ERIC Number: ED563103
Record Type: Non-Journal
Publication Date: 2009
Abstractor: As Provided
Reference Count: 6
Validity of the SAT® for Predicting First-Year Grades: 2009 SAT Validity Sample. Statistical Report No. 2012-2
Patterson, Brian F.; Mattern, Krista D.
In an effort to continuously monitor the validity of the SAT for predicting first-year college grades, the College Board has continued its multi-year effort to recruit four-year colleges and universities (henceforth, "institutions") to provide data on the cohorts of first-time, first-year students entering in the fall semester beginning with 2006 through 2009. Its goal in doing so is to provide clear evidence for the use of the SAT in college admission. Prior research based on the same data collection effort has demonstrated a strong, linear relationship of the SAT section scores with first-year grade point average (FYGPA) in college across a variety of institutional and student characteristics (Kobrin, Patterson, Shaw, Mattern, & Barbuti, 2008; Mattern, Patterson, Shaw, Kobrin, & Barbuti, 2008; Patterson, Mattern, & Kobrin, 2009; Patterson & Mattern, 2011). This study serves as a replication of prior analyses for the most recent cohort of students: those who graduated from high school in the spring of 2009 and subsequently enrolled in a four-year college in the fall of 2009 The present study examined the extent to which four predictors commonly used in college admission were linearly related to FYGPA; in particular, SAT critical reading (SAT-CR), mathematics (SAT-M), and writing (SAT-W), as well as high school grade point average (HSGPA), were considered. Overall FYGPA correlations were approximately equal for the combination of all three SAT sections and HSGPA (r = 0.54, for both correlations). Combining these four predictors led to the strongest linear relationship with FYGPA (r = 0.62), indicating that the SAT added substantially to predictions that relied solely on HSPGA. Among the three SAT sections, SAT-W tended to exhibit the strongest linear relationship with FYGPA (r = 0.52). In addition, many of these patterns held true across institutional characteristics, such as control (i.e., public or private), size, and selectivity, and across student characteristics, such as gender, racial/ethnic identity, best spoken language, household income, and highest parental education level. Finally, analyses of differential prediction for the student characteristics showed that using the three SAT sections to predict FYGPA tended to result in smaller differential prediction in absolute magnitude than when using HSGPA alone. With the exception of a few student subgroups, the differential prediction of FYGPA was reduced the most when using the combination of SAT sections and HSGPA. The following tables are appended: (1) Institutions Providing First-Year Outcomes Data for the 2009 Cohort; (2) Raw Correlations of SAT and HSGPA with FYGPA by Institutional Characteristics; and (3) Raw Correlation of SAT Scores and HSGPA with FYGPA by Subgroups.
Descriptors: College Entrance Examinations, Test Validity, Prediction, Grades (Scholastic), Predictor Variables, Grade Point Average, Correlation, Institutional Characteristics, School Size, Selective Admission, Student Characteristics, Gender Differences, Racial Differences, Ethnic Groups, Language Usage, Family Income, Parent Background, Educational Attainment, High School Students, College Freshmen, Critical Reading, Reading Skills, Mathematics Skills, Writing Skills
College Board. 250 Vesey Street, New York, NY 10281. Tel: 212-713-8000; e-mail: email@example.com; Web site: http://research.collegeboard.org
Publication Type: Numerical/Quantitative Data; Reports - Research
Education Level: Higher Education; Postsecondary Education; High Schools; Secondary Education
Authoring Institution: College Board
Identifiers - Assessments and Surveys: SAT (College Admission Test)