When it comes to hierarchical multiple regression models, choosing the right approach can be a real head-scratcher. I recently came across a Reddit post that got me thinking about this very issue. The poster had run a hierarchical multiple regression with three blocks: demographic variables, empathy, and reflective functioning (RFQ). The RFQ scale, in particular, presented some challenges due to its structural dependency between two dimensions.
The poster tried two approaches for Block 3: entering both RFQ dimensions simultaneously and entering each dimension separately. The results were interesting, with the simultaneous approach yielding fewer significant effects, while the separate models showed more significant results.
This got me thinking: when it comes to reporting the results, which approach is best? Should we prioritize comprehensiveness or significance? Do we present the model with both RFQ dimensions together, or do we opt for separate models with more significant effects? Or do we include both and discuss the differences?
As researchers, we’ve all been there – torn between presenting a more comprehensive picture and highlighting the most significant findings. Ultimately, the choice depends on our research questions, the nature of our data, and the story we want to tell. But it’s essential to be aware of these considerations and to think critically about how we present our results.
So, what do you think? How do you navigate these kinds of decisions in your own research?