Table 4.
Limited-information estimation |
Full-information estimation |
||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Statistic | Slope scaling factor | α | Average interitem polychoric correlation | ω | RMSEA [90% CI] | CFI | TLI | Fitted model AIC | Fitted model RMSEA | Null model RMSEA | TLI |
M | 1 | .96 | .79 | .97 | .01 [.00, .03] | 1.00 | 1.00 | 15,612 | .01 | .44 | 1.00 |
1/2 | .87 | .50 | .91 | .01 [.00, .02] | 1.00 | 1.00 | 21,896 | .01 | .26 | 1.00 | |
1/3 | .77 | .31 | .82 | .01 [.00, .02] | 1.00 | 1.00 | 24,889 | .01 | .15 | 1.00 | |
1/4 | .66 | .21 | .73 | .01 [.00, .02] | 1.00 | 1.00 | 26,129 | .01 | .10 | 1.00 | |
1/5 | .56 | .15 | .63 | .01 [.00, .02] | .99 | 1.00 | 26,482 | .01 | .06 | 1.00 | |
1/6 | .47 | .11 | .55 | .01 [.00, .02] | .99 | 1.01 | 26,376 | .01 | .05 | .92 | |
SD | 1 | .01 | .01 | .00 | .01 [.01, .01] | .00 | .00 | 234 | .01 | .09 | .00 |
1/2 | .01 | .02 | .01 | .01 [.00, .01] | .00 | .00 | 267 | .02 | .06 | .02 | |
1/3 | .01 | .01 | .01 | .01 [.00, .01] | .00 | .01 | 249 | .01 | .03 | .04 | |
1/4 | .02 | .01 | .02 | .01 [.00, .01] | .01 | .01 | 228 | .01 | .02 | .09 | |
1/5 | .02 | .01 | .02 | .01 [.00, .01] | .01 | .02 | 205 | .01 | .02 | .24 | |
1/6 | .03 | .01 | .03 | .01 [.00, .01] | .02 | .04 | 188 | .01 | .01 | .48 |
Note. CFI = comparative fit index; TFI = Tucker–Lewis index; RMSEA = root mean square of error approximation; AIC = Akaike information criterion; CI = confidence interval. Data were generated for 10 items and 1,000 simulees in flexMIRT with 100 replications per slope scaling factor.