Abstract: Techniques, methods, systems, and devices for diagnostic assessment of a respondent are described. The techniques for diagnostic assessment may use a classification model to characterize multiple dimensions of the respondent. The classification model may characterize the respondent as having an attribute (e.g., unobserved variable) that is measured by one or more outcomes (e.g., observed variables). The outcomes may be based on responses received from the respondent to items on an assessment. The classification model may also characterize one or more item parameters associated with the items of the assessment.