Analysis of Covariance Experimental Designs

Statistical Analysis of the Analysis of Covariance Design
Example of the Analysis of Covariance Design

The Analysis of Covariance Design(ANCOVA or ANACOVA) is a noise-reducing experimental design. The basic ANCOVA is a just pretest-posttest randomized experimental design.The notation shown here suggests that the pre-program measure is the same one as the post-program measure, and so we would call this a pretest. But you should note that the pre-program measure doesn't have to be a pretest -- it can be any variable measured prior to the program intervention. It is also possible for a study to have more than one covariate.

Design Notation

The pre-program measure or pretest is sometimes also called a “covariate” because of the way it’s used in the data analysis—we “covary” it with the outcome variable or posttest in order to remove variability or noise. This is what we mean by "adjusting" for the effects of one variable on another in social research. you can use any continuous variable as a covariate, but the pretest is the best, because the pretest is usually the variable that would be most highly correlated. Because it is highly correlated, when you "subtract it out" or "remove it", you're removing more extraneous variability from the post test. The rule in selecting covariates is to select the measure that correlate most highly with the outcome and, for multiple covariates, have little intercorrelation.

In this web page, you can find the "Statistical Analysis of the Analysis of Covariance Design" which shows the principle of ANCOVA, and " Example of the Analysis of Covariance Design" which include the hypothetical example of Analysis of Covariance Experimental Design.

Return to Knowledge Base Home Page