Meta Analysis Page Logo Meta - Analysis:
Methods of Accumulating Results Across Research Domains
Page 16
Home Contents Abstract Introduction Links Contact Info

Correcting for Range Departure

As mentioned previously, the other source of error is range departure. This is the random deviation from rho (the estimated population mean effect size) because of variation due to the restriction, (when selecting a decreased range of scores) or inflation (when selecting extreme scores only) of the range of possible scores on any measure.

To correct for range departure, first compute the ratio of the standard deviation of the individual study to the standard deviation of some reference population, "u". This ratio is used to construct "c", an estimate of the range departure for the individual study that presents the standard deviation information. The formula for calculating c is presented below:

 

Previous Home Contents Intro Top Next