Page 169 - Emotions through the eyes of our closest living relatives- Exploring attentional and behavioral mechanisms
P. 169
Validation of the Pictorial Implicit Association Test
tasks) affects subsequent D-scores on the final task (e.g., due to a priming effect). For this, we first calculated a difference score between performance on the tests by subtracting the D-score from the second task from the first. We then fitted a linear model using First Task (PIAT first or WIAT first) a sum-to-zero coded fixed effect and Difference Score as the dependent variable. Furthermore, see supplemental Figure S5 for our sensitivity power analysis results.
Results
Implicit associations
For the PIAT, we found a significantly positive D-score average of .18 (95% CI [.07, .28], t(139)= 3.34, p = .001, Cohen’s d = .28), meaning that participants associated faces of men of Moroccan descent with negativity, and faces of men of Dutch descent with positivity (Figure 4). In the linear model controlling for Congruency and Location, this effect remained present (D-score: .17, 95% CI [.07, .27], t(137)= 3.36, p = .001). We also found a significant effect of Congruency on D-scores (F(1, 137) =7.03, p = .009),
7
Figure 4. D-score distribution for the online PIAT (left) and online WIAT (right). Positive values represent stronger associations between faces of men of Dutch descent and positive scenes, and faces of men of Moroccan descent and negative scenes.
167
