PSYC 3301 Lecture Notes - Lecture 1: Repeated Measures Design, Random Assignment
Document Summary
Dependent groups designs & the paired samples t-test. Exactly the same participants in all treatment groups. Reduces error variance by holding individual difference/characteristics constant across all treatment groups. Deals with people dropping out of the study in a systematic way. An approximation of the repeated measures design. Two separate samples, but each individual in one sample is matched one-to-one with an individual in the other sample. Matched on one or more variables that are considered especially important for the experiment. A researcher would like to compare two treatment conditions using a sample of 30. A repeated measures test usually is more likely to detect a real treatment effect than an. May not know what to match on. May not have a valid/reliable way to assess matching variable. If there are several matching variables, may be impossible to do. In general, dependent groups designs have 2 sets of scores. One sample is related directly to the other sample.
