ABSTRACT
We begin by presenting the topic of replication, which uses multiple samples. In our rst example, the validation of a theoretical conrmatory factor model using two samples of data is presented. Cross validation is presented next, where a larger sample is randomly split into two subsamples. Then, we present the basics of how to determine the stability of parameter estimates using the bootstrap method. Ideally, a researcher should seek model validation with additional samples of data (replication). The other methods are not as rigorous, but in the absence of replication, provide evidence of model validity-that is, the viability of the theoretical framework suggested by the measurement and/or structural models.
