Table 4. F-ratio of the correlation estimates when correcting with multiple imputation by chained equations and Thorndike’s formulas.

	Selection ratio (SR)
	.1	.2	.3	.4	.5	.6	.7	.8	.9
DRR, artificially dichotomous	2.23**	1.23**	1.03	1.00	1.14**	1.27**	1.66**	1.95**	2.25**
DRR, naturally dichotomous	3.46**	2.08**	2.25**	2.87**	4.24**	5.22**	6.71**	9.19**	10.3**
IRR, artificially dichotomous	1.72**	1.46**	1.34**	1.42**	1.48**	1.69**	1.95**	2.40**	2.72**
IRR, naturally dichotomous	2.41**	2.68**	3.62**	5.02**	6.76**	8.29**	10.4**	12.4**	12.3**

${\widehat{r}}_{\text{b}}$ is the estimate of the biserial correlation coefficient, ${\widehat{r}}_{\text{pb}}$ is the estimate point-biserial correlation coefficient, F-ratio is calculated by the mean square error (MSE) of the estimate using Thorndike’s formula divided by the MSE of the estimate using MICE, ** p < .001.