Abstract
AbstractTwo types of test bias were explored, which may affect standardized achievement test scores on a large scale, possibly including everyone taking a particular test. The biases were referred to as “norming” bias and “induced” bias. A multiple regression analysis was used to compare the scores of 407 eighth graders taking different nationally normed mathematics achievement tests. Students in the experiment taking one particular test (1977 California Achievement Tests) scored much higher than the available information led one to expect (p > .0001). Students of one school taking one of the tests (1982 Iowa Tests) scored lower than expected (p > .03). Plausible explanations for the results in terms of test bias are given.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
