FAQ/p+sp - CBU statistics Wiki

Revision 2 as of 2010-12-15 16:10:39

Clear message
location: FAQ / p+sp

Differences between partial and semi-partial correlations

Consider the artificial example where x perfectly predicts y in each of two groups (r = 1.00 in both) and all the x and y values are higher in one of the groups than the other (as shown on this powerpoint slide [attachment:xygroups.ppt here]). Suppose we wish to see how x correlates with y adjusting for group differences.

We can simply subtract the appropriate group mean from the x value of each observation depending upon what group it is in and correlate this new value with y (ie correlate y with $$x - \bar{x}_text{G}$$). This is the semi-partial correlation (also equal to the signed square root of the change in R-squared when you add x to a model containing group in predicting y) between x and y adjusting for group differences where only x is adjusted for group differences and will be less than the zero-order x,y correlation (around 0.5). If we also adjust y for group differences by subtracting the appropriate y group mean from each y value we obtain the partial correlation between x and y adjusted for group with both x and y adjusted for group and obtain a partial correlation equal to 1.00. The partial correlation of 1.00 follows because there is perfect relationship between x and y in each group.

The statistical test of significance is the same for both partial and semi-partial correlations and is equal to testing the regression coefficient of x in a regression also containing group to predict y. A partial correlation can be used if x differs between groups and is related to y (as in the example above) and a semi-partial correlation if either x only differs between group and is not releated to y or if x does nto differ between groups and is only related to y.