e-Articles (Chapter 15)
Illustrates the use of an analysis of covariance for the express purpose of "reducing the error variance." Also shows how the covarite variable can be different from the dependent variable. (See the section entitled "Research Design" and the 5th-8th paragraphs of "Findings.")
Illustrates how researchers ought to be concerned with ANCOVA's underlying assumptions. In this study, "the data were analyzed for independence, normality, homogeneity of variance, linearity, and parallelism." (See the first 4 paragraphs of "Findings" as well as Tables 1 and 2.)
Using pretest scores on "dullness" as the covariate, a one-way ANCOVA was used to compare an experimental group versus a control group.
In this study, a three-way ANCOVA containing three active factors was used to study altruism and cheating among relatives and strangers. Results yielded a first-order interaction. (See the "Method" and "Results" sections, plus Tables 1 and 2.)
Copyright © 2012
Schuyler W. Huck