NotesFAQContact Us
Collection
Advanced
Search Tips
ERIC Number: ED344910
Record Type: RIE
Publication Date: 1992-Apr
Pages: 34
Abstractor: N/A
Reference Count: N/A
ISBN: N/A
ISSN: N/A
Test Anxiety and Test Performance Under Computerized Adaptive Testing Methods.
Powell, Z. Emily
Little research exists on the psychological impacts of computerized adaptive testing (CAT) and how it may affect test performance. Three CAT procedures were examined, in which items were selected to match students' achievement levels, from the item pool at random, or according to student choice of item difficulty levels. Twenty-four graduate students (5 males and 19 females) at Indiana University (Richmond) were randomly assigned to one of six testing orders formed by the three mastery test approaches and by blocking on native and non-native speakers. While at a computer, students received a description of adaptive testing methods and then a 20-item pre-test anxiety measure. Right after completing each test, students responded to a 10-item in-test anxiety scale, ranked their preferences among tests, and evaluated their performance on each of the tests. No statistically significant mean differences were found among mean student achievement scores or among in-test anxiety means under the three adaptive testing methods. Students reporting higher anxiety scored significantly higher in the matched-selection test. Those preferring the matched-selection and self-selection tests the most were less anxious during those tests. Instead of actual performance, students' perceptions of how well they did were significantly correlated with preference rankings for the tests. The matched-selection tests required significantly fewer items to reach decisions than did the random-selection tests. Ten tables, 1 figure, and a 19-item list of references are included. (Author/SLD)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A
Identifiers: Psychological Influences; Testing Effects
Note: Paper presented at the Annual Meeting of the American Educational Research Association (San Francisco, CA, April 20-24, 1992).