A brief overview of the examples and source code indicates that if one
desires to use this software in an interactive fashion, one can use
the OSCATS_TYPE_ALG_SIMULATE class (simulate.h / simulate.c) to create
a new class / algorithm. This class registers the "administer" signal.
The implementation of the "administer" function would allow this new
algorithm to perform any custom work (like presenting the item to the
examinee and collecting the response). This way, the interactive
application can make use of the rest of the good work you have
implemented in this library.
I would like to find out if the following strategy would work:
1: Initialize an item bank with items and and their difficulty.
2: Use 1PL model and OSCATS_TYPE_ALG_ESTIMATE
3: Implement an algorithm that implements "administer" signal (similar
to OSCATS_TYPE_ALG_SIMULATE). This function may present the user with
the item and collect the response.
4: At the end of the test, find out the est_theta.
Thanks.