My advisor (who doesn't have a thorough knowledge of SEM) has just reviewed my dissertation and has some concerns with my analyses.
He questions whether I should be using more than one scale to represent my manifest variables since their anchor points are different. For example, my latent variable is work outcomes. The indicators predicting this are job sat, org. commit, engagement, and intent to turnover. Two scales have subscales and the anchor points are different for each of them. Does this pose a problem? He thinks I have too many variables representing my latent variable.