What Counts as Reading: Limits to Certainty in Assessment

Abstract

Any assessment of reading ability, processes, or skills is culturally and socially organized. Assessment requires concerted activity between assessor and assessee, teacher and student. Assessment is accomplished using presuppositions, assumptions, expectations, conventions, and rules which are sensible (at least to assessors) as culturally possible, warranted and wise ways of organizing an activity. The very ways that reading activities are culturally and socially organized, however, present problems of assessment that cannot be remedied in complete or certain ways, i.e., in ways adequate to logico-empiricist canons of rigor. These problems are principled problems. They are present, in principle, wherever and however assessment of reading ability, processes, or skills is attempted. These problems have their own origins in our concepts of reading and ways of organizing assessment activities. In the circumstances where reading is assessed, these problems raise the question: What counts as reading, criterially? This article is divided into two parts: the first part is mainly philosophical; the second part, sociological. In part one we formulate the paradigm assumptions of psychometrics, which operate in standardized achievement testing and are guiding ideals in classroom assessment. Content validity is discussed and a discussion is drawn between (1) the grammatical claim that the relation between a task outcome and a target skill is valid, and (2) the