A colleague has proposed this analysis which seems to me helpful and
innovative but completely outside my experience.
He has fitted similar SEM models to different sets of cases; at this stage
I am not certain whether the sets are exclusive or may overlap. He has
recorded the parameters from each fit; again, at this stage I have not
checked whether all the parameters are strictly beta weights or if some
are bidirectional correlations. The proposed analysis is to use the set
of parameter vectors (transposed) as data for a factor analysis (principal
components).
Does this make sense? My feeling is that the parameters are random
variates with a covariance structure, and the PCA may demonstrate which of
the sets of cases support similar models. But I can't find any literature
that describes such a procedure. Comments or hints invited, please.
R. Allan Reese Email: [log in to unmask]
Associate Manager GRI Direct voice: +44 1482 466845
Graduate School Voice messages: +44 1482 466844
Hull University, Hull HU6 7RX, UK. Fax: +44 1482 466436
====================================================================
|