Page 48 - Demo

Chapter 246model on the attentional bias to emotions, Congruency, B = -7.48, 95% CI [-13.06, -1.90], t(8439.91) = -2.63, p = 0.009, as well as the control terms Probe distance, B = 30.11, 95% CI [26.81, 33.41], t(8439.65) = 17.90, p < .001, and Age, B = 3.12, 95% CI [1.87, 4.37], t(94.96) = 4.97, p < 0.001, were significant predictors of reaction times in the model including Autistic traits. Participants reacted faster in congruent trials, as well as when the probe appeared on the side of their dominant hand and when they were younger. We also found a significant interaction between Emotion category and Age. Here, the slowing in reaction times with higher age was more pronounced in trials with angry facial expressions, B = 0.24, 95% CI [0.03, 0.44], t(8446.75) = 2.23, p = .026, and less pronounced in trials with happy facial expressions, B = -0.32, 95% CI [-0.52, -0.11], t(8432.03) = -3.01, p = .003. A summary of the model fit can be found in Table 2 and a closer description of all coefficients in Table S5 in Online Resource 1.Autistic Traits, Social Anxiety Traits and the Attentional Bias to Angry Facial ExpressionsIn line with previous research, the two questionnaire scores, indicating autistic trait and social anxiety traits, were significantly positively correlated in our sample, rs = .30, p < .001. Zooming in on a potential moderating effect of Social anxiety traits on the link between Autistic traits and the attentional bias to angry facial expressions, our model revealed a significant three-way interaction between Congruency, Social anxiety traits and Autistic traits, B = 0.03, 95% CI [0.01, 0.06], t(2056.63) = 2.52, p = .012. An examination of the predicted value plots (see Figure 3) as well as slope comparisons (see Table 3) at three different values on one of the trait dimensions (mean %u2013 1SD, mean, mean + 1SD) suggested that this interaction is likely to be driven by a decrease in attentional bias with higher autistic traits at a relatively %u201chigh%u201d social anxiety trait level (mean + 1SD) and/or an increase in attentional bias with higher social anxiety traits at a relatively %u201clow%u201d autistic trait level (mean %u2013 1SD). In the context of this three-way interaction, the two-way interactions between each trait dimension and Congruency were also approaching significance (see Table 2). However, when running the same model without the three-way interaction added, this was not the case. Similar to the previous models with multiple emotion categories, Congruency, B = 23.36, 95% CI [1.01, 45.72], t(2056.92) = 2.05, p = .041 , as well as the control predictors Age, B = 3.48, 95% CI [2.19, 4.78], t(92.90) = 5.33, p< .001, and Probe distance, B = 36.69, 95% CI [29.94, 43.44], t(2056.97) = 10.66, p < .001, were significant predictors in this model (see Table 2 for an overview of the model fit as well as Table S6 in Online Resource 1 for all coefficients).