Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Interactive Virtual Reality versus Vignette-Based Assessment of Children’s Aggressive Social Information Processing

Interactive Virtual Reality versus Vignette-Based Assessment of Children’s Aggressive Social... This study examined whether interactive Virtual Reality (VR) provides a more ecologically valid assessment of children’s aggressive social information processing (SIP) and aggressive responses than a standard vignette-based assessment. We developed a virtual classroom where children could meet and play games with virtual peers. Participants were boys (N = 184; ages 7–13) from regular education and special education for children with disruptive behavior problems. They reported on their SIP in four scenarios (i.e., two instrumental gain and two provocation scenarios) presented through both interactive VR and vignettes. Teachers reported on children’s real-life aggressive behavior and reactive and proactive motives for aggression. Results demonstrated that children found the interactive VR assessment more emotionally engaging and immersive than the vignette-based assessment. Moreover, compared to vignettes, the interactive VR assessment evoked higher levels of aggres- sive SIP and responses in provocation scenarios only. Results supported the enhanced predictive validity of the interactive VR assessment of children’s aggressive SIP and responses, which predicted children’s real-life aggression above and beyond the vignette-based assessment with 2 to 12% additional explained variance. Similar results were found for children’s real- life reactive and proactive motives for aggression, with 3 to 12% additional variance explained by interactive VR above and beyond vignettes. Interactive VR did not, however, evoke larger individual differences (i.e., variances) in children’s aggressive SIP and responses than vignettes. Together, these findings suggest that interactive VR provides a more ecologically valid method to assess children’s aggressive SIP and responses than hypothetical vignettes. Keywords Social information processing · Aggression · Children · Virtual Reality · Reactive and proactive motives Children are often confronted with challenging social situa- anger, frustration, desire, or jealousy may trigger aggressive tions, such as not being allowed to join a peer group or being cognitions that would not have been triggered without these reprimanded by their teachers or parents. Such situations emotions. For instance, children may only interpret others’ are likely to elicit strong emotions, which may affect chil- behavior as hostile when they feel frustrated, or may only dren’s thinking and responding in these situations (Caporaso justify stealing when they strongly desire an object. Thus, & Marcovitch, 2021; Lemerise & Arsenio, 2000; Reijntjes to better understand, predict, and treat children’s aggressive et  al., 2011). In many children, strong emotions such as behavior, we need to assess how children think in social situations when they are emotionally engaged. Yet tradi- tional methods to assess children’s social information pro- * Rogier E. J. Verhoef cessing (SIP) often use hypothetical stories (i.e., vignettes) R.E.J.Verhoef@uu.nl that are unlikely to elicit strong emotions. We have therefore developed an interactive Virtual Reality (VR) environment Department of Developmental Psychology, Utrecht to assess children’s aggressive SIP and responses. The pre- University, Heidelberglaan 1, 3508 TC Utrecht, The Netherlands sent study examines whether our VR-based assessment of children’s SIP and responses better predicts their real-life Research Institute of Child Development and Education, University of Amsterdam, Nieuwe Achtergracht 127, aggressive behavior compared to a standard, vignette-based 1001 NG Amsterdam, The Netherlands assessment. Centre for Urban Mental Health, University of Amsterdam, Our interactive VR assessment is based on the SIP model Nieuwe Achtergracht 127, 1001 NG Amsterdam, (Crick & Dodge, 1994; Lemerise & Arsenio, 2000). This The Netherlands Vol.:(0123456789) 1 3 Research on Child and Adolescent Psychopathology SIP model proposes that children’s behavioral responses to (Verhoef et al., 2021b), suggesting that interactive VR may social situations result from a sequence of mental process- also enhance the prediction of individual differences in real- ing steps: (1) encoding of social cues, (2) representation life aggressive behavior. The present study capitalizes on of social cues, (3) specification of interactional goals, (4) these findings by examining whether our interactive VR generation of responses, (5) evaluation of responses, and assessment of children’s SIP indeed is (1) more immersive (6) enactment of a selected response. Children’s aggres- and emotionally engaging, and (2) more strongly associated sive behavior has been associated with deviations in each with children’s real-life aggression, compared to a vignette- of these SIP steps, such as biased encoding, making hostile based assessment of children’s SIP. intent attributions, setting interactional goals directed at Another advantage of using interactive VR may be that it revenge or instrumental gain, generating more aggressive allows for more precise assessment of distinct SIP patterns responses, and evaluating aggressive responses and their underlying reactive and proactive aggression (Dodge, 1991). outcomes more positively (for reviews, see: De Castro & Reactive aggression—an impulsive aggressive response to Van Dijk, 2017; Dodge, 2011). Moreover, children with perceived threat or provocation (Dodge, 1991)—may stem aggressive behavior problems are more likely to experience from SIP characterized by excessive anger, heightened sen- anger (De Castro & Van Dijk, 2017), and research suggests sitivity to threatening cues, a tendency to attribute hostile that their SIP is more strongly affected by negative emotions intent to others, and goals directed at self-defense or taking (De Castro et al., 2003). revenge (e.g., Hubbard et al., 2010; Martinelli et al., 2018). Previous work has shown that children’s SIP patterns Such reactive SIP patterns may particularly be triggered in explain substantial variance in their concurrent and future provocation contexts (Hubbard et al., 2010) where children aggressive behavior (e.g., De Castro & Van Dijk, 2017; are refused to join a peer group (i.e., social provocation) or Lansford et al., 2006; Verhoef et al., 2019). Nonetheless, a peer damages their property (i.e., object provocation). In findings vary considerably between studies and SIP meas- contrast, proactive aggression—planned aggressive behav- ures used. A meta-analysis (Verhoef et al., 2019) revealed ior aimed at obtaining a desired outcome (Dodge, 1991)— that the association between aggressive behavior and chil- may stem from SIP characterized by instrumental goals, dren’s hostile intent attributions was stronger in studies using positive outcome expectations of aggression, and positive actual social interactions (d = 1.33) than in studies using evaluations of aggression (e.g., Hubbard et al., 2010). Such vignettes (d = 0.23 to 0.44) or video-game tasks (d = 0.36; proactive SIP patterns may particularly be triggered in Yaros et al., 2014). The small to moderate effect sizes for instrumental gain contexts (Hubbard et al., 2010), where vignettes and video games may be due to a lack of emotional children have the opportunity to steal something (i.e., object engagement (i.e., vignettes may not evoke strong emotions) acquisition) or win a game by (i.e., competition). Although or limited ecological validity (i.e., video games may not reactive and proactive motives for aggression can be mixed resemble real-life social interaction). These findings align (e.g., taking revenge to show who is the boss; Bushman & with theoretical work suggesting that strong emotions such Anderson, 2001), there is ample empirical work to suggest as anger or frustration may trigger aggressive SIP patterns that they often occur in isolation (Polman et al., 2007; Van that are not triggered when children are calm (Anderson & Dijk et al., 2021). Earlier studies, however, have not always Bushman, 2002; Lemerise & Arsenio, 2000; Verhoef et al., found clearly delineated reactive versus proactive SIP pat- 2021a). However, few studies exist that used actual social terns; possibly because their vignette-based assessment did interactions to assess children’s SIP—possibly because using not evoke the specific emotions underlying real-life reactive this method is challenging in terms of standardization and and proactive aggression (e.g., Crick & Dodge, 1996; Dodge ethics (Underwood, 2005). et al., 1997; Oostermeijer et al., 2016; Stoltz et al., 2013). Ideally, SIP assessment would combine highly emotional Our interactive VR may address this issue by immersing engaging, realistic social interactions with adequate stand- children in engaging social interactions with virtual peers, ardization and ethically and practically feasible methodol- where they are actually (not just hypothetically) provoked ogy. To attain this goal, we developed an interactive VR or tempted to use aggression. They may experience anger, classroom where children can walk around freely, talk to frustration, or jealousy, activating the unique SIP patterns virtual peers, and play games, allowing us to present stand- underlying reactive- and proactive aggression. Interactive ardized social events within an engaging environment. As VR then allows children to actually aggress against virtual children are fully immersed in the VR environment, the peers instead of reporting on their hypothetical aggressive peer interactions they have (e.g., their game being ruined responses as with vignettes. Consequently, interactive VR by a peer) may evoke substantial levels of anger, frustration, permits an assessment of children’s outcome expectancies or jealousy. A recent pilot study revealed that our interac- and evaluations regarding their actual behavior instead of tive VR assessment evoked larger individual differences presenting them with hypothetical response options they in aggressive responses than a vignette-based assessment might never carry out in real life. 1 3 Research on Child and Adolescent Psychopathology In sum, the present study examines whether interactive whose behavior problems are so severe that they require VR provides a better assessment of children’s aggressive extra support that cannot be provided in regular education. SIP and responding than a standard, vignette-based assess- In our study, boys from special education were nominated ment. We chose vignettes for this comparison because they by their teacher for frequently showing aggressive behavior are the standard method to assess children’s SIP, and have problems. Boys were excluded if they had an IQ below 80 been shown to yield similarly modest associations with or an Autism Spectrum Disorder (ASD) according to their children’s real-life aggression as other methods, such as casefiles, or had a clinical score on ASD symptoms on the video-game tasks (Verhoef et  al., 2019). Children com- teacher-rated Social Emotional Questionnaire (SEQ; Scholte pleted both an interactive VR-based and a hypothetical & Van der Ploeg, 2007). Schools sent parents an information vignette-based assessment of SIP, and teachers reported on letter in which the study was explained. All parents provided their aggressive behavior. We had three main goals. First, written consent for their child’s participation in the study by we tested whether interactive VR, compared to vignettes, signing the attached informed consent form and returning it would elicit higher levels of emotional engagement (1a) and to their child’s teacher. Boys provided verbal assent. immersion (1b). Consequently, we expected that interac- tive VR would trigger aggressive SIP and response patterns Procedure that are not triggered when children are calm. This should result in larger individual differences (i.e., variances) in Participants were individually tested in a silent room at SIP and aggressive responses (1c), and higher scores on their school by trained graduate students or the first author. aggressive SIP and aggressive responses (1d). Moreover, Graduate students were trained in multiple sessions by the it should result in more congruent SIP and response pat- first author and were supervised during the first two assess- terns, visible as stronger correlations between all SIP and ments to ensure assessment fidelity. The interactive VR- and aggressive response variables in each scenario (1e). Sec- vignette-based SIP assessments both lasted 45 min and were ond, we examined whether interactive VR explained addi- completed on two different days with approximately one tional variance in children’s real-life aggressive behavior week in between. We counterbalanced the order of these reported by teachers, above and beyond the vignette-based assessments across participants to control for order effects. assessment. We examined this both for the assessment of At the end of each assessment, boys reported on their emo- children’s aggressive SIP (2a) and children’s aggressive tional engagement and immersion during the assessment. responses (2b). Third, we examined whether interactive VR Boys received a small monetary reward (€5) for their partici- explained additional variance in teacher-reported reactive pation. Teachers reported on boys’ aggressive behavior and and proactive motives for aggression, above and beyond the filled out the SEQ through online questionnaires (response vignette-based assessment—again, both for aggressive SIP rate = 98%). The study was approved by the Medical Ethics (3a) and aggressive responses (3b). Committee of University Medical Center Utrecht. Method Materials Participants Interactive Virtual Reality Environment Participants were 184 Dutch boys ages 7 to 13  years Participants wore VR glasses to immerse them in the VR (M = 10.22; SD = 1.30). They were recruited from 18 Dutch environment. They could walk around freely (in a demar- primary schools. Schools were from neighbourhoods repre- cated 4 × 4 m space), use controllers that mimicked their sentative of the Dutch population, with on average 9% inhab- hands, and respond in similar fashion as in real life: through itants with a Western migration background (SD = 3%), 13% verbal and physical behavior. The interactive VR environ- with a non-Western migration background (SD = 9%), 21% ment was designed as a virtual school classroom where par- with a lower educational level (SD = 4%), and with 7% of the ticipants could interact and play games with virtual peers households having a low-income (SD = 3%) (Statistics Neth- (for a detailed description of the interactive VR environment, erlands, 2018, 2019). To maximize variance in aggressive see: Verhoef et al., 2021b). We presented the virtual class- behavior, boys high on disruptive behavior problems were room to participants as an actual classroom where standard oversampled by including boys from special education for behavior rules applied (e.g., respecting other children) and disruptive behavior problems (n = 118) and a random sample where they would meet real children from other schools who of boys from regular education (n = 66). In the Netherlands, were also participating in the study. In reality, virtual peers special education for children with disruptive behavior prob- were controlled by the experimenter through default move- lems and/or psychiatric problems is reserved for children ment options and standardized verbal responses. 1 3 Research on Child and Adolescent Psychopathology Participants could play two games: (1) building a tower lose high scores and bonuses), allowing for a clean com- of blocks as high as possible, and (2) throwing five balls parison between assessment methods. We counterbalanced to hit as many cans from a table as possible. We designed the type of game across participants (i.e., participants who our VR assessment around these games to allow for both received the tower game in interactive VR, received the peer-directed aggression (e.g., hitting, name calling) and cans game with vignettes, and vice versa). As in most property-directed aggression (e.g., knocking over the peer’s vignette procedures, participants were told that they would tower). To increase participants’ emotional engagement and listen to stories about everyday social situations with peers to provide experimental control over gains and losses, we and were asked to imagine that each story actually hap- included high scores and bonuses for participants’ perfor- pened to them (Verhoef et al., 2019). mance during the games (e.g., building a high tower). The instructions, game rules, and score count were displayed on Measures a digital school board, which also explained these matters through standardized verbal instructions. Emotional Engagement Virtual Reality Scenarios We assessed children’s emotional engagement during the assessment in two ways. First, we used two items imme- Participants were presented with six VR scenarios in a fixed diately after each assessment to directly capture children’s order: (1) practice scenario, (2) neutral scenario, (3) object emotional engagement during the assessment, aiming to acquisition, (4) competition, (5) social provocation, and (6) minimize the effect of memory on their ratings (i.e., “How object provocation—all centering around one of the games angry did you feel when something bad happened to you in (i.e., the tower or cans game; randomly assigned). The prac- VR/vignettes?” and “How much did you care when some- tice VR scenario served to familiarize participants with the thing bad happened to you in VR/vignettes?”). Children VR environment and game rules by practicing the game responded on a rating scale from 1 (not at all) to 10 (very). without any virtual characters present. The neutral scenario We averaged the two items to create emotional engage- served to familiarize participants with the SIP questions by ment scores for both interactive VR (r = 0.83) and vignettes having them play the game while engaging in neutral small (r = 0.67). Second, to allow children to make a comparison talk with a virtual peer, and asking the SIP questions after- between the VR- and vignette-based assessment, we again wards. Next, participants completed the four experimental administered these two items after they had completed both scenarios, which we based on taxonomies of problematic assessments, but then phrased in comparative form (e.g., for situations for children with aggressive behavior problems the first item: “You have completed both the VR and the sto - (Matthys et  al., 2001). The first two scenarios involved ries. How angry did you feel when something bad happened instrumental gain. In the object acquisition scenario, par- to you in the VR? And in the stories?;” question order was ticipants had the opportunity to steal a block or ball from counterbalanced). We again averaged the two items to create the virtual peer, which would earn them additional points emotional engagement scores for interactive VR (r = 0.74) in the game. In the competition scenario, they could win and vignettes (r = 0.74). the game and thus earn additional points by sabotaging the virtual peer’s progress in the game (i.e., by knocking over the peer’s tower, ruining the virtual peer’s balls). The last Immersion two scenarios involved provocation. In the social provoca- tion scenario, participants were refused to join the game by We assessed children’s immersion during the assessment in two virtual peers. In the object provocation scenario, their two ways. First, we used six items immediately after each game was ruined by a virtual peer. As such, the provocations assessment, which were adapted from the Dutch translation caused them to earn no points. In the two provocation sce- of the Igroup Presence Questionnaire (Schubert et al., 1999). narios, participants could not obtain any points by respond- Two of the six items had low factor loadings (i.e., below ing aggressively. We expected these provocation scenarios 0.60) and were excluded. The four items used were: 1) “I to elicit the strongest emotions, and therefore presented them was totally caught up by the events in VR/vignettes;” 2) “I last to prevent carry-over effects. had the feeling that the events in VR/vignettes were actually happening to me;” 3) “During the VR/vignettes it felt like I Hypothetical Vignettes was actually experiencing the events;” and 4) “The events in VR/vignettes seemed almost real.” Participants rated the For the vignette-based SIP assessment, we developed audi- items on a scale from 1 (strongly disagree) to 5 (strongly otaped vignettes with the exact same content as the VR agree). We averaged across items to create immersion scores scenarios (e.g., describing how participants would gain or for both interactive VR (α = 0.78) and vignettes (α = 0.81). 1 3 Research on Child and Adolescent Psychopathology Second, to allow children to make a comparison between Interaction Goals. Interaction goals were assessed using the VR- and vignette-based assessment, we administered one one open-ended question following each VR-scenario: item after they had completed both assessments, but then “When the other boy did [behavior of other boy], you did phrased in comparative form (i.e., “You have participated [behavior of participant]. What was the reason you did in both the VR- and vignette-based assessment. How much this?” In line with earlier research (De Castro et al., 2012), did you have the feeling that the events in VR were actually the first author coded each answer as revenge goals (e.g., happening to you? And in the stories?”). Children responded “to retaliate,” “because I was angry,” “to defend myself”), on a rating scale from 0 (not at all) to 10 (very). instrumental goals (e.g., “to win the game,” “to show him who’s the boss”), goals underlying non-aggressive behav- ior (e.g., “to become friends,” “to avoid problems”), or no Aggressive SIP and Responses goals (e.g., “I don’t know”). A second rater also coded 35% of the transcriptions. Inter-rater reliability was excellent, We assessed participants’ aggressive SIP and responses in with Cohen’s κ ranging from 0.85-0.96 across scenarios two provocation scenarios and two instrumental gain sce- (M = 0.91, Mdn = 0.91). Scores for revenge goals were cre- narios (both in interactive VR and with vignettes). Initially, ated by assigning 1 to revenge goals codes and 0 to other we planned to create aggregate SIP and response variables codes. Similarly, scores for instrumental goals were created for provocation and instrumental gain contexts. However, by assigning 1 to instrumental goals codes and 0 to other we found low correlations for SIP and response variables codes. between the social provocation and object provocation sce- Aggressive Responses. We assessed participants’ nario (i.e., ranging from 0.37-0.60 in VR and from 0.27- behavioral responses in interactive VR through observa- 0.50 with vignettes) and between the object acquisition and tion. A trained research assistant made detailed descrip- competition scenario (i.e., ranging from 0.34-0.58 in VR and tions of participants’ behavioral responses in each VR- from 0.35-0.48 with vignettes), suggesting that aggressive scenario. The first author coded these descriptions into SIP and behavior may be highly situation specific (Dodge non-aggressive behavior (e.g., prosocial behavior, avoid- et al., 1985; Matthys et al., 2001). Hence, we decided to cre- ance), mild aggressive behavior (e.g., coercion, verbal ate variables for children’s SIP and aggressive responses for aggression), and severe aggressive behavior (e.g., physi- each scenario separately. cal aggression, destructive aggression) following stand- Interactive VR Assessment. We assessed participants’ ard coding procedures (De Castro et al., 2005). If mul- aggressive responses through observation of their behavior tiple codes applied, the highest category was scored. A in VR, and used self-report to assess their anger, intent attri- second rater also coded 35% of the behavioral descrip- butions, goals, outcome expectancies, and response evalua- tions. Inter-rater reliability was excellent, with κ ranging tions at the end of each VR-scenario. In between scenarios, from 0.92–1.00 across scenarios (M = 0.97, Mdn = 0.98). participants kept their VR-glasses on while replying ver- Because frequencies of mild aggressive behavior were low bally to the experimenter’s questions. For procedural clar- or even absent (i.e., 0 to 17% across VR-scenarios and ity, we assessed all SIP questions following all scenarios, vignettes, Mdn = 2%), we created a dichotomous variable even though we were only interested in proactive SIP in by coding mild and severe aggressive behavior as 1 and instrumental scenarios (i.e., instrumental goals, outcome non-aggressive behavior as 0. expectancies, and response evaluation) and reactive SIP in Outcome Expectancies. Outcome expectancies of provocation scenarios (i.e., anger, hostile intent attribution, aggression were assessed using one item following each and revenge goals). VR-scenario: “What did you expect would happen when Anger. Anger was assessed using one item following each you [behavior of participant]?” We coded only answers VR-scenario: “The other boy did [behavior of other boy]. of participants who had actually used aggression in How angry did this make you feel, on a scale from 1, mean- that VR-scenario and assigned missing values to other ing not at all, to 10, meaning very?”. answers. The first author coded each answer as posi- Hostile Intent Attribution. Intent attributions were tive outcome expectancies of aggression (e.g., “I would assessed using two items following each VR-scenario: “The win the game”), or no positive outcome expectancies of other boy did [behavior of other boy]. To what extent did he aggression (e.g., “He would dislike me”). A second rater try to be mean, on a scale from 1, meaning not at all, to 10, also coded 35% of the transcriptions. Inter-rater reliability meaning very?” and “To what extent did he try to hinder was excellent, with κ being 1.00 for each scenario. Scores you, on a scale from 1 to 10?” These two items were mod- for positive outcome expectancies of aggression were cre- erately to highly correlated within each of the four VR sce- ated by assigning 1 to positive outcome expectancies of narios (M = 0.83, Mdn = 0.87, range = 0.67-0.90) and were aggression and 0 to no positive outcome expectancies of therefore averaged within each VR-scenario. aggression. 1 3 Research on Child and Adolescent Psychopathology Response Evaluations. Positive evaluations of aggres- Real‑Life Aggressive Behavior sion were assessed using one item following each VR- scenario: “When the other boy did [behavior of other boy], Teachers completed two questionnaires to assess partici- you did [behavior of participant]. To what extent do you pants’ aggressive behavior in real life. First, teachers filled approve your behavior on a scale from 1, meaning not at out the Aggressive Behavior subscale of the Dutch version all, to 10, meaning very?” We only used scores of children of the Teacher Report Form (TRF; Verhulst et al., 1997). who had actually used aggression in that VR-scenario and They rated 20 items (e.g., “This child threatens others”) on coded other scores as missing. a 3-point Likert scale (1 = not true for this child, 2 = some- Participants’ outcome expectancies and response evalu- what true for this child, or 3 = very often true for this child). ations of aggression were only scored when they displayed Scores were averaged across items (α = 0.96). Second, they aggressive responses, limiting the number of observations filled out the Instrument for Reactive and Proactive Aggres - for these variables. Conversely, other SIP variables (i.e., sion (IRPA; Polman et al., 2009). This instrument differ - anger, hostile intent attributions, revenge goals and instru- entiates between the frequency of aggression on the one mental goals) could be scored irrespectively of whether hand, and the motives underlying aggression on the other participants engaged in aggressive responses, yielding full hand. We used the frequency scale to assess children’s real- data for these variables (see Table 1 for descriptive statis- life aggressive behavior. Teachers rated the frequency of 7 tics of SIP and aggressive response variables). distinct forms of aggressive behavior (i.e., kicking, push- Vignette Assessment. Children reported on their SIP ing, hitting, name calling, arguing, gossiping, and doing following each vignette. We used the same questions and sneaky things) in the previous month on a 5-point Likert coding schemes as used for the interactive VR-assessment, scale (1 = never, 2 = once, 3 = weekly, 4 = multiple times a except that we formulated the questions as hypothetical week, 5 = daily). Scores on these seven items were averaged (e.g., “How angry would you feel…?”) instead of actual (α = 0.90). IRPA frequency scores (M = 1.95, SD = 0.86) and (e.g., “How angry were you…?”). The two items assess- TRF scores (M = 1.67, SD = 0.57) were highly correlated ing intent attributions were averaged within each vignette (r = 0.85). We therefore standardized and averaged them to as they were highly correlated (M = 0.80, Mdn = 0.81, create a single aggressive behavior score. range = 0.68-0.90). Inter-rater reliability (κ) for open- ended questions was based on 35% of transcriptions and Reactive & Proactive Motives for Aggression was excellent for both interaction goals (range = 0.81–1.00, M = 0.91, Mdn = 0.91) and outcome expectancies We assessed reactive and proactive motives for aggression (range = 0.83–1.00, M = 0.94, Mdn = 1.00). We assessed by again using the IRPA (Polman et al., 2009), but this time participants’ anticipated behavioral responses for each the motive scales. For each form of aggression rated above vignette using an open-ended question (i.e., “What would 0, teachers rated 3 reactive motives items (e.g., “Because you do if [social event]?”). Inter-rater reliability was based someone teased or upset him”) and 3 proactive motives on 35% of the transcriptions and was excellent, with κ items (e.g., “To hurt someone or to be mean”) on a 5-point ranging from 0.91–1.00 (M = 0.94, Mdn = 0.93). Likert scale (0 = never, 1 = rarely, 2 = sometimes, 3 = often, Table 1 Descriptive statistics of SIP variables for each scenario between VR and Vignettes (VIG) 1 2 Object Acquisition Competition Social Provocation Object Provocation VR VIG VR VIG VR VIG VR VIG Anger 1.75 (1.78) 1.65 (1.62) 3.13 (2.66) 4.98 (3.09) 5.36 (2.88) 5.13 (2.94) 6.79 (2.76) 7.54 (2.58) Hostile Intent Attribution 1.47 (1.21) 1.59 (1.21) 2.17 (2.18) 5.00 (3.09) 5.23 (3.08) 4.75 (2.94) 7.35 (2.78) 6.65 (3.32) Revenge Goals 1 (1) 1 (1) 6 (3) 15 (8) 51 (28) 19 (10) 92 (51) 69 (38) Instrumental Goals 41 (23) 43 (24) 37 (20) 27 (15) 16 (9) 6 (3) 12 (7) 7 (4) Aggressive Responses 42 (23) 44 (24) 43 (24) 42 (23) 69 (38) 25 (14) 105 (58) 77 (42) Outcome Expectancies 10 (24) 14 (32) 18 (42) 3 (7) 2 (3) 0 (0) 0 (0) 5 (6) Positive Evaluations 5.17 (3.56) 4.32 (2.77) 4.91 (3.71) 4.26 (3.25) 4.04 (3.24) 6.24 (3.38) 5.22 (3.40) 5.49 (3.16) Columns display means (M) and standard deviations (SD) of Anger, Hostile Intent Attribution, and Positive Evaluations, and the number (n) and proportion (%) of children displaying Revenge Goals, Instrumental Goals, Aggressive Responses, and Positive Outcome Expectancies Scores only apply to children who displayed an Aggressive Response for this scenario Based on n = 179 because 5 children had missing data in VR/vignettes Based on n = 178 because 6 children had missing data in VR/vignettes 1 3 Research on Child and Adolescent Psychopathology 4 = always). For aggression frequency items rated 0, motives their aggressive behavior in real life compared to vignettes, scores were missing by design. We calculated reactive and we conducted the same hierarchical regression analyses proactive motives scores by averaging across all reactive as used for our second hypothesis, but then with reactive motives items (i.e., 3 items times 7 forms of aggression; motives as dependent variables for the provocation scenarios α = 0.94) and all proactive motives items (α = 0.95), respec- and proactive motives as dependent variables for the instru- tively. Thus, high scores on reactive (M = 2.75, SD = 0.94) mental gain scenarios. or proactive (M = 2.04, SD = 0.84) motives indicate that if participants engaged in aggressive behavior, they often had reactive or proactive motives. The correlation between Results reactive and proactive motives was non-significant (r = 0.14, p = 0.075). Preliminary Analyses Statistical Analyses Table 1 presents the descriptive statistics for all SIP vari- ables in both VR and vignettes. As most SIP variables were To test our first hypothesis that interactive VR is more skewed, we conducted our analyses using a bootstrapping engaging than vignettes, we considered five aspects. First, procedure with bias-corrected accelerated (BCa) 95% con- we examined whether interactive VR yielded higher mean fidence intervals (CI) based on 5000 resamples. levels of emotional engagement than vignettes, using paired Our VR elicited aggressive responses in 23% to 58% of t-tests. Second, we examined whether participants’ immer- children, depending on the scenario (Table  1). However, sion was higher in VR versus vignettes, also using paired few children who responded aggressively in the VR, also t-tests. Third, we examined whether VR elicited larger responded aggressively in the same scenario in vignettes individual differences in aggressive SIP and aggressive (i.e., 9 to 32% across scenarios, Mdn = 10%; see Supplemen- responses than vignettes. To this end, we used an adaptation tary Material Table S1). As a result, we had insufficient data of the Pittman-Morgan test which replaces Pearson’s r with to compare VR versus vignettes on SIP variables that were Spearman’s rank correlation to account for non-normal data only assessed if children actually responded aggressively (McCulloch, 1987). Fourth, we examined whether interactive (i.e., positive outcome expectancies and positive evaluations VR yielded higher scores on aggressive SIP and aggressive of aggression). We therefore reported descriptive statistics responses than vignettes, using paired t-tests for continuous for these two variables (see Table 1) but excluded them from SIP variables and McNemar’s tests for dichotomous SIP and our main analyses. response variables. Fifth, we examined whether VR yielded stronger correlations among SIP and aggressive responses Children’s Engagement in Interactive VR Versus than vignettes. To do so, we calculated correlations between Vignettes all SIP and response variables for each scenario using Pear- son’s r, Pearson’s π, and Point-Biserial correlations. Next, Emotional Engagement we tested for inequality of the obtained correlation matri- ces using Steiger’s test (1980), which directly compares all As predicted, children reported feeling more emotionally elements of two dependent correlation matrices instead of engaged in VR (M = 5.59, SD = 2.74) than with vignettes comparing each correlation separately. (M = 2.91, SD = 2.37), BCa 95% CI [2.23, 3.10], p < 0.001,, To test our second hypothesis that interactive VR assess- d = 0.92. The same result was found when we asked chil- ment of aggressive SIP (2a) and responses (2b) better pre- dren about their engagement after they had completed both dicts children’s aggressive behavior in real life compared to assessments: their engagement was higher in VR (M = 5.68, vignettes, we examined whether VR explained additional SD = 2.69) than with vignettes (M = 3.05, SD = 2.16), BCa variance in real life aggression above and beyond vignettes, 95% CI [2.23, 3.04], p < 0.001, d = 0.96. but not vice versa. For aggressive SIP, we conducted two hierarchical regression analyses: the first with vignette- Immersion assessed SIP entered at step 1 and VR-assessed SIP at step 2; the second with VR-assessed SIP at step 1 and vignette- As predicted, children reported feeling more immersed in assessed SIP at step 2. For aggressive responses, we repeated VR (M = 4.18, SD = 0.83) than with vignettes (M = 2.57, these analyses with VR- versus vignette-assessed aggressive SD = 1.07), BCa 95% CI [1.44, 1.77], p < 0.001, d = 1.45. responses as predictors. They also reported feeling more immersed in VR (M = 7.96, To test our third hypothesis that interactive VR assess- SD = 2.21) than with vignettes (M = 3.70, SD = 2.37) after ment of aggressive SIP (3a) and responses (3b) better pre- they had completed both assessments, BCa 95% CI [3.85, dicts children’s reactive and proactive motives underlying 4.64], p < 0.001, d = 1.52. 1 3 Research on Child and Adolescent Psychopathology Variances in Aggressive SIP and Responses limited. Steiger’s test revealed that the correlation matrix of aggressive SIP and response variables was significantly We found limited support for our hypothesis that VR elicits higher for VR versus vignettes for the competition scenario, larger variances in SIP variables and aggressive responses χ (1) = 23.33, p < 0.001, but did not significantly differ than vignettes. For the object acquisition and competi- between VR and vignettes for the object acquisition sce- tion scenario respectively, we found no difference in vari- nario, χ (1) = 0.03, p = 0.862, social provocation scenario, ances between VR and vignettes for instrumental goals, χ (6) = 6.58, p = 0.361, and object provocation scenario, t(178) < 0.01, p = 1.000, and t(179) = 1.56, p = 0.120, χ (6) = 6.46, p = 0.374. and aggressive responses, t(178) < 0.01, p = 1.000, and In sum, children reported more emotional engagement t(179) = 0.21, p = 0.835. In the social provocation sce- and immersion in VR than with vignettes. Partial support nario, we did find higher variances in VR versus vignettes was found for VR outperforming vignettes on other aspects: for revenge goals, t(178) = 4.87, p < 0.001, and aggres- It yielded more variance for 2 out of 12 results, higher levels sive responses, t(178) = 4.37, p < 0.001, but not for anger, of aggressive SIP and responses for 6 out of 12 results, and t(178) = -0.28, p = 0.777, and hostile intent attributions, stronger correlations for 1 out of 4 results. t(178) = 0.61, p = 0.544. Last, in the object provocation sce- nario, we found no support for our hypothesis: there were Predicting Real‑Life Aggressive Behavior no differences for anger, t (177) = 0.91, p = 0.364, revenge goals, t(177) = 0.33, p = 0.738, and aggressive responses, Tables 3 and 4 present the results of the hierarchical regres- t(177) = -0.02, p = 0.983, and, contrary to our expectation, sion analyses of aggressive behavior in real life regressed larger variances with vignettes versus VR for hostile intent on aggressive SIP a) and aggressive responses b), first con- attributions, t(177) = -2.08, p = 0.039. ducted with vignettes in Step 1 and VR in Step 2, and next with VR in Step 1 and vignettes in Step 2. Analyses were Levels of Aggressive SIP and Responses conducted for each scenario separately. We tested whether VR yielded more aggressive SIP and Aggressive SIP responses than vignettes (Table 1). Details of the McNe- mar’s test for dichotomous variables can be found in Sup- Children’s aggressive SIP in all four VR scenarios signifi- plementary Material Table S1. Our hypothesis was partly cantly predicted their real-life aggression, with explained supported. For the object acquisition and competition sce- variances at Step 1 ranging from 4 to 13% across scenarios. narios respectively, we found no differences between VR As expected, effects were weaker for vignettes. Children’s and vignettes in instrumental goals, p = 1.000, OR = 1.000, aggressive SIP assessed with vignettes significantly pre - p = 0.154, OR = 1.67, nor aggressive responses, p = 1.000, dicted their real-life aggression at Step 1 in the object acqui- OR = 1.00, p = 0.885, OR = 1.09. For the social provoca- sition scenario (R = 0.03) and social provocation scenario 2 2 tion and object provocation scenarios respectively, chil- (R = 0.05), but not in the competition (R = 0.02) and object dren showed more hostile intent attributions, BCa 95% CI provocation scenario (R = 0.04). Turning to the incremental [0.00, 0.97], p = 0.048, d = 0.15, BCa 95% CI [0.07, 1.29], value of VR, we found that VR entered at Step 2 explained p = 0.025, d = 0.17, revenge goals, p < 0.001, OR = 4.67, significant variance over and above vignettes in all scenarios p = 0.014, OR = 1.85, and aggressive responses, p < 0.001, (i.e., 2% in object acquisition, 5% in competition, 12% in OR = 6.63, p = 0.001, OR = 2.35, in VR than with vignettes. social provocation, and 9% in object provocation). As pre- However, we found no differences in anger between VR and dicted, vignettes did not explain significant variance over vignettes in the social provocation scenario, BCa 95% CI and above VR in any scenario. [-0.24, 0.69], p = 0.354, d = 0.07, and even higher mean lev- els of anger for vignettes versus VR in the object provocation Aggressive Responses scenario, BCa 95% CI [-1.21, -0.28], p = 0.002, d = -0.24. Children’s aggressive responses in all four VR scenarios sig- Correlations between Aggressive SIP Variables nificantly predicted their real-life aggression, with explained and Responses variances at Step 1 ranging from 4 to 12% across scenarios. Similar effects were found for vignettes, with explained We tested whether correlations among aggressive SIP and variances at Step 1 ranging from 4 to 10%. Turning to the response variables were stronger in VR versus vignettes. incremental value of VR, we found that VR entered at Step 2 Table 2 presents all correlations between these variables for explained significant variance over and above vignettes in all each scenario separately. Steiger’s test to compare correla- scenarios (i.e., 2% in object acquisition, 5% in competition, tion matrices showed that support for our hypothesis was 9% in social provocation, and 7% in object provocation). 1 3 Research on Child and Adolescent Psychopathology Table 2 Bivariate Correlations Instrumental gain scenarios 1 2 3 4 5 6 of SIP and response Variables in * * * * * VR and Vignettes with real-life 1. VR: Instrumental Goals 0.98 0.27 0.29 0.19 0.27 aggressive behavior and reactive * * * * * 2. VR: Aggressive Responses 0.91 0.26 0.28 0.20 0.27 and proactive motives for * * * * 3. Vignette: Instrumental Goals 0.23 0.18 0.99 0.17 0.15 aggression in instrumental gain * * * * scenarios (object acquisition 4. Vignette: Aggressive Responses 0.27 0.24 0.76 0.20 0.16 * * * * scenario above the diagonal; 5. Real-Life Aggressive Behavior 0.26 0.29 0.17 0.32 0.58 competition scenario below) * * * * 6. Real-Life Proactive Motives 0.24 0.23 0.14 0.20 0.58 and provocation scenarios Provocation scenarios 1 2 3 4 5 6 7 8 9 10 (social provocation scenario * * * * * * * above the diagonal; object 1. VR: Anger 0.52 0.41 0.41 0.37 0.40 0.21 0.23 0.12 0.06 * * * * * provocation scenario below) 2. VR: Hostile Intent Attribution 0.55 0.45 0.36 0.29 0.40 0.10 0.13 0.05 0.14 * * * * * * * * 3. VR: Revenge Goals 0.41 0.40 0.80 0.09 0.17 0.16 0.19 0.34 0.30 * * * * * * * 4. VR: Aggressive Responses 0.36 0.40 0.86 0.07 0.14 0.16 0.23 0.35 0.29 * * * * * 5. Vignette: Anger 0.30 0.18 0.11 0.10 0.52 0.32 0.41 0.06 -0.12 * * * * 6. Vignette: Hostile Intent Attribution 0.10 0.12 0.14 0.11 0.45 0.30 0.35 -0.05 -0.21 * * * * * * * * 7. Vignette: Revenge Goals 0.19 0.16 0.18 0.23 0.43 0.37 0.86 0.17 0.05 * * * * * * * * 8. Vignette: Aggressive Responses 0.19 0.19 0.24 0.28 0.39 0.39 0.91 0.23 0.09 * * * * * * 0.29 0.27 0.32 0.15 0.07 0.14 0.20 0.46 8. Real-Life Aggressive Behavior 0.20 * * * * * 9. Real-Life Reactive Motives 0.19 0.21 0.29 0.27 -0.04 -0.07 -0.04 0.01 0.46 Correlations of SIP and responses in Instrumental Gain Scenarios are reported in the upper part of the Table (Object Acquisition Scenario above the Diagonal; Competition Scenario below) and correlations of SIP and Responses in Provocation Scenarios in the lower part of the Table (Social Provocation Scenario above the Diagonal; Object Provocation Scenario below All correlations including Instrumental Goals, Revenge Goals or Aggressive Responses are point-biserial correlations, all correlations between Instrumental Goals, Revenge Goals and Aggressive Responses used Pearson’s π, and other correlations used Pearson’s r Indicates that the bootstrap 95% confidence interval did not include zero However, we also found that vignettes at Step 2 explained less pronounced for vignettes. Children’s aggressive SIP significant variance over and above VR in three scenarios, assessed with vignettes significantly predicted their reac- with higher levels of explained variance in the competition tive and proactive motives in the object acquisition scenario 2 2 scenario (i.e., 6%), but lower levels in in social provoca- (R = 0.03) and social provocation scenario (R = 0.06), but tion and object provocation scenarios (i.e., 3% and 2%, not in the competition (R = 0.02) and object provocation respectively). scenario (R < 0.01). Turning to the incremental value of In sum, all eight hierarchical regression analyses regard- VR, we found that VR entered at Step 2 explained signifi- ing children’s real-life aggression supported the incremental cant variance over and above vignettes in all scenarios (i.e., value of VR over vignettes, whereas only three analyses sup- 6% in object acquisition, 5% in competition, 12% in social ported the reverse. provocation, and 11% in object provocation). In contrast, we found that vignettes at Step 2 explained significant variance Predicting Reactive & Proactive Motives over and above VR only in the social provocation scenario (i.e., 8%). Next, we conducted the same set of hierarchical regression analyses as for children’s real-life aggressive behavior, in Aggressive Responses this case predicting children’s reactive and proactive motives for aggression. Detailed results of these analyses are pro- Children’s aggressive responses in all four VR scenarios vided in the Supplementary Materials (Table S2 and S3). significantly predicted their reactive and proactive motives in real life, with explained variances at Step 1 ranging from Aggressive SIP 5 to 9% across scenarios. Effects were weaker for vignettes. Children’s aggressive responses assessed with vignettes As predicted, children’s aggressive SIP in all four VR significantly predicted their reactive and proactive motives scenarios significantly predicted their reactive and proac - in the object acquisition scenario (R = 0.03) and competi- tive motives in real life, with explained variances at Step tion scenario (R = 0.05), but not in the social provocation 2 2 1 ranging from 6 to 10% across scenarios. Effects were (R = 0.01) and object provocation scenario (R < 0.01). 1 3 Research on Child and Adolescent Psychopathology 1 3 Table 3 Hierarchical regression analyses of real-life aggression regressed both on instrumental goals and aggressive responses Object acquisition scenario Competition scenario 2 2 Step Predictor β β SE 95% CI ∆R df F change β β SE 95% CI ∆R df F change 1 Vignette: Instrumental 0.37 0.20 -0.02-0.77 0.03 1,174 4.54 0.40 0.25 -0.07-0.91 0.02 1,175 3.75 Goals * ** 2 Vignette: Instrumental 0.27 0.19 -0.10-0.65 0.02 1,173 4.38 0.23 0.25 -0.26-0.75 0.05 1,174 10.27 Goals ** VR: Instrumental 0.37 0.20 -0.01-0.75 0.58 0.18 0.21-0.94 Goals * * ** *** 1 VR: Instrumental 0.44 0.20 0.04-0.82 0.04 1,174 6.66 0.63 0.19 0.27–1.01 0.07 1,175 13.02 Goals ** 2 VR: Instrumental 0.37 0.20 -0.01-0.75 0.01 1,173 2.30 0.58 0.19 0.22-0.95 0.01 1,174 1.18 Goals Vignette: Instrumental 0.27 0.19 -0.10-0.64 0.23 0.25 -0.21-0.72 Goals * * *** *** 1 Vignette: Aggressive 0.44 0.20 0.07-0.83 0.04 1,174 6.86 0.74 0.20 0.36–1.15 0.10 1,175 19.98 Responses * ** ** 2 Vignette: Aggressive 0.34 0.19 -0.02-0.72 0.02 1,173 4.12 0.61 0.19 0.24–1.00 0.05 1,174 9.19 Responses ** VR: Aggressive 0.35 0.19 -0.00-0.73 0.50 0.17 0.15-0.84 Responses * ** *** *** 1 VR: Aggressive 0.45 0.20 0.08-0.84 0.04 1,174 7.09 0.66 0.18 0.31–1.00 0.08 1,175 16.07 Responses ** *** 2 VR: Aggressive 0.35 0.19 -0.00-0.72 0.02 1,173 3.90 0.50 0.17 0.17-0.83 0.06 1,174 12.39 Responses ** Vignette: Aggressive 0.34 0.19 -0.01-0.71 0.61 0.19 0.24-0.97 Responses Hierarchical Regression Analyses were run for the Two Instrumental Gain Scenarios separately, both with Vignettes and VR Entered First. Model output is based on a non-bootstrapped proce- dure whereas output on separate predictors is based on a bootstrapping procedure * ** *** p < .05; p < .01; p < .001 Research on Child and Adolescent Psychopathology 1 3 Table 4 Hierarchical regression analyses of real-life aggression regressed both on reactive SIP and aggressive responses Social Provocation Object Provocation 2 2 Step Predictor β β SE 95% CI ∆R df F change β β SE 95% CI ∆R df F change 1 Vignette: Anger 0.01 0.03 -0.05-0.08 0.05 3,172 2.80 0.05 0.03 -0.01-0.11 0.04 3,171 2.55 Vignette: Hostile Intent -0.04 0.03 -0.10-0.02 -0.00 0.02 -0.05-0.04 Attribution Vignette: Revenge 0.67 0.31 0.09–1.28 0.22 0.17 -0.10-0.55 Goals *** ** 2 Vignette: Anger 0.02 0.03 -0.05-0.08 0.12 3,169 7.85 0.04 0.03 -0.01-0.10 0.09 3,168 5.73 Vignette: Hostile Intent -0.05 0.03 -0.11-0.02 -0.01 0.02 -0.05-0.03 Attribution Vignette: Revenge 0.50 0.28 -0.03–1.04 0.14 0.17 -0.18-0.47 Goals VR: Anger 0.01 0.03 -0.06-0.07 -0.01 0.03 -0.08-0.06 VR: Hostile Intent -0.03 0.04 -0.10-0.04 0.07 0.03 0.01-0.14 Attribution *** * VR: Revenge Goals 0.79 0.19 0.42–1.13 0.35 0.16 0.02-0.65 *** *** 1 VR: Anger 0.01 0.03 -0.06-0.08 0.13 3,172 8.67 0.01 0.03 -0.06-0.07 0.11 3,171 7.14 VR: Hostile Intent -0.04 0.03 -0.11-0.03 0.07 0.03 0.01-0.13 Attribution *** * VR: Revenge Goals 0.83 0.19 47–1.18 0.35 0.15 0.05-0.63 2 VR: Anger 0.01 0.03 -0.06-0.07 0.03 3,169 2.15 -0.01 0.03 -0.08-0.06 0.02 3,168 1.31 VR: Hostile Intent -0.03 0.04 -0.10-0.04 0.07 0.03 0.01-0.13 Attribution *** * VR: Revenge Goals 0.79 0.19 0.42–1.15 0.35 0.16 0.03-0.63 Vignette: Anger 0.02 0.03 -0.05-0.08 0.04 0.03 -0.01-0.09 Vignette: Hostile Intent -0.05 0.03 -0.12-0.02 -0.01 0.02 -.05-.03 Attribution Vignette: Revenge 0.50 0.28 -0.04–1.10 0.14 0.17 -0.19-0.49 Goals * ** ** ** 1 Vignette: Aggressive 0.68 0.26 0.19–1.21 0.06 1,174 11.12 0.44 0.15 0.16–74 0.05 1,173 9.80 Responses *** * *** 2 Vignette: Aggressive 0.48 0.25 -0.02–1.00 0.09 1,173 18.07 0.30 0.14 0.02-0.58 0.07 1,172 14.04 Responses *** *** VR: Aggressive 0.60 0.15 0.33-0.89 0.53 0.13 0.28-0.79 Responses Research on Child and Adolescent Psychopathology Turning to the incremental value of VR, we found that VR entered at Step 2 explained significant variance over and above vignettes in all scenarios (i.e., 5% in object acquisi- tion, 3% in competition, 8% in social provocation, and 8% in object provocation). In contrast, we found that vignettes at Step 2 explained significant variance over and above VR only in the competition scenario (i.e., 3%). In sum, all eight hierarchical regression analyses regard- ing children’s reactive and proactive motives supported the incremental value of VR over vignettes, whereas only two analyses supported the reverse. Discussion This study tested whether interactive Virtual Reality (VR) provides a more ecologically valid assessment of social information processing (SIP) underlying aggressive behav- ior in children than a standard vignette-based assessment. In line with expectations, children reported that the inter- active VR assessment was more emotionally engaging and immersive than the vignette-based assessment. Moreover, the assessment of children’s aggressive SIP and responses in VR predicted their real-life aggressive behavior and reactive and proactive motives for aggression, above and beyond the vignette assessment. Interactive VR immerses children in emotionally engaging social interactions and enables them to actually aggress against virtual peers—an important difference with vignettes, which ask children to consider their hypo- thetical aggressive responses. Accordingly, interactive VR has evoked higher levels of aggressive SIP and responses in children in provocation scenarios, and improved the predictive validity of their assessed aggressive SIP and responses. These findings support the proposition that emo- tional engagement influences SIP and consequent behavior (Anderson & Bushman, 2002; Lemerise & Arsenio, 2000). Thus, the emotionally engaging nature of our interactive VR assessment seems to have triggered aggressive SIP patterns and responses that may only occur with sufficient emotional engagement. We expected that the engaging nature of interactive VR would also evoke larger individual differences in children’s aggressive SIP and responses, and stronger correlations between children’s aggressive SIP and responses compared to vignettes. However, interactive VR and vignettes gener- ally evoked similar variances in children’s aggressive SIP and responses, and similar correlations between aggres- sive SIP steps and responses. Perhaps, our vignettes val- idly assessed individual differences in children’s “calm” SIP; that is, the way they would reflect on social situations when they do not experience strong emotions. Such “calm” SIP may also differ between children and show similar 1 3 Table 4 (continued) Social Provocation Object Provocation *** *** *** *** 1 VR: Aggressive 0.68 0.15 0.38-0.99 0.12 1,174 23.86 0.62 0.13 0.35-0.87 0.10 1,173 19.80 Responses *** * *** * 2 VR: Aggressive 0.60 0.15 0.32-0.90 0.03 1,173 5.78 0.53 0.13 0.27-0.79 0.02 1,172 4.39 Responses Vignette: Aggressive 0.48 0.25 -0.01-0.98 0.30 0.14 0.02-0.58 Responses Hierarchical Regression Analyses were run for the Two Provocation Scenarios separately, both with Vignettes and VR Entered First. Model output is based on a non-bootstrapped procedure whereas output on separate predictors is based on a bootstrapping procedure * ** *** p < .05; p < .01; p < .001 Research on Child and Adolescent Psychopathology correlations between children’s SIP and responses as their Second, contrary to our predictions, children reported emotional SIP, but would be less suitable to predict chil- similar levels of anger in interactive VR and vignettes, and dren’s real-life aggression. Indeed, our findings showed even more anger with vignettes in the object provocation that interactive VR yielded incremental predictive value scenario. This finding contrasts with our finding that VR above and beyond the vignette-based assessment in all four is more emotionally engaging than vignettes. However, it scenarios, both for the prediction of children’s real-life may also reveal a potential limitation of vignettes: asking aggression (i.e., 2 to 12% additional explained variance) children to reflect on their anticipated anger in a hypothetical and underlying reactive and proactive motives (i.e., 3 to situation could lead them to overestimate how they would 12% additional explained variance). actually feel. Indeed, research has shown that individuals One unexpected pattern in our findings was that interac - generally find it difficult to report on anticipated negative tive VR seemed to outperform vignettes more so for provo- affective states and tend to overestimate them (Robinson & cation scenarios than for instrumental gain scenarios: the Clore, 2002). Although we do not know whether this was incremental predictive value of VR versus vignettes was actually the case, the stronger correlations of VR- versus larger in provocation scenarios (with 7 to 12% increases vignette-assessed anger with children’s real-life aggression in explained variance in children’s real-life aggression) indicate that children are more accurate when reporting on than in instrumental gain scenarios (with 2 to 5% increases their anger in interactive VR. Perhaps, as in interactive VR in explained variance in children’s real-life aggression), children are actually (not just hypothetically) provoked or and only in provocation scenarios children showed more tempted to use aggression, they may experience emotions aggressive SIP and responses in VR versus vignettes. more similar to daily life than the anticipated emotions Although we expected that the engaging nature of inter- assessed with vignettes. active VR would enhance children’s proactive aggressive This study had several strengths. To our knowledge, it is tendencies in instrumental gain scenarios as well (e.g., the first empirical study that used interactive VR to assess because the stakes are higher, so they would experience children’s aggressive SIP and responses and compared its more jealousy or desire), children did not show more pro- external validity directly to a standard vignette-based assess- active SIP and responses in VR versus vignettes. Possi- ment of children’s aggressive SIP and responses. Moreover, bly, the provocation scenarios were more salient than the we maximized clinically meaningful variance in children’s instrumental gain scenarios, because the instrumental gain SIP by recruiting boys from both regular and special educa- of points in the VR constituted no actual gain outside of tion for disruptive behavior problems. The use of interac- the game in the real world. As such, it makes sense that tive VR in a sample with substantial variance in children’s the incremental value of interactive VR was the largest SIP allowed us to test important hypotheses concerning the for children’s aggressive SIP and aggressive responses in validity of VR-based assessment of SIP. provocation scenarios. In sum, interactive VR seems to Our study also had its limitations. First, as few children yield an improved assessment of both children’s reactive responded with aggression in the same scenarios in both VR SIP and proactive SIP patterns and responses compared and vignettes, we were not able to analyze whether inter- to vignettes, but the difference in favor of interactive VR active VR provides an improved assessment of children’s is the largest when measuring children’s reactive SIP pat- positive outcome expectancies and response evaluations of terns and aggression. aggression, as these were assessed only if children had actu- Several findings on separate SIP steps warrant further ally aggressed. Consequently, we tested our hypotheses on discussion. First, children’s interactional goals were the children’s proactive SIP and responses in instrumental gain strongest SIP predictor of their real-life aggression and scenarios using two variables only (i.e., instrumental goals underlying motives, and yielded the largest effect sizes for and aggressive responses). Second, children’s responses levels of aggressive SIP in VR versus vignettes. Moreover, were coded for reliability by the first author, who may have as children’s revenge goals were strongly correlated with been biased because he was aware of the research questions. their anger and hostile intent attributions, they were the only However, inter-rater agreement with a second coder who significant SIP step predicting children’s real-life aggres- was blind to the research questions was excellent, suggest- sion and reactive motives for aggression in most analyses. ing that this bias was limited. Third, as interactive VR is Although such overlap among predictors (i.e., multicolline- obviously more time-consuming and costly to develop than arity) may seem problematic from a statistical point of view, vignettes, we were only able to include four assessment sce- it does make sense conceptually, because children’s interac- narios. Given that children may show aggression in various tional goals seem to be most proximal to their (aggressive) contexts (De Castro & Van Dijk, 2017), it can be assumed behavior and may often derive from preceding SIP steps that using only four scenarios involving playing games with such as anger and hostile intent attributions (Crick & Dodge, peers in a school-setting did not cover the broad range of 1994). social situations known to evoke aggression in children. 1 3 Research on Child and Adolescent Psychopathology Fourth, as children’s SIP and responses were only weakly involve play, settings that allow for the assessment of rela- to moderately associated across scenarios, we conducted our tional aggression (e.g., spreading rumors), or settings that analyses for each scenario separately. Although this finding better allow to examine cooperative behaviors. Last, using aligns with empirical research demonstrating that children’s interactive VR may minimize cognitive load), increasing aggression is situation-specific (e.g., De Castro & Van Dijk, the validity of children’s reported SIP. 2017; Matthys et al., 2001), it prohibited us from testing how In sum, this empirical study demonstrated that interac- reliable our SIP measurements were per type of scenario. tive VR is an improved method to assess children’s aggres- Last, since our study included only boys between 7–13 years sive SIP and behavior compared to a standard vignette- with limited diversity in cultural and socio-economic back- based assessment. The use of VR allows researchers ground, findings cannot directly be generalized to girls, and practitioners to assess aggressive SIP patterns in an older, or younger children, or children from other cultural emotionally engaging, ecologically valid context that is or socio-economic backgrounds than our sample. truly interactive and realistic. Ultimately, interactive VR There are both advantages and disadvantages of using may also facilitate interventions with children, because it interactive VR to assess children’s aggressive SIP and allows for extensive practice with the specific situations responses. One important disadvantage is that interactive relevant to their individual needs, with precise control to nature of VR makes establishing ambiguity of social situ- adapt difficulty and complexity during the intervention. ations more difficult than with vignettes. This interactive Moreover, practitioners may use cooperative contexts nature might enhance the experience of an actual social that yield rewards for specific desirable behaviors (e.g., interaction, however it might also affect ambiguity to some prosocial), reinforcing these behaviors repeatedly through extent (e.g., children who talked a lot with the virtual peer operant conditioning. As such, interactive VR may fur- during the interaction might be prone to attribute non-hostile ther our understanding of the SIP mechanisms underlying intent). Moreover, interactive VR is obviously costly and aggressive behavior problems in children and may enhance time-consuming to develop, and so it is relevant to directly assessment and intervention for children with aggressive compare this method to other assessment methods besides behavior problems. vignettes, such as video game tasks (e.g., Yaros et al., 2014). Supplementary Information The online version contains supplemen- That said, VR has multiple advantages over the use of tary material available at https://doi. or g/10. 1007/ s10802- 021- 00879-w . vignettes. In interactive VR, children are actually pro- voked, tempted to use aggression, and able to aggress Authors’ Contributions The study was designed by all authors. Material against virtual peers, and may therefore experience similar preparation was performed by all authors. Data collection was per- formed by Rogier E.J. Verhoef and trained graduate students. Analyses emotions as in real-life (e.g., anger, frustration), activat- were performed by Rogier E.J. Verhoef. The first draft of the manu- ing similar SIP patterns and responses as in real-world script was written by Rogier E.J. Verhoef and edited by all authors. All interactions. As such, researchers may examine the effect authors commented on previous versions of the manuscript. All authors of a broad range of emotions on children’s SIP; that is, not read and approved the final manuscript. only anger or frustration, but also shame, guilt, fear, desire or sadness. Relatedly, since children actually ‘respond’ Data Availability The data that support the findings of this study are available through the Open Science Framework at https:// doi. org/ 10. in VR, it is possible to include physiological indicators 17605/ OSF. IO/ 7SA6M of children’s arousal, permitting researchers to test more specific hypotheses on the role of emotional arousal in Code Availability The syntax of the analyses run for this study are children’s SIP and responses. Moreover, the large experi- available through the Open Science Framework (see link above). mental control over social stimuli provided by interactive VR (e.g., control over virtual peers’ nonverbal behaviors and emotional expressions) allows researchers to test Compliance with Ethical Standards more specific hypotheses about causal effects of subtle Funding This research was supported by a grant from the Netherlands social cues on children’s SIP and responses than has been Organization for Scientific Research to the last author (grant number feasible thus far. In addition, interactive VR may allow 453–15-004/511). researchers to use a broad variety of emotionally engag- ing contexts known to evoke aggression in children. For Conflicts of Interest We have no known conflict of interest to disclose. example, researchers may present children with more sali- ent cues to evoke proactive SIP and aggression (e.g., by Ethics Approval The study was approved by the Dutch Medical-Ethical Testing Committee Utrecht (METC-Utrecht) and conducted in accord- allowing children to obtain actual gains outside of the VR ance with the 2013 Helsinki Declaration. environment). Relatedly, researchers may also examine children’s SIP in other settings than playing games with Consent to Participate Written informed consent the study was peers, such as settings with parents, settings which do not obtained from parents and children provided verbal assent. 1 3 Research on Child and Adolescent Psychopathology Open Access This article is licensed under a Creative Commons Attri- Dodge, K. A., Lochman, J. E., Harnish, J. D., Bates, J. E., & Pettit, bution 4.0 International License, which permits use, sharing, adapta- G. S. (1997). Reactive and proactive aggression in school chil- tion, distribution and reproduction in any medium or format, as long dren and psychiatrically impaired chronically assaultive youth. as you give appropriate credit to the original author(s) and the source, Journal of Abnormal Psychology, 106(1), 37–51. https:// doi. provide a link to the Creative Commons licence, and indicate if changes org/ 10. 1037/ 0021- 843x. 106.1. 37 were made. The images or other third party material in this article are Dodge, K. A., McClaskey, C. L., & Feldman, E. (1985). Situational included in the article's Creative Commons licence, unless indicated approach to the assessment of social competence in children. otherwise in a credit line to the material. If material is not included in Journal of Consulting and Clinical Psychology, 53(3), 344–353. the article's Creative Commons licence and your intended use is not https:// doi. org/ 10. 1037/ 0022- 006X. 53.3. 344 permitted by statutory regulation or exceeds the permitted use, you will Hubbard, J. A., McAuliffe, M. D., Morrow, M. T., & Romano, L. need to obtain permission directly from the copyright holder. To view a J. (2010). Reactive and proactive aggression in childhood and copy of this licence, visit http://cr eativ ecommons. or g/licen ses/ b y/4.0/ . adolescence: Precursors, outcomes, processes, experiences, and measurement. Journal of Personality, 78(1), 95–118. https://doi. org/ 10. 1111/j. 1467–6494. 2009. 00610.x Lansford, J. E., Malone, P. S., Dodge, K. A., Crozier, J. C., Pettit, G. References S., & Bates, J. E. (2006). A 12-year prospective study of pat- terns of social information processing problems and external- Anderson, C. A., & Bushman, B. J. (2002). Human aggression. izing behaviors. Journal of Abnormal Child Psychology, 34(5), Annual Review of Psychology, 53(1), 27–51. https:// doi. or g/ 709–718. https:// doi. org/ 10. 1007/ s10802- 006- 9057-4 10. 1146/ annur ev. psych. 53. 100901. 13523 1xs Lemerise, E. A., & Arsenio, W. F. (2000). An integrated model of Bushman, B. J., & Anderson, C. A. (2001). Is it time to pull the emotion processes and cognition in social information process- plug on hostile versus instrumental aggression dichotomy? Psy- ing. Child Development, 71(1), 107–118. https:// doi. or g/ 10. chological Review, 108(1), 273–279. https:// doi. org/ 10. 1037/ 1111/ 1467- 8624. 00124 0033- 295X. 108.1. 273 Martinelli, A., Ackermann, K., Bernhard, A., Freitag, C. M., & Caporaso, J. S., & Marcovitch, S. (2021). The effect of taxing situations on Schwenck, C. (2018). Hostile attribution bias and aggression in preschool children’s responses to peer conflict. Cognitive Develop- children and adolescents: A systematic literature review on the ment, 57(6), 100989. https://doi. or g/10. 1016/j. cogde v.2020. 100989 influence of aggression subtype and gender. Aggression and Violent Crick, N. R., & Dodge, K. A. (1994). A review and reformulation of Behavior, 39, 25–32. https:// doi. org/ 10. 1016/j. avb. 2018. 01. 005 social-information-processing mechanisms in children’s social Matthys, W., Maassen, G. H., Cuperus, J. M., & van Engeland, H. adjustment. Psychological Bulletin, 115(1), 74–101. https://doi. (2001). The assessment of the situational specificity of chil- org/ 10. 1037// 0033- 2909. 115.1. 74 dren’s problem behaviour in peer–peer context. Journal of Child Crick, N. C., & Dodge, K. A. (1996). Social information processing Psychology and Psychiatry, 42(3), 413–420. https://doi. or g/10. deficits in reactive and proactive aggression. Child Develop-1111/ 1469- 7610. 00734 ment, 67(3), 993–1002. https:// doi. org/ 10. 2307/ 11318 75 McCulloch, C. E. (1987). Tests for equality of variances with paired De Castro, B. O., & Van Dijk, A. (2017). “It’s gonna end up with a data. Communications in Statistics - Theory and Methods, fight anyway”: Social cognitive processes in children with dis- 16(5), 1377–1391. https://doi. or g/10. 1080/ 03610 92870 88294 45 ruptive behavior disorders. In: Lochman, J. E., & Matthys, W. Oostermeijer, S., Nieuwenhuijzen, M., van de Ven, P. M., Popma, A., (Eds.), The Wiley handbook of disruptive and impulse-control & Jansen, L. M. (2016). Social information processing problems disorders (pp. 237–253). John Wiley & Sons Limited. https:// related to reactive and proactive aggression of adolescents in doi. org/ 10. 1002/ 97811 19092 254. ch15 residential treatment. Personality and Individual Differences, De Castro, B. O., Merk, W., Koops, W., Veerman, J. W., & Bosch, 90, 54–60. https:// doi. org/ 10. 1016/j. paid. 2015. 10. 035 J. D. (2005). Emotions in social information processing and Polman, H., de Castro, B. O., Koops, W., van Boxtel, H. W., & Merk, their relations with reactive and proactive aggression in W. W. (2007). A meta-analysis of the distinction between reac- referred aggressive boys. Journal of Clinical Child and Ado- tive and proactive aggression in children and adolescents. Jour- lescent Psychology, 34(1), 105–116. https:// doi. org/ 10. 1207/ nal of Abnormal Child Psychology, 35(4), 522–535. https://doi. s1537 4424j ccp34 01_ 10org/ 10. 1007/ s10802- 007- 9109-4 De Castro, B. O., Slot, N. W., Bosch, J. D., Koops, W., & Veerman, Polman, H., Orobio de Castro, B., Thomaes, S., & van Aken, M. J. W. (2003). Negative feelings exacerbate hostile attributions (2009). New directions in measuring reactive and proactive of intent in aggressive boys. Journal of Clinical Child and aggression: Validation of a teacher questionnaire. Journal of Adolescent Psychology, 31(1), 56–65. https://doi. or g/10. 1207/ Abnormal Child Psychology, 37(2), 183–193. https:// doi. org/ s1537 4424j ccp32 01_ 0610. 1007/ s10802- 008- 9266-0 De Castro, B., Verhulp, E. E., & Runions, K. (2012). Rage and Reijntjes, A., Thomaes, S., Kamphuis, J., Bushman, B., De Castro, revenge: Highly aggressive boys’ explanations for their B. O., & Teich, M. (2011). Explaining the paradoxical rejec- responses to ambiguous provocation. European Journal of tion-aggression link: The mediating effects of hostile intent Developmental Psychology, 9(3), 331–350. https:// doi. org/ 10. attributions, anger, and decreases in state self-esteem on peer 1080/ 17405 629. 2012. 680304 rejection-induced aggression in youth. Personality and Social Dodge, K. A. (1991). The structure and function of reactive and pro- Psychology Bulletin, 37(7), 955–963. https:// doi. org/ 10. 1177/ active aggression. In D. Pepler & K. H. Rubin (Eds.), The devel-01461 67211 410247 opment and treatment of childhood aggression (pp. 201–218). Robinson, M. D., & Clore, G. L. (2002). Belief and feeling: Evi- Lawrence Erlbaum Associates Inc. dence for an accessibility model of emotional self-report. Psy- Dodge, K. A. (2011). Social information processing patterns as chological Bulletin, 128(6), 934–960. https:// doi. org/ 10. 1037/ mediators of the interaction between genetic factors and life 0033- 2909. 128.6. 934 experiences in the development of aggressive behavior. In P. Scholte, E. M., & van der Ploeg, J. D. (2007). Handleiding Sociaal- Shaver & M. Mikulincer (Eds.), Human aggression and vio- Emotionele Vragenlijst (SEV). Bohn Stafleu van Loghum. lence: Causes, manifestations, and consequences (pp. 165– Schubert, T., Friedmann, F., & Regenbrecht, H. (1999). Embodied 185). American Psychological Association. presence in virtual environments. In R. Paton & I. Neilson 1 3 Research on Child and Adolescent Psychopathology (Eds.), Visual representations and interpretations (pp. 268– Verhoef, R. E. J., Van Dijk, A., & De Castro, B. O. (2021a). A Dual 278). Springer-Verlag. Mode Social Information Processing Model to Explain Individual Statistics Netherlands (2018, 2019). StatLine. Retrieved June 24, Differences in Children’s Aggressive Behavior. Clinical Psycho- 2021, from https:// www. opend ata. cbs. nl logical Science, Advance Online Publication. https:// doi. org/ 10. Steiger, J. H. (1980). Tests for comparing elements of a correlation 1177/ 21677 02621 10163 96 matrix. Psychological Bulletin, 87(2), 245–251. https://doi. or g/ Verhoef, R. E. J., Van Dijk, A., Verhulp, E. E., & De Castro, B. O. 10. 1037/ 0033- 2909. 87 (2021b). Interactive virtual reality assessment of aggressive social Stoltz, S., van Londen, M., Dekovic, M., De Castro, B. O., Prinzie, information processing in boys with behaviour problems: A pilot P., & Lochman, J. E. (2013). Effectiveness of an individual study. Clinical Psychology & Psychotherapy, 28(3), 489–499. school-based intervention for children with aggressive behav-https:// doi. org/ 10. 1002/ cpp. 2620 iour: A randomized controlled trial. Behavioural and Cogni- Verhulst, F. C., Van Der Ende, J., & Koot, H. M. (1997). Handleiding tive Psychotherapy, 41(5), 525–548. https:// doi. org/ 10. 1017/ voor de Teachers's Report Form (TRF). Rotterdam: afd. Kinder- s1352 46581 20005 25 en jeugdpsychiatrie, Sophia Kinderziekenhuis/ AZR/ EUR Underwood, M. K. (2005). Observing anger and aggression among Yaros, A., Lochman, J. E., Rosenbaum, J., & Jimenex-Camargo, L. A. preadolescent girls and boys: Ethical dilemmas and practical solu- (2014). Real-time hostile attribution measurement and aggression tions. Ethics and Behavior, 15(3), 235–245. https:// doi. org/ 10. in children. Aggressive Behavior, 40(5), 409–420. https://do i.or g/ 1207/ s1532 7019e b1503_410. 1002/ ab. 21532 Van Dijk, A., Hubbard, J. A., Deschamps, P. K. H., Hiemstra, W., & Polman, H. (2021). Do distinct groups of reactively and proac- Publisher's Note Springer Nature remains neutral with regard to tively aggressive children exist? Research on Child and Adoles- jurisdictional claims in published maps and institutional affiliations. cent Psychopathology. Advance online publication. https:// doi. org/ 10. 1007/ s10802- 021- 00813-0 Table 4 Hierarchical regression analyses of real-life aggression Verhoef, R. E. J., Alsem, S. C., Verhulp, E. E., & de Castro, B. O. regressed both on reactive SIP and aggressive responses (2019). Hostile intent attribution and aggressive behavior in children revisited: A meta-analysis. Child Development, 90(5), 525–547. https:// doi. org/ 10. 1111/ cdev. 13255 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Abnormal Child Psychology Springer Journals

Interactive Virtual Reality versus Vignette-Based Assessment of Children’s Aggressive Social Information Processing

Loading next page...
 
/lp/springer-journals/interactive-virtual-reality-versus-vignette-based-assessment-of-KtUdHAQ6yp
Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2021
ISSN
0091-0627
eISSN
2730-7174
DOI
10.1007/s10802-021-00879-w
Publisher site
See Article on Publisher Site

Abstract

This study examined whether interactive Virtual Reality (VR) provides a more ecologically valid assessment of children’s aggressive social information processing (SIP) and aggressive responses than a standard vignette-based assessment. We developed a virtual classroom where children could meet and play games with virtual peers. Participants were boys (N = 184; ages 7–13) from regular education and special education for children with disruptive behavior problems. They reported on their SIP in four scenarios (i.e., two instrumental gain and two provocation scenarios) presented through both interactive VR and vignettes. Teachers reported on children’s real-life aggressive behavior and reactive and proactive motives for aggression. Results demonstrated that children found the interactive VR assessment more emotionally engaging and immersive than the vignette-based assessment. Moreover, compared to vignettes, the interactive VR assessment evoked higher levels of aggres- sive SIP and responses in provocation scenarios only. Results supported the enhanced predictive validity of the interactive VR assessment of children’s aggressive SIP and responses, which predicted children’s real-life aggression above and beyond the vignette-based assessment with 2 to 12% additional explained variance. Similar results were found for children’s real- life reactive and proactive motives for aggression, with 3 to 12% additional variance explained by interactive VR above and beyond vignettes. Interactive VR did not, however, evoke larger individual differences (i.e., variances) in children’s aggressive SIP and responses than vignettes. Together, these findings suggest that interactive VR provides a more ecologically valid method to assess children’s aggressive SIP and responses than hypothetical vignettes. Keywords Social information processing · Aggression · Children · Virtual Reality · Reactive and proactive motives Children are often confronted with challenging social situa- anger, frustration, desire, or jealousy may trigger aggressive tions, such as not being allowed to join a peer group or being cognitions that would not have been triggered without these reprimanded by their teachers or parents. Such situations emotions. For instance, children may only interpret others’ are likely to elicit strong emotions, which may affect chil- behavior as hostile when they feel frustrated, or may only dren’s thinking and responding in these situations (Caporaso justify stealing when they strongly desire an object. Thus, & Marcovitch, 2021; Lemerise & Arsenio, 2000; Reijntjes to better understand, predict, and treat children’s aggressive et  al., 2011). In many children, strong emotions such as behavior, we need to assess how children think in social situations when they are emotionally engaged. Yet tradi- tional methods to assess children’s social information pro- * Rogier E. J. Verhoef cessing (SIP) often use hypothetical stories (i.e., vignettes) R.E.J.Verhoef@uu.nl that are unlikely to elicit strong emotions. We have therefore developed an interactive Virtual Reality (VR) environment Department of Developmental Psychology, Utrecht to assess children’s aggressive SIP and responses. The pre- University, Heidelberglaan 1, 3508 TC Utrecht, The Netherlands sent study examines whether our VR-based assessment of children’s SIP and responses better predicts their real-life Research Institute of Child Development and Education, University of Amsterdam, Nieuwe Achtergracht 127, aggressive behavior compared to a standard, vignette-based 1001 NG Amsterdam, The Netherlands assessment. Centre for Urban Mental Health, University of Amsterdam, Our interactive VR assessment is based on the SIP model Nieuwe Achtergracht 127, 1001 NG Amsterdam, (Crick & Dodge, 1994; Lemerise & Arsenio, 2000). This The Netherlands Vol.:(0123456789) 1 3 Research on Child and Adolescent Psychopathology SIP model proposes that children’s behavioral responses to (Verhoef et al., 2021b), suggesting that interactive VR may social situations result from a sequence of mental process- also enhance the prediction of individual differences in real- ing steps: (1) encoding of social cues, (2) representation life aggressive behavior. The present study capitalizes on of social cues, (3) specification of interactional goals, (4) these findings by examining whether our interactive VR generation of responses, (5) evaluation of responses, and assessment of children’s SIP indeed is (1) more immersive (6) enactment of a selected response. Children’s aggres- and emotionally engaging, and (2) more strongly associated sive behavior has been associated with deviations in each with children’s real-life aggression, compared to a vignette- of these SIP steps, such as biased encoding, making hostile based assessment of children’s SIP. intent attributions, setting interactional goals directed at Another advantage of using interactive VR may be that it revenge or instrumental gain, generating more aggressive allows for more precise assessment of distinct SIP patterns responses, and evaluating aggressive responses and their underlying reactive and proactive aggression (Dodge, 1991). outcomes more positively (for reviews, see: De Castro & Reactive aggression—an impulsive aggressive response to Van Dijk, 2017; Dodge, 2011). Moreover, children with perceived threat or provocation (Dodge, 1991)—may stem aggressive behavior problems are more likely to experience from SIP characterized by excessive anger, heightened sen- anger (De Castro & Van Dijk, 2017), and research suggests sitivity to threatening cues, a tendency to attribute hostile that their SIP is more strongly affected by negative emotions intent to others, and goals directed at self-defense or taking (De Castro et al., 2003). revenge (e.g., Hubbard et al., 2010; Martinelli et al., 2018). Previous work has shown that children’s SIP patterns Such reactive SIP patterns may particularly be triggered in explain substantial variance in their concurrent and future provocation contexts (Hubbard et al., 2010) where children aggressive behavior (e.g., De Castro & Van Dijk, 2017; are refused to join a peer group (i.e., social provocation) or Lansford et al., 2006; Verhoef et al., 2019). Nonetheless, a peer damages their property (i.e., object provocation). In findings vary considerably between studies and SIP meas- contrast, proactive aggression—planned aggressive behav- ures used. A meta-analysis (Verhoef et al., 2019) revealed ior aimed at obtaining a desired outcome (Dodge, 1991)— that the association between aggressive behavior and chil- may stem from SIP characterized by instrumental goals, dren’s hostile intent attributions was stronger in studies using positive outcome expectations of aggression, and positive actual social interactions (d = 1.33) than in studies using evaluations of aggression (e.g., Hubbard et al., 2010). Such vignettes (d = 0.23 to 0.44) or video-game tasks (d = 0.36; proactive SIP patterns may particularly be triggered in Yaros et al., 2014). The small to moderate effect sizes for instrumental gain contexts (Hubbard et al., 2010), where vignettes and video games may be due to a lack of emotional children have the opportunity to steal something (i.e., object engagement (i.e., vignettes may not evoke strong emotions) acquisition) or win a game by (i.e., competition). Although or limited ecological validity (i.e., video games may not reactive and proactive motives for aggression can be mixed resemble real-life social interaction). These findings align (e.g., taking revenge to show who is the boss; Bushman & with theoretical work suggesting that strong emotions such Anderson, 2001), there is ample empirical work to suggest as anger or frustration may trigger aggressive SIP patterns that they often occur in isolation (Polman et al., 2007; Van that are not triggered when children are calm (Anderson & Dijk et al., 2021). Earlier studies, however, have not always Bushman, 2002; Lemerise & Arsenio, 2000; Verhoef et al., found clearly delineated reactive versus proactive SIP pat- 2021a). However, few studies exist that used actual social terns; possibly because their vignette-based assessment did interactions to assess children’s SIP—possibly because using not evoke the specific emotions underlying real-life reactive this method is challenging in terms of standardization and and proactive aggression (e.g., Crick & Dodge, 1996; Dodge ethics (Underwood, 2005). et al., 1997; Oostermeijer et al., 2016; Stoltz et al., 2013). Ideally, SIP assessment would combine highly emotional Our interactive VR may address this issue by immersing engaging, realistic social interactions with adequate stand- children in engaging social interactions with virtual peers, ardization and ethically and practically feasible methodol- where they are actually (not just hypothetically) provoked ogy. To attain this goal, we developed an interactive VR or tempted to use aggression. They may experience anger, classroom where children can walk around freely, talk to frustration, or jealousy, activating the unique SIP patterns virtual peers, and play games, allowing us to present stand- underlying reactive- and proactive aggression. Interactive ardized social events within an engaging environment. As VR then allows children to actually aggress against virtual children are fully immersed in the VR environment, the peers instead of reporting on their hypothetical aggressive peer interactions they have (e.g., their game being ruined responses as with vignettes. Consequently, interactive VR by a peer) may evoke substantial levels of anger, frustration, permits an assessment of children’s outcome expectancies or jealousy. A recent pilot study revealed that our interac- and evaluations regarding their actual behavior instead of tive VR assessment evoked larger individual differences presenting them with hypothetical response options they in aggressive responses than a vignette-based assessment might never carry out in real life. 1 3 Research on Child and Adolescent Psychopathology In sum, the present study examines whether interactive whose behavior problems are so severe that they require VR provides a better assessment of children’s aggressive extra support that cannot be provided in regular education. SIP and responding than a standard, vignette-based assess- In our study, boys from special education were nominated ment. We chose vignettes for this comparison because they by their teacher for frequently showing aggressive behavior are the standard method to assess children’s SIP, and have problems. Boys were excluded if they had an IQ below 80 been shown to yield similarly modest associations with or an Autism Spectrum Disorder (ASD) according to their children’s real-life aggression as other methods, such as casefiles, or had a clinical score on ASD symptoms on the video-game tasks (Verhoef et  al., 2019). Children com- teacher-rated Social Emotional Questionnaire (SEQ; Scholte pleted both an interactive VR-based and a hypothetical & Van der Ploeg, 2007). Schools sent parents an information vignette-based assessment of SIP, and teachers reported on letter in which the study was explained. All parents provided their aggressive behavior. We had three main goals. First, written consent for their child’s participation in the study by we tested whether interactive VR, compared to vignettes, signing the attached informed consent form and returning it would elicit higher levels of emotional engagement (1a) and to their child’s teacher. Boys provided verbal assent. immersion (1b). Consequently, we expected that interac- tive VR would trigger aggressive SIP and response patterns Procedure that are not triggered when children are calm. This should result in larger individual differences (i.e., variances) in Participants were individually tested in a silent room at SIP and aggressive responses (1c), and higher scores on their school by trained graduate students or the first author. aggressive SIP and aggressive responses (1d). Moreover, Graduate students were trained in multiple sessions by the it should result in more congruent SIP and response pat- first author and were supervised during the first two assess- terns, visible as stronger correlations between all SIP and ments to ensure assessment fidelity. The interactive VR- and aggressive response variables in each scenario (1e). Sec- vignette-based SIP assessments both lasted 45 min and were ond, we examined whether interactive VR explained addi- completed on two different days with approximately one tional variance in children’s real-life aggressive behavior week in between. We counterbalanced the order of these reported by teachers, above and beyond the vignette-based assessments across participants to control for order effects. assessment. We examined this both for the assessment of At the end of each assessment, boys reported on their emo- children’s aggressive SIP (2a) and children’s aggressive tional engagement and immersion during the assessment. responses (2b). Third, we examined whether interactive VR Boys received a small monetary reward (€5) for their partici- explained additional variance in teacher-reported reactive pation. Teachers reported on boys’ aggressive behavior and and proactive motives for aggression, above and beyond the filled out the SEQ through online questionnaires (response vignette-based assessment—again, both for aggressive SIP rate = 98%). The study was approved by the Medical Ethics (3a) and aggressive responses (3b). Committee of University Medical Center Utrecht. Method Materials Participants Interactive Virtual Reality Environment Participants were 184 Dutch boys ages 7 to 13  years Participants wore VR glasses to immerse them in the VR (M = 10.22; SD = 1.30). They were recruited from 18 Dutch environment. They could walk around freely (in a demar- primary schools. Schools were from neighbourhoods repre- cated 4 × 4 m space), use controllers that mimicked their sentative of the Dutch population, with on average 9% inhab- hands, and respond in similar fashion as in real life: through itants with a Western migration background (SD = 3%), 13% verbal and physical behavior. The interactive VR environ- with a non-Western migration background (SD = 9%), 21% ment was designed as a virtual school classroom where par- with a lower educational level (SD = 4%), and with 7% of the ticipants could interact and play games with virtual peers households having a low-income (SD = 3%) (Statistics Neth- (for a detailed description of the interactive VR environment, erlands, 2018, 2019). To maximize variance in aggressive see: Verhoef et al., 2021b). We presented the virtual class- behavior, boys high on disruptive behavior problems were room to participants as an actual classroom where standard oversampled by including boys from special education for behavior rules applied (e.g., respecting other children) and disruptive behavior problems (n = 118) and a random sample where they would meet real children from other schools who of boys from regular education (n = 66). In the Netherlands, were also participating in the study. In reality, virtual peers special education for children with disruptive behavior prob- were controlled by the experimenter through default move- lems and/or psychiatric problems is reserved for children ment options and standardized verbal responses. 1 3 Research on Child and Adolescent Psychopathology Participants could play two games: (1) building a tower lose high scores and bonuses), allowing for a clean com- of blocks as high as possible, and (2) throwing five balls parison between assessment methods. We counterbalanced to hit as many cans from a table as possible. We designed the type of game across participants (i.e., participants who our VR assessment around these games to allow for both received the tower game in interactive VR, received the peer-directed aggression (e.g., hitting, name calling) and cans game with vignettes, and vice versa). As in most property-directed aggression (e.g., knocking over the peer’s vignette procedures, participants were told that they would tower). To increase participants’ emotional engagement and listen to stories about everyday social situations with peers to provide experimental control over gains and losses, we and were asked to imagine that each story actually hap- included high scores and bonuses for participants’ perfor- pened to them (Verhoef et al., 2019). mance during the games (e.g., building a high tower). The instructions, game rules, and score count were displayed on Measures a digital school board, which also explained these matters through standardized verbal instructions. Emotional Engagement Virtual Reality Scenarios We assessed children’s emotional engagement during the assessment in two ways. First, we used two items imme- Participants were presented with six VR scenarios in a fixed diately after each assessment to directly capture children’s order: (1) practice scenario, (2) neutral scenario, (3) object emotional engagement during the assessment, aiming to acquisition, (4) competition, (5) social provocation, and (6) minimize the effect of memory on their ratings (i.e., “How object provocation—all centering around one of the games angry did you feel when something bad happened to you in (i.e., the tower or cans game; randomly assigned). The prac- VR/vignettes?” and “How much did you care when some- tice VR scenario served to familiarize participants with the thing bad happened to you in VR/vignettes?”). Children VR environment and game rules by practicing the game responded on a rating scale from 1 (not at all) to 10 (very). without any virtual characters present. The neutral scenario We averaged the two items to create emotional engage- served to familiarize participants with the SIP questions by ment scores for both interactive VR (r = 0.83) and vignettes having them play the game while engaging in neutral small (r = 0.67). Second, to allow children to make a comparison talk with a virtual peer, and asking the SIP questions after- between the VR- and vignette-based assessment, we again wards. Next, participants completed the four experimental administered these two items after they had completed both scenarios, which we based on taxonomies of problematic assessments, but then phrased in comparative form (e.g., for situations for children with aggressive behavior problems the first item: “You have completed both the VR and the sto - (Matthys et  al., 2001). The first two scenarios involved ries. How angry did you feel when something bad happened instrumental gain. In the object acquisition scenario, par- to you in the VR? And in the stories?;” question order was ticipants had the opportunity to steal a block or ball from counterbalanced). We again averaged the two items to create the virtual peer, which would earn them additional points emotional engagement scores for interactive VR (r = 0.74) in the game. In the competition scenario, they could win and vignettes (r = 0.74). the game and thus earn additional points by sabotaging the virtual peer’s progress in the game (i.e., by knocking over the peer’s tower, ruining the virtual peer’s balls). The last Immersion two scenarios involved provocation. In the social provoca- tion scenario, participants were refused to join the game by We assessed children’s immersion during the assessment in two virtual peers. In the object provocation scenario, their two ways. First, we used six items immediately after each game was ruined by a virtual peer. As such, the provocations assessment, which were adapted from the Dutch translation caused them to earn no points. In the two provocation sce- of the Igroup Presence Questionnaire (Schubert et al., 1999). narios, participants could not obtain any points by respond- Two of the six items had low factor loadings (i.e., below ing aggressively. We expected these provocation scenarios 0.60) and were excluded. The four items used were: 1) “I to elicit the strongest emotions, and therefore presented them was totally caught up by the events in VR/vignettes;” 2) “I last to prevent carry-over effects. had the feeling that the events in VR/vignettes were actually happening to me;” 3) “During the VR/vignettes it felt like I Hypothetical Vignettes was actually experiencing the events;” and 4) “The events in VR/vignettes seemed almost real.” Participants rated the For the vignette-based SIP assessment, we developed audi- items on a scale from 1 (strongly disagree) to 5 (strongly otaped vignettes with the exact same content as the VR agree). We averaged across items to create immersion scores scenarios (e.g., describing how participants would gain or for both interactive VR (α = 0.78) and vignettes (α = 0.81). 1 3 Research on Child and Adolescent Psychopathology Second, to allow children to make a comparison between Interaction Goals. Interaction goals were assessed using the VR- and vignette-based assessment, we administered one one open-ended question following each VR-scenario: item after they had completed both assessments, but then “When the other boy did [behavior of other boy], you did phrased in comparative form (i.e., “You have participated [behavior of participant]. What was the reason you did in both the VR- and vignette-based assessment. How much this?” In line with earlier research (De Castro et al., 2012), did you have the feeling that the events in VR were actually the first author coded each answer as revenge goals (e.g., happening to you? And in the stories?”). Children responded “to retaliate,” “because I was angry,” “to defend myself”), on a rating scale from 0 (not at all) to 10 (very). instrumental goals (e.g., “to win the game,” “to show him who’s the boss”), goals underlying non-aggressive behav- ior (e.g., “to become friends,” “to avoid problems”), or no Aggressive SIP and Responses goals (e.g., “I don’t know”). A second rater also coded 35% of the transcriptions. Inter-rater reliability was excellent, We assessed participants’ aggressive SIP and responses in with Cohen’s κ ranging from 0.85-0.96 across scenarios two provocation scenarios and two instrumental gain sce- (M = 0.91, Mdn = 0.91). Scores for revenge goals were cre- narios (both in interactive VR and with vignettes). Initially, ated by assigning 1 to revenge goals codes and 0 to other we planned to create aggregate SIP and response variables codes. Similarly, scores for instrumental goals were created for provocation and instrumental gain contexts. However, by assigning 1 to instrumental goals codes and 0 to other we found low correlations for SIP and response variables codes. between the social provocation and object provocation sce- Aggressive Responses. We assessed participants’ nario (i.e., ranging from 0.37-0.60 in VR and from 0.27- behavioral responses in interactive VR through observa- 0.50 with vignettes) and between the object acquisition and tion. A trained research assistant made detailed descrip- competition scenario (i.e., ranging from 0.34-0.58 in VR and tions of participants’ behavioral responses in each VR- from 0.35-0.48 with vignettes), suggesting that aggressive scenario. The first author coded these descriptions into SIP and behavior may be highly situation specific (Dodge non-aggressive behavior (e.g., prosocial behavior, avoid- et al., 1985; Matthys et al., 2001). Hence, we decided to cre- ance), mild aggressive behavior (e.g., coercion, verbal ate variables for children’s SIP and aggressive responses for aggression), and severe aggressive behavior (e.g., physi- each scenario separately. cal aggression, destructive aggression) following stand- Interactive VR Assessment. We assessed participants’ ard coding procedures (De Castro et al., 2005). If mul- aggressive responses through observation of their behavior tiple codes applied, the highest category was scored. A in VR, and used self-report to assess their anger, intent attri- second rater also coded 35% of the behavioral descrip- butions, goals, outcome expectancies, and response evalua- tions. Inter-rater reliability was excellent, with κ ranging tions at the end of each VR-scenario. In between scenarios, from 0.92–1.00 across scenarios (M = 0.97, Mdn = 0.98). participants kept their VR-glasses on while replying ver- Because frequencies of mild aggressive behavior were low bally to the experimenter’s questions. For procedural clar- or even absent (i.e., 0 to 17% across VR-scenarios and ity, we assessed all SIP questions following all scenarios, vignettes, Mdn = 2%), we created a dichotomous variable even though we were only interested in proactive SIP in by coding mild and severe aggressive behavior as 1 and instrumental scenarios (i.e., instrumental goals, outcome non-aggressive behavior as 0. expectancies, and response evaluation) and reactive SIP in Outcome Expectancies. Outcome expectancies of provocation scenarios (i.e., anger, hostile intent attribution, aggression were assessed using one item following each and revenge goals). VR-scenario: “What did you expect would happen when Anger. Anger was assessed using one item following each you [behavior of participant]?” We coded only answers VR-scenario: “The other boy did [behavior of other boy]. of participants who had actually used aggression in How angry did this make you feel, on a scale from 1, mean- that VR-scenario and assigned missing values to other ing not at all, to 10, meaning very?”. answers. The first author coded each answer as posi- Hostile Intent Attribution. Intent attributions were tive outcome expectancies of aggression (e.g., “I would assessed using two items following each VR-scenario: “The win the game”), or no positive outcome expectancies of other boy did [behavior of other boy]. To what extent did he aggression (e.g., “He would dislike me”). A second rater try to be mean, on a scale from 1, meaning not at all, to 10, also coded 35% of the transcriptions. Inter-rater reliability meaning very?” and “To what extent did he try to hinder was excellent, with κ being 1.00 for each scenario. Scores you, on a scale from 1 to 10?” These two items were mod- for positive outcome expectancies of aggression were cre- erately to highly correlated within each of the four VR sce- ated by assigning 1 to positive outcome expectancies of narios (M = 0.83, Mdn = 0.87, range = 0.67-0.90) and were aggression and 0 to no positive outcome expectancies of therefore averaged within each VR-scenario. aggression. 1 3 Research on Child and Adolescent Psychopathology Response Evaluations. Positive evaluations of aggres- Real‑Life Aggressive Behavior sion were assessed using one item following each VR- scenario: “When the other boy did [behavior of other boy], Teachers completed two questionnaires to assess partici- you did [behavior of participant]. To what extent do you pants’ aggressive behavior in real life. First, teachers filled approve your behavior on a scale from 1, meaning not at out the Aggressive Behavior subscale of the Dutch version all, to 10, meaning very?” We only used scores of children of the Teacher Report Form (TRF; Verhulst et al., 1997). who had actually used aggression in that VR-scenario and They rated 20 items (e.g., “This child threatens others”) on coded other scores as missing. a 3-point Likert scale (1 = not true for this child, 2 = some- Participants’ outcome expectancies and response evalu- what true for this child, or 3 = very often true for this child). ations of aggression were only scored when they displayed Scores were averaged across items (α = 0.96). Second, they aggressive responses, limiting the number of observations filled out the Instrument for Reactive and Proactive Aggres - for these variables. Conversely, other SIP variables (i.e., sion (IRPA; Polman et al., 2009). This instrument differ - anger, hostile intent attributions, revenge goals and instru- entiates between the frequency of aggression on the one mental goals) could be scored irrespectively of whether hand, and the motives underlying aggression on the other participants engaged in aggressive responses, yielding full hand. We used the frequency scale to assess children’s real- data for these variables (see Table 1 for descriptive statis- life aggressive behavior. Teachers rated the frequency of 7 tics of SIP and aggressive response variables). distinct forms of aggressive behavior (i.e., kicking, push- Vignette Assessment. Children reported on their SIP ing, hitting, name calling, arguing, gossiping, and doing following each vignette. We used the same questions and sneaky things) in the previous month on a 5-point Likert coding schemes as used for the interactive VR-assessment, scale (1 = never, 2 = once, 3 = weekly, 4 = multiple times a except that we formulated the questions as hypothetical week, 5 = daily). Scores on these seven items were averaged (e.g., “How angry would you feel…?”) instead of actual (α = 0.90). IRPA frequency scores (M = 1.95, SD = 0.86) and (e.g., “How angry were you…?”). The two items assess- TRF scores (M = 1.67, SD = 0.57) were highly correlated ing intent attributions were averaged within each vignette (r = 0.85). We therefore standardized and averaged them to as they were highly correlated (M = 0.80, Mdn = 0.81, create a single aggressive behavior score. range = 0.68-0.90). Inter-rater reliability (κ) for open- ended questions was based on 35% of transcriptions and Reactive & Proactive Motives for Aggression was excellent for both interaction goals (range = 0.81–1.00, M = 0.91, Mdn = 0.91) and outcome expectancies We assessed reactive and proactive motives for aggression (range = 0.83–1.00, M = 0.94, Mdn = 1.00). We assessed by again using the IRPA (Polman et al., 2009), but this time participants’ anticipated behavioral responses for each the motive scales. For each form of aggression rated above vignette using an open-ended question (i.e., “What would 0, teachers rated 3 reactive motives items (e.g., “Because you do if [social event]?”). Inter-rater reliability was based someone teased or upset him”) and 3 proactive motives on 35% of the transcriptions and was excellent, with κ items (e.g., “To hurt someone or to be mean”) on a 5-point ranging from 0.91–1.00 (M = 0.94, Mdn = 0.93). Likert scale (0 = never, 1 = rarely, 2 = sometimes, 3 = often, Table 1 Descriptive statistics of SIP variables for each scenario between VR and Vignettes (VIG) 1 2 Object Acquisition Competition Social Provocation Object Provocation VR VIG VR VIG VR VIG VR VIG Anger 1.75 (1.78) 1.65 (1.62) 3.13 (2.66) 4.98 (3.09) 5.36 (2.88) 5.13 (2.94) 6.79 (2.76) 7.54 (2.58) Hostile Intent Attribution 1.47 (1.21) 1.59 (1.21) 2.17 (2.18) 5.00 (3.09) 5.23 (3.08) 4.75 (2.94) 7.35 (2.78) 6.65 (3.32) Revenge Goals 1 (1) 1 (1) 6 (3) 15 (8) 51 (28) 19 (10) 92 (51) 69 (38) Instrumental Goals 41 (23) 43 (24) 37 (20) 27 (15) 16 (9) 6 (3) 12 (7) 7 (4) Aggressive Responses 42 (23) 44 (24) 43 (24) 42 (23) 69 (38) 25 (14) 105 (58) 77 (42) Outcome Expectancies 10 (24) 14 (32) 18 (42) 3 (7) 2 (3) 0 (0) 0 (0) 5 (6) Positive Evaluations 5.17 (3.56) 4.32 (2.77) 4.91 (3.71) 4.26 (3.25) 4.04 (3.24) 6.24 (3.38) 5.22 (3.40) 5.49 (3.16) Columns display means (M) and standard deviations (SD) of Anger, Hostile Intent Attribution, and Positive Evaluations, and the number (n) and proportion (%) of children displaying Revenge Goals, Instrumental Goals, Aggressive Responses, and Positive Outcome Expectancies Scores only apply to children who displayed an Aggressive Response for this scenario Based on n = 179 because 5 children had missing data in VR/vignettes Based on n = 178 because 6 children had missing data in VR/vignettes 1 3 Research on Child and Adolescent Psychopathology 4 = always). For aggression frequency items rated 0, motives their aggressive behavior in real life compared to vignettes, scores were missing by design. We calculated reactive and we conducted the same hierarchical regression analyses proactive motives scores by averaging across all reactive as used for our second hypothesis, but then with reactive motives items (i.e., 3 items times 7 forms of aggression; motives as dependent variables for the provocation scenarios α = 0.94) and all proactive motives items (α = 0.95), respec- and proactive motives as dependent variables for the instru- tively. Thus, high scores on reactive (M = 2.75, SD = 0.94) mental gain scenarios. or proactive (M = 2.04, SD = 0.84) motives indicate that if participants engaged in aggressive behavior, they often had reactive or proactive motives. The correlation between Results reactive and proactive motives was non-significant (r = 0.14, p = 0.075). Preliminary Analyses Statistical Analyses Table 1 presents the descriptive statistics for all SIP vari- ables in both VR and vignettes. As most SIP variables were To test our first hypothesis that interactive VR is more skewed, we conducted our analyses using a bootstrapping engaging than vignettes, we considered five aspects. First, procedure with bias-corrected accelerated (BCa) 95% con- we examined whether interactive VR yielded higher mean fidence intervals (CI) based on 5000 resamples. levels of emotional engagement than vignettes, using paired Our VR elicited aggressive responses in 23% to 58% of t-tests. Second, we examined whether participants’ immer- children, depending on the scenario (Table  1). However, sion was higher in VR versus vignettes, also using paired few children who responded aggressively in the VR, also t-tests. Third, we examined whether VR elicited larger responded aggressively in the same scenario in vignettes individual differences in aggressive SIP and aggressive (i.e., 9 to 32% across scenarios, Mdn = 10%; see Supplemen- responses than vignettes. To this end, we used an adaptation tary Material Table S1). As a result, we had insufficient data of the Pittman-Morgan test which replaces Pearson’s r with to compare VR versus vignettes on SIP variables that were Spearman’s rank correlation to account for non-normal data only assessed if children actually responded aggressively (McCulloch, 1987). Fourth, we examined whether interactive (i.e., positive outcome expectancies and positive evaluations VR yielded higher scores on aggressive SIP and aggressive of aggression). We therefore reported descriptive statistics responses than vignettes, using paired t-tests for continuous for these two variables (see Table 1) but excluded them from SIP variables and McNemar’s tests for dichotomous SIP and our main analyses. response variables. Fifth, we examined whether VR yielded stronger correlations among SIP and aggressive responses Children’s Engagement in Interactive VR Versus than vignettes. To do so, we calculated correlations between Vignettes all SIP and response variables for each scenario using Pear- son’s r, Pearson’s π, and Point-Biserial correlations. Next, Emotional Engagement we tested for inequality of the obtained correlation matri- ces using Steiger’s test (1980), which directly compares all As predicted, children reported feeling more emotionally elements of two dependent correlation matrices instead of engaged in VR (M = 5.59, SD = 2.74) than with vignettes comparing each correlation separately. (M = 2.91, SD = 2.37), BCa 95% CI [2.23, 3.10], p < 0.001,, To test our second hypothesis that interactive VR assess- d = 0.92. The same result was found when we asked chil- ment of aggressive SIP (2a) and responses (2b) better pre- dren about their engagement after they had completed both dicts children’s aggressive behavior in real life compared to assessments: their engagement was higher in VR (M = 5.68, vignettes, we examined whether VR explained additional SD = 2.69) than with vignettes (M = 3.05, SD = 2.16), BCa variance in real life aggression above and beyond vignettes, 95% CI [2.23, 3.04], p < 0.001, d = 0.96. but not vice versa. For aggressive SIP, we conducted two hierarchical regression analyses: the first with vignette- Immersion assessed SIP entered at step 1 and VR-assessed SIP at step 2; the second with VR-assessed SIP at step 1 and vignette- As predicted, children reported feeling more immersed in assessed SIP at step 2. For aggressive responses, we repeated VR (M = 4.18, SD = 0.83) than with vignettes (M = 2.57, these analyses with VR- versus vignette-assessed aggressive SD = 1.07), BCa 95% CI [1.44, 1.77], p < 0.001, d = 1.45. responses as predictors. They also reported feeling more immersed in VR (M = 7.96, To test our third hypothesis that interactive VR assess- SD = 2.21) than with vignettes (M = 3.70, SD = 2.37) after ment of aggressive SIP (3a) and responses (3b) better pre- they had completed both assessments, BCa 95% CI [3.85, dicts children’s reactive and proactive motives underlying 4.64], p < 0.001, d = 1.52. 1 3 Research on Child and Adolescent Psychopathology Variances in Aggressive SIP and Responses limited. Steiger’s test revealed that the correlation matrix of aggressive SIP and response variables was significantly We found limited support for our hypothesis that VR elicits higher for VR versus vignettes for the competition scenario, larger variances in SIP variables and aggressive responses χ (1) = 23.33, p < 0.001, but did not significantly differ than vignettes. For the object acquisition and competi- between VR and vignettes for the object acquisition sce- tion scenario respectively, we found no difference in vari- nario, χ (1) = 0.03, p = 0.862, social provocation scenario, ances between VR and vignettes for instrumental goals, χ (6) = 6.58, p = 0.361, and object provocation scenario, t(178) < 0.01, p = 1.000, and t(179) = 1.56, p = 0.120, χ (6) = 6.46, p = 0.374. and aggressive responses, t(178) < 0.01, p = 1.000, and In sum, children reported more emotional engagement t(179) = 0.21, p = 0.835. In the social provocation sce- and immersion in VR than with vignettes. Partial support nario, we did find higher variances in VR versus vignettes was found for VR outperforming vignettes on other aspects: for revenge goals, t(178) = 4.87, p < 0.001, and aggres- It yielded more variance for 2 out of 12 results, higher levels sive responses, t(178) = 4.37, p < 0.001, but not for anger, of aggressive SIP and responses for 6 out of 12 results, and t(178) = -0.28, p = 0.777, and hostile intent attributions, stronger correlations for 1 out of 4 results. t(178) = 0.61, p = 0.544. Last, in the object provocation sce- nario, we found no support for our hypothesis: there were Predicting Real‑Life Aggressive Behavior no differences for anger, t (177) = 0.91, p = 0.364, revenge goals, t(177) = 0.33, p = 0.738, and aggressive responses, Tables 3 and 4 present the results of the hierarchical regres- t(177) = -0.02, p = 0.983, and, contrary to our expectation, sion analyses of aggressive behavior in real life regressed larger variances with vignettes versus VR for hostile intent on aggressive SIP a) and aggressive responses b), first con- attributions, t(177) = -2.08, p = 0.039. ducted with vignettes in Step 1 and VR in Step 2, and next with VR in Step 1 and vignettes in Step 2. Analyses were Levels of Aggressive SIP and Responses conducted for each scenario separately. We tested whether VR yielded more aggressive SIP and Aggressive SIP responses than vignettes (Table 1). Details of the McNe- mar’s test for dichotomous variables can be found in Sup- Children’s aggressive SIP in all four VR scenarios signifi- plementary Material Table S1. Our hypothesis was partly cantly predicted their real-life aggression, with explained supported. For the object acquisition and competition sce- variances at Step 1 ranging from 4 to 13% across scenarios. narios respectively, we found no differences between VR As expected, effects were weaker for vignettes. Children’s and vignettes in instrumental goals, p = 1.000, OR = 1.000, aggressive SIP assessed with vignettes significantly pre - p = 0.154, OR = 1.67, nor aggressive responses, p = 1.000, dicted their real-life aggression at Step 1 in the object acqui- OR = 1.00, p = 0.885, OR = 1.09. For the social provoca- sition scenario (R = 0.03) and social provocation scenario 2 2 tion and object provocation scenarios respectively, chil- (R = 0.05), but not in the competition (R = 0.02) and object dren showed more hostile intent attributions, BCa 95% CI provocation scenario (R = 0.04). Turning to the incremental [0.00, 0.97], p = 0.048, d = 0.15, BCa 95% CI [0.07, 1.29], value of VR, we found that VR entered at Step 2 explained p = 0.025, d = 0.17, revenge goals, p < 0.001, OR = 4.67, significant variance over and above vignettes in all scenarios p = 0.014, OR = 1.85, and aggressive responses, p < 0.001, (i.e., 2% in object acquisition, 5% in competition, 12% in OR = 6.63, p = 0.001, OR = 2.35, in VR than with vignettes. social provocation, and 9% in object provocation). As pre- However, we found no differences in anger between VR and dicted, vignettes did not explain significant variance over vignettes in the social provocation scenario, BCa 95% CI and above VR in any scenario. [-0.24, 0.69], p = 0.354, d = 0.07, and even higher mean lev- els of anger for vignettes versus VR in the object provocation Aggressive Responses scenario, BCa 95% CI [-1.21, -0.28], p = 0.002, d = -0.24. Children’s aggressive responses in all four VR scenarios sig- Correlations between Aggressive SIP Variables nificantly predicted their real-life aggression, with explained and Responses variances at Step 1 ranging from 4 to 12% across scenarios. Similar effects were found for vignettes, with explained We tested whether correlations among aggressive SIP and variances at Step 1 ranging from 4 to 10%. Turning to the response variables were stronger in VR versus vignettes. incremental value of VR, we found that VR entered at Step 2 Table 2 presents all correlations between these variables for explained significant variance over and above vignettes in all each scenario separately. Steiger’s test to compare correla- scenarios (i.e., 2% in object acquisition, 5% in competition, tion matrices showed that support for our hypothesis was 9% in social provocation, and 7% in object provocation). 1 3 Research on Child and Adolescent Psychopathology Table 2 Bivariate Correlations Instrumental gain scenarios 1 2 3 4 5 6 of SIP and response Variables in * * * * * VR and Vignettes with real-life 1. VR: Instrumental Goals 0.98 0.27 0.29 0.19 0.27 aggressive behavior and reactive * * * * * 2. VR: Aggressive Responses 0.91 0.26 0.28 0.20 0.27 and proactive motives for * * * * 3. Vignette: Instrumental Goals 0.23 0.18 0.99 0.17 0.15 aggression in instrumental gain * * * * scenarios (object acquisition 4. Vignette: Aggressive Responses 0.27 0.24 0.76 0.20 0.16 * * * * scenario above the diagonal; 5. Real-Life Aggressive Behavior 0.26 0.29 0.17 0.32 0.58 competition scenario below) * * * * 6. Real-Life Proactive Motives 0.24 0.23 0.14 0.20 0.58 and provocation scenarios Provocation scenarios 1 2 3 4 5 6 7 8 9 10 (social provocation scenario * * * * * * * above the diagonal; object 1. VR: Anger 0.52 0.41 0.41 0.37 0.40 0.21 0.23 0.12 0.06 * * * * * provocation scenario below) 2. VR: Hostile Intent Attribution 0.55 0.45 0.36 0.29 0.40 0.10 0.13 0.05 0.14 * * * * * * * * 3. VR: Revenge Goals 0.41 0.40 0.80 0.09 0.17 0.16 0.19 0.34 0.30 * * * * * * * 4. VR: Aggressive Responses 0.36 0.40 0.86 0.07 0.14 0.16 0.23 0.35 0.29 * * * * * 5. Vignette: Anger 0.30 0.18 0.11 0.10 0.52 0.32 0.41 0.06 -0.12 * * * * 6. Vignette: Hostile Intent Attribution 0.10 0.12 0.14 0.11 0.45 0.30 0.35 -0.05 -0.21 * * * * * * * * 7. Vignette: Revenge Goals 0.19 0.16 0.18 0.23 0.43 0.37 0.86 0.17 0.05 * * * * * * * * 8. Vignette: Aggressive Responses 0.19 0.19 0.24 0.28 0.39 0.39 0.91 0.23 0.09 * * * * * * 0.29 0.27 0.32 0.15 0.07 0.14 0.20 0.46 8. Real-Life Aggressive Behavior 0.20 * * * * * 9. Real-Life Reactive Motives 0.19 0.21 0.29 0.27 -0.04 -0.07 -0.04 0.01 0.46 Correlations of SIP and responses in Instrumental Gain Scenarios are reported in the upper part of the Table (Object Acquisition Scenario above the Diagonal; Competition Scenario below) and correlations of SIP and Responses in Provocation Scenarios in the lower part of the Table (Social Provocation Scenario above the Diagonal; Object Provocation Scenario below All correlations including Instrumental Goals, Revenge Goals or Aggressive Responses are point-biserial correlations, all correlations between Instrumental Goals, Revenge Goals and Aggressive Responses used Pearson’s π, and other correlations used Pearson’s r Indicates that the bootstrap 95% confidence interval did not include zero However, we also found that vignettes at Step 2 explained less pronounced for vignettes. Children’s aggressive SIP significant variance over and above VR in three scenarios, assessed with vignettes significantly predicted their reac- with higher levels of explained variance in the competition tive and proactive motives in the object acquisition scenario 2 2 scenario (i.e., 6%), but lower levels in in social provoca- (R = 0.03) and social provocation scenario (R = 0.06), but tion and object provocation scenarios (i.e., 3% and 2%, not in the competition (R = 0.02) and object provocation respectively). scenario (R < 0.01). Turning to the incremental value of In sum, all eight hierarchical regression analyses regard- VR, we found that VR entered at Step 2 explained signifi- ing children’s real-life aggression supported the incremental cant variance over and above vignettes in all scenarios (i.e., value of VR over vignettes, whereas only three analyses sup- 6% in object acquisition, 5% in competition, 12% in social ported the reverse. provocation, and 11% in object provocation). In contrast, we found that vignettes at Step 2 explained significant variance Predicting Reactive & Proactive Motives over and above VR only in the social provocation scenario (i.e., 8%). Next, we conducted the same set of hierarchical regression analyses as for children’s real-life aggressive behavior, in Aggressive Responses this case predicting children’s reactive and proactive motives for aggression. Detailed results of these analyses are pro- Children’s aggressive responses in all four VR scenarios vided in the Supplementary Materials (Table S2 and S3). significantly predicted their reactive and proactive motives in real life, with explained variances at Step 1 ranging from Aggressive SIP 5 to 9% across scenarios. Effects were weaker for vignettes. Children’s aggressive responses assessed with vignettes As predicted, children’s aggressive SIP in all four VR significantly predicted their reactive and proactive motives scenarios significantly predicted their reactive and proac - in the object acquisition scenario (R = 0.03) and competi- tive motives in real life, with explained variances at Step tion scenario (R = 0.05), but not in the social provocation 2 2 1 ranging from 6 to 10% across scenarios. Effects were (R = 0.01) and object provocation scenario (R < 0.01). 1 3 Research on Child and Adolescent Psychopathology 1 3 Table 3 Hierarchical regression analyses of real-life aggression regressed both on instrumental goals and aggressive responses Object acquisition scenario Competition scenario 2 2 Step Predictor β β SE 95% CI ∆R df F change β β SE 95% CI ∆R df F change 1 Vignette: Instrumental 0.37 0.20 -0.02-0.77 0.03 1,174 4.54 0.40 0.25 -0.07-0.91 0.02 1,175 3.75 Goals * ** 2 Vignette: Instrumental 0.27 0.19 -0.10-0.65 0.02 1,173 4.38 0.23 0.25 -0.26-0.75 0.05 1,174 10.27 Goals ** VR: Instrumental 0.37 0.20 -0.01-0.75 0.58 0.18 0.21-0.94 Goals * * ** *** 1 VR: Instrumental 0.44 0.20 0.04-0.82 0.04 1,174 6.66 0.63 0.19 0.27–1.01 0.07 1,175 13.02 Goals ** 2 VR: Instrumental 0.37 0.20 -0.01-0.75 0.01 1,173 2.30 0.58 0.19 0.22-0.95 0.01 1,174 1.18 Goals Vignette: Instrumental 0.27 0.19 -0.10-0.64 0.23 0.25 -0.21-0.72 Goals * * *** *** 1 Vignette: Aggressive 0.44 0.20 0.07-0.83 0.04 1,174 6.86 0.74 0.20 0.36–1.15 0.10 1,175 19.98 Responses * ** ** 2 Vignette: Aggressive 0.34 0.19 -0.02-0.72 0.02 1,173 4.12 0.61 0.19 0.24–1.00 0.05 1,174 9.19 Responses ** VR: Aggressive 0.35 0.19 -0.00-0.73 0.50 0.17 0.15-0.84 Responses * ** *** *** 1 VR: Aggressive 0.45 0.20 0.08-0.84 0.04 1,174 7.09 0.66 0.18 0.31–1.00 0.08 1,175 16.07 Responses ** *** 2 VR: Aggressive 0.35 0.19 -0.00-0.72 0.02 1,173 3.90 0.50 0.17 0.17-0.83 0.06 1,174 12.39 Responses ** Vignette: Aggressive 0.34 0.19 -0.01-0.71 0.61 0.19 0.24-0.97 Responses Hierarchical Regression Analyses were run for the Two Instrumental Gain Scenarios separately, both with Vignettes and VR Entered First. Model output is based on a non-bootstrapped proce- dure whereas output on separate predictors is based on a bootstrapping procedure * ** *** p < .05; p < .01; p < .001 Research on Child and Adolescent Psychopathology 1 3 Table 4 Hierarchical regression analyses of real-life aggression regressed both on reactive SIP and aggressive responses Social Provocation Object Provocation 2 2 Step Predictor β β SE 95% CI ∆R df F change β β SE 95% CI ∆R df F change 1 Vignette: Anger 0.01 0.03 -0.05-0.08 0.05 3,172 2.80 0.05 0.03 -0.01-0.11 0.04 3,171 2.55 Vignette: Hostile Intent -0.04 0.03 -0.10-0.02 -0.00 0.02 -0.05-0.04 Attribution Vignette: Revenge 0.67 0.31 0.09–1.28 0.22 0.17 -0.10-0.55 Goals *** ** 2 Vignette: Anger 0.02 0.03 -0.05-0.08 0.12 3,169 7.85 0.04 0.03 -0.01-0.10 0.09 3,168 5.73 Vignette: Hostile Intent -0.05 0.03 -0.11-0.02 -0.01 0.02 -0.05-0.03 Attribution Vignette: Revenge 0.50 0.28 -0.03–1.04 0.14 0.17 -0.18-0.47 Goals VR: Anger 0.01 0.03 -0.06-0.07 -0.01 0.03 -0.08-0.06 VR: Hostile Intent -0.03 0.04 -0.10-0.04 0.07 0.03 0.01-0.14 Attribution *** * VR: Revenge Goals 0.79 0.19 0.42–1.13 0.35 0.16 0.02-0.65 *** *** 1 VR: Anger 0.01 0.03 -0.06-0.08 0.13 3,172 8.67 0.01 0.03 -0.06-0.07 0.11 3,171 7.14 VR: Hostile Intent -0.04 0.03 -0.11-0.03 0.07 0.03 0.01-0.13 Attribution *** * VR: Revenge Goals 0.83 0.19 47–1.18 0.35 0.15 0.05-0.63 2 VR: Anger 0.01 0.03 -0.06-0.07 0.03 3,169 2.15 -0.01 0.03 -0.08-0.06 0.02 3,168 1.31 VR: Hostile Intent -0.03 0.04 -0.10-0.04 0.07 0.03 0.01-0.13 Attribution *** * VR: Revenge Goals 0.79 0.19 0.42–1.15 0.35 0.16 0.03-0.63 Vignette: Anger 0.02 0.03 -0.05-0.08 0.04 0.03 -0.01-0.09 Vignette: Hostile Intent -0.05 0.03 -0.12-0.02 -0.01 0.02 -.05-.03 Attribution Vignette: Revenge 0.50 0.28 -0.04–1.10 0.14 0.17 -0.19-0.49 Goals * ** ** ** 1 Vignette: Aggressive 0.68 0.26 0.19–1.21 0.06 1,174 11.12 0.44 0.15 0.16–74 0.05 1,173 9.80 Responses *** * *** 2 Vignette: Aggressive 0.48 0.25 -0.02–1.00 0.09 1,173 18.07 0.30 0.14 0.02-0.58 0.07 1,172 14.04 Responses *** *** VR: Aggressive 0.60 0.15 0.33-0.89 0.53 0.13 0.28-0.79 Responses Research on Child and Adolescent Psychopathology Turning to the incremental value of VR, we found that VR entered at Step 2 explained significant variance over and above vignettes in all scenarios (i.e., 5% in object acquisi- tion, 3% in competition, 8% in social provocation, and 8% in object provocation). In contrast, we found that vignettes at Step 2 explained significant variance over and above VR only in the competition scenario (i.e., 3%). In sum, all eight hierarchical regression analyses regard- ing children’s reactive and proactive motives supported the incremental value of VR over vignettes, whereas only two analyses supported the reverse. Discussion This study tested whether interactive Virtual Reality (VR) provides a more ecologically valid assessment of social information processing (SIP) underlying aggressive behav- ior in children than a standard vignette-based assessment. In line with expectations, children reported that the inter- active VR assessment was more emotionally engaging and immersive than the vignette-based assessment. Moreover, the assessment of children’s aggressive SIP and responses in VR predicted their real-life aggressive behavior and reactive and proactive motives for aggression, above and beyond the vignette assessment. Interactive VR immerses children in emotionally engaging social interactions and enables them to actually aggress against virtual peers—an important difference with vignettes, which ask children to consider their hypo- thetical aggressive responses. Accordingly, interactive VR has evoked higher levels of aggressive SIP and responses in children in provocation scenarios, and improved the predictive validity of their assessed aggressive SIP and responses. These findings support the proposition that emo- tional engagement influences SIP and consequent behavior (Anderson & Bushman, 2002; Lemerise & Arsenio, 2000). Thus, the emotionally engaging nature of our interactive VR assessment seems to have triggered aggressive SIP patterns and responses that may only occur with sufficient emotional engagement. We expected that the engaging nature of interactive VR would also evoke larger individual differences in children’s aggressive SIP and responses, and stronger correlations between children’s aggressive SIP and responses compared to vignettes. However, interactive VR and vignettes gener- ally evoked similar variances in children’s aggressive SIP and responses, and similar correlations between aggres- sive SIP steps and responses. Perhaps, our vignettes val- idly assessed individual differences in children’s “calm” SIP; that is, the way they would reflect on social situations when they do not experience strong emotions. Such “calm” SIP may also differ between children and show similar 1 3 Table 4 (continued) Social Provocation Object Provocation *** *** *** *** 1 VR: Aggressive 0.68 0.15 0.38-0.99 0.12 1,174 23.86 0.62 0.13 0.35-0.87 0.10 1,173 19.80 Responses *** * *** * 2 VR: Aggressive 0.60 0.15 0.32-0.90 0.03 1,173 5.78 0.53 0.13 0.27-0.79 0.02 1,172 4.39 Responses Vignette: Aggressive 0.48 0.25 -0.01-0.98 0.30 0.14 0.02-0.58 Responses Hierarchical Regression Analyses were run for the Two Provocation Scenarios separately, both with Vignettes and VR Entered First. Model output is based on a non-bootstrapped procedure whereas output on separate predictors is based on a bootstrapping procedure * ** *** p < .05; p < .01; p < .001 Research on Child and Adolescent Psychopathology correlations between children’s SIP and responses as their Second, contrary to our predictions, children reported emotional SIP, but would be less suitable to predict chil- similar levels of anger in interactive VR and vignettes, and dren’s real-life aggression. Indeed, our findings showed even more anger with vignettes in the object provocation that interactive VR yielded incremental predictive value scenario. This finding contrasts with our finding that VR above and beyond the vignette-based assessment in all four is more emotionally engaging than vignettes. However, it scenarios, both for the prediction of children’s real-life may also reveal a potential limitation of vignettes: asking aggression (i.e., 2 to 12% additional explained variance) children to reflect on their anticipated anger in a hypothetical and underlying reactive and proactive motives (i.e., 3 to situation could lead them to overestimate how they would 12% additional explained variance). actually feel. Indeed, research has shown that individuals One unexpected pattern in our findings was that interac - generally find it difficult to report on anticipated negative tive VR seemed to outperform vignettes more so for provo- affective states and tend to overestimate them (Robinson & cation scenarios than for instrumental gain scenarios: the Clore, 2002). Although we do not know whether this was incremental predictive value of VR versus vignettes was actually the case, the stronger correlations of VR- versus larger in provocation scenarios (with 7 to 12% increases vignette-assessed anger with children’s real-life aggression in explained variance in children’s real-life aggression) indicate that children are more accurate when reporting on than in instrumental gain scenarios (with 2 to 5% increases their anger in interactive VR. Perhaps, as in interactive VR in explained variance in children’s real-life aggression), children are actually (not just hypothetically) provoked or and only in provocation scenarios children showed more tempted to use aggression, they may experience emotions aggressive SIP and responses in VR versus vignettes. more similar to daily life than the anticipated emotions Although we expected that the engaging nature of inter- assessed with vignettes. active VR would enhance children’s proactive aggressive This study had several strengths. To our knowledge, it is tendencies in instrumental gain scenarios as well (e.g., the first empirical study that used interactive VR to assess because the stakes are higher, so they would experience children’s aggressive SIP and responses and compared its more jealousy or desire), children did not show more pro- external validity directly to a standard vignette-based assess- active SIP and responses in VR versus vignettes. Possi- ment of children’s aggressive SIP and responses. Moreover, bly, the provocation scenarios were more salient than the we maximized clinically meaningful variance in children’s instrumental gain scenarios, because the instrumental gain SIP by recruiting boys from both regular and special educa- of points in the VR constituted no actual gain outside of tion for disruptive behavior problems. The use of interac- the game in the real world. As such, it makes sense that tive VR in a sample with substantial variance in children’s the incremental value of interactive VR was the largest SIP allowed us to test important hypotheses concerning the for children’s aggressive SIP and aggressive responses in validity of VR-based assessment of SIP. provocation scenarios. In sum, interactive VR seems to Our study also had its limitations. First, as few children yield an improved assessment of both children’s reactive responded with aggression in the same scenarios in both VR SIP and proactive SIP patterns and responses compared and vignettes, we were not able to analyze whether inter- to vignettes, but the difference in favor of interactive VR active VR provides an improved assessment of children’s is the largest when measuring children’s reactive SIP pat- positive outcome expectancies and response evaluations of terns and aggression. aggression, as these were assessed only if children had actu- Several findings on separate SIP steps warrant further ally aggressed. Consequently, we tested our hypotheses on discussion. First, children’s interactional goals were the children’s proactive SIP and responses in instrumental gain strongest SIP predictor of their real-life aggression and scenarios using two variables only (i.e., instrumental goals underlying motives, and yielded the largest effect sizes for and aggressive responses). Second, children’s responses levels of aggressive SIP in VR versus vignettes. Moreover, were coded for reliability by the first author, who may have as children’s revenge goals were strongly correlated with been biased because he was aware of the research questions. their anger and hostile intent attributions, they were the only However, inter-rater agreement with a second coder who significant SIP step predicting children’s real-life aggres- was blind to the research questions was excellent, suggest- sion and reactive motives for aggression in most analyses. ing that this bias was limited. Third, as interactive VR is Although such overlap among predictors (i.e., multicolline- obviously more time-consuming and costly to develop than arity) may seem problematic from a statistical point of view, vignettes, we were only able to include four assessment sce- it does make sense conceptually, because children’s interac- narios. Given that children may show aggression in various tional goals seem to be most proximal to their (aggressive) contexts (De Castro & Van Dijk, 2017), it can be assumed behavior and may often derive from preceding SIP steps that using only four scenarios involving playing games with such as anger and hostile intent attributions (Crick & Dodge, peers in a school-setting did not cover the broad range of 1994). social situations known to evoke aggression in children. 1 3 Research on Child and Adolescent Psychopathology Fourth, as children’s SIP and responses were only weakly involve play, settings that allow for the assessment of rela- to moderately associated across scenarios, we conducted our tional aggression (e.g., spreading rumors), or settings that analyses for each scenario separately. Although this finding better allow to examine cooperative behaviors. Last, using aligns with empirical research demonstrating that children’s interactive VR may minimize cognitive load), increasing aggression is situation-specific (e.g., De Castro & Van Dijk, the validity of children’s reported SIP. 2017; Matthys et al., 2001), it prohibited us from testing how In sum, this empirical study demonstrated that interac- reliable our SIP measurements were per type of scenario. tive VR is an improved method to assess children’s aggres- Last, since our study included only boys between 7–13 years sive SIP and behavior compared to a standard vignette- with limited diversity in cultural and socio-economic back- based assessment. The use of VR allows researchers ground, findings cannot directly be generalized to girls, and practitioners to assess aggressive SIP patterns in an older, or younger children, or children from other cultural emotionally engaging, ecologically valid context that is or socio-economic backgrounds than our sample. truly interactive and realistic. Ultimately, interactive VR There are both advantages and disadvantages of using may also facilitate interventions with children, because it interactive VR to assess children’s aggressive SIP and allows for extensive practice with the specific situations responses. One important disadvantage is that interactive relevant to their individual needs, with precise control to nature of VR makes establishing ambiguity of social situ- adapt difficulty and complexity during the intervention. ations more difficult than with vignettes. This interactive Moreover, practitioners may use cooperative contexts nature might enhance the experience of an actual social that yield rewards for specific desirable behaviors (e.g., interaction, however it might also affect ambiguity to some prosocial), reinforcing these behaviors repeatedly through extent (e.g., children who talked a lot with the virtual peer operant conditioning. As such, interactive VR may fur- during the interaction might be prone to attribute non-hostile ther our understanding of the SIP mechanisms underlying intent). Moreover, interactive VR is obviously costly and aggressive behavior problems in children and may enhance time-consuming to develop, and so it is relevant to directly assessment and intervention for children with aggressive compare this method to other assessment methods besides behavior problems. vignettes, such as video game tasks (e.g., Yaros et al., 2014). Supplementary Information The online version contains supplemen- That said, VR has multiple advantages over the use of tary material available at https://doi. or g/10. 1007/ s10802- 021- 00879-w . vignettes. In interactive VR, children are actually pro- voked, tempted to use aggression, and able to aggress Authors’ Contributions The study was designed by all authors. Material against virtual peers, and may therefore experience similar preparation was performed by all authors. Data collection was per- formed by Rogier E.J. Verhoef and trained graduate students. Analyses emotions as in real-life (e.g., anger, frustration), activat- were performed by Rogier E.J. Verhoef. The first draft of the manu- ing similar SIP patterns and responses as in real-world script was written by Rogier E.J. Verhoef and edited by all authors. All interactions. As such, researchers may examine the effect authors commented on previous versions of the manuscript. All authors of a broad range of emotions on children’s SIP; that is, not read and approved the final manuscript. only anger or frustration, but also shame, guilt, fear, desire or sadness. Relatedly, since children actually ‘respond’ Data Availability The data that support the findings of this study are available through the Open Science Framework at https:// doi. org/ 10. in VR, it is possible to include physiological indicators 17605/ OSF. IO/ 7SA6M of children’s arousal, permitting researchers to test more specific hypotheses on the role of emotional arousal in Code Availability The syntax of the analyses run for this study are children’s SIP and responses. Moreover, the large experi- available through the Open Science Framework (see link above). mental control over social stimuli provided by interactive VR (e.g., control over virtual peers’ nonverbal behaviors and emotional expressions) allows researchers to test Compliance with Ethical Standards more specific hypotheses about causal effects of subtle Funding This research was supported by a grant from the Netherlands social cues on children’s SIP and responses than has been Organization for Scientific Research to the last author (grant number feasible thus far. In addition, interactive VR may allow 453–15-004/511). researchers to use a broad variety of emotionally engag- ing contexts known to evoke aggression in children. For Conflicts of Interest We have no known conflict of interest to disclose. example, researchers may present children with more sali- ent cues to evoke proactive SIP and aggression (e.g., by Ethics Approval The study was approved by the Dutch Medical-Ethical Testing Committee Utrecht (METC-Utrecht) and conducted in accord- allowing children to obtain actual gains outside of the VR ance with the 2013 Helsinki Declaration. environment). Relatedly, researchers may also examine children’s SIP in other settings than playing games with Consent to Participate Written informed consent the study was peers, such as settings with parents, settings which do not obtained from parents and children provided verbal assent. 1 3 Research on Child and Adolescent Psychopathology Open Access This article is licensed under a Creative Commons Attri- Dodge, K. A., Lochman, J. E., Harnish, J. D., Bates, J. E., & Pettit, bution 4.0 International License, which permits use, sharing, adapta- G. S. (1997). Reactive and proactive aggression in school chil- tion, distribution and reproduction in any medium or format, as long dren and psychiatrically impaired chronically assaultive youth. as you give appropriate credit to the original author(s) and the source, Journal of Abnormal Psychology, 106(1), 37–51. https:// doi. provide a link to the Creative Commons licence, and indicate if changes org/ 10. 1037/ 0021- 843x. 106.1. 37 were made. The images or other third party material in this article are Dodge, K. A., McClaskey, C. L., & Feldman, E. (1985). Situational included in the article's Creative Commons licence, unless indicated approach to the assessment of social competence in children. otherwise in a credit line to the material. If material is not included in Journal of Consulting and Clinical Psychology, 53(3), 344–353. the article's Creative Commons licence and your intended use is not https:// doi. org/ 10. 1037/ 0022- 006X. 53.3. 344 permitted by statutory regulation or exceeds the permitted use, you will Hubbard, J. A., McAuliffe, M. D., Morrow, M. T., & Romano, L. need to obtain permission directly from the copyright holder. To view a J. (2010). Reactive and proactive aggression in childhood and copy of this licence, visit http://cr eativ ecommons. or g/licen ses/ b y/4.0/ . adolescence: Precursors, outcomes, processes, experiences, and measurement. Journal of Personality, 78(1), 95–118. https://doi. org/ 10. 1111/j. 1467–6494. 2009. 00610.x Lansford, J. E., Malone, P. S., Dodge, K. A., Crozier, J. C., Pettit, G. References S., & Bates, J. E. (2006). A 12-year prospective study of pat- terns of social information processing problems and external- Anderson, C. A., & Bushman, B. J. (2002). Human aggression. izing behaviors. Journal of Abnormal Child Psychology, 34(5), Annual Review of Psychology, 53(1), 27–51. https:// doi. or g/ 709–718. https:// doi. org/ 10. 1007/ s10802- 006- 9057-4 10. 1146/ annur ev. psych. 53. 100901. 13523 1xs Lemerise, E. A., & Arsenio, W. F. (2000). An integrated model of Bushman, B. J., & Anderson, C. A. (2001). Is it time to pull the emotion processes and cognition in social information process- plug on hostile versus instrumental aggression dichotomy? Psy- ing. Child Development, 71(1), 107–118. https:// doi. or g/ 10. chological Review, 108(1), 273–279. https:// doi. org/ 10. 1037/ 1111/ 1467- 8624. 00124 0033- 295X. 108.1. 273 Martinelli, A., Ackermann, K., Bernhard, A., Freitag, C. M., & Caporaso, J. S., & Marcovitch, S. (2021). The effect of taxing situations on Schwenck, C. (2018). Hostile attribution bias and aggression in preschool children’s responses to peer conflict. Cognitive Develop- children and adolescents: A systematic literature review on the ment, 57(6), 100989. https://doi. or g/10. 1016/j. cogde v.2020. 100989 influence of aggression subtype and gender. Aggression and Violent Crick, N. R., & Dodge, K. A. (1994). A review and reformulation of Behavior, 39, 25–32. https:// doi. org/ 10. 1016/j. avb. 2018. 01. 005 social-information-processing mechanisms in children’s social Matthys, W., Maassen, G. H., Cuperus, J. M., & van Engeland, H. adjustment. Psychological Bulletin, 115(1), 74–101. https://doi. (2001). The assessment of the situational specificity of chil- org/ 10. 1037// 0033- 2909. 115.1. 74 dren’s problem behaviour in peer–peer context. Journal of Child Crick, N. C., & Dodge, K. A. (1996). Social information processing Psychology and Psychiatry, 42(3), 413–420. https://doi. or g/10. deficits in reactive and proactive aggression. Child Develop-1111/ 1469- 7610. 00734 ment, 67(3), 993–1002. https:// doi. org/ 10. 2307/ 11318 75 McCulloch, C. E. (1987). Tests for equality of variances with paired De Castro, B. O., & Van Dijk, A. (2017). “It’s gonna end up with a data. Communications in Statistics - Theory and Methods, fight anyway”: Social cognitive processes in children with dis- 16(5), 1377–1391. https://doi. or g/10. 1080/ 03610 92870 88294 45 ruptive behavior disorders. In: Lochman, J. E., & Matthys, W. Oostermeijer, S., Nieuwenhuijzen, M., van de Ven, P. M., Popma, A., (Eds.), The Wiley handbook of disruptive and impulse-control & Jansen, L. M. (2016). Social information processing problems disorders (pp. 237–253). John Wiley & Sons Limited. https:// related to reactive and proactive aggression of adolescents in doi. org/ 10. 1002/ 97811 19092 254. ch15 residential treatment. Personality and Individual Differences, De Castro, B. O., Merk, W., Koops, W., Veerman, J. W., & Bosch, 90, 54–60. https:// doi. org/ 10. 1016/j. paid. 2015. 10. 035 J. D. (2005). Emotions in social information processing and Polman, H., de Castro, B. O., Koops, W., van Boxtel, H. W., & Merk, their relations with reactive and proactive aggression in W. W. (2007). A meta-analysis of the distinction between reac- referred aggressive boys. Journal of Clinical Child and Ado- tive and proactive aggression in children and adolescents. Jour- lescent Psychology, 34(1), 105–116. https:// doi. org/ 10. 1207/ nal of Abnormal Child Psychology, 35(4), 522–535. https://doi. s1537 4424j ccp34 01_ 10org/ 10. 1007/ s10802- 007- 9109-4 De Castro, B. O., Slot, N. W., Bosch, J. D., Koops, W., & Veerman, Polman, H., Orobio de Castro, B., Thomaes, S., & van Aken, M. J. W. (2003). Negative feelings exacerbate hostile attributions (2009). New directions in measuring reactive and proactive of intent in aggressive boys. Journal of Clinical Child and aggression: Validation of a teacher questionnaire. Journal of Adolescent Psychology, 31(1), 56–65. https://doi. or g/10. 1207/ Abnormal Child Psychology, 37(2), 183–193. https:// doi. org/ s1537 4424j ccp32 01_ 0610. 1007/ s10802- 008- 9266-0 De Castro, B., Verhulp, E. E., & Runions, K. (2012). Rage and Reijntjes, A., Thomaes, S., Kamphuis, J., Bushman, B., De Castro, revenge: Highly aggressive boys’ explanations for their B. O., & Teich, M. (2011). Explaining the paradoxical rejec- responses to ambiguous provocation. European Journal of tion-aggression link: The mediating effects of hostile intent Developmental Psychology, 9(3), 331–350. https:// doi. org/ 10. attributions, anger, and decreases in state self-esteem on peer 1080/ 17405 629. 2012. 680304 rejection-induced aggression in youth. Personality and Social Dodge, K. A. (1991). The structure and function of reactive and pro- Psychology Bulletin, 37(7), 955–963. https:// doi. org/ 10. 1177/ active aggression. In D. Pepler & K. H. Rubin (Eds.), The devel-01461 67211 410247 opment and treatment of childhood aggression (pp. 201–218). Robinson, M. D., & Clore, G. L. (2002). Belief and feeling: Evi- Lawrence Erlbaum Associates Inc. dence for an accessibility model of emotional self-report. Psy- Dodge, K. A. (2011). Social information processing patterns as chological Bulletin, 128(6), 934–960. https:// doi. org/ 10. 1037/ mediators of the interaction between genetic factors and life 0033- 2909. 128.6. 934 experiences in the development of aggressive behavior. In P. Scholte, E. M., & van der Ploeg, J. D. (2007). Handleiding Sociaal- Shaver & M. Mikulincer (Eds.), Human aggression and vio- Emotionele Vragenlijst (SEV). Bohn Stafleu van Loghum. lence: Causes, manifestations, and consequences (pp. 165– Schubert, T., Friedmann, F., & Regenbrecht, H. (1999). Embodied 185). American Psychological Association. presence in virtual environments. In R. Paton & I. Neilson 1 3 Research on Child and Adolescent Psychopathology (Eds.), Visual representations and interpretations (pp. 268– Verhoef, R. E. J., Van Dijk, A., & De Castro, B. O. (2021a). A Dual 278). Springer-Verlag. Mode Social Information Processing Model to Explain Individual Statistics Netherlands (2018, 2019). StatLine. Retrieved June 24, Differences in Children’s Aggressive Behavior. Clinical Psycho- 2021, from https:// www. opend ata. cbs. nl logical Science, Advance Online Publication. https:// doi. org/ 10. Steiger, J. H. (1980). Tests for comparing elements of a correlation 1177/ 21677 02621 10163 96 matrix. Psychological Bulletin, 87(2), 245–251. https://doi. or g/ Verhoef, R. E. J., Van Dijk, A., Verhulp, E. E., & De Castro, B. O. 10. 1037/ 0033- 2909. 87 (2021b). Interactive virtual reality assessment of aggressive social Stoltz, S., van Londen, M., Dekovic, M., De Castro, B. O., Prinzie, information processing in boys with behaviour problems: A pilot P., & Lochman, J. E. (2013). Effectiveness of an individual study. Clinical Psychology & Psychotherapy, 28(3), 489–499. school-based intervention for children with aggressive behav-https:// doi. org/ 10. 1002/ cpp. 2620 iour: A randomized controlled trial. Behavioural and Cogni- Verhulst, F. C., Van Der Ende, J., & Koot, H. M. (1997). Handleiding tive Psychotherapy, 41(5), 525–548. https:// doi. org/ 10. 1017/ voor de Teachers's Report Form (TRF). Rotterdam: afd. Kinder- s1352 46581 20005 25 en jeugdpsychiatrie, Sophia Kinderziekenhuis/ AZR/ EUR Underwood, M. K. (2005). Observing anger and aggression among Yaros, A., Lochman, J. E., Rosenbaum, J., & Jimenex-Camargo, L. A. preadolescent girls and boys: Ethical dilemmas and practical solu- (2014). Real-time hostile attribution measurement and aggression tions. Ethics and Behavior, 15(3), 235–245. https:// doi. org/ 10. in children. Aggressive Behavior, 40(5), 409–420. https://do i.or g/ 1207/ s1532 7019e b1503_410. 1002/ ab. 21532 Van Dijk, A., Hubbard, J. A., Deschamps, P. K. H., Hiemstra, W., & Polman, H. (2021). Do distinct groups of reactively and proac- Publisher's Note Springer Nature remains neutral with regard to tively aggressive children exist? Research on Child and Adoles- jurisdictional claims in published maps and institutional affiliations. cent Psychopathology. Advance online publication. https:// doi. org/ 10. 1007/ s10802- 021- 00813-0 Table 4 Hierarchical regression analyses of real-life aggression Verhoef, R. E. J., Alsem, S. C., Verhulp, E. E., & de Castro, B. O. regressed both on reactive SIP and aggressive responses (2019). Hostile intent attribution and aggressive behavior in children revisited: A meta-analysis. Child Development, 90(5), 525–547. https:// doi. org/ 10. 1111/ cdev. 13255 1 3

Journal

Journal of Abnormal Child PsychologySpringer Journals

Published: Oct 14, 2021

Keywords: Social information processing; Aggression; Children; Virtual Reality; Reactive and proactive motives

References