Page 173 - Emotions through the eyes of our closest living relatives- Exploring attentional and behavioral mechanisms

Validation of the Pictorial Implicit Association Test
opposed results in the order effects, which suggests that order effects likely occurred due to noise or random differences between the groups of participants. Our data therefore do not clearly support an effect of block sequence on IAT scores. Order effects remain a topic of debate in the IAT literature, and thus deserve more attention in future studies.
Psychometrics (internal consistency)
The PIAT shows an acceptable internal consistency in children, and good internal consistency in adults. The result of the child PIAT is somewhat lower than internal consistencies revealed in a comparable PIAT used in children (e.g., α = 86.5 on average across two PIATs (Cvencek et al., 2011)), which could be explained by the more noisy setting in which the PIAT was distributed. Generally, our results are in line with the limited amount of studies using image-based IATS that report internal consistency values (Brand et al., 2014; Palfai et al., 2016; Slabbinck et al., 2011). Importantly, despite the less-controlled circumstances in which we conducted the experiment, the PIAT reveals ethnicity-based implicit associations consistent with previous findings (Dunham et al., 2008; Greenwald et al., 1998).
Choice of stimulus material
For all IATs, we created several different versions of the task that differed on the
stimuli that represented the attribute dimensions (in Experiment 1: PIAT Version 1,
2 and 3), or differed in the order of critical block presentation and the location of the
concepts and attributes on the screen (All IATS in Experiment 1 and 2). We found that 7 children showed a lower D-score average in version 1 of the PIAT in Experiment 1, and
that the spread in D-scores was higher in this version than in the other versions. This effect was, however, not present in the adult PIAT nor in the online PIAT and WIAT. It is important that the differences in the attribute dimensions are clear (i.e., clearly negative and positive) as the IAT effect relies heavily on responses that do not require a lot of deliberation (Lane et al., 2007), which is why we used different types of stimuli that were rated by children in Experiment 1 on valence and or whether they were low or high in arousal. We found that the ratings for the different stimuli were very close to each other (see Table 1c in supplements), thus it is unclear what it is about the stimuli used in task version 1 that resulted in lower D-score averages compared to the other versions. The IAT represents a family of instruments where differences in e.g., choice of stimuli can result in entirely different results, even when the versions were built with the aim to measure the same underlying construct (Foroni & Bel-Bahar,
171