A Contribution to the Question of Assessing the Difficulty of Test Items

In pedagogical research as well as in educational practice didactic tests are mostly used for the purpose of checking up on learner’s progress or achievement or for examination purposes. Standardized tests can be used for both group and individual pedagogical diagnosis. If the primary task of didactic tests is to help create objective yardsticks for evaluating and marking learners’ results, then the test-writer must possess a combined knowledge of pedagogy, methodology of the subject being taught and mathematical statistics. The criterion of the appropriateness of the test is given by some of its parameters resulting from the item or test analysis. These parameters include ascertaining the statistical difficulty of test items as well as the difficulty of the test as a whole and that of subtests. For pedagogical reasons test items are arranged in the order of increasing difficulty. The number of points awarded for the correct response to each item can be made proportional to the degree of difficulty of the item; this technique is used especially in tests with a small number of items, e. g. in tests serving research purposes. The article also describes a graphical method and a numerical one showing how to allot different scores to different items. A correlation analysis shows that in scoring correctly answered items the difficulty factor may be disregarded if there is a greater number of test items, which are arranged in the order of increasing difficulty.