Premium
RUMORS REGARDING THE DEATH OF THE EQUIPERCENTILE ASSUMPTION MAY HAVE BEEN GREATLY EXAGGERATED
Author(s) -
TALLMADGE G. KASTEN
Publication year - 1985
Publication title -
journal of educational measurement
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.917
H-Index - 47
eISSN - 1745-3984
pISSN - 0022-0655
DOI - 10.1111/j.1745-3984.1985.tb01047.x
Subject(s) - psychology , selection (genetic algorithm) , norm referenced test , norm (philosophy) , statistics , econometrics , mathematics education , computer science , criterion referenced test , standardized test , economics , mathematics , political science , law , artificial intelligence
Powers, Slaughter, and Helmick (1983) recently analyzed selection, pretest, and posttest scores collected from large numbers of students in two cohorts. These analyses led them to conclude that the equipercentile assumption underlying norm‐referenced evaluation methodology is “inappropriate.” Re‐examination of the data suggests that there is strong support for the validity of the equipercentile assumption in the selection and pretest scores they present. The observed “gains” from pre‐ to posttests are better attributed to stakeholder bias, posttests that match the curriculum content too closely, or a combination of these two factors than to inappropriateness of the equipercentile assumption. Annual testing where the posttest from one year also serves as the pretest for the next is suggested as a promising solution to both of the cited threats to the internal validity of norm‐referenced evaluations.