Series

Elements of Computer Graphics

Acceleration. Part I. Uniformly Accelerated Motion: Kinematics & Geometry

4.2 Continuous-Time Systems and Processes Problem Definition Let the state variable representation of a linear system be

Christos Papadimitriou & Luca Trevisan November 22, 2016

Final Spring 2007


KEY. Math 334 Midterm I Fall 2008 sections 001 and 003 Instructor: Scott Glasgow

Math 10B: Mock Mid II. April 13, 2016

Dynamic Programming 11/8/2009. Weighted Interval Scheduling. Weighted Interval Scheduling. Unweighted Interval Scheduling: Review

Module: Principles of Financial Econometrics I Lecturer: Dr Baboo M Nowbutsing

Two Coupled Oscillators / Normal Modes

R t. C t P t. + u t. C t = αp t + βr t + v t. + β + w t

Key Chemistry 102 Discussion #4, Chapter 11 and 12 Student name TA name Section. ; u= M. and T(red)=2*T(yellow) ; t(yellow)=4*t(red) or

1.6. Slopes of Tangents and Instantaneous Rate of Change

Lecture 4 Notes (Little s Theorem)

ANOVA INTERPRETING. It might be tempting to just look at the data and wing it

Newtonian Relativity

Kinematics. introduction to kinematics 15A

Crash course in interpretting NMR spectra for lab. NMR = the workhorse of characterization tools reveals connectivity & alkyl chains

CORRELATION. suppose you get r 0. Does that mean there is no correlation between the data sets? many aspects of the data may a ect the value of r

Chapter 15 Lasers, Laser Spectroscopy, and Photochemistry

Chapter 2. Models, Censoring, and Likelihood for Failure-Time Data

The Arcsine Distribution

ACE 562 Fall Lecture 4: Simple Linear Regression Model: Specification and Estimation. by Professor Scott H. Irwin

Integration of the equation of motion with respect to time rather than displacement leads to the equations of impulse and momentum.

Transcription:

Inrodcion o Saisics in sychology SY 1 rofessor Greg Francis Lecre 1 correlaion Did I damage my dagher s eyes? CORRELATION wo ariables may be relaed SAT scores, GA hors in herapy, self-eseem grade on homeworks, grade on exams nmber of risk facors, probabiliy of geing AIDS heigh, poins in baskeball... how do we show he relaionship? scaergrams SCATTERGRAMS plo ale of one ariable agains he ale of he oher ariable Semeser hors of mahemaics 3 1 3 4 6 Academic abiliy 3 RELATIONSHIS Idenifying hese ypes of relaionships is one of he key isses in saisical analysis Consider a 1999 sdy ha repored a relaionship beween he se of nighlighs in a child s room and he endency of he child o need glasses My dagher slep wih a nighligh? COMLICATIONS Clearly here is a relaionship beween sing a nighligh and needing glasses Howeer, i s no clear wha he nare of he relaionship inoles I cold be ha he exra ligh somehow inflences he child s eyes and cases he need for glasses Or i cold be ha needing glasses will somehow co-occr wih he se of a nighligh (e.g., children who need glasses will wan a nigh ligh, or heir parens will wan a nighligh) Finding a relaionship is necessary for esablishing casaion, b i is no enogh OSITIVE CORRELATION Firs, we need o ndersand how o qanify he exisence of a relaionship. Increases in he ale of one ariable end o occr wih increases in he ale of he oher ariable SAT scores and exam scores Final Exam Score 6 4 Sden Sde Sden 8 4 4 6 6 7 SAT Score 4 6

NEGATIVE CORRELATION Increases in he ale of one ariable end o occr wih decreases in he ale of he oher ariable emperare and nmber of people wih frosbie ERFECT CORRELATIONS perfec posiie correlaion 1 - -1-1 - 1 NO CORRELATION no correlaion balance of larger and smaller ales 1 - -1-1 - 1 Nmber of frosbie cases 4 3 1-6 -4 - Degrees (Fahrenhei) perfec negaie correlaion 1 - -1-1 - 1 - -1-1 7 8 9 CORRELATION COEFFICIENT qaniaie measre of correlaion bonded beween 1. & +1. correlaion coe cien of -1. indicaes perfec negaie correlaion correlaion coe cien of +1. indicaes perfec posiie correlaion correlaion coe cien of. indicaes no correlaion ales in beween gie ordinal measres of relaionship earson prodc-momen correlaion coe cien one correlaion coe cien for qaniaie daa (he mos imporan one) degree o which X and Y ary ogeher r = degree o which X and Y ary separaely 1. z-scores. Deiaion scores 3. Raw scores 4. Coariance seeral formlas all gie he same resl! z SCORES Two seps 1. Coner raw scores ino z scores. Find he mean of cross-prodcs 1 11 1

z SCORES wha does his calclaion do? sppose yo hae wo disribions ha hae a posiie correlaion hen a large ale of X will be aboe X and hae a posiie z x score and a corresponding Y will be aboe Y and hae a posiie z y score Ths he cross-prodc will be posiie also a small ale of X will be below X and hae a negaie z x score and he corresponding Y will be below Y and hae a negaie z y score Ths will again be posiie o find he aerage, sm all he prodcs (posiie nmbers) we diide by sill a posiie nmber! exacly he opposie is re for negaiely correlaed disribions hen a large ale of X will be aboe X and hae a posiie z x score and a corresponding Y will be below Y and hae a negaie z y score Ths will be negaie 1 while a small ale of X will be below X and hae a negaie z x score and he corresponding Y will be aboe Y and hae a posiie z y score Ths will again be negaie o find he aerage, sm all he prodcs (negaie nmbers) we diide by sill a negaie nmber! DEVIATION FORMULA i is awkward o coner o z scores we can ge he same nmber wih deiaion scores x = X y = Y X Y deiaion score formla r xy = xy s x y RAW SCORE FORMULA i is awkward o calclae deiaion scores r xy = raw score formla n XY X Y 4 n X ( X) 3 4 n Y ( Y ) 3 16 17 18

COVARIANCE FORMULA coariance = s xy = (X X)(Y Y ) aerage cross-prodc of deiaion scores (similar o ariance) earson r rns o o be: r xy = s xy s x s y where s x and s y are he sandard deiaions of heir respecie disribions Final Exam Score 6 4 Sden X Y 9 68 71 6 4 4 68 64 49 4 6 6 8 9 61 6 43 4 44 38 1 38 37 1 4 6 3 Sde Sden 8 sandard score formla = 1.67 =.9 X Y z xz y 9 68.63 1.64 1.3 -.1.3 -. 71 6 1.88 1.34. 4 4-1.34 -.94 1.6 68 64 1.1 1.4 1.87 49 4 -.46 -.64.9 6 6.3.4. 8 9.48.74.36 61 6.84.4.38 43 4-1.3 -.94.97 44 38 -.97-1.33 1.9 1 -. -.1.3 38 37-1.6-1.43.9 1 4 -. -.94.4 6 3.3.1. X =81 Y =77 z x =. z y =. z xz y =1.67 4 4 6 6 7 SAT Score 19 1 deiaion score formla xy r xy = p x y = 3. r =.93 (46.)(9.7) X Y x y xy 9 68 61. 16.3 18.33 -. 3.3-49.4 71 6 181..3 448.93 4 4-19. -9.47 11.63 68 64 6. 1.3 189.38 49 4-44. -6.47 84.68 6 6 31. 4.3.43 8 9 46. 7.3 346.38 61 6 81. 4.3 366.93 43 4-99. -9.47 937.3 44 38-94. -.47 166.18 1-19. -1.47 7.93 38 37-14. -.47 8.38 1 4-4. -9.47 7.8 6 3 31. 1.3 47.43 X =81 Y =77 x =. y =. xy =3. x =46. and y =9.7 raw score formla n XY X Y rxy = r hn X ( X) ih n Y ( Y ) i (1)(448) (81)(77) q [(1)(4478) (81) ][(1)(4116) (77) ] =.93 X Y XY 9 68 446 86 71 6 4647 4 4 171 68 64 43 49 4 6 6 3164 8 9 34 61 6 3444 43 4 187 44 38 167 1 7 38 37 6 1 4 6 3 994 X =81 Y =77 XY =448 X =4478and Y =4116 3 coariance formla r xy = s xy 88.86 = s x s y (96.3)(1.11) =.93 where, s xy = xy = 3 =88.86 s x = s y = x = y = 46 =96.3 9.7 =1.11 4

CORRELATION r measres correlaion beween wo ariables no js any wo ariables 1. The wo ariables ms be paired obseraions.. Variables ms be qaniaie (ineral or raio scale). CONCLUSIONS correlaion scaergrams earson r formlas NEXT TIME facors a ecing r inerpreing r and Is here a link beween IQ and problem soling abiliy? 6 7