ERIC Number: ED194583
Record Type: RIE
Publication Date: 1980-Aug
Reference Count: 0
Design Coding and Interpretation in Multiple Regression.
Lunneborg, Clifford E.
The multiple regression or general linear model (GLM) is a parameter estimation and hypothesis testing model which encompasses and approaches the more familiar fixed effects analysis of variance (ANOVA). The transition from ANOVA to GLM is accomplished, roughly, by coding treatment level or group membership to produce a set of predictor or dependent variables. How treatments should be coded is controversial. The present paper redevelops the independence of hypothesis testing and of estimation (point or interval) from the coding of variables. Further, this redevelopment, based on the simple concepts of a design template (a k x k matrix in the k groups design) and of an expectation template (a k element vector), permits mean contrasts or parameter constraints to be directly stated and easily translated into a natural and common computational form. The technique is developed in three parts. First, the essentials of the GLM are reviewed. Then, forms for stating hypothesis or estimating contrasts are developed which will be independent of coding. Finally, the approach is illustrated for the three Serlin and Levin designs, for unbalanced designs, and for designs with missing cells. (Author/CP)
Publication Type: Reports - Research
Education Level: N/A
Authoring Institution: Washington Univ., Seattle. Educational Assessment Center.
Identifiers: General Linear Model
Note: A version of this paper was presented at the Annual Meeting of the American Psychological Association (88th, Montreal, Quebec, Canada, September 1-5, 1980).