Table 4.

Simulation Results Relative to the No-Purification Method.

				Purification
				Partial		Full
Impact	% DIF items	Sample size	Test length	P	TIE	P	TIE
0	0	500	15	—	−0.003	—	0.001
—	—	—	30	—	−0.008	—	−0.006
—	—	1,000	15	—	−0.005	—	0.000
—	—	—	30	—	−0.005	—	−0.003
—	20	500	15	−0.014	−0.036	0.056	−0.034
—	—	—	30	−0.039	−0.051	0.016	−0.046
—	—	1,000	15	−0.010	−0.109	0.053	−0.100
—	—	—	30	−0.030	−0.111	0.008	−0.105
—	33	500	15	−0.014	−0.127	0.126	−0.120
—	—	—	30	−0.079	−0.151	0.032	−0.149
—	—	1,000	15	−0.010	−0.333	0.064	−0.329
—	—	—	30	−0.028	−0.344	0.063	−0.343
−0.5	0	500	15	—	−0.001	—	0.007
—	—	—	30	—	−0.007	—	−0.002
—	—	1,000	15	—	−0.003	—	0.001
—	—	—	30	—	−0.004	—	−0.003
—	20	500	15	−0.010	−0.021	0.113	−0.012
—	—	—	30	−0.063	−0.048	0.032	−0.043
—	—	1,000	15	0.000	−0.090	0.067	−0.085
—	—	—	30	−0.028	−0.108	0.018	−0.103
—	33	500	15	−0.020	−0.081	0.130	−0.071
—	—	—	30	−0.037	−0.095	0.071	−0.086
—	—	1,000	15	−0.038	−0.283	0.096	−0.280
—	—	—	30	−0.016	−0.289	0.101	−0.288

Note. DIF = differential item functioning.