I've recently started learning the Confirmatory Factor Analysis (CFA) method.
The textbook(Timothy A. Brown) shows some examples of model constructs that are either underidentified, just-identified or overidentified. I understand these concepts as whether the number of parameters I'm trying to estimate through CFA is larger-equal-or smaller than the number of data points I've acquired via the covariance matrix.
I thought this concept of overidentified model was equivalent to the concept of overdetermined model concept I picked up from linear algebra and multiple regression classes, so just to clarify I asked my professor whether multiple regression could be considered as "overidentified" model, but he told me it was just-identified. I know that multiple regression in matrix form of Ax=b usually has more rows at the A matrix than the rows in x, which is the coefficients we're trying to estimate via least-squares method.
So I'm inclined to believe that "~identified" and "~determined" are two different concepts. How are these concepts different, given that both compare number of parameters to estimate and the number of data points we have?