Though item bias statistics are widely recommended for use in test development and analysis, problems arise in their interpretation. This research evaluates logistic test models and computer simulation methods for providing a frame of reference for interpreting item bias statistics. Specifically, the intent was to produce simulated sampling distributions of item bias statistics under the no-bias hypothesis, for use in determining cut-off points to provide guidelines for interpreting item bias statistics obtained with actual test data. In this case, potential sex bias was studied in the item responses of 937 Cleveland ninth graders to 75 items from the 1985 Cleveland Reading Competency Test. Results supported the basic data simulation approach used in the study. Real and simulated distributions for three item bias statistics (area between characteristic curves, root mean squared differences between curves, and the Mantel-Haenszel statistic) when bias was not present were very similar. The minor differences found between the distributions had little effect on the interpretation of item bias statistics obtained with actual data. Seven steps for applying the method of computer-simulated baseline statistics in test development settings were outlined. (Author/LPG)