Respected Researchers
I ran the panel regression in Stata and got some strange results. The direct effect of one of my IV on DV is negative and insignificant. When i tested the indirect effect (mediation) the sign of the relationship remains negative but it is statistically significant.
in simple word, the direct effect effect is insignificant and negative, indirect effect is significant and negative.
how i can interpret these results.
I think its not a suppressor ( a case in which positive direct and negative indirect relationship occurs).
Negative but statistically insignificant result implies that independent and dependent variables are not associated with eachother.
For mediating effect you can say that when we add some other variables (e.g. mediating variable) in regression analysis, we find negative but statistically significant relation among two variables. This negative but significant result was caused due to the addition of mediating variables.
- Badges
- Science topic
- Similar topics
- Philosophy
- Esthetics
- Interpretation