Examining the Stability of SAT Predictive Relationships Across Cohorts and Over Time - Key Findings

Sep 14, 2026

Study investigates whether national SAT validity findings generalize beyond the 2017 entering college cohort and whether SAT scores predict longer-term outcomes beyond first-year GPA and second-year retention. Findings show SAT validity for predicting first-year GPA remains stable and strong across 2017 and 2018 cohorts, with correlations remaining similar. SAT also provides predictive value comparable to first-year outcomes for cumulative GPA through the second year and retention to the third year, supporting broader institutional use in admissions, placement, scholarships/honors, instructional support, and advising.

LanguageEnglish

CategoryExam

FormatPDF

Pages4

Reading Time10 min

AccessFree to download

Key Takeaways

Purpose

Replicating SAT Validity Evidence: 2018 Entering College Cohort

Sample

Results

SAT Relationships with Longer-Term College Outcomes: 2017 Entering College Cohort

June 2021 Examining the Stability of SAT Predictive Relationships Across Cohorts and Over Time Jessica P. Marini, Paul A. Westrick, and Emily J. Shaw Key Takeaways • The SAT remains a stable and strong predictor of college success. o Correlations between SAT scores and firstyear GPA in the 2018 entering college cohort are strong and consistent with those in the 2017 entering college cohort. o SAT scores are as predictive of cumulative GPA through second-year of college and retention to the third-year, as they are of firstyear GPA and retention to the second-year. • These study results allow institutions to more confidently use SAT scores in admissions, placement, scholarship/honors program, instructional support, and advising decisions on campus , as we know SAT scores predict college success beyond the first-year. Purpose The purpose of this study is to understand how well the national SAT Validity Study results (Westrick et al. , 2019) , which is based on the entering class of 2017, generalize to other cohorts of students and how well the SAT predicts longer term outcomes beyond first-year grade point average (FYGPA) and second-year retention. Results show that the validity of the SAT for predicting FYGPA remains stable and strong, and that the SAT is as predictive of the longer-term college outcomes (cumulative GPA through second-year and retention to thirdyear) analyzed in this study as it is of first-year college outcomes. Findings support the continued use of SAT scores in understanding student readiness for college and for making more informed decisions related to admissions, placement, scholarship/honors program, instructional support, and advising on campus. Replicating SAT Validity Evidence: 2018 Entering College Cohort Sample We collected data on 237,883 first-year students attending 146 four-year institutions of varying size, selectivity, control (public/private) and region of the country. The students in the sample had to have SAT scores, a self-reported high school grade point average (HSGPA), and an FYGPA provided by their institution. Results We were interested in comparing correlational analyses of SAT scores and first-year college performance from the 2017 entering college cohort – the first to have taken the redesigned SAT, with the results from another cohort of students, namely the 2018 entering college cohort. Table 1 below shows that all correlations increased slightly from the 2017 entering college cohort to the 2018 entering college cohort. In particular, we can see that the SAT maintains its strong predictive validity across cohorts (r= .54 in 2018 and .51 in 2017) . Table 1. Corrected (Raw) Correlations of Predictors with FYGPA | Predictors(s) | 2018 Cohort
Correlations | 2017 Cohort
Correlations | | --- | --- | --- | | SAT, HSGPA | .64 ( .43) | .61 ( .42) | | HSGPA | .57 ( .34) | .53 ( .33) | | SAT | .54 ( .33) | .51 ( .32) | | SAT ERW | .52 ( .30) | .49 ( .29) | | SAT Math | .51 ( .28) | .47 ( .27) | Note. n = 237,883 . References to "SAT" on its own include SAT ERW and SAT Math sections. To better visualize the information added by the SAT above HSGPA for understanding FYGPA within the 2018 entering cohort, we can examine Figure 1 below. Controlling for HSGPA, student performance in college increases by SAT score band. For example, among students with an “A” HSGPA, students with SAT Total scores between 600 and 790 had a mean FYGPA of 2 .34, but students with SAT Total scores between 1400 and 1600 had a mean FYGPA of 3 .57, more than a full letter grade higher than the students with the lower scores but in the same HSGPA group. Figure 1. Mean FYGPA by HSGPA and SAT Total Score Bands Note. Results are reported for categories with at least 15 students. To further demonstrate the added informational value of SAT scores above HSGPA in predicting college performance, Figure 2 shows the probability of a student earning a FYGPA of 2.50 or higher in colleg

FAQs

What is the main purpose of the study?

To determine how well the national SAT Validity Study results generalize to other student cohorts and how effectively the SAT predicts longer-term outcomes beyond first-year GPA and second-year retention.

What does the study find about SAT validity across cohorts?

SAT validity for predicting first-year GPA remains stable and strong across the 2017 and 2018 entering college cohorts, with correlations staying similar.

How does SAT performance relate to longer-term outcomes like retention?

SAT remains useful for understanding retention to the second year above and beyond high school GPA, showing positive relationships when controlling for HSGPA.

Sep 14, 2026

Uploaded by Ava Thompson