ABSTRACT
This chapter proposes a panel multitrait-multimethod (MTMM) design in which the repetitions of questions are spread temporally, over different waves of interviewing, in combination with an adapted MTMM model. It considers this to be a design perfectly suited to longitudinal studies. The covariances or correlations between the resulting measurements are then ordered in a matrix that is analyzed with a confirmatory factor model. The purpose of the analysis is to obtain estimates of measurement quality: validity, method effects, and reliability. However, for the model and for the whole class of multiple-indicator MTMM models that they discuss, even more indicators are needed at each point in time in the classical MTMM design. In MTMM models, the improper solutions are very often due to overfactoring, in which case they can easily be solved by restricting the non-significant loadings of one trait or one method factor to zero.
