skip to main content skip to footer

A Study of the Relationship Among Speed and Power, Aptitude Test Scores, and Ethnic Identity SAT

Evans, Franklin R.
Publication Year:
Report Number:
ETS Research Report
Document Type:
Page Count:
Subject/Key Words:
College Board, Analysis of Variance (ANOVA), Aptitude Tests, Minority Groups, Performance Factors, Scholastic Aptitude Test (SAT), Test Bias, Timed Tests


This is the fourth report in a series of research studies to determine the effects of altering time-per-item as it relates to the test candidate's ethnic identity, residence (rural or urban), and sex. Four specially designed SAT®-like tests were administered to volunteer samples of Black, Hispanic, and White college-bound juniors in 12 high schools in urban and rural settings. Within each high school, the volunteers were randomly assigned to one of three time-limit conditions: speeded (20 minutes per section), normal (30 minutes per section), and unspeeded (40 minutes per section). Results of several independent analyses of variance (ANOVAs) did not reveal any differential performance related to residency or sex. Significant effects were detected for test-time limits and ethnic identity. No significant two-way interactions involving any of the four factors (time, ethnic identity, residence, or sex) were detected. Two significant findings were observed in the item analysis of the selected sections (verbal, math). There is some evidence that minority volunteers were employing less than optimal test taking strategies. The average section scores of the volunteers still working near the end of the test section were often lower than the average sections scores for all candidates. Secondly, when alternate form reliability estimates were compared for minorities, the estimates of reliabilities for wrong scores were higher than those for rights or formula scores. Possible explanations of these phenomena are that the tests were too difficult for the volunteers or that the volunteers who persisted were being attracted to incorrect options by way of misinformation.

Read More