Abstract
This paper describes a regression approach to experimental design. Typically, the educational researcher is familiar with the usual analysis of variance techniques and is unaware of the versatility of multiple regression for problem solving. The specific approach used in this paper assumes only a familiarity with experimental design and access to a general purpose multiple regression program which is likely to be on hand at any computer installation. Examples of the t-test, 1-way analysis of variance, and the treatment X subjects design from a regression standpoint are given.

This publication has 5 references indexed in Scilit: