Some Shortcomings of Soliciting Students' Self-Reported SAT Scores

Richard E. Mayer
Andrew T. Stull
Julie Campbell
Kevin Almeroth
Bruce Bimber
Dorothy Chun
Alan Knight

The authors analyzed self-reported SAT scores and actual SAT scores 
for 285 college students. Students overestimated their actual SAT 
scores by an average of 23 points (SD = 65, d = .72), with 7% 
under-reporting, 53% reporting accurately, and 40% over-reporting, 
indicating a systematic bias towards over-reporting.  The amount of 
over-reporting was greater for lower-scoring than higher-scoring 
students, was greater for upper division than lower division students, 
and was equivalent for men and women. There was a strong correlation 
between self-reported and actual SAT scores (r = .87), indicating 
high validity of students' memories of their scores.  Results are 
consistent with a motivated distortion hypothesis.  Caution is 
suggested in using self-reported SAT scores in educational research.