Table 2.
Fit statistics, parameters for each item of the GBS-A for gender groups, and DIF analysis of discrimination and severity parameters across genders.
Males | Females | aDIF | bDIF | ||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Item | DSM-5 criterion | S-χ2 (df) | p | a (SE) | b1 (SE) | b2 (SE) | S-χ2 (df) | p | a (SE) | b1 (SE) | b2 (SE) | χ2 | df | p | χ2 | df | p |
1 | Tolerance | 11.15 (12) | 0.517 | 3.56 (0.54) | 1.52 (0.13) | 2.71 (0.27) | 3.89 (7) | 0.793 | 3.27 (0.82) | 1.92 (0.19) | 3.07 (0.41) | 0.1 | 1 | 0.707 | 0.3 | 2 | 0.869 |
2 | Withdrawal | 16.19 (17) | 0.512 | 3.00 (0.43) | 1.37 (0.12) | 2.17 (0.20) | 19.71 (10) | 0.032 | 2.82 (1.22) | 1.91 (0.25) | 2.78 (0.48) | 0.2 | 1 | 0.622 | 1.8 | 2 | 0.414 |
3 | Loss of control | 20.30 (16) | 0.207 | 3.42 (0.60) | 1.56 (0.13) | 2.27 (0.21) | 7.69 (8) | 0.465 | 2.20 (0.42) | 2.15 (0.24) | 3.33 (0.51) | 1.0 | 1 | 0.311 | 0.7 | 2 | 0.708 |
4 | Preoccupation | 36.07 (19) | 0.012 | 1.98 (0.25) | 1.16 (0.11) | 2.04 (0.20) | 10.62 (13) | 0.644 | 1.65 (0.29) | 1.77 (0.22) | 2.61 (0.34) | 0.0 | 1 | 0.979 | 3.4 | 2 | 0.186 |
5 | Escape | 15.68 (16) | 0.477 | 2.34 (0.36) | 1.59 (0.15) | 2.76 (0.30) | 6.04 (5) | 0.304 | 2.40 (0.41) | 1.63 (0.16) | 3.73 (0.63) | 0.8 | 1 | 0.386 | 5.2 | 2 | 0.074 |
6 | Chasing | 15.40 (16) | 0.497 | 1.86 (0.22) | 0.75 (0.09) | 2.53 (0.24) | 15.90 (10) | 0.102 | 1.35 (0.22) | 0.96 (0.14) | 3.40 (0.48) | 0.5 | 1 | 0.467 | 2.0 | 2 | 0.368 |
7 | Lying | 15.36 (18) | 0.638 | 2.58 (0.35) | 1.35 (0.12) | 2.31 (0.22) | 6.30 (8) | 0.614 | 2.29 (0.47) | 2.04 (0.22) | 3.45 (0.52) | 0.1 | 1 | 0.809 | 5.9 | 2 | 0.053 |
8 | Risked/lost relationships, opportunities | 16.01 (14) | 0.312 | 3.27 (0.76) | 1.58 (0.15) | 2.46 (0.28) | 14.74 (8) | 0.064 | 3.34 (0.44) | 1.73 (0.13) | 2.63 (0.24) | 0.1 | 1 | 0.719 | 4.3 | 2 | 0.117 |
9 | Bail-out | 25.12 (21) | 0.241 | 1.78 (0.23) | 1.42 (0.15) | 2.74 (0.30) | 13.68 (11) | 0.251 | 1.85 (0.32) | 1.71 (0.21) | 3.22 (0.45) | 0.9 | 1 | 0.339 | 0.7 | 2 | 0.714 |
Parameters were computed under the GRM model (a, discrimination; b, severity). df, degrees of freedom; SE, standard error. Due to the large sample size α was fixed at 0.01.