What could be the possible reasons for getting a negative Moderation when it is expected on the basis of the Literature as a positive moderation effect?
The most simple explanation - and the first thing I'd check - is a sign error in my programming, such as a variable/scale being inverted for some reason (e.g. measuring "emotional stability" vs. "neuroticism").
If that's not it, it there might be any number of substantive reasons: You might want to consider both procedural variables, or perhaps participants representing a different population than most other studies - really any difference between your sample (e.g. age, culture, naivety or non-naievety to the research question) or design (framing, anchoring, motivation) and those generally reported in the literature. For your data, you're probably the best-equipped person to gauge which of those factors could plausibly account for the difference.
Lastly, and perhaps most frustratingly, depending on how robust the postulated effect is , either finding might be spurious.
Thank you Julia Englert for adding your reply! I have been contemplating on possible reasons as you suggested; however, couldn't reach upon any justification. Nonetheless, much appreciation for your response!
I shall go through the data again.
So your method/model is not right. You have to correct it because it sounds all presumptions of the model are not met.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA