Statistical Tests
t-test: This statistical test determines whether there is a statistically significant difference between the means of 2 groups and uses numerical DV.
Chi-Square: This statistical test determines whether there is a statistically significant difference between the outcomes of 2 or more groups and uses categorical DV.
Regression: This statistical test determines whether there is a statistically significant difference between the means of 2 or more groups and uses numerical DV, controlling for additional independent variables.
Tests of Moderation
Moderator: A variable that tests the differential effect of the intervention based on a third variable, such as diagnostic group, timepoint, etc. Tests for whom an intervention works best.
Interaction: The interaction term is part of the regression analysis. This term represents the moderator.
Denoted by anything that represents multiplication ( * X or x).
- Interpretation if the interaction term is significant: There is a differential effect of the intervention based on the moderator.
- Interpretation if the interaction term is not significant: There is no differential effect of the intervention based on the moderator.
Tests of Mediation
Mediator: A variable that explains the relationship between the independent variable and the dependent variable. (Think: Why?)
To test if there is a statistically significant mediator, we would complete the following steps. First, we would test for the association between the IV and the mediator (a). Then, we would test for the association between the mediator and the DV (b). Next, we would test of the association between the IV and the DV (c). Finally, we would test the mediation model and compare whether c is different from c’. If there is a mediator, we would expect c’ to be closer to zero than c.
