Dear All,

I have a problem in one of my longitudinal studies. I have investigated relationships between a few variables, and found some confusing results:

1) In cross-sectional analyses (correlations) variables X and Y are positively correlated (r = .40 - .50);

2) Changes in these two variables across time are also positively correlated to each other (r = .25), what means increase in variable X is related to an increase in variable Y.

3) But... in Cross-Lagged Panel Model analyses (including various variables, not only these two), X in T1 predicts lower Y in T2.

Do you have any ideas about the possible statistical explanation? Other results of CLPM are in line with expectations and findings from cross-sectional analyses.

Kind regards,

Kamil