Why are you interested in structural equation modeling?
Because SEM is an intuitive way of analyzing data (I even prefer to carry out ANOVA like analyses in a SEM framework).
I like that one can model model mean and covariance structure in one and the same statistical model.