Classroom Testing Procedures and Student Anxiety.

An attempt to reduce test anxiety was made by pro viding increased information to students a out the types of questions likely to appear on end-of-unit ex aminations. Two experimental testing formats, and a combination of the two, were compared with the typical practice of providing no supplemental information. The experimental formats were as follows: (1) During the class session prior to the test, a copy of a previous test, given to a class in an earlier semester but covering the same content, was given to each student. Students were told that similar, but not identical, questions could be expected on the upcoming test. (2) Students were of fered two opportunities to be tested over the same course unit at consecutive class sessions. They were re quired to take one, although they could take both, of the tests. Their score of record would be the better of the two scores on the tests. The tests contained com pletely different sets of questions, but did cover the same unit of study. Table 1 summarizes the testing design as it was employed over a full year's classes in Introductory Sociology. Each testing condition was used at least three times during the year. For example, the combination of experimental formats, (t)TT, was employed with Unit 1 for both of the Fall 1981 classes, with Unit 2 for the