Hello Fernando,

In the first example you give, do you mean correlation of scores from a measure of organizational climate with scores from a measure of job satisfaction for: (a) workers; (b) managers; (c) members of the board of directors; (d) perceptions of each from clients or customers of the organization; (e) multiple levels; or (f) something else altogether?

If you were interested say, in just workers, then ordinary correlation measures would likely address your question. Of course, if each measure yields multiple scores, or is based on multiple responses, there may be a need to evaluate the dimensionality of each measure (perhaps by using factor analysis).

If you're interested in trying to assess a model to explain possible differences in one or more of the measures then yes, perhaps SEM could be applied.

The second example, "external customer," isn't clear to me as to either what that might represent as a variable, or what it is you wish to estimate as being associated with it. Could you elaborate?

Good luck with your work.

Hello David

First, I thanked your response.

Well, my aim is to determine if there is a relationship between Organizational climate (OC), job satisfaction (JS), and external user satisfaction (EUS). But I have a problem. I think so, how I could relate organizational climate, and job satisfaction with the satisfaction of the external user. The relationship between OC and JS could be easier to analyze but add another variable (EUS). This variable belongs to another population, so what test could be helpful to find any relationship between them?

Finally, I have two populations and three variables. This is the reason I think to apply SEM.

Thanks.