Ecological validity of a Norwegian version of the Awareness of Social Inference Test

The ecological validity of the Norwegian version of The Awareness of Social Inference Test (N-TASIT), a test of social cognition, was explored in individuals with (<i>n</i> = 101) and without (<i>n</i> = 50) traumatic brain injury (TBI). Linear regression analyses revealed that for participants with moderate to severe TBI, N-TASIT performance was associated with proxy-reported (R² = 0.05, F(1, 60) = 4.21, <i>p</i> = 0.045), but not self-reported, social skills assessed with the Social Skills after TBI Questionnaire. Adding demographic and cognitive variables to the regressions did not increase explanatory power. When the TBI and non-TBI sample were collapsed, this result was present in both Virtual reality (VR) and the 2D version. For the VR version, N-TASIT performance was the only significant variable, while for 2D, only processing speed was predictive. There were no differences in experienced presence in the VR version between individuals with TBI and non-injured individuals, but the TBI group had a lower predisposition to engage in media experiences. Finally, adding individual differences in immersive tendency to the regression model increased the explained variance in self-reported social skills, but only in VR. The results indicate modest ecological validity and support the use of the VR version of N-TASIT in assessment of social cognitive impairment after TBI.