One of my variables has a low Cronbach's Alpha. I continued to run an planned contrast to test my hypothesis and the results were significant.
I understand that the results should be interpreted lightly. My question is however, what does it mean if the Cronbach's Alpha is low yet significant results have been found?
Best Answer
Cronbach's alpha measures internal consistency. Low alpha indicates that the variable comprises unrelated, or 'weakly related', items. How low is Cronbach's alpha for your variable? In cases of very low internal consistency, a 'significant result' (what did you test?) may not be interpretable as reflecting a significant effect over all variable items, but only over a specific fraction of them. Furthermore, if items are unrelated a 'positive' change in some of them may be masked by a 'negative' change in unrelated others.