Speededness As A Source Of Test Bias For Non-Native English Speakers On The College Level Academic Skills Test

Date of Award




Degree Name

Doctor of Philosophy (Ph.D.)


Educational Research


This study investigates the effects of increased test taking time on the performance of native and non-native speakers of English on the College Level Academic Skills Test (CLAST). This test was developed by the Florida Department of Education to assess the performance of College Sophomores on a set of skills mandated by State Board of Education Rule. The test consists of four sub-sections: Computation, Reading, Writing, and Essay. A student must obtain passing scores on all sub-sections of the test prior to receiving an Associate of Arts Degree or progressing to the Upper Division in a Community College or State University in Florida.The 659 students who sat for the March 1985 Administration of the CLAST at a large urban Community College and a large urban State University were placed into administrative groups that were given additional testing time on different sub-tests. The experimental design was a 2 x 3 factorial design, fixed effects model, with three levels of time ((1) no extra time; (2) extra 10 minutes and (3) extra 20 minutes) and two levels of native language ((1) English and (2) Not English). The main effects due to time were found to be significant on the Reading, Writing, and Essay sub-test, but no differential effects were discovered. The Computation sub-test demonstrated no main effect due to extra time for Native English examinees but evidenced a significant benefit for non-native examinees, resulting in a significant time/native language interaction. The level of semantic difficulty for the logic and word problems on the Computation subtest is postulated as the cause of the time/native language interaction observed. The main effect for ethnic origin was significant for all sub-tests.


Education, Tests and Measurements

Link to Full Text


Link to Full Text