The Reliability,Discrimination and Difficulty of Word-Knowledge Tests Employing Multiple Choice Items Containing Three,Four,or Five Alternatives

Abstract

The multiple-choice item found on standardized and teacher-made tests typically requires the examinee to choose among either four or five alternative answers. The present study offers some evidence favouring three-option over four-option items.

Word knowledge tests containing three, four, and five-option items were developed and administered to pupils in ten upper primary classes from schools in the Adelaide metropolitan area. In each of the classes, item discrimination was better and reliability was higher for the three-option test compared to the test employing four-option items. The five-option test was also superior to the four-option test. Differences between the three and five-option tests on reliability and item discrimination were not significant.

References

The writer wishes to acknowledge the help of a small group of his senior undergraduate students who assisted with the testing and the analysis of results.

Costin

The optimal number of alternatives in multiple-choice achievement tests: Some empirical evidence for a mathematical proof. Educational and Psychological Measurement, 1970, 30, 353–358.

Tversky

On the optimal number of alternatives at a choice point. Journal of Mathematical Psychology. 1964, 1, 386–391.

Costin drew some conclusions regarding test “power” from his obtained difficulty indices. In the present study, the difficulty indices of the tests are simply reported and compared as such.

Johnson

A. P.

Notes on a suggested index of item validity: The U-L index. Journal of Educational Psychology, 1951, 42, 499–504.

This simple index, which has been shown to be a valid alternative to more elaborate procedures (see Findley, W. G. “A rationale for evaluation of item discrimination statistics”. Educational and Psychological Measurement, 1956, 16, 175–180), is now more commonly known as the D-index and will be referred to as such in the present article.

Hoyt

Test reliability estimated by analysis of variance. Psychometrika, 1941, 6, 153–160.

Siegel

Nonparametric Statistics for the Behavioral Sciences. New York: McGraw-Hill, 1956.

Langley

Practical Statistics. New York: Drake Publishers, 1971, 227–228.