Table 4 Goodness of fit measures of applying the Rasch IRT, MIRT, and testlet models and the bifactor models on the EFL reading comprehension tests
Examination | Model | Goodness of fit measures | ||
|---|---|---|---|---|
Log-likelihood | AIC | BIC | ||
2016 | Rasch IRT | − 16,288.60 | 32,599.21 | 32,668.98 |
Rasch MIRT | − 16,380.62 | 32,785.23 | 32,861.35 | |
Rasch testlet | − 16,237.48 | 32,501.00 | 32,583.00 | |
Confirmatory Bifactor | − 16,098.53 | 32,257.00 | 32,447.00 | |
Exploratory Bifactor | − 16,112.51 | 32,283.00 | 32,467.00 | |
2017 | Rasch IRT | − 14,226.57 | 28,475.14 | 28,541.99 |
Rasch MIRT | − 14,693.90 | 29,411.80 | 29,484.72 | |
Rasch testlet | − 14,206.54 | 28,439.00 | 28,518.00 | |
Confirmatory Bifactor | − 14,056.71 | 28,173.00 | 28,356.00 | |
Exploratory Bifactor | − 14,067.30 | 28,193.00 | 28,369.00 | |
2018 | Rasch IRT | − 22,116.68 | 44,255.37 | 44,325.90 |
Rasch MIRT | − 22,936.52 | 45,897.05 | 45,973.99 | |
Rasch testlet | − 22,085.99 | 44,198.00 | 44,281.00 | |
Confirmatory Bifactor | − 21,880.97 | 43,822.00 | 44,014.00 | |
Exploratory Bifactor | − 21,881.98 | 43,822.00 | 44,008.00 | |