Agenda: Recap. Lecture. Chapter 12. Homework. Chapt 12 #1, 2, 3 SAS Problems 3 & 4 by hand. Marquette University MATH 4740/MSCS 5740

Similar documents
Class 23. Daniel B. Rowe, Ph.D. Department of Mathematics, Statistics, and Computer Science. Marquette University MATH 1700

Class 27. Daniel B. Rowe, Ph.D. Department of Mathematics, Statistics, and Computer Science. Marquette University MATH 1700

Important Formulas. Expectation: E (X) = Σ [X P(X)] = n p q σ = n p q. P(X) = n! X1! X 2! X 3! X k! p X. Chapter 6 The Normal Distribution.

Lecture 8: Non-parametric Comparison of Location. GENOME 560, Spring 2016 Doug Fowler, GS

Common Large/Small Sample Tests 1/55

Describing the Relation between Two Variables

Lecture 7: Non-parametric Comparison of Location. GENOME 560, Spring 2016 Doug Fowler, GS

Class 27. Daniel B. Rowe, Ph.D. Department of Mathematics, Statistics, and Computer Science. Marquette University MATH 1700

Stat 200 -Testing Summary Page 1

Lecture 7: Non-parametric Comparison of Location. GENOME 560 Doug Fowler, GS

Grant MacEwan University STAT 252 Dr. Karen Buro Formula Sheet

Chapters 5 and 13: REGRESSION AND CORRELATION. Univariate data: x, Bivariate data (x,y).

1 Inferential Methods for Correlation and Regression Analysis

TABLES AND FORMULAS FOR MOORE Basic Practice of Statistics

Chapter 13: Tests of Hypothesis Section 13.1 Introduction

Goodness-of-Fit Tests and Categorical Data Analysis (Devore Chapter Fourteen)

Linear Regression Models

Agreement of CI and HT. Lecture 13 - Tests of Proportions. Example - Waiting Times

STAT431 Review. X = n. n )

UNIVERSITY OF TORONTO Faculty of Arts and Science APRIL/MAY 2009 EXAMINATIONS ECO220Y1Y PART 1 OF 2 SOLUTIONS

MidtermII Review. Sta Fall Office Hours Wednesday 12:30-2:30pm Watch linear regression videos before lab on Thursday

Statistics 20: Final Exam Solutions Summer Session 2007

Pearson Edexcel Level 3 Advanced Subsidiary and Advanced GCE in Statistics

MOST PEOPLE WOULD RATHER LIVE WITH A PROBLEM THEY CAN'T SOLVE, THAN ACCEPT A SOLUTION THEY CAN'T UNDERSTAND.

Rule of probability. Let A and B be two events (sets of elementary events). 11. If P (AB) = P (A)P (B), then A and B are independent.

Comparing Two Populations. Topic 15 - Two Sample Inference I. Comparing Two Means. Comparing Two Pop Means. Background Reading

Good luck! School of Business and Economics. Business Statistics E_BK1_BS / E_IBA1_BS. Date: 25 May, Time: 12:00. Calculator allowed:

Stat 319 Theory of Statistics (2) Exercises

Direction: This test is worth 150 points. You are required to complete this test within 55 minutes.

A Relationship Between the One-Way MANOVA Test Statistic and the Hotelling Lawley Trace Test Statistic

Chapter 13, Part A Analysis of Variance and Experimental Design

Continuous Data that can take on any real number (time/length) based on sample data. Categorical data can only be named or categorised

This is an introductory course in Analysis of Variance and Design of Experiments.

STATISTICAL INFERENCE

2 1. The r.s., of size n2, from population 2 will be. 2 and 2. 2) The two populations are independent. This implies that all of the n1 n2

Sample Size Determination (Two or More Samples)

Final Review. Fall 2013 Prof. Yao Xie, H. Milton Stewart School of Industrial Systems & Engineering Georgia Tech

Lecture 33: Bootstrap

TABLES AND FORMULAS FOR MOORE Basic Practice of Statistics

FACULTY OF MATHEMATICAL STUDIES MATHEMATICS FOR PART I ENGINEERING. Lectures

Regression, Inference, and Model Building

[ ] ( ) ( ) [ ] ( ) 1 [ ] [ ] Sums of Random Variables Y = a 1 X 1 + a 2 X 2 + +a n X n The expected value of Y is:

A quick activity - Central Limit Theorem and Proportions. Lecture 21: Testing Proportions. Results from the GSS. Statistics and the General Population

( )( ) χ = where Ei = npi. χ = where. : At least one proportion differs significantly. H0 : The two classifications are independent H

Simple Regression. Acknowledgement. These slides are based on presentations created and copyrighted by Prof. Daniel Menasce (GMU) CS 700

Chapter 22. Comparing Two Proportions. Copyright 2010 Pearson Education, Inc.

SDS 321: Introduction to Probability and Statistics

Final Examination Solutions 17/6/2010

S Y Y = ΣY 2 n. Using the above expressions, the correlation coefficient is. r = SXX S Y Y

Parameter, Statistic and Random Samples

Lecture 5: Parametric Hypothesis Testing: Comparing Means. GENOME 560, Spring 2016 Doug Fowler, GS

Sampling Distributions, Z-Tests, Power

z is the upper tail critical value from the normal distribution

ST 305: Exam 3 ( ) = P(A)P(B A) ( ) = P(A) + P(B) ( ) = 1 P( A) ( ) = P(A) P(B) σ X 2 = σ a+bx. σ ˆp. σ X +Y. σ X Y. σ X. σ Y. σ n.

TAMS24: Notations and Formulas

GG313 GEOLOGICAL DATA ANALYSIS

Recall the study where we estimated the difference between mean systolic blood pressure levels of users of oral contraceptives and non-users, x - y.

Interval Estimation (Confidence Interval = C.I.): An interval estimate of some population parameter is an interval of the form (, ),

STA Learning Objectives. Population Proportions. Module 10 Comparing Two Proportions. Upon completing this module, you should be able to:

Formulas and Tables for Gerstman

EXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY

Comparing your lab results with the others by one-way ANOVA

Confidence Level We want to estimate the true mean of a random variable X economically and with confidence.

Tests of Hypotheses Based on a Single Sample (Devore Chapter Eight)

(all terms are scalars).the minimization is clearer in sum notation:

Chapter 22. Comparing Two Proportions. Copyright 2010, 2007, 2004 Pearson Education, Inc.

Overview. p 2. Chapter 9. Pooled Estimate of. q = 1 p. Notation for Two Proportions. Inferences about Two Proportions. Assumptions

Question 1: Exercise 8.2

Hypothesis Testing. Evaluation of Performance of Learned h. Issues. Trade-off Between Bias and Variance

TMA4245 Statistics. Corrected 30 May and 4 June Norwegian University of Science and Technology Department of Mathematical Sciences.


This chapter focuses on two experimental designs that are crucial to comparative studies: (1) independent samples and (2) matched pair samples.

October 25, 2018 BIM 105 Probability and Statistics for Biomedical Engineers 1

Mathematical Notation Math Introduction to Applied Statistics

A statistical method to determine sample size to estimate characteristic value of soil parameters

UCLA STAT 110B Applied Statistics for Engineering and the Sciences

Chapter 20. Comparing Two Proportions. BPS - 5th Ed. Chapter 20 1

3/3/2014. CDS M Phil Econometrics. Types of Relationships. Types of Relationships. Types of Relationships. Vijayamohanan Pillai N.

t distribution [34] : used to test a mean against an hypothesized value (H 0 : µ = µ 0 ) or the difference

Topic 9: Sampling Distributions of Estimators

A Question. Output Analysis. Example. What Are We Doing Wrong? Result from throwing a die. Let X be the random variable

1036: Probability & Statistics

Two sample test (def 8.1) vs one sample test : Hypotesis testing: Two samples (Chapter 8) Example 8.2. Matched pairs (Example 8.6)

Properties and Hypothesis Testing

BIOS 4110: Introduction to Biostatistics. Breheny. Lab #9

Data Analysis and Statistical Methods Statistics 651

Resampling Methods. X (1/2), i.e., Pr (X i m) = 1/2. We order the data: X (1) X (2) X (n). Define the sample median: ( n.

April 18, 2017 CONFIDENCE INTERVALS AND HYPOTHESIS TESTING, UNDERGRADUATE MATH 526 STYLE

Introductory statistics

Regression. Correlation vs. regression. The parameters of linear regression. Regression assumes... Random sample. Y = α + β X.

Math 140 Introductory Statistics

Lecture 10: Performance Evaluation of ML Methods

Statistics. Chapter 10 Two-Sample Tests. Copyright 2013 Pearson Education, Inc. publishing as Prentice Hall. Chap 10-1

(7 One- and Two-Sample Estimation Problem )

Open book and notes. 120 minutes. Cover page and six pages of exam. No calculators.

Sample Size Estimation in the Proportional Hazards Model for K-sample or Regression Settings Scott S. Emerson, M.D., Ph.D.

Lecture 3. Properties of Summary Statistics: Sampling Distribution

Grant MacEwan University STAT 151 Formula Sheet Final Exam Dr. Karen Buro

V. Nollau Institute of Mathematical Stochastics, Technical University of Dresden, Germany

One-Sample Test for Proportion

Transcription:

Ageda: Recap. Lecture. Chapter Homework. Chapt #,, 3 SAS Problems 3 & 4 by had. Copyright 06 by D.B. Rowe

Recap.

6: Statistical Iferece: Procedures for μ -μ 6. Statistical Iferece Cocerig μ -μ Recall yes Variaces kow? Idepedet populatios? o Case 4 yes Case Z Case o F test Case 3 30 Z 30 t, 30, 30, 30, 30 Z t Z t 3

6: Statistical Iferece: Procedures for μ -μ 6. Statistical Iferece Cocerig μ -μ Depedet Populatios: Case 4: Test Statistic Cofidece Iterval & ukow 30 s X 30 Z d d d Xd Z / s / & ukow 30 30 df t X d / d d s d X i d t / di X i X i Xd di sd ( d ) i X d s d i 4

: Noparametric Tests. The Sig Test (Two depedet Samples Test) If d i s are ot N(μ d,σ ), & <30 the we kow that X d is ot N(μ d,σ /), o CLT, ad that X d X d d μ d is ot N(0,σ ), o CLT, ad that / ( ) s X d d t= s / d d is ot N(0,), o CLT, ad that is ot χ (-), but approximately ~t(-)! Studies have show that if X i s are from moud shaped dist, the approximatio ot bad. This used i Chapt 6 Case 4 Robust to mild departures from ormality. 5

: Noparametric Tests. The Sig Test (Two depedet Samples Test) For the Sig Test, we still calculate differeces d i =X i X i, but ot the sample mea, X d. We determie whether the differeces are positive +, egative, or zero 0. If the medias of the two distributios are truly the same, the we should have a similar allocatio of + ad. d i i 6

: Noparametric Tests. The Sig Test (Two depedet Samples Test) H 0 : p=( or )0.5 (this is a media test). Let X be the umber of + sigs. If there is o differece i medias, i.e. H 0 true, the X has a biomial distributio with p=0.5. Sigificace from biomial. Assumptios For Test:. d i s are idepedet (each trial idepedet). d i s from same distributio (p ot chagig) 7

: Noparametric Tests. The Sig Test (Two depedet Samples Test) If the medias of the two distributios are truly the same, the we should have a similar allocatio of + ad. Example.: Pre/Post Blood Pressures H 0 : p 0.5, =0, x=7. Assume H 0 true. How likely is it to observe 7 or more successes out of 0 whe p=0.5? P(X 7)=0.79. Reject H 0 if p-value 0.05. p-value 8

: Noparametric Tests. The Sig Test (Two depedet Samples Test) Example.: was H 0 : p 0.5 vs. H 0 : p>0.5. by calculatig P(X x), reject H 0 if p-value<α. p-value But we ca also test H 0 : p 0.5 vs. H 0 : p<0.5. by calculatig P(X x), reject H 0 if p-value<α. p-value But we ca also test H 0 : p=0.5 vs. H 0 : p 0.5. by calculatig P(X x)+p(x x), reject H 0 if p-value<α. p-value 9

: Noparametric Tests. The Wilcoxo Siged-Rak Test (Two Depedet Samples Test) Example.: Pre/Post Blood Pressures H 0 : p 0.5, =0, x=7. Assume H 0 true. T=40.5. The max possible T is 45. ad the media is.5. Sice T>37, reject H 0! 0

: Noparametric Tests. The Wilcoxo Siged-Rak Test (Two Depedet Samples Test) Geerally the exact critical value table is ot used, a approximate Z statistic is used, Z ( ) T 4 ( )( ) 4 ad the usual ormal Z table is used. ( ) ET ( ) ( )( ) var( T) 4 6

: Noparametric Tests. The Wilcoxo Siged-Rak Test Example.: Pre/Post Blood Pressures H 0 : p 0.5, =0, x=7. Assume H 0 true. T=40.5. The media is.5. (Two Depedet Samples Test) Z ( ) T 4 ( )( ) 4 Z 40.5.5 9(9 )( 9 ) 4.3 Sice Z>.645, reject H 0!

Chapter : Noparametric Tests Daiel B. Rowe, Ph.D. Departmet of Mathematics, Statistics, ad Computer Sciece 3

6: Statistical Iferece: Procedures for μ -μ Itroductio Recall X Four Cases: Case. Two idepedet populatios populatio variaces are kow, ( & kow & ). Case. Two idepedet populatios populatio variaces are ukow but assumed to be equal, ( & ukow & =). Case 3. Two idepedet populatios populatio variaces ukow ad possibly uequal, ( & ukow & ). Case 4. Two depedet populatios data are assumed to be matched or paired ( & kow or ukow). 4

6: Statistical Iferece: Procedures for μ -μ 6. Statistical Iferece Cocerig μ -μ Recall yes Variaces kow? Idepedet populatios? o Case 4 yes Case Z Case o F test Case 3 30 Z 30 t, 30, 30, 30, 30 Z t Z t 5

6: Statistical Iferece: Procedures for μ -μ 6. Statistical Iferece Cocerig μ -μ Recall Idepedet Populatios: Case : Test Statistic & ukow X X Z 30 30 S p & ukow 30 30 ( ) ( ) ( X X) ( ) t S p S df ( ) s ( ) s P 6

6: Statistical Iferece: Procedures for μ -μ 6. Statistical Iferece Cocerig μ -μ Recall Idepedet Populatios: Case 3: Test Statistic & ukow X X 30 Z s s 30 & ukow 30 30 ( ) ( ) ( X X) ( ) t s s Next larger umber s s df s / s / 7

: Noparametric Tests.3 The Wilcoxo Rak Sum Test (Two Idepedet Samples Test) We kow that if X i s~n(μ, ) & X i s~n(μ, ), the we kow X ~N(μ, / ) & X~N(μ, / ), ad Case :, 30 ( X X ) ~ N(( ), / / ), ad ( X X ) ( ) ~ N(0, / / ), ad ( X X) ( ) ~ N (0,), ad 8

: Noparametric Tests.3 The Wilcoxo Rak Sum Test idepedet that ( X X) ( ) ~ N (0,) s ( ) ( ) s (Two Idepedet Samples Test) ~ ( ), ad ~ ( ), ad, ad Case : ( ) s ( ) s ~ ( ), 30 9

: Noparametric Tests.3 The Wilcoxo Rak Sum Test that (Two Idepedet Samples Test) ( X X) ( ) ~ N (0,) Case :, ad, 30 ( ) s ( ) s ( X X) ( ) ~ ( ) t t Z S p ~ ( ) S if large. ( ) s ( ) s P 0

: Noparametric Tests.3 The Wilcoxo Rak Sum Test (Two Idepedet Samples Test) If X i s ot~n(μ, ) & X i s ot~n(μ, ), the X Case : ot~n(μ, / ) & X ot~n(μ, / ), ad, 30 ( X X ) ot ~ N(( ), / / ), ad ( X X ) ( ) ot ~ N(0, / / ), ad ( X X ) ( ) ot ~ N(0,), ad

: Noparametric Tests.3 The Wilcoxo Rak Sum Test idepedet (Two Idepedet Samples Test) ( X X ) ( ) that ot ~ N(0,), ad ( ) s ( ) s ot ( ) s ( ) s ~ ( ), ad ot ~ ( ), ad ot Case : ~ ( ), 30

: Noparametric Tests.3 The Wilcoxo Rak Sum Test (Two Idepedet Samples Test) ( X X ) ( ) that ot ~ N(0,), ad, 30 ( ) s ( ) s ot ~ ( ), but approximately ( X X) ( ) t ~ t ( ) S p Case : Studies have show that if X i s are from moud shaped dist, the approximatio ot bad. This used i Chapt 6 Case 3

: Noparametric Tests.3 The Wilcoxo Rak Sum Test (Two Idepedet Samples Test) Whe & <30 ad we are cocered that ( X X) ( ) t does ot have a Studet-t distributio, S p the we eed to resort to a oparametric test. H 0 : MD MD =0 ot H 0 : µ -µ =0 Case :, 30 4

: Noparametric Tests.3 The Wilcoxo Rak Sum Test (Two Idepedet Samples Test) For the Wilcoxo Rak Sum Test with ad, Rak all the observatios i icreasig order. Assig ties their average rak. Let S=the sum of the raks for the sample of size. The sum of all raks is (+)/, = +. If there is o differece, the we expect some small ad some large raks so S (+)/4! 5

: Noparametric Tests.3 The Wilcoxo Rak Sum Test (Two Idepedet Samples Test) S=the sum of the raks for the sample of size. Table Not i Book. http://faculty.fiu.edu/~mcguckd/wilcoxo%0rak%0sum%0table.pdf 6

: Noparametric Tests.3 The Wilcoxo Rak Sum Test (Two Idepedet Samples Test) Example.3: Health Status H 0 : MD MD =0 H : MD MD 0 α=0.05 S =6 S =0 If there is o differece, the we expect to see some low ad some high raks i each group. Sum of all raks=36. Expect S 8. S=6. Accordig to the Table, sice S 5, do ot reject. 7

: Noparametric Tests.3 The Wilcoxo Rak Sum Test (Two Idepedet Samples Test) Geerally the exact critical value table is ot used, a approximate Z statistic is used, Z ( ) S ( ) ad the usual ormal Z table is used. E( S) ( ) ( ) var( S) 4 3 8

: Noparametric Tests.3 The Wilcoxo Rak Sum Test (Two Idepedet Samples Test) Example.3: Health Status H 0 : MD MD =0 H : MD MD 0 α=0.05 4(4 4 ) 6 Z 0.58 4 4(4 4 ) We kow that sice Z <.96, do ot reject. Z ( ) S ( ) 9

9: Aalysis of Variace Recall ANOVA is a geeralizatio of these methods ad is for testig for differeces i three or more populatio meas. i.e. H 0 : μ = μ = μ 3 H : meas ot all equal (at least two are differet) (Ca be ad ; or ad 3 ; or ad 3 ; or,, ad 3 ) More geerally, H 0 : μ = μ = = μ k, k is # of populatio meas comparig H : meas ot all equal (at least two are differet) I ANOVA there are o iequality hypotheses. No vs > or vs <. 30

9: Aalysis of Variace X X Assumptios for valid applicatio of ANOVA ) k idepedet populatios ) Radom samples from each of k > populatios 3) Large samples ( i 30) or ormal populatios 4) Equal populatio variaces k 3

: Noparametric Tests.4 The Kruskal-Wallis Test (k Idepedet Samples Test) The K-W Test is a oparametric alterative to ANOVA. Rak all the observatios ad replace by raks, R ij. Assig ties their average rak. rak i i populatio j Let R i =the sum of the raks for the ith sample of size i. k R. ( ) i k ( ) H i s Rij s i i 4 i j 4 Reject H 0 for large H. If there ( ) s are o ties 3

: Noparametric Tests.4 The Kruskal-Wallis Test (k Idepedet Samples Test) The K-W Test is a oparametric alterative to ANOVA. There are exact tables available that deped o k ad i s. http://faculty.virgiia.edu/kruskal-wallis/ 33

: Noparametric Tests.4 The Kruskal-Wallis Test (k Idepedet Samples Test) The K-W Test is a oparametric alterative to ANOVA. H 0 : MD =MD =MD 3 =MD 4, H : At least two MDs. H 0.703 CV 7.38 Reject H 0, at least two MDs. R ( ) k. i H s i i 4 s k i k i Rij i j s R. ( ) i 349. i 4 ( ) 4 3.636 34

: Noparametric Tests.4 The Kruskal-Wallis Test (k Idepedet Samples Test) Geerally the exact critical value table is ot used. If the i are 5, the H has a approximate χ distributio uder the ull hypothesis. H k R. ( ) i s i i 4 ad the usual χ table is used. Reject H 0 if H> ( k ). ~ ( k ) 35

: Noparametric Tests.4 The Kruskal-Wallis Test (k Idepedet Samples Test) The K-W Test is a oparametric alterative to ANOVA. R ( ) k. i H s i i 4 H 0 : MD =MD =MD 3 =MD 4, H : At least two. H 0.703 ( k ) 7.8 Reject H 0, at least two MDs. s 3.636 36

0: Correlatio ad Regressio 0. Correlatio Aalysis 0.. The Sample Correlatio Coefficiet Recall The sample correlatio is r also give by r cov( XY, ) var( X) var( Y) ( x y ) ( x )( y ) i i i i i i i [ ( xi ) ( xi ) ][ ( yi ) ( yi ) ] i i i i cov( X, Y) ( X i X )( Yi Y ) i var( X ) ( X i X ) i var( Y) ( Yi Y ) i 37

: Noparametric Tests.5 Spearma Correlatio (Correlatio Betwee Variables) The Spearma rak correlatio betwee X ad Y is to covert the actual observed values to raks R X ad R y, the calculate the correlatio betwee R X ad R y. r s cov( R, R ) x var( R ) var( R ) x y y cov( R, R ) ( R R )( R R ) x y xi x yi y i var( R ) ( R R ) x xi x i var( R ) ( R R ) y yi y i 38

: Noparametric Tests.5 Spearma Correlatio (Correlatio Betwee Variables) We ca perform a hypothesis test o r s. H 0 : ρ s =0 vs. ρ s 0. r s cov( R, R ) x var( R ) var( R ) x y y Reject H 0 if r s large. 39

: Noparametric Tests.5 Spearma Correlatio There is a exact table. r s cov( R, R ) x var( R ) var( R ) x y y https://www.york.ac.uk/depts/maths/tables/spearma.pdf 40

: Noparametric Tests.5 Spearma Correlatio But the exact table is ot geerally used. t r ~ ( ) s t r s H 0 : ρ s = 0 vs. H : ρ s 0, α=0.05 Reject if t >t.05 (df-) 4

: Noparametric Tests Questios? Homework: Chapter #,, 3, SAS Problems 3 & 4 by had. 4