SAT Score Predictor

A note on what this model can and can't tell you.

This estimator is calibrated to the published causal effects of SAT prep from Weatherholtz et al. (2020, College Board Technical Report, n=545,640), Briggs (2009), Powers & Rock (1999), and Appelrouth et al. (2015). The headline "115 points from 20+ hours of Khan Academy" claim from the 2017 press release was a raw association — the same authors corrected it to a causal effect of roughly +21–39 points in their 2020 follow-up. Most published causal estimates of test prep cluster around 20–40 points total.

The biggest predictor of your next SAT score isn't your study plan — it's your last SAT score (~85% of variance in the largest published dataset). The estimated range here is intentionally wide because hours-of-study only explain 5–15% of the variation in actual student improvement. The rest comes from individual differences this tool can't measure: motivation, test anxiety, baseline ability, the quality of how you actually use your study time, and luck on test day.

Use this as a sanity check on your study plan, not a guarantee.