Overview
The inverse variance weight method assigns a weight to each effect size prior to averaging. Given that a sample variance is inversely related to the sample size, the inverse of variance is then used to weight the effects in each study.
Systematic error occurs when the studies in a meta analysis are of differing quality. When this occurs, the weights assigned to each study may not be optimal, without taking into account the quality issues.