This paper explains how commonality analysis (CA) can be conducted using a specific Statistical Analysis System (SAS) procedure and some simple computations. CA is used in educational and social science research to partition the variance of a dependent variable into its constituent predicted parts. CA determines the proportion of explained variance that is unique to a predictor variable and the proportion that is common to two or more predictors. Whereas the ordering of the predictors using stepwise regression may lead to faulty data interpretations, CA is a method by which all possible predictor combinations are tested to determine the model that best explains predicted variance. Data from a study of life satisfaction (LS) among 198 elderly residents in 17 Texas nursing homes illustrate procedures for conducting CA with regression results. The subjects completed a LS questionnaire to determine if their self-reports of LS differed from those of the elderly living outside of nursing homes. Eight subscale components and the number of years in the nursing home were analyzed by regression to determine which variable best predicted nursing home satisfaction. Meaning was the dominant factor in predicting nursing home satisfaction and accounted for about 80% of all explained variance in the sample. In addition, a SAS computer program for obtaining all possible R-squared values is discussed as an efficient method of implementing the required analyses. CA offers a fairly straightforward method of analysis when no more than four independent variables are of interest. Three tables of data are presented, and the R-squares of LS scales are included. (SLD)