[Review of the book The Constant and Variable Errors of Educational Measurements

Abstract

Reviews the book, The Constant and Variable Errors of Educational Measurements by Walter S. Monroe (1923). In this brief monograph the author discusses constant and variable errors in educational measurements, and points out their influence on such derived measures as the average, median, standard deviation, and coefficient of correlation. In the course of the discussion there is presented evidence of some five different sources of constant errors in educational measurements. A careful examination of the scores of any test will reveal evidence of the presence of variable errors of measurement. This becomes quite apparent when the same test is repeated after a short interval and the two scores of the same individual pupils are compared. What should be the effect of a knowledge of the existence of such errors upon the movement for scientific measurement m education? It is not the author's purpose to advise the dicontinuance of the use of educational tests. It is rather to point out to those using such tests the limitations of the measures yielded by their use. It is certainly worth while for both the administrator and the school statistician to recognize the possible presence of one or both types of errors in every application of the standard tests. (PsycInfo Database Record (c) 2025 APA, all rights reserved)