Analyzing Within-Subjects Experiments

by Cotton; John W.

• ISBN13: 9780805828047
• ISBN10: 0805828044
• Edition: 1st
• Format: Hardcover
• Copyright: 1998-01-01
• Publisher: Psychology Pres

Summary

Most behavioral scientists know two important concepts -- how to analyze continuous data from randomly assigned treatment groups of subjects and how to assess practice effects for a single group of subjects given a constant treatment at each of several stages of practice. However, except in the case of the repeated measures Latin square design, researchers are not facile in analyzing data from different subjects receiving different treatments at various times in an experiment. This book helps fill the void.

Table of Contents

PREFACE	xv
Chapter 1 AN ORIENTATION TO WITHIN-SUBJECTS DESIGNS	1	(11)
Introduction	1	(1)
Types of Within-Subjects Designs	2	(3)
Background on Analysis of Within-Subjects Designs	5	(3)
Tools of the Present Book: ANOVA and the SAS(R) GLM Statistical Package	8	(1)
The Idea of the General Linear Model	9	(3)
Chapter 2 TWO-WAY EXPERIMENTAL PLANS: SPLITPLOT AND RANDOMIZED BLOCK DESIGNS	12	(35)
Introduction to Models With Practice or Treatment Effects, but Not Both	12	(1)
Analyzing One-Group Split-Plot Data	12	(20)
Analyzing Multigroup Data Without Time-Related Effects	32	(12)
Method C: Randomization Tests Appropriate in Randomized Block Designs Without Random Selection of Subjects	44	(1)
Things to Remember	45	(1)
Exercises	46	(1)
Chapter 3 ANALYZING DATA FROM A RANDOMIZED BLOCK DESIGN EXPERIMENT THAT MAY EXHIBIT TIME-RELATED EFFECTS	47	(32)
Introduction	47	(1)
Crossover Design Approach to Mental Rotation Data	47	(28)
Other Tests of Experimental Effects When Sphericity Does Not Hold	75	(2)
Things to Remember	77	(1)
Exercises	78	(1)
Chapter 4 INTERPRETING ESTIMABILITY INFORMATION AND REPORTED ESTIMATES OF PARAMETERS IN SAS(R) GLM PROGRAMS	79	(13)
Introduction	79	(1)
Interpreting Estimability Tables in SAS(R) GLM Output	80	(5)
"Solution" Versus "Estimate" Commands	85	(4)
More About Expected Mean Squares	89	(1)
Things to Remember	90	(1)
Exercises	91	(1)
Chapter 5 ANALYZING DATA FROM WITHIN-SUBJECT FACTORIAL DESIGNS, TAKING INTO ACCOUNT STAGE-OF-PRACTICE EFFECTS	92	(42)
Introduction	92	(1)
Analyzing a 2 X 2 Within-Subjects Example	93	(21)
Analyzing a 2 X 2 Within-Subjects Example Without Complete Balance	114	(14)
Judging If a Given Within-Subjects Factorial Design Will Lead to Interpretable Results	128	(3)
Things to Remember	131	(2)
Exercises	133	(1)
Chapter 6 PRETEST-POSTTEST CONTROL GROUP DESIGNS: COMPARING DIFFERENT TREATMENT GROUPS AFTER PRETESTING	134	(29)
Introduction	134	(1)
Method A: Analyzing Data From a Simple Design--AA Versus AB Sequences	135	(4)
Method B: Comparing A(3)A(3) and A(3) B(3) Sequences	139	(6)
Generalizing This Design	145	(2)
Exercises	147	(1)
Optional Theory, Examples, and Exercises Involving BMDP(TM), SPSS(R), and SYSTAT(TM)	147	(14)
Things to Remember	161	(2)
Chapter 7 SWITCHING TREATMENTS IN BLOCKS: A(m)A(m), A(m)B(m), B(m)A(m), OR B(m)B(m) PATTERNS WITH m STAGES	163	(21)
Introduction	163	(1)
Method A: Analyzing Data From an AA Versus AB Versus BA Versus BB Experiment	164	(9)
Method B: Analyzing Data From Three-Stage Blocks: A(3)A(3), A(3)B(3), B(3)A(3), and B(3)B(3) Sequences	173	(9)
Final Thoughts	182	(1)
Things to Remember	182	(1)
Exercises	183	(1)
Chapter 8 ANALYZING DATA FROM VARIANTS OF ALTERNATING TREATMENT DESIGNS	184	(29)
Introduction	184	(1)
Analyzing Data From an ABAB Versus BABA Reversal Design	185	(16)
Analyzing Data From a More General Design: A(m1)B(m2)A(m3)B(m4) Versus B(m1)A(m2)B(m3)A(m4)	201	(9)
Final Thoughts	210	(1)
Things to Remember	211	(1)
Exercises	212	(1)
Chapter 9 DATA ANALYSIS FOR MULTIPLE-BASELINE DESIGNS	213	(17)
Introduction	213	(1)
Method A: Analyzing Data From a Multiple-Baseline Experiment Interpreted by a Crossover Design Model	214	(5)
Method B: Computing Efficiencies and Variances of Estimators (in XXX Units) for the Current Example	219	(2)
Related Multiple-Baseline Designs	221	(7)
Final Thoughts	228	(1)
Things to Remember	228	(1)
Exercises	229	(1)
Chapter 10 DATA ANALYSIS FOR DUAL-BALANCED MULTIPLE-BASELINE DESIGNS	230	(14)
Introduction	230	(2)
A Hypothetical Dual-Balanced Multiple-Baseline Experiment on Elation Versus Depression Effects	232	(7)
Related Designs	239	(3)
Final Thoughts	242	(1)
Things to Remember	242	(1)
Exercises	243	(1)
Chapter 11 BLOCK-RANDOMIZATION EXPERIMENTS WITH MULTIPLE TREATMENTS, EACH ONCE PER BLOCK OF STAGES	244	(16)
Introduction	244	(1)
Method A: Data Analysis Assuming Classical Carryover Effects	244	(8)
Method B1: Data Analysis Assuming Transitional Carryover Effects	252	(2)
Method B2: Finding Direct Plus Transitional Carryover Effects	254	(1)
Final Thoughts	255	(2)
Things to Remember	257	(1)
Exercises	257	(3)
Chapter 12 ANALYZING DATA FROM AN ABBA VERSUS BAAB COUNTERBALANCED DESIGN	260	(16)
Introduction	260	(2)
Method A: Data Analysis Assuming Linear Stage Effects	262	(7)
Method B: Data With Possibly Nonlinear Stage Effects	269	(5)
Final Thoughts	274	(1)
Things to Remember	274	(1)
Exercises	275	(1)
Chapter 13 SHOULD "OPTIMAL DESIGNS" BE PREFERRED IN BEHAVIOR SCIENCE CROSSOVER EXPERIMENTS?	276	(21)
Introduction	276	(1)
Optimal Design for Two-Treatment Experiments	277	(8)
Optimal Design With More Than Two Treatments	285	(2)
Possible Difficulties With Some Optimal Designs	287	(4)
Final Thoughts: Comparison of Efficiencies of Different Classes of Possible Psychological Designs	291	(3)
Things to Remember	294	(2)
Exercises	296	(1)
Appendix 1 A LITTLE ABOUT MATRICES AND VECTORS	297	(11)
Introduction	297	(1)
Some Matrix Operations and Properties	297	(11)
Appendix 2 USING THE GAUSS MATRIX PROGRAMMING LANGUAGE	308	(11)
A GAUSS(TM) Program for Estimating Parameters for ABBA Versus BAAB Design Data With Linear Slope	308	(6)
A GAUSS(TM) Program for Finding Contrasts for ABBA Versus BAAB Design Data With Linear Slope	314	(1)
Relation to SAS(R) Definitions of Contrasts Based on Table 12.2	315	(2)
Finding Contrasts for Data Without Linear Slope	317	(1)
Use of GAUSS(TM) in Efficiency Estimation	318	(1)
GAUSS(TM) Is Just One Option	318	(1)
REFERENCES	319	(8)
AUTHOR INDEX	327	(4)
SUBJECT INDEX	331