| Abstract: | This article is concerned with an approach to disconfounding effects in repeated measure designs with multiple groups. An extension of latent curve analysis (MEREDITH and TISAK, 1990; RAO, 1958, 1965) is described. The method permits estimation and testing of intervention effects separately from temporal effects. It is based on restricted factor analysis of individual change profiles and utilizes basis curves representing group patterns of change. The approach is illustrated on data from a two-group intervention study. |
