Premium
Empirical versus subjective procedures for identifying gender differences in science test items
Author(s) -
Sudweeks Richard R.,
Tolman Richard R.
Publication year - 1993
Publication title -
journal of research in science teaching
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 3.067
H-Index - 131
eISSN - 1098-2736
pISSN - 0022-4308
DOI - 10.1002/tea.3660300103
Subject(s) - test (biology) , psychology , statistical analysis , empirical research , statistics , mathematics , paleontology , biology
When judgmental and statistical procedures are both used to identify potentially gender‐biased items in a test, to what extent do the results agree? In this study, both procedures were used to evaluate the items in a statewide, 78‐item, multiple‐choice test of science knowledge. Only one item was flagged by the sensitivity reviewers as being potentially biased, but this item was not flagged by the statistical procedure. None of the nine items flagged by the Mantel‐Haenszel procedure were flagged by the sensitivity reviewers. Eight of the nine statistically flagged items were differentially easier for males. Four of these eight measured the same category of objectives. The authors conclude that both judgmental and statistical procedures provide useful information and that both should be used in test construction. They caution readers that content‐validity issues need to be addressed when making decisions based on the results of either procedure.