Sequential model for appraising instructional superiority of a revised material |
| |
Authors: | Subramaniam Kandaswamy |
| |
Institution: | (1) Educational Testing Service, 60201 Evanston, IL |
| |
Abstract: | The effectiveness of a revised instructional material compared with its prototype usually is determined by subjects’ performance
on criterion-referenced tests. The alternative method outlined in this article - sequential analysis - appears to have several
advantages over the traditional method. For one thing, it permits estimation of how much better (or worse) the revised material
might be, in terms of both educational and statistical significance. In addition, it provides advance estimates of the number
of subjects required (usually 50% fewer than traditionally), allows immediate analysis of data, and offers a scheme for controlling
both Type I and Type II errors. |
| |
Keywords: | |
本文献已被 SpringerLink 等数据库收录! |
|