THE RATING OF CONTENT DIFFICULTY OF ITEMS IN CRITERION-REFERENCED TESTS AS APPRAISED JOINTLY BY EXPERTS

In this paper we first give the concept of content difficulty of items in Criterion^*- referenced tests. Secondly, we discuss the method of experts’ judgments to estimate item content difficulties through an actual example.
The results statistical analysis show that for the example the agreement for the whole set of 8 judges is very high. So the arithmetic mean of the individual ratings can be used as the raring of item content difficulties. Further, we indicate that in practical work the average rating of two judges might be satisfactory.