Two questions and a half on Conquest...

[DMarengo]

a. It is said that Conquest can fit 2pl models. Could you provide an example syntax to run a 2pl model in Conquest?

b. Are Conquest EAP scores based on pattern-scoring or  (raw-score) summed-scoring? Why do I get different EAP scores for subjects with the same raw score when fitting a within-item multidimensional model?

[Eveline Gebhardt]

a. model     item+item*step !scoresfree;

b. EAP scores are the are the averages of individuals posterior distributions (from which plausible values are drawn). Raw scores are not sufficient statistics for these like they are for MLE and WLEs

My appologies for the very slow response!