| This study used four statistical item bias analysis strategies, and a panel of French bilingual/English experts to determine French cross-cultural validity of the Maine Educational Assessment test (MEA), a program which grew out of the Maine Educational Reform Act of 1984. It is administered to Maine Students in grades 4, 8, and 11, in the subject areas of reading, mathematics, writing, science, social studies, and humanities.; An analysis was done of eighth grade pupil performance in test year 1988-89, in the areas of the 100 common reading and mathematics items that all pupils take. The statistical strategies used were: (1) Scheuneman's modified chi-square procedure (SSX{dollar}sp2{dollar}); (2) Rudner and Convey's TID-45{dollar}sp0{dollar} item difficulty p-value; (3) The Rasch one-parameter latent trait model; (4) The Mantel-Haenszel procedure.; Item response comparisons were made of two of Maine's pupil populations: 336 French bilingual/English fluent speakers from the communities of St. Agatha, Madawaska, Lewiston, Ft. Kent, Van Buren, and Caribou, and 336 monolingual English speaking pupils randomly selected and paired to insure identifiably equal abilities on the basis of equal overall test scores.; Findings indicated that eight out of fifty mathematics items (16%), and nine out of fifty reading items (18%) from the common questions asked of all eighth grade pupils in the two subject areas analyzed indicated differential validity. However, the identified items also indicated differential functioning in that they did not favor either language group in a material way. In fact this differential functioning finding suggests that the French bilingual/English fluent minority children have virtually equal chances of achieving identical scores on the common mathematics and reading questions as did the monolingual English pupils.; Lower mean scoring by bilingual/English fluent pupils, as indicated in the "1988-89 Maine Educational Assessment Grade 8 Mean Comparisons," is evidently not caused by item bias in the MEA as far as the French bilingual/English fluent pupils are concerned. Either the French pupils are not affected as are other cross-cultural groups, or other reasons are at play as to the origin of the lower means of Maine's French bilingual pupils. As indicated in a survey done in preparation for this study, mis-identification of bilingual/English fluent pupils is surely a primary cause. |
