Table 4.
Correlations Between Parallel and 2PL IRT Fits, Data From 2PL IRT Model.
| Correlation in discrimination parameter |
Correlation in difficulty parameter |
Correlation in factor scores |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| 2PL data vs. parallel fit | Minimum | Median | Maximum | Minimum | Median | Maximum | Minimum | Median | Maximum |
| Case 1 | • | • | • | 1.000 | 1.000 | 1.000 | 0.995 | 0.997 | 0.998 |
| Case 2 | • | • | • | 1.000 | 1.000 | 1.000 | 0.995 | 0.997 | 0.999 |
| Case 3 | • | • | • | 1.000 | 1.000 | 1.000 | 0.995 | 0.998 | 0.999 |
| Case 4 | • | • | • | 1.000 | 1.000 | 1.000 | 0.996 | 0.998 | 0.998 |
| Case 5 | • | • | • | 1.000 | 1.000 | 1.000 | 0.996 | 0.998 | 0.999 |
| Case 6 | • | • | • | 1.000 | 1.000 | 1.000 | 0.997 | 0.998 | 0.999 |
Note. 2PL IRT = two-parameter logistic item response theory. • = Discrimination parameter is not available in the 1PL IRT model, and loadings are not unique in the parallel model. With these models, the correlation is not calculated.