skip to main content skip to footer

Speeded Tests and Power Tests -- An Empirical Study of Validities

Lord, Frederic M.
Publication Year:
Report Number:
ETS Research Bulletin
Document Type:
Page Count:
Subject/Key Words:
Office of Naval Research, Comparative Testing, Power Tests, Test Reliability, Test Speededness, Test Validity, Timed Tests


This study addresses the question of how speeded a given test should be in order to maximize the validity for predicting a given criteria. It compares the validates of certain power (unspeeded) tests, certain moderately speeded tests, and certain highly speeded tests for predicting certain course grades. Examinees were 649 entering freshmen at one institution. It is concluded that, with the present data, the speeded tests are more valid for short administration times and the unspeeded (power) tests are more valid for long administration times. These conclusions were reached regarding the relative validates of the tests at the three levels of speededness studied; no attempt was made to infer the precisely optimum number of items for a given time limit. (JGL)

Read More