Standardized Test Scores Vs Term‑End GPA

Here’s a report on the predictive power and correlation between standardized test scores (SAT/ACT) and term‑end GPA, focusing on the most recent, reliable studies with important caveats about institutional context and contradictory findings.

Term‑End GPA Vs Standardized Test Scores

Ivy‑Plus Colleges (2025 Study)

A March 2025 NBER working paper analyzing Ivy‑Plus institutions found:

• Standardized test scores (SAT/ACT) predict first‑year college GPA with a slope four times higher than high school GPA, controlling for socioeconomic variables [1]
• Note: This finding is specific to elite institutions (8 Ivy League schools plus Chicago, Duke, MIT, Stanford) and may not generalize to other college types [1]

Omitting test score submission correlates with a FY GPA equivalent to SAT ~1307 [2]

Contradictory Evidence – University of Chicago System

Research on broader student populations shows different patterns:

• A study of 55,084 Chicago Public Schools graduates found high school GPAs are five times stronger than ACT scores at predicting college graduation [3]
• “No connection between students’ ACT scores and eventual college graduation” at many high schools [3]

Term‑End GPA Vs Standardized Test Scores Correlation Coefficients Table (2025)

Measure	Predictor	Criterion	Pearson’s r (approx.)	Institution Type
Ivy‑Plus first‑year GPA	SAT/ACT	FY GPA	~0.32–0.35 [1]	Elite institutions only
Ivy‑Plus first‑year GPA	HSGPA	FY GPA	~0.08–0.09 [1]	Elite institutions only
General college population	HSGPA	Graduation	Stronger predictor [3]	Non-elite institutions
General college population	ACT	Graduation	Weaker/no connection [3]	Non-elite institutions

Standardized Test Scores vs GPA: Mixed Evidence

Ivy‑Plus Context (Limited Generalizability)

• Test scores show stronger prediction at elite institutions where grade compression occurs [1]
• Critical Limitation: University of Chicago researcher Elaine Allensworth noted this research studies “students attending Ivy Plus colleges” rather than the general college population, questioning its broader applicability [3]

Broader Institutional Evidence

• Bates College Long-term Study: 20+ year analysis of ~7,000 students found minimal differences between test submitters and non-submitters [4]
  • Graduation rate difference: 0.1%
  • GPA difference: 0.05 points

HSGPA’s Predictive Strength & Grade Inflation

Recent ACT Research Findings

• High school GPA has become a less consistent predictor while ACT Composite scores have shown relative stability [5]
• HSGPA reliability has declined over time due to grade inflation [5]

Grade Inflation Evidence

• High school GPA rose from 3.44 to 3.59 between 2017-2021 [6]
• ACT scores decreased from 22.34 to 21.90 over the same period [6]
• This divergence complicates direct comparisons [6]

Socioeconomic Factors & Predictive Validity

Socioeconomic Controls

• NBER’s Ivy‑Plus study controlled for race, income, gender, and first‑generation status [1]
• Even after controls, test scores retained predictive validity at elite institutions [1]

Expert Analysis

• Economist John Friedman’s research found that “standardized test scores are more predictive than high school grades, even when comparing students of the same family income, race and gender” [7]

Institutional Response

• Dartmouth task force concluded standardized tests better predict first‑year GPA and dropout likelihood at their institution, with grade inflation diminishing GPA’s usefulness [8]

Term‑End GPA vs. Standardized Scores: Comparative Analysis

Metric	Standardized Test Scores	High School GPA	Context
Elite Institution Prediction	Strong (r ≈ 0.32–0.35) [1]	Weak (r ≈ 0.08–0.09) [1]	Ivy-Plus only
General Institution Prediction	Weaker/Variable [3]	Stronger (5x predictive power) [3]	Broader populations
Long-term Success	Minimal difference [4]	Minimal difference [4]	20+ year studies
Robustness to controls	High at elite schools [1]	Variable by context	Institution-dependent
Susceptibility to inflation	Lower (standardized scaling)	High (documented increase) [6]	Both measures affected
Generalizability	Limited to elite contexts [1]	Broader applicability [3]	Critical limitation

What Current Data Reveals?

Key Findings with Important Caveats:

1. Context Matters: Test score predictive power varies dramatically by institutional type [1][3]
   • Stronger at elite institutions with grade compression
   • Weaker or negligible at many non-elite institutions
2. Contradictory Evidence: Multiple large-scale studies show different patterns than the Ivy-Plus research [3][4]
3. Grade Inflation Affects Both Measures: Both HSGPA and test score relationships have changed over time [5][6]
4. Limited Generalizability: Most publicized research focuses on elite institutions, which educate <1% of college students [1]

Conclusion

Current Research Findings:

• At elite institutions: Standardized test scores show stronger predictive power than HSGPA for first-year GPA [1]
• At non-elite institutions: HSGPA often shows stronger predictive power, especially for graduation outcomes [3]
• Long-term studies: Minimal differences between test submitters and non-submitters across various success metrics [4]
• Grade inflation: Affects both measures, complicating historical comparisons [5][6]

Critical Limitations:

• Most research supporting test score superiority focuses on elite institutions with unique characteristics [1]
• Substantial contradictory evidence exists for broader college populations [3]
• Predictive patterns differ for first-year GPA vs. graduation outcomes [1][3]

For Academic Forecasting: Both metrics provide value, but their relative importance depends heavily on institutional context, student population, and specific outcomes measured.

Claims about the universal superiority of either measure are not supported by the full body of evidence [1][3][4].

References

[1] Friedman, J., Sacerdote, B., Staiger, D., & Tine, M. (2025). Standardized Test Scores and Academic Performance at Ivy-Plus Colleges. NBER Working Paper w33570. Retrieved from https://www.nber.org/papers/w33570

[2] Opportunity Insights. (2024). Standardized Test Scores and Academic Performance at Ivy-Plus Colleges. Retrieved from https://opportunityinsights.org/paper/test-scores/

[3] University of Chicago News. (2024). Test scores don’t stack up to GPAs in predicting college success. Retrieved from https://news.uchicago.edu/story/test-scores-dont-stack-gpas-predicting-college-success

[4] Bates College. (2005-2025). More Than Your Score: 20-year Bates College study of optional SATs finds no differences. Retrieved from https://www.bates.edu/admission/more-than-your-score-20-year-bates-college-study-of-optional-sats-finds-no-differences/

[5] ACT Leadership Blog. (September 2024). New Post-COVID Study Finds High School GPA is Becoming a Less Reliable Indicator for First-Year College GPA. Retrieved from https://leadershipblog.act.org/2024/09/gpa-test-score-predictive-validity.html

[6] Education Week. (October 2024). Grades and Standardized Test Scores Aren’t Matching Up. Here’s Why. Retrieved from https://www.edweek.org/teaching-learning/grades-and-standardized-test-scores-arent-matching-up-heres-why/2024/10

[7] GBH News. (January 23, 2024). Standardized tests aren’t biased, says new data—but scores reflect society’s biases. Retrieved from https://www.wgbh.org/news/education-news/2024-01-23/standardized-tests-arent-biased-says-new-data-but-scores-reflect-societys-biases

[8] Dartmouth President’s Office. (February 5, 2024). Reactivating the SAT/ACT requirement for Dartmouth undergraduate admissions. Retrieved from https://president.dartmouth.edu/news/2024/02/reactivating-satact-requirement-dartmouth-undergraduate-admissions