A Reply to Marx

Abstract

It might be profitable if I attempt to answer briefly what I judge to be Marx' three main criticisms of my paper: (i) misconstrues the role of the 'criterion' in criterion-referenced measurement, (2) accepts too readily the nature of the cognitive processes presumed to underlie different types of test items, (3) his tacit assumption instructional feedback plays a role as a reinforcer. In answer to the first criticism it should be recalled the initial discussion took place within a framework of public school and extension of the curricula, and there is little question those responsible for a core curriculum, at least in British Columbia, view it in terms of mastery requirements for most students. Consequently, the use of criterion referenced which suggests that judgements concerning an individual's achievement are dichotomous would not be inconsistent with the context of the discussion. However, I believe a rereading of the paper will show the focus in this case was not on standards of achievement but on a well-defined behavioral domain. Whether or not several performance standards are established with regard to the learning outcomes specified in a core curriculum is irrelevant to the thrust of the argument. I am simply suggesting because a core curriculum usually defines rather carefully the knowledge and skills expected of the students, this logically requires a criterion-referenced interpretation of performance on test items which reflect as closely as possible those learning outcomes described in core curriculum. This brings me to Marx' second criticism I accept too readily the nature of the cognitive processes presumed to underlie different types of test items. While this is not completely true, I will state, without hesitation, different types of test item, for example, completion versus multiple choice, frequently do require different cognitive processes. There are three ways to support this statement: by findings on item difficulty, by introspection, and by studies designed to measure this cognitive difference. Consider first the simple case of the recall of a specific fact in a sentence-completion item versus the recognition of this same fact in a