In your experience, is it necessary to collect the same data in both measuring points in longitudinal studies and if not, what criteria or reasoning do you follow? For example, if personality traits are considered relatively stable (and present predictors in our models), could we omit them in the second wave?