NotesFAQContact Us
Search Tips
ERIC Number: ED457714
Record Type: Non-Journal
Publication Date: 2001
Pages: 26
Abstractor: N/A
Reference Count: N/A
Predicting Student Performances at a Minority Professional School. AIR 2001 Annual Forum Paper.
Chen, Chau-Kuang; Campbell, Vickie C.; Suleiman, Ahmad
This study attempted to build the best fitting prediction model, with high predictive validity, for success at a minority professional school. The resulting model could be used to document the effectiveness of academic programs and applicant screening. Data for 216 medical students from the college were used to evaluate the usefulness of the United States Medical Licensure Examination (USMLE) Step 1 pass status and test scores for this purpose. The USMLE models were constructed through the application of logistic and linear regression models. These models appeared reasonable and workable for the college because the significant predictors, the Medical College Admission Test (MCAT) scores, medical school freshman grade point average, sophomore course performance, and financial aid work-study dollars, were identified and included in the prediction equation. The measure of model goodness of fit and the overall prediction accuracy of the USMLE mode were reasonably high, and the assessment of the underlying assumptions of linear regression showed that linearity, normality, and independence were not violated. Ranking the predicted USMLE Step 1 scores and pass/fail status for prospective test takers, the administrators of the medical school could identify a small group of potential at-risk students and enable them to participate in mandatory board review or tutorial programs. Medical students themselves could use the prediction results to determine the optimum time to take the licensure examinations. (Contains 4 tables and 11 references.) (Author/SLD)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A
Identifiers - Assessments and Surveys: United States Medical Licensing Examination