Skip to main content
. 2013 Mar;74(2):320–328. doi: 10.15288/jsad.2013.74.320

Table 4.

Results of two-group item response theory analysis

Symptom Constrained parameter No. of free parameters Log-likelihood Δχ2 df P Sample size- adjusted BIC
Base model 31 −6,922.92 13,965.44
Withdrawal Loading 30 −6,923.69 1.60 1 .2060 13,962.13
Threshold 30 −6,928.01 11.65 1 .0006* 13,968.13
Caused problems Loading 30 −6,923.63 1.31 1 .2521 13,961.99
Threshold 30 −6,938.48 24.05 1 <.0001* 13,997.66
Role failure Loading 30 −6,923.04 0.20 1 .6578 13,963.86
Threshold 30 −6,929.87 13.98 1 .0002* 13,979.96
Hazardous use Loading 30 6,926.58 5.65 1 .0174 13,965.06
Threshold 30 6,935.48 26.88 1 <.0001* 13,990.00
Social problems Loading 30 6,924.02 2.15 1 .1424 13,962.21
Threshold 30 6,925.89 7.56 1 .0059* 13,969.93
Final model- All factor loadings constrained 26 −6,929.00 11.52 5 .0420 13,958.30

Notes: BIC = Bayesian information criterion; bolded text indicates best-fitting model.

*

Indicates significant result at p < .01.