INFLUENCE OF PSYCHOPATHIC IMPULSIVE-IRRESPONSIBLE TRAITS ON REACTIVE AGGRESSION AFTER RECEIVING NEGATIVE SOCIAL FEEDBACK IN EARLY ADULTHOOD

(1)

Bachelor Thesis Julie Tournier

LEIDEN UNIVERSITY AND VUMC AMSTERDAM

24-01-2020

INFLUENCE OF PSYCHOPATHIC

IMPULSIVE-IRRESPONSIBLE TRAITS

ON REACTIVE AGGRESSION AFTER

RECEIVING NEGATIVE SOCIAL

FEEDBACK IN EARLY ADULTHOOD

(2)

Abstract

How we react towards social feedback is important for our social network and social life. In some situations, individuals react aggressively toward people who gave them negative social feedback, this is called reactive aggression. People with high presence of psychopathic impulsive-irresponsible traits are more likely to react impulsively and develop persisting anti-social behaviour. In the current study the relation of impulsive-irresponsible traits on reactive aggression will be studied utilizing the Social Network Aggression Task. The Social Network Aggression Task measures reactive aggression after receiving social feedback from a peer. In this task participants are presented with social feedback (negative, neutral or positive) along with pictures of peers who gave the feedback on their personal profile. Afterwards,

participants are supposed to send a noise blast to the peer, by pressing a button, as measurement of reactive aggression. The length and latency after different feedback conditions will be compared between participants with either presence or absence of the impulsive-irresponsible traits. Presence of the impulsive-irresponsible traits are measured with the self-report questionnaire: Youth Psychopathic Traits Inventory. Analyses showed that the Social Aggression Network Task is a valid method to measure different reactive aggression after social feedback. However, there are no influence of impulsive-irresponsible traits on reactive aggression found for both noise blast duration and latency.

(3)

Introduction

Almost every day people receive social feedback from another person. Everyone reacts differently to social feedback. The way we respond to each other and to feedback from other peers is important for our social life. People have the urge to be accepted by others and are motivated to receive social acceptance. Therefore, negative feedback sometimes leads to an aggressive reaction which is, in some situations, not always appropriate. When someone interprets a peers’ social evaluation as harmful to themselves and reacts aggressively, it’s termed as reactive aggression (Crick & Dodge, 1996). Reactive aggression can influence someone’s social network and have negative impact on their social life. Therefore, it is important to investigate how differences in reaction toward social feedback emerge. We hypothesize an association between psychopathic impulsive-irresponsible traits and reactive aggression after peer rejection. Because people with presence of impulsive-irresponsible traits are more likely to react impulsive and with temper. Besides this, a previous study discovered an association between reactive aggression after peer rejection and hyperactivity-impulsivity (Evans, Fite, Hendrickson, Rubens, & Mages, 2015). These findings suggest that impulsiveness and irresponsibility is a possible predictor of reactive aggression after negative peer feedback. In the current study we investigate (1) the relation between negative social feedback and reactive aggression and (2) whether impulsive-irresponsible traits can predict reactive aggression.

First, we will discuss the social information processing model (SIP), to explain how social feedback is processed. According to the social information processing (SIP) model, social information is processed in certain steps. First: the interpretation of the social cue, second: clarification of the meaning, third: the response access or construction, fourth: the response

(5)

decision and at last behavioural enactment (Crick & Dodge, 1996). These steps are presented in the model in figure 1 (Crick, N. R., Dodge, 1994).

Figure 1 Social Information Processing model. (Crick, N. R., Dodge, 1994)

Differences in reaction depend on the stimulus and the way of processing the steps. Concerning reactive aggression, it is hypothesized that the steps of encoding and attribution of intent affect the divergent behaviour (de Castro, 2004). Following this model, it can be hypothesized that the rejection from a peer is experienced as a social stressor. This stressor results in an aggressive reaction towards the peer and could result in anti-social behaviour (Dodge & Pettit, 2003). The SIP model is a relevant theory to compare different behaviour and development of children. A later study discovered that the development of anti-social behaviour is very different (Odgers et al., 2008). This indicates that a model predicting the development of anti-social behaviour is yet to be found. It is therefore in considerable value to

(6)

look for possible predictors, such as reactive aggression, which could foresee the development of anti-social behaviour. In other words, the way we process and react to social situation and feedback is very divergent and could result in anti-social behaviour. The question remains where differences in reaction after a negative social evaluation derive from. Achterberg and colleagues (2016) designed an experimental study where different reactions after social feedback can be analysed, which will be used in this current study. Achterberg and colleagues investigated the connection between reactive aggression regulation and social feedback on a behavioural and neuronal level in using the Social Network Aggression Task (SNAT). In this experiment, participants had to fill in a personal profile. Afterwards, they received positive, neutral or negative feedback from other peers on their profile. To examine the reaction and possible reactive aggression, the participants were asked to send a noise blast to the peers who evaluated them. Besides this, the brain activity was measured using functional magnetic resonance imaging (fMRI) to examine neural mechanisms behind the response. It was found that some participants had high reactive aggression towards the peers who gave a negative evaluation. They also found that positive vs. neutral feedback resulted in increased activity in the striatum and ventral medial prefrontal cortex (mPFC) and negative vs. neutral feedback resulted in increased activity in the dorsal mPFC and insula. In addition, Achterberg and colleagues found an indication of good aggression regulation which is associated with increased activity in the lateral prefrontal cortex (PFC) after sending a short noise blast after receiving negative social feedback (Achterberg, vanDuijvenvoorde, Bakermans-Kranenburg, & Crone, 2016). In conclusion, these findings suggest that people react differently to negative feedback. Achterberg and colleagues (2016) speculates in their conclusion that people who are better regulating behavioural impulses have lower impulsive response. This leads us to the

(7)

second goal of our study, to check whether impulsiveness has an influence on reactive aggression after receiving social feedback.

As mentioned earlier, there is a connection between hyperactivity-impulsiveness and reactive aggression after receiving negative social feedback from a peer in adolescence (Evans et al., 2015). Adding up to this, the lateral prefrontal cortex (PFC), which is responsible for impulse control, is important in modulating different aggression responses (Achterberg et al., 2016). These results suggest a possible link between the presence of impulsive behaviour and reactive aggression. Impulsive behaviour is higher in people with higher presence of irresponsible psychopathic traits, which is why we will examine the presence of impulsive-irresponsible traits. The impulsive-impulsive-irresponsible traits are a subscale of the youth psychopathic personality traits (Andershed, Hodgins, & Tengström, 2007). Aside from this, the presence of psychopathic traits are found to increase the risk of persisting anti-social behaviour (Cohn et al., 2015). Additionally, psychopathic traits are precursors for psychopathy (Andershed et al., 2007), and are addressed differently by numerous studies. Psychopathy is categorised as a mental disorder and consequently classified as a mental illness (Insel et al., 2010). Studies have created a three-factor model of psychopathic personality which consists of the following three traits: callous-unemotional, grandiose-manipulative and impulsive-irresponsible (Pape et al., 2015). The psychiatric traits could be examined using questionnaires, for example, the Youth Psychopathic Traits Inventory (YPI). In the current study we will use the YPI to examine whether early adults have high or low presence of the impulsive-irresponsible traits and to investigate whether these traits effect on the reaction after receiving negative social feedback. We examine this in early adults because first, we replicate the design from Achterberg et al. (2016), which uses a similar age sample (age range: 18-27). Second, it is interesting to test this age because most results found a stabilisation of psychopathic traits from late adolescence to early

(8)

adulthood (Hemphälä, Kosson, Westerman, & Hodgins, 2015). We will use the SNAT from Achterberg and colleagues (2016). With the SNAT reactive aggression is measured by how fast and how long a noise blast is given after receiving a negative social evaluation from a peer.

In other words, the current study will investigate the following question: Does the presence of psychopathic impulsive-irresponsible traits have influence on reactive aggression after receiving negative social feedback? We will investigate this by replicating the study from Achterberg et al. (2016) to check the different reactions on social feedback on a behavioural level by examining how aggressive the response is and how fast participants respond. Besides this we will investigate if impulsive-irresponsible traits influence the reaction after receiving social feedback. It is hypothesized that the group with a higher impulsive-irresponsible trait will have more trouble with aggression regulation. In other words, they are expected to give a harder noise blast towards negative social feedback from the peer. Additionally, it is also expected that they would give this noise blast quickly after receiving negative social evaluation.

(9)

Methods

Participants

Forty participants participated in this study, eleven participants were removed from the sample because, they did not fill in the Youth Traits Inventory and/or there were other missing values. This resulted in a total of twenty-nine participants. The sample consisted of 17 females and 12 males (M= 23,21 years, SD= 3.04). Participants were recruited via social media or in public areas (gyms) in Leiden. All participants had to sign the informed consent and spoke Dutch fluently. The participants’ intelligence quotient (IQ) was measured with two tests of the Wechsler Intelligence Scale for Adults, third education (WAIS-IV): ‘similarities’ and ‘block design’ (Wechsler & Corporation., 1997). The outcomes of the estimated IQ scores were between 85 and 130 (M= 106, SD=17.10).

Procedure

Psychopathic traits questionnaire

The Impulsive-irresponsible trait was measured using the Youth Psychopathic Traits Inventory (YPI). The YPI is self-report questionnaire consisting 50 questions, it measures 3 different psychopathic traits including subscales: (1) the Grandiose/Manipulative factor (subscales: Dishonest Charm, Grandiosity, Lying, and Manipulation); (2) the

Callous/Unemotional factor (subscales Callousness, Unemotionality, and Remorse- lessness); and (3) the Impulsive/Irresponsible factor (subscales Impulsiveness, Thrill-Seeking, and Irresponsibility) (Andershed, Gustafson, Kerr, & Stattin, 2002). The questions consist of a 4-point scale where 1 represents ‘does not apply at all’ and 4 represented ‘applies very well’. Questions that scored for impulsive-irresponsible were subtracted (questions: 1, 3, 4, 5, 9, 13

(10)

,16 ,18, 22, 26, 29, 32, 34, 40 and 42). Scores were summed up for every participant

(maximum score=60). The sample was dived in the low and high impulsive-irresponsible using a median split.

Social Network Aggression task

During this study, we used an adapted version of the Social Network Aggression Task (SNAT, Achterberg et al., 2016). The experiment was executed in a magnetic resonance imaging (MRI) scanner, functional (T2*) images were obtained. Before participants completed the task during an MRI session, they were asked to fill out a personal profile (Appendix A). The profile consisted of personal statements like “my favourite sports is .. “or

“my biggest wish is..“.

During the experiment in the scanner, participants reportedly received negative, neutral or positive feedback from same-age peers on their profile which were randomly assigned. Negative feedback was presented with a red thumb down meaning the peer didn’t like their profile. Positive feedback was presented with a green thumb up which would mean the peer liked their profile. Neutral feedback was presented with a grey dot which presents the peer did not know what to think of the profile (Figure 2a). After the feedback, the participants were asked to send a hypothetical noise blast by pressing a button, where a longer button press corresponds to a louder noise blast. This was visualized with a volume bar (Figure 2b). The participants were asked to imagine that they would send the noise blast towards their peer, but they wouldn’t actually receive it. Previous studies showed that imagining sending a noise blast results in enough aggression compared to really sending the noise blast (Konijn, Bijvank, & Bushman, 2007). As shown in Figure 2c, the feedback is coupled with a picture of the peer which was, taken from a well-validated database (Moor,

(11)

Van Leijenhorst, Rombouts, Crone, & Van Der Molen, 2010). The pictures of both male and female peers were randomly assigned to a type of feedback. Before the real experiment, the participants familiarized with the design of the experiment by means of six practice trials. One trial of the SNAT is designed as shown in Figure 2.c: first, there is a fixation screen (500 ms) which is followed by the judgement of the peer (negative, neutral or positive) (2500 ms). After the judgement, there is again a fixation screen (3000-5000 ms) followed by the volume bar where the participant is supposed to give the noise blast (5000 ms). The trial is ended with again a fixation screen (0-11550 ms). The total experiment exists of 3 blocks, where

every block consists of 30 trials.

Figure 2c Design and timing of the Social Network Aggression Task.

Figure 2 The Social Network Aggression Task. (a) illustration from the positive, neutral and negative feedback. (b)

Illustration of the volume bar while giving the noise blast. (c) Timing from one trial from the SNAT. (Achterberg et

al., 2016)

Figure 2a Different feedback types: positive, neutral and negative

Figure 2b Illustration of the volume bar with each representing a louder noise blast.

(12)

Manipulation checks

On the day of the study, the participants first filled out several control questions. To check whether they liked or disliked the noise blast they would send later. The participants had to rate the noise blast two times: “How do you like the noise blast” and “How much do

you dislike the noise blast”. Rating was done with a 7-point Likert scale where 1 represented

‘very little’ and 7 represented ‘very much’.

The second manipulation checks are the exit question to check how the participants experienced receiving social feedback from a peer. After the SNAT, the participants were asked three exit questions where they had to indicate how much they liked to receive a positive, neutral and negative feedback on a 7-point scale. 1 represented ‘very little’ and 7 represented ‘very much’.

Behavioural analyses

Manipulation analyses

We analysed the manipulation checks using Rstudio Version 1.2.5001. For the noise blast ratings, we calculated the mean and standard deviation. The exit questions will be analysed using a repeated measures ANOVA. This measured whether there is a difference between the ratings on neutral, negative or positive feedback. The rating will be compared within subjects for the three different conditions. To check if the ratings were the same for both high and low I-I groups, we will conduct a repeated measures ANOVA to check if there are differences found within high and low I-I groups. If assumptions of sphericity are not met, the Greenhouse-Geisser correction was applied. We chose the conservative correction

(13)

measures ANOVA (cf. P<.05) were followed up by paired t-tests with a Bonferroni correction to investigate which groups differed.

SNAT analyses

First, we conducted a repeated measures ANOVA to check for differences in duration of noise blast and latency of the noise blast dependent on the type of feedback (negative-neutral, negative-positive, neutral-positive). If assumptions of sphericity were not met, the Greenhouse-Geisser correction was applied. Significant differences in the repeated measures ANOVA (cf. P<.05) were followed up by paired t-tests with a Bonferroni correction to check which conditions differed.

Psychopathic traits analyses

We subsequently performed the same ANOVAs with Impulsive-Irresponsible traits added as a between subjects’ factor. The scores from the impulsive-irresponsible (I-I) items of the YPI are taken together for every participant. Based on these sums scores, the sample was divided in a low impulsive-irresponsible and high impulsive-irresponsible group based on a median split. After we conducted a repeated measures ANOVA for differences in noise blast and latency within the participants for different feedback condition and between high I-I and low I-I groups. As mentioned before if needed, we used the same Greenhouse-Geisser correction and post-hoc paired t-test (cf. P<0.5) with a Bonferroni correction to investigate which groups differed.

(14)

Results

Behavioural analyses

Noise blast manipulation task

The participants evaluated how much they liked and disliked the noise blast, it appeared they did not like the sound (M=1.97, SD=1.48, range=1-6). Additionally, they disliked the noise blast a lot (M=5.90, SD=1.42, range=1-7). We also conducted this for the high and low impulsive-irresponsible groups to check if there are differences found. The impulsive-irresponsible groups were separated with a median split (median=30). This resulted in the high I-I group: participants who scored higher than 30 (n=13) and the low I-I group: participants who scored lower than 30 (n=16). Both groups did not like the sound, the high impulsive-irresponsible group (M=1.82, SD=1.33) rated the noise blast comparable to the low impulsive-irresponsible group (M=2, SD=1.4). Additionally, both the high I-I group (M=5.91, SD=1.81) and low I-I group (M=5.69, SD=1.25) rated the noise blast as very unpleasant.

Social feedback manipulation task

To check how the participants experienced receiving negative, neutral or negative social feedback we conducted a manipulation task. We analysed the exit questions with a repeated measures ANOVA. This showed an overall significant difference between the conditions F(2,50)=6.66, P=.0068 (GG corrected). A Post Hoc test paired t-test (Bonferroni corrected) showed that people disliked negative feedback (M=3.81, SD=.49) over receiving positive feedback (M=4.5, SD=.49, P=.029). Besides this, they preferred getting positive feedback over neutral feedback (M=3.96, SD=.60, P=.031). However, there was no significant difference between neutral and negative feedback (P=.882). To check if people rated this

(15)

with impulsive-irresponsible (I-I) traits (2: High vs. Low) and condition (3: Negative vs. Positive vs. Neutral) as independent variables. There is no significant for impulsive-irresponsible traits on the ratings F(1,22)=1.37, P=.25. There is also no interaction effect found for impulsive-irresponsible traits between conditions F(2,44)=.21, P=.71 (GG corrected).

Noise blast duration

A repeated measures ANOVA was performed, to check if a difference in length of the noise blast given after a negative, neutral or positive feedback for the complete sample. These results are presented in figure 2. We found an overall significant difference between the feedback types F(2,56)=26.28, P<0.001 (GG corrected). In the paired t-test (Bonferroni corrected) we found that a noise blast was given significantly longer in response to negative feedback (M=1126.16, SD=763.26) compared to the neutral feedback (M=716.45, SD=519.76,

P<.001). Furthermore, a significantly longer noise blast was given after neutral feedback

compared to positive feedback (M=457.93, SD=357.72, P=.004). The participants also gave a significantly longer noise blast after negative feedback compared to after the positive feedback (P<.001).

(16)

Figure 3 Duration of noise blast in milliseconds after receiving a neutral (M=716.45, SD=519.76), negative

(M=1126.16, SD=763.26) or positive (M=457.93, SD= 357.72) feedback.

Latency

To examine whether there was a difference between reaction times, we did a repeated measures ANOVA for latency. The latency in milliseconds for neutral (M=520.24,

SD=104.78), negative (M=516.16, SD=95.31) and positive (M=606.47, SD=96.97) feedback is

shown in Figure 4. No significant difference between the conditions was observed F(2,56)=.63, P=.53.

(17)

Figure 4 Latency (reaction time) in milliseconds after receiving a neutral (M=520.24, SD=104.78), negative

(M=516.16, SD=95.31) or positive (M=606.47, SD=96.97).

Impulsive-irresponsible traits of the YPI

After the overall analyses, we conducted a 2 x 3 mixed design ANOVA based with impulsive-irresponsible (I-I) traits (2: High vs. Low) and condition (3: Negative vs. Positive vs. Neutral) as independent variables. We conducted these analyses for noise blast (figure 5) and latency (figure 6). We compared differences in noise blast duration after a neutral, negative or positive feedback between the high and low I-I groups, this was also checked in between the feedback conditions. There is no main effect found for impulsive-irresponsible

F(1,27)=1.96, P=.17. There is also no interaction effect found between

impulsive-irresponsible and condition F(2,54)=.37, P=.60 (GG corrected). Besides this we also checked if the low I-I group would react faster or slower with a noise blast after receiving neutral, negative or positive feedback compared to the high I-I group. Again, there is a main effect for

(18)

impulsive-irresponsible F(1,27)=.03, P=.087. Besides this there is also no interaction effect found for impulsive-irresponsible and condition F(2,54)=.41, P=.67. The same analyses were executed without outliers, since this showed similar results and no significant effect, we decided to include the outliers in the results.

(19)

Figure 6 Latency for given noise blast after neutral, negative or positive feedback for high and low

(20)

Discussion

The first goal of this study was to replicate the study from Achterberg and colleagues (2016). The second goal of the study was to test the role of impulsive-irresponsible traits on reactive aggression after receiving social feedback from a peer. As hypothesized the results from the current study are similar like the results found in Achterberg and colleagues (2016). We found that when participants received negative social feedback, they would send a longer, which corresponds to a hypothetical higher, noise blast compared to when they received neutral to negative social feedback. This also accounts for positive vs. neutral and positive vs. negative. In this study we also measured latency, which measured how fast the noise blast was sent after receiving a social feedback. There were no differences between the different conditions feedback. This implies that the kind of social feedback does not affect the reaction time. The second aim of the study was to check whether the presence of impulsive-irresponsible traits would influence the reaction after receiving a social feedback. We hypothesized that higher presence of impulsive-irresponsible traits would result in reactive aggression. This would mean a longer noise blast and shorter latency. There was no significant difference found for the duration of the noise blast sent between the group that scored high and the group who scored low on impulsive-irresponsible traits. Moreover, there was no significant difference found between the two groups regarding latency and the noise blast. There was also no interaction effect discovered of impulsive-irresponsible traits between the three different feedback types on the latency, how fast they sent the noise blast. This also accounts for the length of the noise blast there was no significant interaction effect found. Additionally, both high and low I-I groups scored the same in the manipulation checks, indicating that there the groups experienced receiving social feedback as the same.

(21)

Besides this, no difference was found on rating the noise blast between the high and low I-I groups. The results of the current study indicate that impulsive-irresponsible psychopathic traits are not related to reactive aggression towards social feedback. In conclusion, high or low presence of impulsive-irresponsible psychopathic traits are not related to reactive aggression after receiving social feedback.

There are some limitations concerning the study. First, an unexpected result in the manipulation checks was observed. There was no significant difference found in the rating of the how much participants liked receiving a neutral or negative social feedback. Interestingly, participants sent a longer noise blast after receiving negative feedback compared to neutral feedback while as mentioned before, they experienced the feedback as the same. Possible explanations could be that some of the participants did not understand the exit questions and filled in something wrong. Since this is not found in the study from Achterberg and colleagues (2016), we think these results are to limitation in our study, and not due to the design of the Social Network Aggression Task. A second limitation is that the impulsive-irresponsible high and low groups were formed using median split, to make the YPI scores categorical. It is questionable whether categorizing is preferable over using a continuous scale. By categorizing, participants with a score of 29 or 31 are placed in different groups even though the scored are not very different. In the sample there are no extreme low or high scores and a lot of participants scored around 30. This could explain why there are no differences found between high and low impulsive-irresponsible traits. Adding up, the YPI is a well validated questionnaire, however we subtracted the part of the questionnaire measuring impulsive-irresponsible traits. It is questionable if the validation still accounts in this situation. Another possible limitation concerning the SNAT is the reliability of the feedback. Despite the fact that there was no suspicion about the feedback being randomly assigned, it might be

(22)

possible that some participants thought it was a setup task. This may have led to a biased response if you think the feedback is not really given by the people in the picture. It is suggested to test for this, for higher ecological validity in future usage of the SNAT.

As mentioned before, we did not find an influence of impulsive-irresponsible traits on reactive aggression. The different reactions after a social feedback are probably due to the interpretation and not due to some impulsive reaction. This is what we also concluded when the impulsive-irresponsible low and high groups were compared. This supports the social information processing model (Crick, N. R., Dodge, 1994), which hypothesized that reactive aggression comes from divergent processing of the first two steps in the SIP (Figure 1). Since there were no high scores for impulsive-irresponsible traits in this study, it is suggested repeat this study with sample with higher probability of having high prevalence of impulsive-irresponsible traits. Therefore, it is suggested to compare this a sample with people who score high for psychopathic traits who are at risk to develop persisting anti-social behaviour (Cohn et al., 2015). We suggest to look for specific markers and to elaborate the social information processing model. A more elaborated information processing model would contribute to a better understanding of how people react differently towards negative feedback.

Summing up, this study supports earlier findings that the Social Network Aggression Task is a good task to measure different reactions after receiving different kinds of social feedback from a peer. However, there is no evidence found for impulsive-irresponsible traits which could explain reactive aggression. Concerning the limitations, future studies are suggested to further explore these findings to gain more knowledge about aggressive aggression and find possible predictors.

(23)

References

Achterberg, M., van Duijvenvoorde, A. C. K., Bakermans-Kranenburg, M. J., & Crone, E. A. (2016). Control your anger! The neural basis of aggression regulation in response to negative social feedback. Social Cognitive and Affective Neuroscience, 11(5), 712–720. https://doi.org/10.1093/scan/nsv154

Andershed, H., Gustafson, S. B., Kerr, M., & Stattin, H. (2002). The usefulness of self-reported psychopathy-like traits in the study of antisocial behaviour among non-referred

adolescents. European Journal of Personality, 16(5), 383–402. https://doi.org/10.1002/per.455

Andershed, H., Hodgins, S., & Tengström, A. (2007). Convergent Validity of the Youth Psychopathic Traits Inventory (YPI). Assessment, 14(2), 144–154.

https://doi.org/10.1177/1073191106298286

Bathke, A. C., Schabenberger, O., Tobias, R. D., & Madden, L. V. (2009). Greenhouse-Geisser

Adjustment and the ANOVA-Type Statistic: Cousins or Twins?

https://doi.org/10.1198/tast.2009.08187

Cohn, M. D., Pape, L. E., Schmaal, L., Brink, W. Van Den, Wingen, G. Van, Vermeiren, R. R. J. M., … Popma, A. (2015). Differential Relations Between Juvenile Psychopathic Traits and

Resting State Network Connectivity. 2405(February), 2396–2405.

https://doi.org/10.1002/hbm.22779

Crick, N. R., Dodge, K. A. (1994). Review of Social Information Processing in Children Social Adjustment_1994. Psychological Bulletin, pp. 74–101.

(24)

Proactive Aggression Child Development. 993–1002.

de Castro, B. O. (2004). The development of social information processing and aggressive behaviour: Current issues. European Journal of Developmental Psychology, 1(1), 87–102. https://doi.org/10.1080/17405620444000058

Dodge, K. A., & Pettit, G. S. (2003). A biopsychosocial model of the development of chronic conduct problems in adolescence. Developmental Psychology, 39(2), 349–371.

https://doi.org/10.1037/0012-1649.39.2.349

Evans, S. C., Fite, P. J., Hendrickson, M. L., Rubens, S. L., & Mages, A. K. (2015). The Role of Reactive Aggression in the Link Between Hyperactive–Impulsive Behaviors and Peer Rejection in Adolescents. Child Psychiatry and Human Development, 46(6), 903–912. https://doi.org/10.1007/s10578-014-0530-y

Hemphälä, M., Kosson, D., Westerman, J., & Hodgins, S. (2015). Stability and predictors of psychopathic traits from mid-adolescence through early adulthood. Scandinavian

Journal of Psychology, 56(6), 649–658. https://doi.org/10.1111/sjop.12257

Insel, T., Cuthbert, B., Garvey, M., Heinssen, R., Pine, D. S., Quinn, K., … Wang, P. (2010). Research Domain Criteria (RDoC): Toward a New Classification Framework for Research on Mental Disorders. American Journal of Psychiatry, 167(7), 748–751.

https://doi.org/10.1176/appi.ajp.2010.09091379

Konijn, E. A., Bijvank, M. N., & Bushman, B. J. (2007). I Wish I Were a Warrior: The Role of

Wishful Identification in the Effects of Violent Video Games on Aggression in Adolescent

Boys. https://doi.org/10.1037/0012-1649.43.4.1038

(25)

(2010). Social Neuroscience Do you like me? Neural correlates of social evaluation and

developmental trajectories. https://doi.org/10.1080/17470910903526155

Odgers, C. L., Moffitt, T. E., Broadbent, J. M., Dickson, N., Hancox, R. J., Harrington, H., … Caspi, A. (2008). Female and male antisocial trajectories: From childhood origins to adult

outcomes. https://doi.org/10.1017/S0954579408000333

Pape, L. E., Cohn, M. D., Caan, M. W. A., van Wingen, G., van den Brink, W., Veltman, D. J., & Popma, A. (2015). Psychopathic traits in adolescents are associated with higher

structural connectivity. Psychiatry Research - Neuroimaging, 233(3), 474–480. https://doi.org/10.1016/j.pscychresns.2015.07.023

Wechsler, D., & Corporation., P. (1997). WAIS-III : administration and scoring manual :

(26)