How to find correlation among dependent variables? If I want to find how strongly a dependent variable is related to another dependent variables in a study, do I make use of multiple regression?
The reason I am asking is because the book mentions that for instance to conduct a standard regression, I will need to input all the independent variables and the corresponding dependent variable into SPSS. Not sure the procedure if I need to find a relationship between two or more dependent variables.
Kindly clarify. Thanks
 A: If you want the simple bivariate correlation(s), just leave out the other variables (the ones you're treating as independent presently) and calculate the correlation(s) between the variables you're interested in. It doesn't necessarily matter that they're dependent variables for your other analyses (unless you want to take those predictive models into account). Bivariate correlations don't require designation of in/dependent variables.
If you want to remove the variance that's explained by the independent variables from your dependent variables, you can use partial (or semipartial, depending on what you're really after) correlations, structural equation modeling, and probably several other methods. It sounds like you're looking for something in that family of analyses, but in case you're actually looking to predict multiple dependent variables from your independent variables, multivariate-regression is the variation on multiple regression that handles that. Structural equation modeling also works for that purpose (I think the SEM gurus would say it subsumes these analyses, technically), and canonical correlation is similar too.
