Hello everyone:
What does it mean when a variable is significant in one step but then in the next step it's no longer significant? For example, in this article (https://cyberleninka.org/article/n/1102972.pdf) neuroticism is a significant predictor of well-being in the first step when the Big Five personality traits are introduced, but in the second step neuroticism is no longer a significant predictor when new variables are added. The authors in the discussion seem to interpret step 1 and step 2 of the model independently (concluding that neuroticism is a predictor of well-being even though it is not in the last step of the model). Would this be an acceptable way to interpret a hierarchical regression?
Thank you!
Yes and no. When a regression coefficient drops notably (which happens when it drops from significant to NS), then it means that the true relationship is not carried (uniquely) by neuroticism. If the new variables are significant, then they are the ones underlying the relationship. It does not mean that neuroticism can't be used to predict well-being. But it does mean that neuroticism per se is not having a direct effect. Examining the regression table suggests to me that the effects observed in Step 1 were 'really' due to neuroticism's association with self-confidence and helplessness. However, without a full correlation table, path analysis, or examination of the neuroticism sub-scales, I can't argue that definitively.
If I understood your question correctly, interpretation would be that in the model where neuroticism stops being significant, implies that whatever variables were introduced, they accounted for most of the variance (or at least the variance that neuroticism had previously accounted for) in the DV. This would suggest that these 2 new variables are more important in their relationship to the DV then neuroticism alone. More analysis would need to be done to know for sure (SEM for example would give direct and indirect effects of variables).
- Badges
- Science topic
- Similar topics
- Philosophy
- Esthetics
- Interpretation