NotesFAQContact Us
Search Tips
Back to results
ERIC Number: ED558092
Record Type: Non-Journal
Publication Date: 2014
Pages: 20
Abstractor: ERIC
AP® Potential Predicted by PSAT/NMSQT® Scores Using Logistic Regression. Statistical Report 2014-1
Zhang, Xiuyuan; Patel, Priyank; Ewing, Maureen
College Board
AP Potential™ is an educational guidance tool that uses PSAT/NMSQT® scores to identify students who have the potential to do well on one or more Advanced Placement® (AP®) Exams. Students identified as having AP potential, perhaps students who would not have been otherwise identified, should consider enrolling in the corresponding AP course if they have the interest and motivation to do so. To date, several studies have been conducted to evaluate the validity of using PSAT/NMSQT scores to predict success on AP Exams (Camara & Millsap, 1998; Ewing, Camara, & Millsap, 2006; Ewing, Camara, Millsap & Milewski, 2007). Results have shown that PSAT/NMSQT scores were moderately to strongly correlated with scores on most AP Exams and that self-reported high school grade point average (HSGPA), grades in related subjects, and total years of study in related subjects accounted for very little additional variance in AP Exam performance once PSAT/NMSQT performance was taken into account. The methodology used to develop prior versions of AP Potential was empirical; that is, it involved pooling test data across schools and computing expectancy tables showing the percentage of test-takers earning passing scores on AP Exams at various levels of PSAT/NMSQT performance. The study reported here has two purposes. The first is to switch from the empirical approach of building expectancy tables to the use of logistic regression models. Logistic regression models provide more flexibility when evaluating and additional variables if desired and can be to determine the PSAT/NMSQT score associated with a particular probability of success (e.g., 50%, 70%). The second purpose for the study is to update AP Potential predictions based on more current PSAT/NMSQT and AP score data. Drawn from 6,091,075 high school sophomores and juniors who took the PSAT/NMSQT in 2007 and 2008, the sample included 1,835,806 students who took the PSAT/NMSQT in October 2007 and October 2008 and then completed one or more AP Exams 19 months later in May 2009 and May 2010, respectively. PSAT/NMSQT scores were found once again to be moderately to strongly correlate with scores on AP Exams with the exception of AP world language and AP Studio Art exams. Using logistic regression models, the PSAT/NMSQT cut scores associated with different probabilities of success on each AP Exam were updated. Users can easily locate the minimum PSAT/NMSQT score for an individual student given an expected chance of success on a particular AP Exam, or derive the chance of success on different AP Exams given a student's specific PSAT/NMSQT scores.
College Board. 250 Vesey Street, New York, NY 10281. Tel: 212-713-8000; e-mail:; Web site:
Publication Type: Reports - Research; Numerical/Quantitative Data
Education Level: High Schools; Secondary Education; Higher Education; Postsecondary Education
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: College Board
Identifiers - Assessments and Surveys: National Merit Scholarship Qualifying Test; SAT (College Admission Test)