Matched pairs designs
Students may not readily see any problems with the matched pairs design. Table 12.2 can serve as a basis for discussing how the gains in power from using the matched design come with the following potential costs:
- one has to find an appropriate matching variable, test participants on a measure of that variable, pair them up on scores on that measure
- the matching task may tip participants off to what the experiment is about
- results may only generalize to those who have been pretested then received treatment.
Note also that students may have the following misconceptions:
- matched pairs designs have internal validity because of matching and
- matched pairs designs can be analyzed via the independent groups t test
Your discussion of matched pairs designs can lead into a discussion within-subject designs. Table 12.3 (p. 379)can be used to help make this transition.
Back to Chapter 12 Main Menu