How to analyse pre and post test data when the outcome measure is Yes/No data? I am doing  research on 20 participants and the participants will work on a lab module.
There will be a pretest before the lab module and post test after the lab module on concepts related to flowchart, pseudo code , etc.,
The possible outcomes for the question is Yes/No.
Which statistical method should i use to analyze the pre - post test yes/no data?
 A: *

*If you have many yes-no items that form a scale, then you'd probably do a paired samples t-test.

*If you have small number of yes-no items that form a scale, then you might use a non-parametric test like Wilcoxon T Test.

*If you have only a single item, then you might use something like a McNemar test.

A: Control groups can be classified into active and passive control.
Active control group gets pseudo lab module which is assumed to have no effect on the test score, and passive control group gets nothing, it gets just pretest and posttest, no intervention. From active control group, you can see if there's placebo effect.
Given that your scores are nearly continuous and that you have either one of control groups, you could do ANCOVA to see if there's significant change of score compared to control group.(Given that regression slope of pretest on posttest is same. It's been asserted through several papers that repeated-ANOVA is not so much recommendable because of power problem.)
If you really believe that the score doesn't change, you might think it's no use to
have control group. But even if people gets nothing between the two tests, they can't be
the same score. It means the reliability of the test is less than 1. So, without control group, you wouldn't be able to say that the score change is due to less-than-1-reliability or lab-module effects.
