ERIC Number: ED265185
Record Type: RIE
Publication Date: 1985-Nov
An Investigation of the Relationship between Item Arrangement and Test Performance.
Chissom, Brad; Chukabarah, Prince C. O.
The comparative effects of various sequences of test items were examined for over 900 graduate students enrolled in an educational research course at The University of Alabama, Tuscaloosa. experiment, which was conducted a total of four times using four separate tests, presented three different arrangements of 50 multiple-choice items: (1) increasing order of item difficulty; (2) descending order of difficulty; and (3) random sequence. Difficulty of the items ranged from .25 to 1.00, and test reliability ranged from .63 to .82 (coefficient Alphas). The scores represented the number of correct items; there was no correction for guessing. No time limits were set, and subjects completed all items. Frequency analyses were conducted, and the descriptive statistics indicated that there were no major differences in mean scores across the experimental replications as well as among the test forms within each replication. Results of the analyses of variance also indicated that the group means were not significantly different. It was concluded that this different arrangement of test items, based on item difficulty, neither relates to power test performance, nor affects examinees' test performance in a positive or a negative manner. However, rearranging test items on different forms may continue to be a useful way to discourage cheating. (GDC)
Publication Type: Speeches/Meeting Papers; Reports - Research
Education Level: N/A
Authoring Institution: N/A
Note: Paper presented at the Annual Meeting of the Mid-South Educational Research Association (14th, Biloxi, MS, November 6-8, 1985).