|
Second, correct the variance of the relationship within the
population for measurement error and range departure using the means
and variances of the reliability and range correction factors a, b
and c:

This is an estimate of the variance of rho. If there is no reason
to expect a serious amount of range variation across studies, as in the
case of most psychological research, then the correction procedure for the
range departure may be omitted (Hunter et al., 1982).
|