Answer: Multi-informant data can be used to classify people as deviant vs. nondeviant by selecting a criterion appropriate for the user’s aims and for the available data. For example, if the OASR and two OABCLs have been completed for each person, a user can elect to classify people as deviant on a particular scale if 2 of the 3 forms yield scores in the borderline or clinical range on that syndrome. Multi-informant data can also be combined quantitatively by converting raw scores to standard scores within a sample. Most statistical software has options for converting raw scores to standard scores with convenient parameters such as mean = 100, SD = 15. The user can then average the standard scores to create a composite cross-informant score for each person. The composite scores can be analyzed statistically. Because the OASR and OABCL scores for a particular scale are assumed to measure a particular latent variable, structural equation modeling can also be used to analyze cross-informant data.