A rationale for relating test scores to objectives

Abstract

Questions of educational and statistical validity raised by multiple assessment techniques are examined in terms of educational objectives, errors and distributions of scores. A method of scaling is proposed and its relation to normalizing procedures discussed in recent papers in this journal is indicated. A hypothetical situation provides the framework for the elaboration of the ideas presented.

Notes

† Now a lecturer at Darwin Community College, Casuarina N.T., Australia.

Additional information

Notes on contributors

D. C. Godwin†

† Now a lecturer at Darwin Community College, Casuarina N.T., Australia.