Training Issue -- Outcome Testing. Most participants reported problems reviewing match results in order to validate a current linkage model or improve a linkage model for use in future matches. This process is both more automatic and more accurate in Version 8.0. Dialogs for analyzing comparisons with tolerances (Tolerances) and comparisons with dependent outcomes (Outcomes) have been redesigned so that it should be easier to conduct tests and review test results. The outcome test examples show that AccLoc (accident location) and StdDate (accident date) have significantly dependent comparison outcomes on Unmatched pairs (p < 0.01), requiring that Agree X be set to 0.91 to correct for the dependency. On the other hand, StdDobD (date of birth day) and StdDobM (date of birth month) have significantly dependent comparison outcomes on Matched pairs (p < 0.01), requiring Agree X to be set to 0.99 and Disagree X to 0.63 to correct for the dependency.