Research Article

A Simulation Study to Assess the Effect of the Number of Response Categories on the Power of Ordinal Logistic Regression for Differential Item Functioning Analysis in Rating Scales

Table 5

The type I error of OLR model under different combinations when .

ConditionsRatioDIF = 0.5DIF = 1

1 0.030.040.040.070.080.09
0.040.040.040.070.080.08
0.030.040.040.050.070.07

20.040.040.040.080.080.09
0.040.040.040.060.070.08
0.040.040.040.060.070.08

30.040.040.040.070.060.07
0.030.030.030.040.050.06
0.030.030.030.040.050.06

40.030.030.030.050.060.06
0.030.020.020.050.060.06
0.030.030.030.050.060.06

50.050.050.060.090.110.12
0.040.050.050.080.090.10
0.040.050.050.070.080.09

60.030.030.030.050.060.06
0.030.030.030.050.060.05
0.030.030.030.050.060.05

70.050.070.070.100.130.14
0.050.060.070.100.120.13
0.050.060.060.080.100.10

80.070.080.080.120.140.16
0.060.070.080.110.130.14
0.050.060.060.090.110.12

90.070.070.080.100.110.14
0.070.070.070.090.110.13
0.060.060.070.090.100.11

: number of response categories.
: and , : and , and : and .
The conditions are described in Table 1.