Effects of Item Disclosure on TOEFL Performance [microform] / Gordon A. Hale and Others

To ascertain how the Test of English as a Foreign Language (TOEFL) would be affected if candidates had access to some of the items before administration of a test containing those items, a number of specially constructed TOEFL forms were made available to 945 foreign students in intensive English language programs. The students were later administered a special TOEFL consisting of items from those forms and a TOEFL consisting of undisclosed items. A significant disclosure effect was determined when performance was contrasted on these two tests. Scores for the test containing disclosed items were greater than those for the test containing undisclosed items, indicating that students studied the disclosed forms and recalled specific items for them. Student questionnaire data suggested that the time spent studying the disclosed material was relatively unaffected by the number of forms given the students. Additional analyses indicated disclosure effects for items in forms that were not discussed in class as well as for items that were discussed. A disclosure effect was observed for each of three main language groups and for every item type, which attests to the robustness of this effect. (Author/PN)