Summary : The use of contrasts in data analysis : How to test specific hypotheses in psychological research Omnibus tests with multiple degrees of freedom test relatively vague questions whereas most hypotheses that we derive from our theoretical models make quite precise predictions. In order to respond to this level of precision, we suggest to test specific contrasts rather than to perform omnibus tests. Two conditions have to be satisfied before one can claim that a given contrast is a parsimonious description of the observed means : 1/ the contrast itself has to explain a significant part of the variance, and 2/ if one statistically controls for the effects of this contrast, the residual variance has to be non-significant. Using specific examples, this article presents the analyses that allow researchers to test these two conditions in different experimental designs.
No takes yet. Share an insight, caveat, or question.
Bräuer et al. (2005) studied this question.
Synapse has enriched 3 closely related papers on similar clinical questions. Consider them for comparative context: