Page 46 - Demo

Chapter 244the 95% credible interval, and (III) the probability of direction (pd). The 95% CrI indicates the range within which the effect falls with 95% probability, while the pd indicates the proportion of the posterior distribution that is of the median%u2019s sign (Makowski et al., 2019). ResultsPre-registered AnalysesGeneral Attentional Bias to Emotions As expected, we found a significant congruency effect in the general model on the reaction times in the dot-probe task, B = -5.79, 95% CI [-9.80, -1.78], t(8697.27) = -2.83, p = .005 (see Figure 2A). Confirming an attentional bias to emotional expressions, participants were on average 5.79 ms faster when the dot appeared behind the emotional expression and 5.79 ms slower when the dot appeared behind the neutral expression than their average reaction times in trial with an emotion-neutral pair. There was, however, no evidence for an enhanced attentional bias to angry compared to other facial expressions, that is, no significant interaction between Emotion category and Congruency. In addition to the general congruency effect, we also observed significant effects of the included control predictor terms. Unsurprisingly, reaction times depended on Age, B = 2.94, 95% CI [1.78, 4.10], t(98.96) = 5.04, p < .001. While older participants overall reacted more slowly to the dot probe than younger participants, this effect was less marked in trials including a happy face, B = -0.27, 95% CI [-0.47, -0.07], t(8708.16) = -2.70, p = .007, as revealed by the significant interaction between Age and Emotion category. Lastly, Probe distance significantly predicted reaction times, with slower responses to far probes (i.e., a mismatch between handedness and probe location) compared to close probes (i.e., handedness and probe location matched), B = 30.94, 95% CI [27.62, 34.25], t(8703.67) = 18.30, p < .001. For an overview of the model fit, see Table 1 as well as Table S3 in Online Resource 1 for the coefficients of all factor levels. Social Anxiety Traits and the Attentional Bias to EmotionsAgainst our hypothesis, we did not find any alteration in the overall or emotionspecific attentional bias related to social anxiety trait levels. Further, the congruency effect which we observed in the general model did not reach significance in the model including Social anxiety traits (p = .166). Yet, the effect of Age on reaction