For the same dataset and for the same dependent variable, I find Slope of overall trajectory is not significantly different from zero in latent growth models (suggesting firm performance DOES NOT improve over time), whereas I find some independent variables or their interactions are significant in traditional panel techniques like fixed effects, random effects (or mixed effects) (suggesting firm performance DOES improve).
I know these are different estimation techniques and explain change in firm performance in different ways. However, I am somehow confused with the contradictory implications on whether firm performance improves or not.
I bought the book <regression>. Unfortunately, it seems latent growth models are not covered. But I am still glad that I got the book at hand. I have also read many other books on latent growth models but fail to find materials that could help make sense of the contradictory implications from these two different sets of techniques.
If you could recommend any book or article that bridges the two set of techniques, that would be great.
If you could even comment on the matter, I would really appreciate it! Thanks!
Like xtreg, fe; xtreg, re. These are commands in Stata. They fit fixed effects or random effects models for panel data. I don't think we can conclude that these panel models show firm performance improve OVER TIME. This is why I did not use OVER TIME after "DOES improve". In these panel models, I think we can only conclude an association between independent variables and dependent variable--firm performance. However, it shows that as the independent variables increase, firm performance improves. In the meanwhile, the Slope from latent growth models is not significantly different zero, showing no improvement over time. I have a hard time to make sense of the supposed-to-be inconsistency.
By the way, the book I mentioned in my last post is Regression And Mediation Analysis Using Mplus. I hope that you and your coauthors could write more books on Mplus. The Mplus users guide is good but sometimes not detailed enough. I am really looking forward to reading more books of yours.