If you have any feedback for the author, E-mail me
REQUIRED TEXT: Experimental Methodology eighth edition by Larry B. Christensen, Allyn and Bacon, 2001 ISBN #0-205-30832-5 ------------------------------------------------------------ Chapter 9 Experimental Research Design Faulty Design: One-Group After-Only: one group of participants experiences the IV and is measured on the DV only afterward – almost no value One-Group Before & After(Pre- & Post-test): not totally worthless but differential rates of maturation, control of environmental variables Nonequivalent posttest-only: treatment group compared with a non-matched non -treatment group. must assign randomly True Design: Adequately tests hypotheses Generalizable beyond actual participants Controls extraneous variables Control group of similar individuals as experimental is a source of comparison and controls rival hypotheses i.e. history and maturation Advantages to Pretesting Participants: Matching increases sensitivity Is there room before ceiling effect? Ascertain participant's initial position on a variable before treatment Insure comparability of Exp and Con groups Show evidence of change when compared to posttest But pretesting costs time and money and sensitizes participants to IV True Designs: Between-Participants After-Only: Participants are randomly chosen and randomly assigned to exp & con groups or matched across groups. Simple: more than one level of ONE IV Factorial: more than one level of TWO or more IVs – Main Effects and Interactions, p249 e.g. p250 Desire to interact measured from nondepressed, dysphoric or depressed individuals receiving positive, neutral & negative feedback. Advantages: test more than one hypothesis, control confounding variables, greater precision, study interactions. But increases # of participants, difficulty of presenting combinations of levels, difficult to interpret higher-order interactions => 3 Iv's Within-Participants After-Only design: Repeated-Measures – participants serve as their own controls, fewer participants, but confounding influence of sequencing effect so must counterbalance. Mixed Model: Combine between factorial, e.g. gender, IQ, depressed or not by type of feedback within the same participants. Before-After designs: controls for intragroup history, maturation, instrumentation, regression to the mean, and selection bias (random assign to groups). Between Designs: Studying discrete levels of IV when previous levels of IV or re- measurements of DV would confound results (not the progression of increments in the DV after various levels of IV, e.g. learning Within Designs: Increases sensitivity and preferred when neither sequencing nor carry-over effects will confound results. ------------------------------------------------------------ My office is in room 116C of HGH (826-4571) and I invite you to consult with me on any matter during my office hours or any time that is agreeable to you and me.Go back to the beginning
Copyright © 2002, Dr. John M. Morgan, All rights
reserved -
This page last edited Nov 25, 2002
If you have any feedback for the author, E-mail
me