Dear all,
I am a bachelor students investigating the moderating effect of positive and negative relationship quality between siblings on the relationship between number of siblings and Theory of Mind understanding. I wondered that how should I report these findings with following consideration.
1. Should I report each model of the printout? I have 3 models in each moderation analysis. Do I have to repeatedly say three of the models are not significant one by one?
2. After saying the models are not significant, should I report the non-significant coefficients of each variables one by one?
Thanks a lot
A good start may be to look up APA style manual (6th edition) and its advice on how to report regression models. It doesn't give a clear instruction on how to report every possible outcome, so you have to decide what would be the most informative way of presenting your results to a reader.
Significance based of p values is not the only (or the best) way to make claims about your results. In regression, you get the coefficients, which in themselves are effect sizes. Ask the big question: what kind of an effect size would be meaningful for me in this study? - and present your results with this in mind. If nothing is meaningful, you can make some generalised phrases, like "all Betas < X.X", or if you stick to p values, all p < .05, etc.
Also, look up some papers which used regression models and see how it was done there. Read the results and ask yourself whether they provide enough information for you to understand what happened in the data. If so, use a similar way to report your results. It's all about reporting the analysis and its results in an informative and digestible way.
- Badges
- Science topic
- Similar topics
- Linguistics
- Philology
- Language
- English