The overarching aim of this study was to examine how repeated experiences of victimization in 8- to 12-year-olds relate to neural processing of and behavioral responses to social exclusion. Based on the literature, we hypothesized that victimization experiences would relate to stronger responses in the insula, ACC and lPFC during social exclusion (vs. inclusion). Based on the SIP model, we hypothesized that victimization experiences would positively correlate with retaliation goals as reflected by intentions to punish excluders. Repeated and recent victimization were not related to neural processing of social exclusion, although there were some indications that insula activity during social exclusion (vs. inclusion) correlated positively with recent victimization experiences. Similarly, neural processing of social exclusion was not related to intentions to punish, although there was some evidence that insula activity during social inclusion (vs. exclusion) related negatively to intentions to punish. Finally, we showed that only repeated (and not recent) victimization experiences related to increased intentions to punish excluders.
Peer Victimization and Behavioral Responses to Social Exclusion
We found that children generally experience exclusion as aversive, as reflected in lower mood and need satisfaction after social exclusion elicited through the Cyberball paradigm. However, only repeated victimization related positively to intentions to punish excluders. Generally, all children are inclined to punish excluders (Gunther Moor et al.,
2012; Will et al.,
2015; Will, Crone, et al.,
2016a; Will et al.,
2016b), yet in previous research peer adversity was not related to increased punishment intentions (Will, Crone, et al.,
2016a; Will et al.,
2016b). Possibly, more intense experiences like victimization (compared to peer rejection) drive the association. Another possibility is that the intention to punish may be driven by need satisfaction. Everyone feels worse after being excluded, however, children who feel less satisfied during inclusion (and after debriefing) were more inclined to punish excluders. Hence, increased intentions to punish others might depend more on their general state (general feelings of belonging) than on specific incidents. This interpretation would be in line with our finding that repeated victimization related positively to intentions to punish excluders, yet recent victimization did not. To further substantiate these results, future research could include intentions to punish includers and neutral others (Gunther Moor et al.,
2012) and relate this to victimization experiences. This way it can be examined whether the finding of repeated victimization and intention to punish generalizes over different types of peers involved in interactions. Furthermore, as first impressions typically last (Lee et al.,
2016), excluders could make a comeback in a follow-up game where they act as includers, to see whether the intentions to punishment last when peers adjust their behavior. With these suggested extensions to research, the social intricacies in relation to (repeated) victimization experiences can be further unraveled.
Neural Responses to Social Exclusion
In examining the neural correlates of social exclusion, we focused both on explicit exclusion (i.e., not receiving a ball in the exclusion block vs. receiving a ball in the inclusion block) and on incidental exclusion (i.e., not receiving a ball in the inclusion block vs. receiving a ball in the inclusion block). We found indications that explicit social exclusion related to heightened neural responses in the (bilateral) paracentral lobule and the (bilateral) occipital lobe, which is only partly in line with prior research (Mwilambwe-Tshilobo & Spreng,
2021; Vijayakumar et al.,
2017). Contrary to our expectations, explicit social exclusion did not result in neural activation in the insula or lateral PFC, although uncorrected whole-brain analyses did show activation in these regions (untresholded statistical maps of the whole-brain contrasts are available on Neurovault (
https://identifiers.org/neurovault.collection:18019)). Another prior finding that was not replicated in the current study is the involvement of ACC during social exclusion (e.g., Cheng et al.,
2020; Will et al.,
2016). Possibly this is due to the few severe chronic victimized children in our sample, as Will, van Lier and colleagues compared severe long-term rejected over six years vs. stably accepted adolescents. Alternatively, the explanation may lie in the examined age groups. The children in our study were younger than in, for example, Will, van Lier and colleagues’ study, and ACC activity during exclusion has been found predominantly in (young) adult samples and not in developmental samples (Vijayakumar et al.,
2017), highlighting the importance of studies into (young) developmental samples.
Similarly, when we examined neural processing of incidental exclusion (i.e., not receiving a ball in the inclusion block vs. receiving a ball in the inclusion block), only an occipital cluster showed heightened activity. Finding an activated occipital cluster was not unique, as there was also an activated occipital cluster during explicit exclusion, and likely reflects visual attention to the stimulus (moving ball and figures). Previous findings for this contrast have been inconsistent, as a study with participants in the same age range as the current study found no heightened activity clusters (Asscheman et al.,
2019); yet two other studies with adolescents found similar activation clusters as during explicit exclusion (de Water et al.,
2017; Will et al.,
2016b). The studies with activated clusters had participants that were in the (12–17) adolescent age range, so perhaps they were at their prime for social sensitivity. In contrast, the participants in our study were still in (late) childhood (8–12 years) and may not have been as socially sensitive yet. Replication of these specific contrasts (incidental and explicit exclusion) in different (age) samples is needed to determine whether incidental exclusion is processed more similarly to explicit exclusion, to understand how the brain processes incidental instances of exclusion over the course of late childhood to adolescence.
We also examined explicit exclusion vs. incidental exclusion (i.e., not receiving a ball in the exclusion block vs. not receiving a ball in the inclusion block), which is very rarely examined in the literature. In line with previous research, we found that explicit exclusion was associated with increased activity in the occipital and striatal regions (Will et al.,
2016), indicating that explicit exclusion is
not processed the same as incidental exclusion. Other studies have also shown activity in parts of the cingulate cortex (Asscheman et al.,
2019; Schulz et al.,
2022). The striatal region has been shown to play an important role in (social) learning and prediction errors (Báez-Mendoza & Schultz,
2013; Schönberg et al.,
2007). Moreover, the activated striatal circuitry also included the amygdala and IFG. The IFG has been associated with processing unpleasantness (negative vs. neutral stimuli) (Sambuco et al.,
2020), and has been implicated in cognitive control. This suggests that the negative social experiences during the exclusion block (i.e., explicit exclusion) might be more salient and possibly requires more cognitive control than the (negative) experiences during the inclusion block (i.e., incidental exclusion).
Finally, we examined neural correlates of being included (i.e., receiving balls in the inclusion block vs. not receiving balls in the exclusion block). These contrasts resulted in widespread neural activation including the (right) insula/striatal circuitry, (left) lateral PFC and bilateral SMA, in line with prior findings showing that inclusion experiences involve (pre)motor activity and reward processes (Gunther Moor et al.,
2012; Puetz et al.,
2014; Schulz et al.,
2022). Whereas the anterior insula has been commonly found more activated by negative stimuli (Büchel et al.,
1998; Carretié et al.,
2009; Mériau et al.,
2009), the middle insula has been found to be increasingly activated by more pleasant stimuli (Bartels & Zeki,
2004). A previous study found that putamen and insula activity could act as tracers of social cue accuracy to receive rewards (Henco et al.,
2020). Another study found insula/striatal activity when receiving positive feedback from peers, especially for very low or very high levels of adversity (Rudolph et al.,
2021). Together with these previous findings, our results suggest that the insula/striatal activation during inclusion may be related to experiencing a pleasant social interaction. Our inclusion ball vs. incidental exclusion contrast led to similar activity clusters, which strengthens this interpretation.
Neural Responses to Social Exclusion, Peer Victimization and Intention to Punish
In ROI analyses, we examined whether children with victimization experiences showed differential neural processing in regions previously associated with social exclusion, i.e., the insula, dACC, IFG and dlPFC. Repeated and recent victimization were not related to neural processing of explicit exclusion or incidental exclusion in these regions, which may be partly explained by the relatively low variability in victimization scores in the sample. However, while it did not survive multiple comparison corrections, the univariate results gave some indications that recent victimization related positively to activity in the insula during explicit exclusion (both relative to inclusion and incidental exclusion), which is in line with previous research on victimization and social exclusion (Kiefer et al.,
2021; McIver et al.,
2018). Given that the insula has been related to threat processing and cognitive control functions (Puiu et al.,
2020; Sambuco et al.,
2020; Tops & Boksem,
2011), it is possible that being socially excluded is a more intense experience for children who are victimized, requiring more cognitive control. This interpretation was paralleled by our behavioral findings that children with higher victimization scores had more intentions to punish excluders. Importantly, the interactions in our study were with unfamiliar peers. Therefore, interactions with personally familiar peers, such as classmates (and bullies), are likely to elicit even stronger responses. Currently, neuroscientific research on exclusion by familiar peers is lacking. It is very difficult to include personally familiar peers in experimental paradigms, especially those with negative relationships such as disliked peers and bullies (see Güroǧlu & Veenstra,
2021 for a discussion). Nevertheless, given that the effects did not survive multiple comparison corrections and had small to medium effect sizes (Cohen,
1969; Richardson,
2011), future research should replicate our findings and also aim to examine the influence of familiarity with peers as bullying is often done by familiar peers.
In exploratory analyses, we tested whether neural responses during social exclusion were related to intention to punish excluders in the Cyberball game. Results did not reveal associations between neural processing across ROIs and intention to punish. However, univariate tests of the separate ROIs provided some indications that increased activity in the insula during explicit exclusion (vs. incidental exclusion) was related to increased intention to punish excluders. This finding is in line with prior studies showing that insula activation was related to increased aggression in late childhood (Achterberg et al.,
2020), and decreased prosocial behaviors in childhood and adulthood (Schreuders et al.,
2018; van der Meulen et al.,
2018).