Table 1 Mean of precision of the estimation of \( {\mathrm{r}}_{\mathrm{y}{\mathrm{x}}_1\mid \mathrm{a}} \) (predictive validity) from different methods when 30%, 20%, and 10% of applicants are selected with RMSE and confidence interval
From: Correcting the predictive validity of a selection test for the effect of indirect range restriction
30% Selection | 20% Selection | 10% Selection | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
mean | SD | RMSE | CI 95% | mean | SD | RMSE | CI 95% | mean | SD | RMSE | CI 95% | |
Thorndike’s case A | −0.07 | .05 | .085 | −0.16 – 0.02 | −0.10 | .07 | .119 | −0.23 – 0.02 | −0.14 | .10 | .167 | −0.34 – 0.04 |
Thorndike’s case C | 0.00 | .04 | .043 | −0.09 – 0.07 | −0.01 | .06 | .065 | −0.14 – 0.10 | −0.02 | .10 | .102 | −0.24 – 0.15 |
EM | 0.00 | .04 | .039 | −0.09 – 0.07 | −0.01 | .06 | .058 | −0.13 – 0.10 | −0.02 | .09 | .092 | −0.23 – 0.12 |
MICE | −0.01 | .04 | .040 | −0.09 – 0.07 | −0.01 | .06 | .059 | −0.14 – 0.09 | −0.03 | .09 | .097 | −0.24 – 0.12 |