The result pattern for Asian Canadians showed some inconsistency across studies. European Canadians recognized White faces better than East Asian faces in both studies. In contrast, Asian Canadians showed a better memory for Asian faces in Study 1 (personality), whereas they better recognized White faces in Study 2 (university affiliation), suggesting a culture-specific effect of group membership. Generally, when observing and memorizing a sequence of many faces, one's cognitive capacities are quickly exceeded, so one has to prioritize the targets. In the absence of other veridical information about a target's relevance we usually refer to contextual information to reduce situational uncertainty (Kaspar, 2013; Kaspar and Krull, 2013) and ambiguity (Kaspar, in press). Correspondingly, color as a contextual sign for one's university affiliation might have helped to identify and separate targets of high (ingroup) and low (outgroup) priority in European Canadians, reflected in a better recognition of ingroup faces. In contrast, social categorization in terms of university affiliation might have primed thoughts of acculturation processes in Asian Canadians instead. Given the central argument of motivationally driven prioritization of visual input (Hugenberg et al., 2010), it is conceivable that the first-generation Asian Canadians tested by Ng and colleagues have been sensitized toward acculturation processes, leading to an increased motivation to accurately process White faces representing the host culture's population. In contrast, when color indicated the personality type (Study 1), such cognitive processes were rather unlikely and hence cross-race effects were visible in both European and Asian Canadians. Consequently, in order to verify the validity of the reported results, the study should be replicated with Asians living in a respective country without a personal involvement in acculturation processes. More generally, it should be considered that test-taking behavior can be influenced by participants' relation to the culture in which the test is administered (Cofresi and Gorman, 2004).

The study by Ng and colleagues shares one methodological limitation with several other studies investigating group effects on face recognition performance, because it remains open whether group membership affects cognitive processes in the learning or/and recognition phase. In addition to the judgmental responses investigated by Ng et al., other research suggests that the recognition bias for ingroup faces reflects a motivationally driven top-down modulation of cognitive processes, leading to an increased activity of the fusiform face area in the recognition phase (Van Bavel et al., 2011). Thus, group membership might affect selective attention and cognitive processing in terms of encoding face attributes when observing target faces in the learning phase. Alternatively, group membership might (also) increase the motivation to thoroughly compare memory content with current visual input in the recognition phase. To compare these alternatives, future studies should extend the traditional study design used by Ng et al. by two additional conditions, as illustrated in Figure 1.

Furthermore, it is important to pay attention when selecting face stimuli for cross-cultural comparisons. Previous studies revealed that attractive faces can intensify approach motivation (Zebrowitz et al., 2015) and that face recognition performance may depend on the emotion displayed (Di Domenico et al., 2015; Righi et al., 2015). Given cultural differences in the judgment of facial attractiveness (Sorokowski et al., 2013) and in the interpretation of emotional facial expressions (Yuki et al., 2007), the attractiveness and emotional valence of same-race and other-race faces have to be validated by all investigated culture groups to prevent potential confounds.

Finally, let us assume the validity of Ng et al.'s focal hypothesis stating that Asians preferentially define their ingroups by preexisting relationships established through friendship and family. If so, it might be that we will nonetheless find ingroup favoritism regarding face recognition in Asians when a minimal group manipulation is applied to a set of faces sampled from the pool of the participant's family or group of friends. Indeed, such a “second order” effect appears to be an inherent part of the assumption that ingroup membership is initially determined by personal relationships in (some) Asian cultures, while this does not make the prioritization of multiple targets superfluous in the face of strongly limited cognitive capacities.

The author confirms being the sole contributor of this work and approved it for publication.