forum.smartpls.com

HI
I'm wondering if pls-pm allows for repeated measurement. for example, I have four period of measurements for two indicators (T1->T2->T3->T4). I suppose that the course of the first indictors would be predicted by the course of the second indicator.
Further, how could I estimate the sample size required in such a longitudinal setting?
thanks a lot
carole

We have not seen much research dealing with such data. If you have responses of the same respondents for the different points of time. This kind of analysis should be technically possible. Alternatively, you may compare you model at different points of time and compare the results (e.g., by conducting a PLS multigroup analysis).

Best regards
Christian

you might want to check this paper,
Mittal, V.; Kumar, P.; Tsiros, M. (1999): „Attribute-Level Performance, Satisfaction, and Behavioral Intentions over Time: A Consumption-System Approach“. In: Journal of Marketing. 63 (2), S. 88–101, DOI: 10.2307/1251947.

it's not PLS (from cross reading it (again) i guess it's covariance-based SEM) but the approach might be similar to yours.

hope that helps