This article identifies the problem of lack of congruence (i. e., "fit") between the assumptions that underly measurement-based approaches to assessment and the movement that has become known as whole language. A "natural" theory of language assessment is proposed and described. Using research from classrooms, the authors describe how some Australian teachers have put this theory of language assessment into practice. Finally, the question of the "scientific rigor" of classroom research is addressed. The authors argue that the traditional criteria of scientific rigor (i. e., reliability, validity, objectivity, etc.) cannot be applied to data collected under naturalistic conditions. Instead they argue that naturalistic analogues of these criteria are available and that teachers can be taught to develop strategies that increase the trustworthiness and credibility of the information they collect.
[1]
M. Halliday.
Learning How to Mean: Explorations in the Development of Language
,
1975
.
[2]
Gilbert A. Clark,et al.
Evaluating the Arts in Education: A Responsive Approach
,
1976
.
[3]
E. Guba,et al.
Effective Evaluation: Improving The Usefulness Of Evaluation Results Through Responsive And Naturalistic Approaches
,
1981
.
[4]
Don Holdaway,et al.
The Foundations of Literacy
,
1984
.
[5]
E. Guba,et al.
Naturalistic inquiry: Beverly Hills, CA: Sage Publications, 1985, 416 pp., $25.00 (Cloth)
,
1985
.
[6]
Brian Cambourne,et al.
The Whole Story: Natural Learning and the Acquisition of Literacy in the Classroom
,
1988
.