Comparing Three or More Groups: Multiple Comparisons vs Planned Comparisons Robert Boudreau, PhD Co-Director of Methodology Core PITT-Multidisciplinary.

Slides:



Advertisements
Similar presentations
Intro to ANOVA.
Advertisements

Dr. AJIT SAHAI Director – Professor Biometrics JIPMER, Pondicherry
Lesson #24 Multiple Comparisons. When doing ANOVA, suppose we reject H 0 :  1 =  2 =  3 = … =  k Next, we want to know which means differ. This does.
Section Copyright © 2014, 2012, 2010 Pearson Education, Inc. Lecture Slides Elementary Statistics Twelfth Edition and the Triola Statistics Series.
Chapter 10 Section 2 Hypothesis Tests for a Population Mean
Regression Part II One-factor ANOVA Another dummy variable coding scheme Contrasts Multiple comparisons Interactions.
Analysis of Variance (ANOVA) Statistics for the Social Sciences Psychology 340 Spring 2010.
1. Estimation ESTIMATION.
POST HOC COMPARISONS A significant F in ANOVA tells you only that there is a difference among the groups, not which groups are different. Post hoc tests.
Review: What influences confidence intervals?
ANOVA Determining Which Means Differ in Single Factor Models Determining Which Means Differ in Single Factor Models.
Comparing Means.
Intro to Statistics for the Behavioral Sciences PSYC 1900
POST HOC COMPARISONS What is the Purpose?
Analyses of K-Group Designs : Omnibus F & Pairwise Comparisons ANOVA for multiple condition designs Pairwise comparisons and RH Testing Alpha inflation.
C82MCP Diploma Statistics School of Psychology University of Nottingham 1 Overview of Lecture Independent and Dependent Variables Between and Within Designs.
Post-hoc Tests for ANOVA Explaining significant differences in 1-way ANOVA.
K-group ANOVA & Pairwise Comparisons ANOVA for multiple condition designs Pairwise comparisons and RH Testing Alpha inflation & Correction LSD & HSD procedures.
Comparing Means.
Today Concepts underlying inferential statistics
Linear Contrasts and Multiple Comparisons (Chapter 9)
Chapter 12 Inferential Statistics Gay, Mills, and Airasian
Quiz 12  Nonparametric statistics. 1. Which condition is not required to perform a non- parametric test? a) random sampling of population b) data are.
Chapter 12: Analysis of Variance
Sets, Combinatorics, Probability, and Number Theory Mathematical Structures for Computer Science Chapter 3 Copyright © 2006 W.H. Freeman & Co.MSCS SlidesProbability.
Intermediate Applied Statistics STAT 460
1 Multiple Comparison Procedures Once we reject H 0 :   =   =...  c in favor of H 1 : NOT all  ’s are equal, we don’t yet know the way in which.
When we think only of sincerely helping all others, not ourselves,
Regression Part II One-factor ANOVA Another dummy variable coding scheme Contrasts Multiple comparisons Interactions.
January 31 and February 3,  Some formulae are presented in this lecture to provide the general mathematical background to the topic or to demonstrate.
Sociology 5811: Lecture 14: ANOVA 2
Introduction to Biostatistics, Harvard Extension School, Fall, 2005 © Scott Evans, Ph.D.1 Analysis of Variance (ANOVA)
ANOVA (Analysis of Variance) by Aziza Munir
Everyday is a new beginning in life. Every moment is a time for self vigilance.
Binomial Probability Distribution
Chapter 10: Analyzing Experimental Data Inferential statistics are used to determine whether the independent variable had an effect on the dependent variance.
Regression Part II One-factor ANOVA Another dummy variable coding scheme Contrasts Multiple comparisons Interactions.
Statistical Hypotheses & Hypothesis Testing. Statistical Hypotheses There are two types of statistical hypotheses. Null Hypothesis The null hypothesis,
Analysis of Variance (ANOVA) Can compare the effects of different treatments Can make population level inferences based on sample population.
Analysis of Variance (ANOVA) Brian Healy, PhD BIO203.
Slide Slide 1 Section 8-4 Testing a Claim About a Mean:  Known.
Marshall University School of Medicine Department of Biochemistry and Microbiology BMS 617 Lecture 13: One-way ANOVA Marshall University Genomics Core.
Chapter 12 Introduction to Analysis of Variance PowerPoint Lecture Slides Essentials of Statistics for the Behavioral Sciences Eighth Edition by Frederick.
Chapter 13 For Explaining Psychological Statistics, 4th ed. by B. Cohen 1 Chapter 13: Multiple Comparisons Experimentwise Alpha (α EW ) –The probability.
IMPORTANCE OF STATISTICS MR.CHITHRAVEL.V ASST.PROFESSOR ACN.
MARE 250 Dr. Jason Turner Analysis of Variance (ANOVA)
Quiz 11  Analysis of Variance (ANOVA)  post hoc tests.
Psych 230 Psychological Measurement and Statistics Pedro Wolf November 18, 2009.
Statistics for the Social Sciences Psychology 340 Spring 2009 Analysis of Variance (ANOVA)
Chapter 11: The ANalysis Of Variance (ANOVA)
Inference for Proportions Section Starter Do dogs who are house pets have higher cholesterol than dogs who live in a research clinic? A.
Simple ANOVA Comparing the Means of Three or More Groups Chapter 9.
Chapters Way Analysis of Variance - Completely Randomized Design.
Stats/Methods II JEOPARDY. Jeopardy Estimation ANOVA shorthand ANOVA concepts Post hoc testsSurprise $100 $200$200 $300 $500 $400 $300 $400 $300 $400.
MARE 250 Dr. Jason Turner Analysis of Variance (ANOVA)
Educational Research Inferential Statistics Chapter th Chapter 12- 8th Gay and Airasian.
Chapter 12 Introduction to Analysis of Variance
Evaluating Hypotheses. Outline Empirically evaluating the accuracy of hypotheses is fundamental to machine learning – How well does this estimate accuracy.
Pairwise comparisons: Confidence intervals Multiple comparisons Marina Bogomolov and Gili Baumer.
PROBABILITY AND COMPUTING RANDOMIZED ALGORITHMS AND PROBABILISTIC ANALYSIS CHAPTER 1 IWAMA and ITO Lab. M1 Sakaidani Hikaru 1.
Inference for Proportions
Multiple Comparisons Q560: Experimental Methods in Cognitive Science Lecture 10.
Analysis of Treatment Means
1-Way Analysis of Variance - Completely Randomized Design
I. Statistical Tests: Why do we use them? What do they involve?
ANOVA Determining Which Means Differ in Single Factor Models
Chi2 (A.K.A X2).
Analysis of Treatment Means
Confidence Intervals.
1-Way Analysis of Variance - Completely Randomized Design
Presentation transcript:

Comparing Three or More Groups: Multiple Comparisons vs Planned Comparisons Robert Boudreau, PhD Co-Director of Methodology Core PITT-Multidisciplinary Clinical Research Center for Rheumatic and Musculoskeletal Diseases

First a simple thought experiment Flip a fair coin 100 times: Let H=# heads H = 0,1,2, …, 100 are the possible outcomes H = 0,1,2, …, 100 are the possible outcomes H has a binomial distribution with known probs H has a binomial distribution with known probs Prob[ 40 < H < 60 ] very close to 0.95 Prob[ 40 < H < 60 ] very close to 0.95  Prob [ H ≤ 40 ] + P[ H ≥ 60] =

First a simple thought experiment Flip a fair coin 100 times: Let H=# heads H = 0,1,2, …, 100 are the possible outcomes H = 0,1,2, …, 100 are the possible outcomes H has a binomial distribution with known probs H has a binomial distribution with known probs Prob[ 40 < H < 60 ] very close to 0.95 Prob[ 40 < H < 60 ] very close to 0.95  Prob [ H ≤ 40 ] + P[ H ≥ 60] = Experiment: 20 people flip their own coin 100 times Q: Approx how many will get 40 or fewer heads or 60+ heads? or 60+ heads?

First a simple thought experiment Flip a fair coin 100 times: Let H=# heads H = 0,1,2, …, 100 are the possible outcomes H = 0,1,2, …, 100 are the possible outcomes H has a binomial distribution with known probs H has a binomial distribution with known probs Prob[ 40 < H < 60 ] very close to 0.95 Prob[ 40 < H < 60 ] very close to 0.95  Prob [ H ≤ 40 ] + P[ H ≥ 60] = Experiment: 20 people flip their own coin 100 times Q: Approx how many will get less than 40 heads or 60+ heads? Answer: One or 60+ heads? Answer: One

First a simple thought experiment Flip a fair coin 100 times: Let H=# heads H = 0,1,2, …, 100 are the possible outcomes H = 0,1,2, …, 100 are the possible outcomes H has a binomial distribution with known probs H has a binomial distribution with known probs Prob[ 40 < H < 60 ] very close to 0.95 Prob[ 40 < H < 60 ] very close to 0.95  Prob [ H ≤ 40 ] + P[ H ≥ 60] = Experiment: 20 people flip their own coin 100 times Q: Approx how many will get less than 40 heads or 60+ heads? Answer: One (1/20 = 5%) or 60+ heads? Answer: One (1/20 = 5%)

First a simple thought experiment Experiment: 20 people flip their own coin 100 times One (1/20=0.05) will flip an unusually small or unusually large # heads (on average) One (1/20=0.05) will flip an unusually small or unusually large # heads (on average) Q: Can we conclude that this person “X” flips an “unfair” coin, or was this explainable by “chance”?

Controlling Experiment-wise Error Experiment: 20 people flip their own coin 100 times  Person X’s confidence interval didn’t cover 0.5 Q: What alpha level should be used so that 95% of the time all 20 confidence intervals each cover 0.5? (i.e. so that the correct conclusion is drawn about (i.e. so that the correct conclusion is drawn about every single coin) every single coin)

Controlling Experiment-wise Error Experiment: 20 people flip their own coin 100 times  Person X’s confidence interval didn’t cover 0.5 Q: What alpha level should be used so that 95% of the time all 20 confidence intervals each cover 0.5? (i.e. so that the correct conclusion is drawn about (i.e. so that the correct conclusion is drawn about every single coin) every single coin)  Equivalent to drawing a “wrong” conclusion about at least one of the coins only 5% of the time (Experiment-wise Type I error)

Controlling Experiment-wise Error Q: What alpha level should be used so that there’s a 95% probability that all 20 confidence intervals each cover 0.5? (aka Experiment-wise correct conclusion) Experiment-wise α=0.05, solve for comparison-wise α*: α = Prob[ At least one C.I. misses 0 ] α = Prob[ At least one C.I. misses 0 ] = 1 – Prob[ All C.I.’s cover 0 ] = 1 – Prob[ All C.I.’s cover 0 ] = 1 – (1 – α* ) 20 = 1 – (1 – α* ) 20 Sidak: Comparison-wise α* = 1 – (1 – α) 1/n n=20 “comparisons”: α* = 1 – (1-.05) 1/20 = n=20 “comparisons”: α* = 1 – (1-.05) 1/20 =

Controlling Experiment-wise Error Q: What alpha level should be used so that there’s a 95% probability that all 20 confidence intervals each cover 0.5? Sidak: Comparison-wise α* = 1 – (1 – α) 1/n n=20 “comparisons”: α* = 1 – (1-.05) 1/20 = n=20 “comparisons”: α* = 1 – (1-.05) 1/20 = Bonferroni: α* = α/n ( 0.05/20=0.0025)

Controlling Experiment-wise Error Mathematically: α/n < 1 – (1 – α) 1/n Mathematically: α/n < 1 – (1 – α) 1/n Bonferroni < Sidak (i.e. higher α-level) But usually very close  Sidak slightly more powerful Bonferroni works in all situations to guarantee control of experimentwise error (but may be conservative) Bonferroni works in all situations to guarantee control of experimentwise error (but may be conservative) Sidak (derived assuming independence) can under- control in presence of high correlations Sidak (derived assuming independence) can under- control in presence of high correlations

Comparison of Adverse Effect of 4 Drugs on Systolic BP

Unadjusted pairwise t-tests (α = 0.05 each comparison) critical value of t=

Pairwise t-tests (Bonferroni) critical value of t= Pairwise t-tests (Bonferroni) critical value of t=

Pairwise t-tests (Sidak) critical value of t= Pairwise t-tests (Sidak) critical value of t=

Comparison of critical values Scheffe: * Designed for arbitrary post-hoc testing * Controls experimentwise error for all possible simultaneous comparisons and contrasts

Comparison of Adverse Effect of 4 Drugs on Systolic BP (v2) Note: For Drug 4, I’ve subtracted 6 from the previous values s s

Comparison of Adverse Effect of 4 Drugs on Systolic BP (v2) ANOVA F-test

Unadjusted pairwise t-tests (v2) (α = 0.05 each comparison) critical value of t=

Pairwise t-tests (Bonferroni) (v2) critical value of t= Pairwise t-tests (Bonferroni) (v2) critical value of t=

Pairwise t-tests (Sidak) (v2) critical value of t= Pairwise t-tests (Sidak) (v2) critical value of t=

Tukey’s Studentized Range Test Related in concept to Scheffe’s Method Related in concept to Scheffe’s Method Designed for all pairwise comparisons exclusively Designed for all pairwise comparisons exclusively (recall: Scheffe applies to all possible simultaneous pairwise comparisons and contrasts) pairwise comparisons and contrasts) Exact experimentwise error coverage if sample sizes equal Exact experimentwise error coverage if sample sizes equal Critical values smaller than Bonferroni or Sidak Critical values smaller than Bonferroni or Sidak  More powerful in finding differences

Pairwise t-tests (Tukey) (v2) critical value of t= Pairwise t-tests (Tukey) (v2) critical value of t=

Comparison of Adverse Effect of 4 Drugs on Systolic BP

Dunnett’s Method (Comparison vs a Control) Related in concept to Scheffe and Tukey Methods Related in concept to Scheffe and Tukey Methods Designed for pairwise comparisons vs a single control exclusively Designed for pairwise comparisons vs a single control exclusively Exact experimentwise error coverage of those comparisons if sample sizes equal Exact experimentwise error coverage of those comparisons if sample sizes equal Critical values smaller than Bonferroni, Sidak or Tukey Critical values smaller than Bonferroni, Sidak or Tukey  More powerful in finding differences vs control

Comparison vs Control (Dunnett) (v2) critical value of t= Comparison vs Control (Dunnett) (v2) critical value of t=

Controlling for Multiple Comparisons in Exploratory Analyses Caterina Rosano, Howard J. Aizenstein, Stephanie Studenski, Anne B. Newman. Caterina Rosano, Howard J. Aizenstein, Stephanie Studenski, Anne B. Newman. A Regions-of-Interest Volumetric Analysis of Mobility Limitations in Community-Dwelling Older Adults. Journal of Gerontology: Medical Sciences 2007 A Regions-of-Interest Volumetric Analysis of Mobility Limitations in Community-Dwelling Older Adults. Journal of Gerontology: Medical Sciences 2007

Controlling for Multiple Comparisons in Exploratory Analyses A Regions-of-Interest Volumetric Analysis of Mobility Limitations in Community-Dwelling Older Adults. Journal of Gerontology: Medical Sciences 2007 A Regions-of-Interest Volumetric Analysis of Mobility Limitations in Community-Dwelling Older Adults. Journal of Gerontology: Medical Sciences 2007

Controlling for Multiple Comparisons in Exploratory Analyses

c

Thank you ! Any Questions? Robert Boudreau, PhD Co-Director of Methodology Core PITT-Multidisciplinary Clinical Research Center for Rheumatic and Musculoskeletal Diseases