A Statistical Study of the Reliability of Scale Grading

Abstract

Introduction [partial]: The purpose of this paper is to determine the reliability of a series of grades obtained by the use of the Ayres Measuring Scale for Handwriting. In order to make clearer the meaning of the results of the study, which are given in terms of the scale, a brief description of the scale and also of the samples graded, will precede the report of the study itself. The Ayres scale was constructed by measuring the speed of reading of a large number of samples of childrens' writing by ten different persons. Eight degrees of legibility are represented by typical samples selected from the entire number graded by the ten persons. The eight steps are supposed to be equal and are designated by 20, 30, 40, - 90. Each step has three styles of writing of approximately equal quality. The three styles of writing are vertical, medium slant, and full slant. Though the scale was constructed on the basis of legibility, in practice it is used, just as is the Thorndike scale, to measure the quality of a sample rather than its legibility alone. To grad a sample, one is instructed to "slide it along the scale until a writing of corresponding quality is found."