Page 154 - Emotions through the eyes of our closest living relatives- Exploring attentional and behavioral mechanisms
P. 154

                                Chapter 7
to represent the concept categories, but retained words for describing the attribute dimensions (e.g., “good” and “bad”) in order to test implicit attitudes in 6- and 10-year- olds (Baron & Banaji, 2006, but also see a pictorial and touchscreen adaptation by Thomas et al., 2007). In the pre-school IAT, Cvencek and colleagues (2011) used schematic representations of smiling and sad faces to indicate the attribute dimensions “good” and “bad”, and pictures of flowers and insects for the concepts. Furthermore, during the task pictures and (spoken) words were alternated (Cvencek et al., 2011). In another child-friendly IAT, only pictures of White and Black faces were used, together with line drawings of happy and sad faces (Rutland et al., 2005). Instead of pushing buttons that were mapped with the left and right superordinate categories, children had to move their mouse towards the target locations. Using a touchscreen, another adaptation of the IAT used pictures of faces that were shown in the middle of the screen with a happy and sad emoticon on the left and right side on the bottom of the screen. In critical blocks, children were told whether to press the “happy” or “sad” emoticon (Qian et al., 2016; Setoh et al., 2019). Notwithstanding the high usefulness of these tasks, a commonality in the adaptations is that child participants still received (extensive) instructions (for instance, extra verbal instruction during critical blocks (Qian et al., 2016; Setoh et al., 2019)), or the tasks combined words with pictures (Baron & Banaji, 2006; Cvencek et al., 2011). To this end, we focused on creating an entirely non-verbal pictorial version of the IAT (from hereon: PIAT) with stimuli that are proven to be interpreted similarly across cultures, and that requires a minimal amount of instructions to complete.
The aim of our study was to assess our PIAT’s validity in comparison to its classic counterpart. In Experiment 1, we test the PIAT in a large, diverse population consisting of Dutch adults and children that are visiting a zoo with two aims in mind, i) to assess whether our version of the PIAT can indeed measure implicit attitudes, and ii) to assess whether it is suitable for use in a heterogeneous sample in a more naturalistic environment (i.e., not in the lab). In Experiment 2, we directly compare the performance of an online version of our PIAT to an online word IAT (WIAT) with the aim to further validate it as a tool for measuring implicit attitudes. We have chosen to measure implicit attitudes towards different ethnicities, as these have also been extensively studied in traditional IATs (Baron & Banaji, 2006; Greenwald et al., 1998). In both experiments, we assess participants’ implicit attitudes towards individuals of Moroccan and Dutch descent using images from validated face and emotional images databases (Lang et al., 2007; Langner et al., 2010). We chose for these ethnicities because negative opinions about individuals of Moroccan descent
152































































































   152   153   154   155   156