A Meta-Analysis of the "Erasing Race" Effect in the United States and Some Theoretical Considerations.

The possible loss in significance might stem from interconversion of effects between d and r (such conversions are known to very slightly bias values; Schmidt and Hunter, 2015 ) and from rounding down to two decimal places (so as to homogenize the reporting of each effect across studies), both of which may make marginally significant d values reported in one study marginally non-significant when recomputed for meta-analysis.

also in 1,493 other papers