Dear all,
I have 2 IVs, 5 DVs, One mediator, and 5 Control Variables. I want to see whether the means of these variables are significantly different across 4 quarters. I find that both IVs are not significantly different, Some DVs are also not significantly different, others and mediator is significantly different, should I conclude that longitudinal analysis of the relationship between variables is not possible? That is why I will just use the cross-section, but no yearly analysis to analyze predictive relation with regression. Please advise on this approach.
The absence of mean differences across time for a given variable does not mean that the different variables could not be longitudinally related. Therefore, it could still make sense to examine, for example, longitudinal mediation using path analysis or multiple regression.
Thanks Dr. Christian Geiser for coming to my post. This is a burning question for me. I want to analyse the relations based on cross-section, but have longitudinal observations. If I use SMART PLS, I am not sure it assumes any clustering in the pls-algorithm. I used all panel data in smartpls without any clustering/ grouping. This gives me results for cross-sections(of course, not for quarters). To cut the length of analysis, I want to keep it this way if it is correct. But , I want to know on what grounds can I justify that I used data from quarter 1 to quarter 4 for the same individuals and for the same variables, but I pooled the quarters .... or what else can I write to justify if ANOVA is not a justification? Please advise
Asma Malik
It appears from your initial post that you are interested in studying mediation (indirect variable effects). Longitudinal data is much preferred for studying mediation. There are numerous problems associated with the use of cross-sectional data for testing indirect effects. See, for example,
Maxwell, S. E., & Cole, D. A. (2007). Bias in cross-sectional analyses of longitudinal mediation. Psychological Methods, 12(1), 23–44. https://doi.org/10.1037/1082-989X.12.1.23
Article Bias in Cross-Sectional Analyses of Longitudinal Mediation: ...
Since you seem to have longitudinal data available, I would recommend making use of it rather than restricting your analyses to cross-sectional data.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA