An error-prone model (EPM) to predict financial aid applicants who are likely to misreport on Basic Educational Opportunity Grant (BEOG) applications was developed, based on interviews conducted with a quality control sample of 1,791 students during 1978-1979. The model was designed to identify corrective methods appropriate for different types of applicants. During interviews, applicants provided documentation regarding the information they provided on the Student Eligibility Report (SER). An analytical method known as sequential search analysis or automatic interaction detection (AID) was used to classify applicants into groups. About 32 percent of the students reported themselves as independent and almost 68 percent as dependent on the SER. The sample included 52.3 percent misreporters: 20.1 percent underclaimers and 28.4 percent overclaimers. Ten dependent and seven independent groups emerged from the model. The 17 groups are discussed, and data for each group are provided on Student Eligibility Index values, assets, taxes, and percent of error. Misreporting patterns of the student groups are analyzed, and edits for underclaimers are briefly discussed. Recommendations for program improvement and for further research are offered, and the research methodology is detailed. (SW)