Key points are not available for this paper at this time.
Reliability has traditionally been taken for granted as a necessary but insufficient condition for validity in assessment use. My purpose in this article is to illuminate and challenge this presumption by exploring a dialectic between psychometric and hermeneutic approaches to drawing and warranting interpretations of human products or performances. Reliability, as it is typically defined and operationalized in the measurement literature (e.g., American Educational Research Association AERA, American Psychological Association, Feldt & Brennan, 1989 ), privileges standardized forms of assessment. By considering hermeneutic alternatives for serving the important epistemological and ethical purposes that reliability serves, we expand the range of viable high-stakes assessment practices to include those that honor the purposes that students bring to their work and the contextualized judgments of teachers.
Pamela Moss (Tue,) studied this question.