Skip to main content
. 2022 Apr 11;32(9):1679–1689. doi: 10.1007/s00787-022-01980-1

Table 2.

Hierarchical Regression Analyses for Structural and Pragmatic Language Skills with predictors CU Traits and Anxiety

Structural language Pragmatic language
β b [95% CI] R2 β b [95% CI] R2
Age −0.09 −1.43 [−3.72, 0.86] 0.28** 1.42 [0.54, 2.30]
Ethnicity 0.23*** 9.72 [4.23, 15.22] 0.20* 2.80 [0.69, 4.91]
SES 0.19** 2.06 [0.57, 3.54] 0.27*** 0.99 [0.42, 1.56]
Offender status −0.59*** −25.25 [−31.97, −18.53] −0.53*** −7.52 [−10.10, −4.94]
ICU 0.01 0.04 [−0.31, 0.38] 0.1 0.08 [−0.05, 0.22]
Anxiety 0.12 0.61 [−0.05, 1.28] 0.47*** 0.06 0.10 [−0.16, 0.36] 0.28***
ICU × anxiety −0.12 −0.07 [−0.15, 0.01] 0.48 −0.33*** −0.07 [−0.10, −0.04] 0.38***

Variables. Age (centered); Ethnicity (dichotomized with weighted effect size: Indigenous Australian = −0.54; non-Indigenous Australian = 0.46); SES (centered); Offender Status (dichotomized with weighted effect coding: Youth offender = 0.38; Non-offender = −0.62); CU Traits (ICU total scale score; centered); Anxiety (YSR anxious-depressed sub-scale score; centered); Language: Structural (CELF-4, Core Language Score; centered); Pragmatic (TASIT, Social Inference Minimal, sum of scores; centered)

***p ≤ .001; **p ≤ .01; *p ≤ .05

β standardized beta, b unstandardized beta, CI confidence interval