[ ] Add some fit measures by default? Maybe determined by the simulation study
[ ] Automatically include respondent probabilities
Need to balance "analysis of this data set" vs. "calibrating a model for future use." If we're just calibrating to score future data, then we are probably less interested in the respondent probabilities for the specific data set used to estimate the model.
Add some fit measures by default?