SAMPLING VARIANCE IN THE CORRELATION-COEFFICIENT UNDER INDIRECT RANGERESTRICTION - IMPLICATIONS FOR VALIDITY GENERALIZATION

The authors conducted Monte Carlo simulations to investigate whether i ndirect range restriction (IRR) on 2 variables X and Y increases the s ampling error variability In the correlation coefficient between them. The manipulated parameters were (a) IRR on X and Y (i.e., direct rest riction on a third variable Z), (b) population correlations rho(XY), r ho(XZ), and rho(YZ), and (c) sample size. IRR increased the sampling e rror variance in r(XY) to values as high as 8.50% larger than the anal ytically derived expected values. Thus, in the presence of IRR, validi ty generalization users need to make theory-based decisions to ascerta in whether the effects of IRR are artifactual or caused by situational -specific moderating effects.