Psychometric Properties of the Reactive-Proactive Aggression Questionnaire Among a Sample of Detained and Community Girls

Abstract

Research on adolescent aggression often distinguishes between reactive and proactive aggression, but there is limited knowledge regarding the application or measurement of these subtypes of aggression among female youth, especially forensic samples of female youth. Drawing on 377 girls (103 selected from forensic settings and 274 selected from school settings) from Portugal, the current study is the first to simultaneously examine the psychometric properties of the Reactive-Proactive Aggression Questionnaire (RPQ) in incarcerated girls and community girls. The results support the use of the RPQ in terms of its two-factor structure, measurement invariance, internal consistency, convergent validity, discriminant validity, and known-groups validity. Significant associations were found with several criterion-related variables such as age of criminal onset, age of first problem with the law, conduct disorder, crime severity, violent crimes, and alcohol and drug use. Findings are discussed in terms of the use of the RPQ among incarcerated girls and community girls.

Keywords

reactive aggression proactive aggression female adolescents assessment validation

Aggression can be characterized as a heterogeneous construct with proactive and reactive aggression as two commonly referenced subtypes of aggression (Dodge, 1991; Dodge & Coie, 1987). While closely related (e.g., Dodge & Coie, 1987), proactive and reactive aggression can be distinguished based on the underlying motivation behind the aggressive act (see Merk, de Castro, Koops, & Matthys, 2005, for a review). Proactive aggression is planned and unprovoked aggression that is motivated by secondary goals such as dominance or instrumental gain, whereas reactive aggression is impulsive or unplanned, and occurs as a response to a perceived or actual threat or insult (Chase, O’Leary, & Heyman, 2001; Dodge, Lochman, Harnish, Bates, & Pettit, 1997). Although recognized as distinct subtypes of aggression, proactive and reactive aggression often coexist and contribute to an individual’s overall level of aggression (Rosell & Siever, 2015).

Despite questions regarding its utility (Bushman & Anderson, 2001), the reactive and proactive distinction remains prominent among scholars, and theories have been developed to explain the distinct causal pathways leading to each form (see Kempes, Matthys, de Vries, & van Engeland, 2005, for a review). Specifically, Crick and Dodge (1996) theorized that disruptions in social-information processing can lead to both reactive and proactive aggression depending on the distinct developmental stage that the disruption occurs. Further solidifying reactive and proactive aggression as separate dimensions of aggression, empirical research has identified unique neurobiological (Bass & Nussbaum, 2010; Levi, Nussbaum, & Rich, 2010; Urheim et al., 2014) and environmental (Brendgen, Vitaro, Boivin, Dionne, & Pérusse, 2006) etiological factors underpinning each of them. For instance, proactive and reactive aggression differ with regard to their underlying emotional processing. That is, proactive aggression has demonstrated associations with affective deficits such as reduced skin conductance and heart rate (Hubbard et al., 2002), while reactive aggression has been linked to poor emotion regulation and low frustration tolerance (Vitaro, Brendgen, & Tremblay, 2002). Likewise, a number of studies have identified external correlates that are unique to these two subtypes of aggression (e.g., Ang, Huan, & Florell, 2014). For example, unlike proactive aggression, reactive aggression tends to show positive associations with emotional dysregulation and internalizing problems such as anxiety and depression (Fite, Stoppelbein, & Greening, 2009). Conversely, proactive aggression tends to be associated with externalizing or antisocial behavior and traits associated with psychopathy (Card & Little, 2006; Cima & Raine, 2009; Fite et al., 2009).

Stability estimates of these two subtypes of aggression also support separate developmental trajectories. For instance, longitudinal research using a twin sample supports separate developmental trajectories for these two subtypes of aggression that differ with regard to stability. Specifically, Tuvblad, Raine, Zheng, and Baker (2009) found that reactive aggression decreased with age, whereas proactive aggression remained relatively stable from childhood into adolescence for both sexes (Tuvblad et al., 2009). However, another study using a large, nontwin sample of youth found that proactive aggression actually increased for boys across adolescence but seemed to remain relatively stable for girls (Fung, Raine, & Gao, 2009). In examining genetic and environmental influences of aggression, the stability in reactive aggression is thought to be the result of both genetic and nonshared environments, whereas stability in proactive aggression is primarily attributed to genetic influences (Tuvblad et al., 2009). Other research in twin samples has found that genetic influences are involved in proactive and reactive aggression for boys, but shared and nonshared environmental influences account for both subtypes of aggression in girls (Baker, Raine, Liu, & Jacobson, 2008). Taken together, both empirical and theoretical evidence support the distinction between these two subtypes of aggression. In addition, this distinction could have important implications for identification and treatment of aggression and antisocial outcomes (Barker et al., 2010) that could differ depending on gender. Therefore, it is imperative that reliable and valid measures are constructed and tested that capture this heterogeneity.

Although various methods for assessing aggression exist, the Reactive-Proactive Aggression Questionnaire (RPQ) is a well-validated and widely used measure that captures both subtypes (Raine et al., 2006). For instance, internal reliability of the RPQ has been substantiated across various samples (Borroni, Somma, Andershed, Maffei, & Fossati, 2014; Cima & Raine, 2009; Raine et al., 2006; Seals, Sharp, Ha, & Michonski, 2012), and results from factor analyses provide consistent support for the two-factor structure of the RPQ (e.g., Baker et al., 2008; Raine et al., 2006). Likewise, the two dimensions of the RPQ have shown distinct associations with external criteria. For example, unlike the Reactive scale, the Proactive scale demonstrates associations with externalizing behaviors such as serious delinquency and initiating fights (Raine et al., 2006), as well as substance use and engaging in risky sex (Miller & Lynam, 2006). The Proactive scale of the RPQ is also positively related to measures of psychopathy, while the Reactive subscale tends to be unrelated to such measures (Cima & Raine, 2009; Raine et al., 2006; Seals et al., 2012; van Baardewijk, Vermeiren, Stegge, & Doreleijers, 2011). In addition, while both the reactive and proactive aggression scales of the RPQ are positively related to narcissism, the Proactive scale shows stronger, positive associations compared with the Reactive scale (Ang & Raine, 2009; Seah & Ang, 2008).

In sum, the RPQ demonstrates good construct validity and external validity. However, research is limited, in that the RPQ has primarily been validated among community samples of adolescent boys or mixed gender samples. Thus, few studies have examined the construct validity of the RPQ among adolescent girls, especially detained adolescent girls. Using a different measure of aggression, research by Marsee and Frick (2007) found that poor emotion regulation was uniquely associated with reactive aggression, while callous-unemotional (CU) traits and positive outcome expectations were associated with proactive aggression among a sample of detained female adolescents. Similar associations have also been detected in samples of children (e.g., Dodge et al., 1997; Vitaro, Barker, Boivin, Brendgen, & Tremblay, 2006), and together this research is consistent with Dodge’s social-information processing model (Crick & Dodge, 1996). More recently, among a sample of detained female adolescents, it was found that levels of proactive aggression were higher among those with both conduct disorder and elevated levels of CU traits compared with those with conduct disorder alone (van Damme, Colins, & Vanderplasschen, 2016).

Overall, several constructs (e.g., psychopathy, narcissism, impulsivity, substance use, and internalizing symptoms) demonstrate divergent associations with regard to the aggression subtypes, and may shed further light on the distinction between reactive and proactive aggression in adolescent girls. However, few studies have examined unique associations between proactive and reactive aggression as captured by the RPQ, and important constructs such as anxiety and empathy. It may be particularly important to examine how these associations manifest among girls, given differences in onset of aggression in girls relative to boys (e.g., Moffitt, Caspi, Rutter, & Silva, 2001) and the increasing rates of aggression and violence among adolescent girls over the last two decades (Moretti, Catchpole, & Odgers, 2005).

Sex Differences in Aggression

Extant research shows that starting at the age of 4, boys are overall more aggressive than girls, both physically and verbally (see Archer [2004] for a review). These gender differences remain stable during childhood and adolescence with girls being 2 to 4 times less likely to develop aggressive behavior problems or receive a diagnosis of conduct disorder (Prior, Smart, Sanson, & Oberklaid, 1993). In contrast, evidence for sex differences in reactive and proactive aggression is equivocal. Results from some studies have concluded that sex differences do exist (e.g., Fossati et al., 2009), while others report no differences between boys and girls in proactive or reactive aggression (e.g., Raine, Fung, Portnoy, Choy, & Spring, 2014). For instance, in a community sample of children and adolescents aged 6 to 12, Kempes, Matthys, Maassen, van Goozen, and van Engeland (2006) found no differences in parent report of proactive aggression, but parents rated boys significantly higher on reactive aggression than girls (Kempes et al., 2006). Alternatively, among a community sample of Greek Cypriot adolescents aged 12 to 18, boys demonstrated higher rates of proactive aggression, but no differences were found between boys and girls in reactive aggression (Fanti, Frick, & Georgiou, 2009). Another study, utilizing a large community sample of Turkish children aged 9 to 14, found that boys scored higher on both proactive and reactive aggression scales compared with girls (Baş & Yurdabakan, 2012). Finally, among a sample of clinic-referred children and adolescents aged 5 to 18, no gender differences in parent reports of reactive or proactive aggression were found (Connor, Steingard, Anderson, & Melloni, 2003).

Despite the equivocal nature of relative sex differences in reactive and proactive aggression, no sex differences have been found in the factor structure of the RPQ across various age ranges (e.g., Baker et al., 2008; Cima, Raine, Meesters, & Popma, 2013) and cultures (Fossati et al., 2009; Fung et al., 2009), suggesting that this instrument is appropriate for assessing aggression in both boys and girls. However, while past research has included small subsamples of detained female adolescents (e.g., Cima et al., 2013), no study to date has explicitly examined the psychometric properties of the RPQ in a sample of detained female adolescents, nor have the psychometric properties of the RPQ been compared across different settings (i.e., forensic vs. school).

Given the gender differences in aggression observed in prior research, it is important that research examines the psychometric properties of the RPQ among certain groups (i.e., females) and across different settings (i.e., forensic and school). Such research can help determine the appropriateness of using the RPQ to assess aggression among certain populations as well as whether findings from research utilizing the RPQ based on samples of other groups generalize to girls, particularly detained girls. Despite the importance of such research evaluating the RPQ, few studies exist that have evaluated the RPQ among this particular group (i.e., detained adolescent girls).

Current Study

The current study attempts to address some of the gaps in the literature regarding the RPQ and its subscales in several ways. First, the main goal of the current study is to validate the RPQ among a detained sample of female Portuguese youth. This sample provides a unique context for validating the RPQ and examining distinctions between reactive and proactive aggression more generally. Cultural differences in the socialization of youth and differences in developmentally appropriate behavioral norms, particularly gender differences, may lead to measurement differences in aggression between boys and girls (e.g., Bergeron & Schneider, 2005). Thus, the psychometric properties including internal consistency, factor structure, and convergent and criterion-related validity of the RPQ among a unique sample of detained Portuguese girls will be examined. Specifically, the current study examines distinct associations between the Reactive and Proactive scales with theoretically relevant criteria (e.g., psychopathic personality traits, narcissism, conduct disorder, age at first arrest, crime severity, physical violence, and substance use). In addition, this study attempts to fill a gap in the literature by examining the distinct associations between the Reactive and Proactive scales of the RPQ with empathy and anxiety.

It was hypothesized that (a) the original two-factor structure would best fit the RPQ using confirmatory factor analytic methods, and would show measurement invariance and good internal consistency; (b) the RPQ would exhibit expected associations with theoretically relevant outcomes used to assess aspects of convergent validity (e.g., psychopathic traits, impulsivity) and discriminant validity (i.e., empathy, anxiety); (c) the RPQ scores would be positively associated with relevant variables used to assess aspects of criterion validity, such as conduct disorder, age of crime onset, age of first problem with the law, increased crime severity, use of physical violence, alcohol abuse, and drug use; and (d) the RPQ would show known-groups validity, with the forensic sample obtaining higher scores than the school sample.

Method

Participants

The sample was comprised of 377 girls (N = 377; M_age = 16.23 years; SD = 1.38 years; range = 14-19 years) recruited in forensic and school contexts. Of this total, 103 participants (M_age = 16.41 years; SD = 1.19 years; range = 14-18 years) formed the forensic sample, and 274 participants (M_age = 16.17 years; SD = 1.44 years; range = 14-19 years) formed the school sample. The school sample was recruited from public schools of the Lisbon, Algarve, and Coimbra regions. The forensic sample was recruited from three juvenile detention centers managed by the Portuguese Ministry of Justice that admit female detainees. They were all detained by the court’s decision, the harshest sentence the court can give to teenagers. Among the detained girls, the mean age of crime onset was 12.50 years (SD = 1.56), and mean age of first criminal problems with the law was 13.27 years (SD = 1.55). Most were detained before they were 16 years old (M = 15.90, SD = 1.04) as a result of committing serious and violent crimes (e.g., robbery, assault).

The participants were mainly White Europeans (forensic sample = 59.2%; school sample = 90.1%) from an urban background (forensic sample = 97.1%; school sample = 100%) with a low socioeconomic status (SES; forensic sample = 60.2%; school sample = 39.1%). No significant differences were found between the two samples in terms of age (F = 2.265, p = .133). Significant differences were found regarding years of education (F = 275.577, p ≤ .001) with the school sample having more years of education (M = 9.17 years; SD = 1.26) than the forensic sample (M = 6.66 years; SD = 1.42). Significant differences were also found regarding ethnicity (χ² = 47,872, p ≤ .001), urban versus rural background (χ² = 8.045; p ≤ .05), and SES (U = 10,407.00; p ≤ .001).

Measures

The RPQ (Raine et al., 2006) is a self-report measure that was designed to distinguish between reactive and proactive aggression. The RPQ consists of 23 items rated on a 3-point ordinal scale (never = 0, sometimes = 1, often = 2). A total of 11 items assess reactive aggression (e.g., “Reacted angrily when provoked by others”) and 12 items assess proactive aggression (e.g., “Hurt others to win a game”). Summed scores provide a measure of reactive or proactive aggression, as well as total aggression. Higher scores indicate higher levels of aggression. The RPQ has been validated in samples of children (Baker et al., 2008), adolescents (Raine et al., 2006), and young adults (Cima et al., 2013). Internal consistency for adolescents has previously been reported as α = .86 for proactive aggression, α = .84 for reactive aggression, and α = .90 for total aggression (Raine et al., 2006). The Portuguese version of the RPQ was used (Pechorro, Ray, Raine, Marôco, & Gonçalves, 2015). Internal consistency reliability statistics for the RPQ in the current sample are presented below.

The Antisocial Process Screening Device–Self-Report (APSD-SR; Frick & Hare, 2001; Muñoz & Frick, 2007) is a multidimensional 20-item measure designed to assess psychopathic traits in adolescents. It was modeled after the Psychopathy Checklist–Revised (PCL-R; Hare, 2003). Each item is scored on a 3-point ordinal scale (never = 0, sometimes = 1, often = 2). The total score, as well as each dimension score, is obtained by adding the respective items. Some studies (e.g., Frick, O’Brien, Wootton, & McBurnett, 1994) reported two main factors (callous-unemotional and impulsivity/conduct problems), while others (e.g., Frick, Barry, & Bodin, 2000) reported three main factors: callous-unemotional, narcissism, and impulsivity. Higher scores are indicative of an increased presence of psychopathic traits. The Portuguese validation of the APSD-SR was used (Pechorro, Hidalgo, Nunes, & Jiménez, 2016; Pechorro, Marôco, Poiares, & Vieira, 2013). The internal consistency for the current study was α = .77.

The Youth Psychopathic Traits Inventory (YPI; H. Andershed, Kerr, Stattin, & Levander, 2002) is a 50-item self-report measure designed to assess the core personality traits of the psychopathic personality constellation in youth aged 12 years and older. Each item is scored on an ordinal 4-point Likert-type scale (ranging from does not apply at all = 1 to applies very well = 4). The YPI consists of 10 subscales (with five items each) designed in line with Cooke and Michie’s (2001) three-dimensional conceptualization of the psychopathy construct, namely, the Grandiose-Manipulative dimension, the Callous-Unemotional dimension, and the Impulsive-Irresponsible dimension. More specifically, the Grandiose-Manipulative dimension consists of the Dishonest Charm, Grandiosity, Lying, and Manipulation subscales; the Callous-Unemotional dimension consists of the Callousness, Unemotionality, and Remorselessness subscales; and the Impulsive-Irresponsible dimension consists of the Impulsivity, Thrill-seeking, and Irresponsibility subscales. Higher scores reflect an increased presence of psychopathic traits. The Portuguese version of the YPI was used (Pechorro, Andershed, Ray, Marôco, & Gonçalves, 2015; Pechorro, Ribeiro da Silva, Andershed, Rijo, & Gonçalves, 2016). The internal consistency for the current study was α = .94.

The Inventory of Callous-Unemotional Traits (ICU; Essau, Sasagawa, & Frick, 2006; Kimonis et al., 2008) is a 24-item self-report scale designed to assess callous and unemotional (CU) traits in youth derived from the Callous-Unemotional subscale of the Antisocial Process Screening Device (APSD; Frick & Hare, 2001). Each item is scored on a 4-point scale (ranging from not at all true = 0 to definitely true = 3). Scores are calculated by reverse scoring the positively worded items, and then summing the items to obtain a total score. Prior research has identified three independent factors: callousness, unemotional, and uncaring (e.g., Kimonis et al., 2008). All items also loaded onto a general callous-unemotional factor. Higher scores are indicative of increased CU traits. The Portuguese validation of the ICU was used (Pechorro, Hawes, Gonçalves, & Ray, 2016; Pechorro, Ray, Barroso, Marôco, & Gonçalves, 2016). The internal consistency for the total score of the ICU in the current study was α = .86.

The Narcissistic Personality Inventory-13 (NPI-13; Gentile et al., 2013) is a short form of the widely used Narcissistic Personality Inventory (NPI; Raskin & Terry, 1988). The NPI-13 consists of 13 pairs of statements where one of each pair is indicative of narcissism and the other is not (coded 1 or 0, respectively). Respondents are asked to select the one that they most agree with. The NPI-13 provides both a total score and three subscale scores: Leadership/Authority, Grandiose Exhibitionism, and Entitlement/Exploitativeness. Higher scores represent an increased presence of narcissistic personality traits. The Portuguese version of the NPI-13, especially adapted for use with adolescents, was used (Pechorro, Gentile, Ray, Nunes, & Gonçalves, 2016). The internal consistency for the current study, estimated by Kuder–Richardson coefficient (i.e., α for dichotomous items), was .82.

The Barratt Impulsiveness Scale version 11 (B1S-11; Patton, Stanford, & Barratt, 1995; Stanford et al., 2009) is a 30-item self-report questionnaire designed to measure impulsiveness. Each item is scored on a 4-point ordinal scale (ranging from rarely/never = 1 to almost always/always = 4). The BIS-11 contains six subscales, which correspond to the six first-order factors: (a) attention, (b) cognitive instability, (c) motor, (d) perseverance, (e) self-control, and (f) cognitive complexity. These six first-order factors converge into three second-order factors, namely, (a) attentional impulsiveness (Attention and Cognitive Instability dimensions), (b) motor impulsiveness (Motor and Perseverance dimensions), and (c) nonplanning impulsiveness (Self-Control and Cognitive Complexity dimensions). The total score, as well as each dimension score, is obtained by adding the respective items. Higher scores on the BIS-11 reflect higher levels of impulsiveness. A Portuguese version of the BIS-11, especially adapted for use with adolescents, was used (Pechorro, Ayala-Nunes, Nunes, Maia, & Gonçalves, 2016; Pechorro, Ayala-Nunes, Ray, Nunes, & Gonçalves, 2016; Pechorro, Marôco, Ray, & Gonçalves, 2015). The internal consistency for the current study was α = .85.

The Basic Empathy Scale (BES; Jolliffe & Farrington, 2006) is a 20-item self-report measure designed to assess empathy in youth. The BES was developed as a concise and coherent scale with the aim of measuring two distinct factors: affective empathy and cognitive empathy. Each item is scored on a 5-point ordinal scale (ranging from strongly disagree = 1 to strongly agree = 5). Scores are calculated by reverse scoring the positively worded items, and then summing the items to obtain the total score and the factor scores. Higher scores indicate an increased presence of the associated characteristics. The Portuguese version of the BES was used (Pechorro, Ray, Salas-Wright, Marôco, & Gonçalves, 2015). The internal consistency for the current study was α = .88.

The Social Anxiety Scale for Adolescents (SAS-A; La Greca & Lopez, 1998) is an 18-item self-report scale designed to assess subjective experience of social anxiety in adolescents. Each item is rated on a 5-point scale (ranging from not at all = 1 to all the time = 5). Three distinct subscales have been identified: the Fear of Negative Evaluation (FNE) subscale reflects fears, concerns, or worries regarding negative evaluations from peers; the Social Avoidance and Distress–New (SAD-New) subscale reflects social avoidance and distress with new social situations or unfamiliar peers; the Social Avoidance and Distress–General (SAD-General) subscale reflects more generalized or pervasive social distress, discomfort, and inhibition. Scores are obtained by summing the ratings for the items comprising each subscale. The Portuguese validation of the SAS-A was used (Pechorro, Ayala-Nunes, Nunes, Marôco, & Gonçalves, 2016). Internal consistency for the present study was α = .93.

The Sellin-Wolfgang Index of Crime Seriousness (ICS; Wolfgang et al., as cited in White et al., 1994) guided the delinquency seriousness classification of the official court reports. Level 0 consists of no delinquency. Level 1 consists of minor delinquency committed at home such as stealing small amounts of money from mother’s purse. Level 2 consists of minor delinquency outside the home, including shoplifting something worth less than 5 euros, vandalism, and minor fraud (e.g., not paying bus fare). Level 3 consists of moderately serious delinquency such as any theft more than 5 euros, gang fighting, carrying weapons, and joyriding. Level 4 consists of serious delinquency such as car theft and breaking and entering. Level 5 consists of having performed at least two of each of the behaviors in the previous level.

A Conduct Disorder (CD) scale was also created based on the 15 items used to assess CD (see, for example, Skilling, Quinsey, & Craig, 2001). The 15 dichotomous items (coded no = 0, yes = 1) were summated to obtain a total continuous score. Thus, higher scores indicate a higher number of positively endorsed indicators of CD. Based on the Kuder–Richardson coefficient, the internal consistency of the CD scale was .89.

In addition, a questionnaire was constructed to describe the sociodemographic and criminal characteristics of the participants. This questionnaire included variables such as participants’ age, ethnic group, geographic classification of residence (rural vs. urban), level of schooling completed, SES, parental marital status, previous use of physical violence (coded no = 0, yes = 1), alcohol abuse, and cannabis use (these last two coded as 5-point ordinal scales from almost never/never = 0 to almost always/always = 4). SES was measured by considering both parental level of education and profession, appropriate to the Portuguese context. Diagnostic and Statistical Manual of Mental Disorders’ (5th ed.; DSM-5; American Psychiatric Association, 2013) Conduct Disorder diagnosis (CD) was assessed only for the forensic sample by the first and last authors of this article, using the official diagnostic criteria (i.e., the standard method described in the DSM-5). The prevalence of CD found in the current forensic sample (85.4%) was higher than those typically found among forensic samples composed of female youth (Sevecke & Kosson, 2010).

Procedure

Authorization to validate the RPQ in Portugal was obtained from the first author of the questionnaire (Raine et al., 2006). Appropriate procedures (e.g., avoiding item bias or differential item functioning) were followed during the translation and retroversion (Hambleton, Merenda, & Spielberger, 2005). The initial translation from English into Portuguese was completed by the first and last authors of this article, who made sure that young people would be able to properly understand the meaning of the items. The questionnaire was then independently translated back into English by a native English speaker with considerable professional experience in translating psychology-related scientific texts. No significant differences were found between the back-translation and the original version, demonstrating that the translated items had the same or very similar meanings as the original English items.

Authorization to assess detained youth was obtained from the General Directorate of Reintegration and Prison Services of the Portuguese Ministry of Justice (DGRSP-MJ). The detainees were informed about the nature of the study and asked to voluntarily participate. The participation rate was approximately 89%. Motives for not participating included refusal to participate (6%), inability to participate due to not understanding the Portuguese language (4%), and inability to participate due to security issues (1%). Authorization to assess youth in the school context was obtained from the General Directorate of Education of the Portuguese Ministry of Education (DGE-ME), and parental permission was obtained for all children. The participants, students from public schools of the Lisbon, Algarve, and Coimbra regions, were informed about the nature of the study and asked to voluntarily participate. The participation rate was approximately 84%. Participants who were unwilling or unable to collaborate were excluded.

All the participants were informed about the nature of the study and gave their informed consent for inclusion before they participated voluntarily. The research was approved by the Ethics Committees of the DGRSP-MJ (Code 122/DSPRE/2013) and DGE-ME (Code 0338400001). Parental permission was obtained for all underage children, and informed consent was obtained from participants who were 18 years of age. The measures were administered using individual face-to-face interviews in an appropriate setting (forensic sample) and in an appropriate classroom group setting (school sample) using a paper-and-pencil method for collecting the data. Some of the information (e.g., sociodemographic variables) was obtained from self-reports, and institutional files were also used to complement the information obtained (e.g., prior criminal activity and detentions).

Data Analysis

The data were analyzed using SPSS v22 (IBM SPSS, 2013) and EQS 6.2 (Bentler & Wu, 2008). The factor structure of the Portuguese language version of the RPQ was assessed using confirmatory factor analysis (CFA) performed in EQS 6.2 (Bentler & Wu, 2008; Byrne, 2006) with the robust estimation methods. Goodness-of-fit indices were calculated, including Satorra–Bentler chi-square/degrees of freedom, comparative fit index (CFI), incremental fit index (IFI), and root mean square error of approximation (RMSEA). A chi-square/degrees of freedom value <5 is considered acceptable, a value ≤2 is considered good, and = 1 very good (Marôco, 2014; West, Taylor, & Wu, 2012). A CFI ≥ .90 and RMSEA ≤ .08 indicate adequate fit, whereas a CFI ≥ .95 and RMSEA ≤ .06 indicate good model fit (Byrne, 2006). The incremental fit index, also known as Bollen’s IFI, is relatively insensitive to sample size; values that exceed .90 are regarded as acceptable.

The CFA was performed on the original scale items, and only items with standardized loading above 0.30 were retained because factor loadings are generally considered to be meaningful when they exceed that value (e.g., Crocker & Algina, 2008). No modification indexes were used to improve the measurement model. Polychoric correlations with robust methodologies were used to perform the CFA on the ordinal items because they provide more accurate estimates for the correlations between ordinal items than Pearson’s correlations (Byrne, 2006). Weak (metric) and strong (scalar) invariance were evaluated, and the S-Βχ² difference test determined if the constraints significantly deteriorated the model (Millsap & Olivera-Aguilar, 2012). The SBDIFF program was used to perform this difference test (http://homepages.abdn.ac.uk/j.crawford/pages/dept/sbdiff.htm).

ANOVAs were used to compare scale variables, Mann–Whitney’s U test was used to compare ordinal variables, and the chi-square test was used to compare nominal variables. Pearson’s correlations were used to analyze associations between scale variables, and Spearman’s correlations were used to analyze associations between ordinal variables and scale variables (Leech, Barrett, & Morgan, 2015). Correlations were considered low if below .20, moderate if between .20 and .50, and high if above .50. Cronbach’s alphas (considered satisfactory if above .70), mean interitem correlations (MIICs; considered good if within the .15-.50 range), and corrected item-total correlation ranges (CITCRs; considered adequate if above .20) were used to assess reliability (Clark & Watson, 1995; Nunnally, & Bernstein, 1994).

Results

The first step in examining the psychometric properties of the Portuguese version of the RPQ among female youth was to attempt to replicate, using CFA, the different factor structures proposed for this instrument (Raine et al., 2006). The following goodness-of-fit indexes were obtained for the different CFA models using the total sample: one-factor model (S-Βχ²/ df = 2.81; IFI = .94; CFI = .94; RMSEA = .07 [.06-.08]; Akaike information criterion [AIC] = 186.99); two-factor first-order intercorrelated model (S-Bχ² / df = 2.48; IFI = .96; CFI = .96; RMSEA = .06 [.06-.07]; AIC = 110.17); and two-factor second-order model (S-Bχ² / df = 4.55; IFI = .75; CFI = .75; RMSEA = .10 [.09-.10]; AIC = 580.77). The strongest support in terms of goodness-of-fit indexes was found for the two-factor first-order intercorrelated model, which also presented the lowest AIC. Figure 1 shows the item loadings for the two-factor intercorrelated structure estimated with the Maximum Likelihood (ML) robust method using the total sample. All items had loadings above a minimum recommended value of 0.30 (Crocker & Algina, 2008), and thus, none were removed from the model. It is worth noting that Item 7, “Had temper tantrums,” had the lowest loading of 0.35.

Figure 1:

RPQ Two-Factor Maximum Likelihood Robust Structure With Standardized Item Loadings

The next step was testing for measurement invariance among the school sample and the forensic sample using the two-factor first-order intercorrelated model, for which we were able to find strong support in terms of goodness-of-fit indexes (see Table 1). The standardized item loadings were all above 0.30. The ΔS-Bχ² (df) values were statistically nonsignificant in the comparison of the nested models regarding weak invariance (factor loadings constrained) and strong invariance (factor loadings and factor covariances constrained). The ΔCFI between the models was below the .01 cutoff. This suggests that the constraints specified do hold and leads us to assume that the models do share equivalence across the two groups (Byrne, 2006).

Table 1:

Test for Invariance of the RPQ Goodness-of-Fit Statistics

Model	S-Βχ² (df)	ΔS-Βχ (df)	CFI	RMSEA [90% CI]
1. No constraints (configural model)	965.23 (458)	—	.93	.08 [0.07, 0.08]
2. Factor loadings constrained	985.29 (479)	8.83 (21)^ns	.93	.08 [0.07, 0.08]
3. Factor loadings and factor covariances constrained	976.19 (480)	28.09 (22)^ns	.93	.07 [0.07, 0.08]

Note. RPQ = Reactive-Proactive Aggression Questionnaire; S-Βχ² (df) = Satorra–Bentler chi-square (degrees of freedom); CFI = robust comparative fit index; RMSEA = robust root mean square error of approximation; CI = confidence interval; ns = nonsignificant at the .05 level.

Table 2 displays the intercorrelations among the RPQ total and its dimensions, Cronbach’s alpha, MIIC, and CITCR. With regard to the intercorrelations between the Proactive and Reactive scales, the findings suggest that they are strongly correlated. In addition, the RPQ and its subscales evidenced good internal reliability across all three measures of internal consistency.

Table 2:

Pearson’s Correlations Matrix, Cronbach’s Alpha, Mean Interitem Correlation, and Corrected Item-Total Correlation Range

Samples/RPQ	RPQ total	RPQ Rea	RPQ Pro	Alpha	MIIC	CITCR
Forensic
RPQ total	1			.87	.23	.29-.75
RPQ Rea	.87***	1		.80	.27	.25-.66
RPQ Pro	.91***	.59***	1	.81	.26	.29-.69
School
RPQ total	1			.88	.27	.21-.70
RPQ Rea	.91***	1		.80	.27	.25-.61
RPQ Pro	.93***	.70***	1	.85	.34	.43-.74
Total
RPQ total	1			90	.28	.25-.69
RPQ Rea	.91***	1		.81	.28	.30-.59
RPQ Pro	.92***	.68***	1	.87	.35	.47-.72

Note. RPQ = Reactive-Proactive Aggression Questionnaire; RPQ Rea = Reactive dimension; RPQ Pro = Proactive dimension; alpha = Cronbach’s alpha; MIIC = mean interitem correlation; CITCR = corrected item-total correlation range.

***

p < .001.

The convergent validity of the RPQ total and its dimensions with APSD-SR, YPI, ICU, and NPI is presented in Table 3. As shown in the table, the correlations were in the expected direction and were moderate to high in magnitude. Interestingly, once the RPQ Proactive scale is taken into account many of the correlations for the RPQ Reactive scale become nonsignificant. This only occurred among the forensic sample and was not the case in the school sample, however. Specifically, the correlations between the RPQ Reactive scale and the APSD-SR, YPI, ICU, and NPI-13 were no longer significant once the RPQ Proactive scale was taken into account. Alternatively, the positive correlations remained significant for the RPQ Proactive scale even when accounting for the RPQ Reactive scale. Table 4 also shows the correlations between the RPQ and the BES and the SAS-A. However, across all three samples none of the correlations were significant at the bivariate and partial levels.

Table 3:

Convergent Validity With the APSD-SR, YPI, ICU, and NPI, and Discriminant Validity With the BES and SAS-A

Samples/Measures	RPQ total	RPQ Reactive	RPQ Proactive
Forensic sample
APSD-SR	.48***	.39*** (.15^ns)	.47* (.32)
YPI	.50***	.36*** (.08^ns)	.52* (.41*)
ICU	.30**	.24* (.08^ns)	.29** (.22*)
BIS-11	.47***	.41*** (.22*)	.43*** (.25*)
NPI-13	.54***	.41*** (.13^ns)	.54* (.41*)
BES	.21^ns	.21^ns (.18^ns)	.14^ns (.04^ns)
SAS-A	.10^ns	−.06^ns (−.12^ns)	.20^ns (.13^ns)
School sample
APSD-SR	.59***	.53* (.36*)	.51* (.31*)
YPI	.45***	.39* (.22*)	.42* (.28*)
ICU	.42***	.39* (.27*)	.33*** (.16*)
BIS-11	.44***	.39* (.24*)	.38* (.22*)
NPI-13	.38***	.35* (.23*)	.31*** (.15*)
BES	−.09^ns	−.10^ns (−.09^ns)	−.05^ns (.01^ns)
SAS-A	.01^ns	−.01^ns (−.03^ns)	.02^ns (.03^ns)
Total sample
APSD-SR	.67***	.60* (.32*)	.62* (.36*)
YPI	.62***	.52* (.19*)	.62* (.42*)
ICU	.47***	.44* (.24*)	.42* (.19*)
BIS-11	.54***	.50* (.25*)	.50* (.26*)
NPI-13	.53***	.46* (.19*)	.51* (.30*)
BES	.05^ns	.02^ns (−.04^ns)	.07^ns (.07^ns)
SAS-A	−.04^ns	−.07^ns (−.09^ns)	−.01^ns (.06^ns)

Note. APSD-SR = Antisocial Process Screening Device–Self-Report; YPI = Youth Psychopathic traits Inventory; ICU = Inventory of Callous-Unemotional Traits; NPI-13 = Narcissistic Personality Inventory 13 items short version; BES = Basic Empathy Scale; SAS-A = Social Anxiety Scale for Adolescents; RPQ = Reactive-Proactive Aggression Questionnaire; ns = nonsignificant; BIS-11 = Barratt Impulsiveness Scale-11. Partial correlations controlling for the remaining dimension of the RPQ are given in parentheses.

p < .05. **p < .01. ***p < .001.

Table 4:

Correlations and Partial Correlations With Criterion-Related Variables

Samples/Variables	RPQ total	RPQ Reactive	RPQ Proactive
Forensic sample
CD symptoms	.52***	.27** (−.16^ns)	.62* (.60*)
CD diagnosis	.29**	.17^ns (−.03^ns)	.33 (.29)
ACO	−.46***	−.34*** (−.09^ns)	−.47* (−.35*)
AFPL	−.55***	−.43*** (−.15^ns)	−.54* (−.40*)
ICS	.32**	.20* (−.01^ns)	.35* (.29)
PPV	.33**	.27** (.10^ns)	.32** (.21*)
Alcohol use	.27**	.12^ns (−.11^ns)	.33** (.23*)
Cannabis use	.17^ns	.02^ns (−.07^ns)	.28** (.19*)
School sample
CD symptoms	.50***	.39* (.18)	.50* (.37*)
ACO	.01^ns	−.07^ns (−.12^ns)	.06^ns (.11^ns)
AFPL	.21^ns	−.08^ns (−.01^ns)	.06^ns (.02^ns)
ICS	.25***	.18** (.03^ns)	.28* (.23*)
PPV	.10^ns	.05^ns (−.04^ns)	.15* (.14*)
Alcohol use	.41***	.40* (.30*)	.30*** (.15*)
Cannabis use	.29***	.32* (.27*)	.17** (.07^ns)
Total sample
CD symptoms	.75***	.56*** (.03^ns)	.80* (.70*)
ACO	−.47***	−.36*** (−.07^ns)	−.47* (−.33*)
AFPL	−.58***	−.44*** (−.07^ns)	−.58* (−.43*)
ICS	.69***	.53*** (.09^ns)	.72* (.58*)
PPV	.64***	.49*** (.05^ns)	.69* (.55*)
Alcohol use	.53***	.47*** (−.08^ns)	.50* (.29*)
Cannabis use	.57***	.49*** (−.09^ns)	.57* (.31*)

Note. RPQ = Reactive-Proactive Aggression Questionnaire; CD = DSM-5 Conduct Disorder diagnosis; CD symptoms = CD symptoms scored as a scale; ns = nonsignificant; ACO = age of crime onset; AFPL = age of first problem with the law; ICS = Index of Crime Seriousness; PPV = previous physical violence. Partial correlations controlling for the remaining dimension of the RPQ are given in parentheses.

p < .05. **p < .01. ***p < .001.

The correlations of the RPQ and its dimensions with CD, age of crime onset, age of first problem with the law, crime seriousness, and use of physical violence are presented in Table 4. In general, the correlations for the RPQ total were in the expected directions; however, there were some discrepancies in the correlations across the two samples. For instance, the RPQ total was unrelated to cannabis use among the forensic sample, yet was positively related among the school sample. In addition, the RPQ was unrelated to age of crime onset, age of first problem with the law, and previous physical violence among the school sample, yet these correlations were significant for the forensic sample. It is also interesting that similar patterns in the correlations emerged for the RPQ Reactive and Proactive scales across the different samples. In addition, it appears that the RPQ Proactive scale has a much stronger association with these criteria given that the partial correlations remain significant for this scale, yet many of the partial correlations are nonsignificant for the RPQ Reactive scale, and this appears to be specific to the forensic sample.

Criterion validity using ANOVAs revealed that statistical significant differences were found when comparing the forensic sample and the school sample (see Table 5). Specifically, the forensic sample scored significantly higher on both proactive and reactive aggression compared with the school sample.

Table 5:

Descriptive Statistics and ANOVAs for the RPQ

Comparisons	Forensic M (SD)	School M (SD)	F (p)	Effect size (power)
RPQ total	18.27 (6.30)	8.98 (4.32)	265.27; ≤.001	η_p² = .41 (1.0)
RPQ Reactive	11.14 (3.20)	7.35 (2.93)	119.10; ≤.001	η_p² = .24 (1.0)
RPQ Proactive	7.14 (3.85)	1.64 (1.98)	329.56; ≤.001	η_p² = .47 (1.0)

Note. RPQ = Reactive-Proactive Aggression Questionnaire.

Discussion

The present study examined the psychometric properties of the RPQ among a sample of forensic and community adolescent girls. Results of the factor analysis confirmed cross-cultural generalizability and factor structure of the RPQ for these Portuguese girls. CFA suggested that the two-factor first-order intercorrelated model was the best-fitting model, which is consistent with prior research (e.g., Baker et al., 2008; Cima et al., 2013; Raine et al., 2006). Furthermore, strong measurement invariance was found among the forensic sample and the school sample, consistent with a recent study among Singaporean adolescents which provided empirical support for the invariance of the two-factor model of the RPQ across gender (Ang, Huan, Li, & Chang, 2016). In addition, because testing for measurement invariance is still not frequent, this study as a whole adds to the psychometric literature on the RPQ.

Within both samples of forensic and community adolescents, internal consistency was good for the two subscales and total RPQ score (Kaplan & Saccuzzo, 2013). Adequate homogeneity was present between the items with MIICs falling within the recommended value range of .15 to .50 (Clark, & Watson, 1995). In addition, while the CFA provided support for reactive and proactive subscales, the intercorrelations between the subscales were fairly strong for both the forensic and school samples (r = .59 and .70, respectively), suggesting that they commonly occur together within the same individual. As expected, the forensic sample scored significantly higher on both proactive and reactive aggression than the school sample. In addition, the effect sizes for these differences were much stronger for proactive aggression (η_p² = .47) than for reactive aggression (η_p² = .24). These findings suggest that the RPQ may be more sensitive to proactive aggression among more severe samples of adolescent girls. Of course, one possible explanation is that while both subtypes of aggression are more prevalent among forensic populations of girls, there is more variation in proactive aggression among this group. Given the higher levels of proactive aggression among adolescents with psychopathic traits, namely, CU traits, compared with antisocial youth without such traits (e.g., Frick, Ray, Thornton, & Kahn, 2014), the RPQ may be useful for identifying a particularly severe subgroup of antisocial youth.

Evidence of adequate convergent and discriminant validity (American Educational Research Association, American Psychological Association, & National Council for Measurement in Education, 2014) was present with the RPQ demonstrating the expected associations with external correlates in both the forensic and school samples. Specifically, the RPQ subscales of proactive and reactive aggression were significantly and positively associated with self-reported measures of psychopathy, CU traits, impulsivity, and narcissism in both the forensic and school samples, although many of the correlations with reactive aggression became nonsignificant after controlling for proactive aggression. No significant associations emerged between the RPQ subscales and self-reported measures of empathy (Vachon, Lynam, & Johnson, 2014) or anxiety (Pechorro, Ayala-Nunes, Nunes, Marôco, & Gonçalves, 2016) in either sample evidencing discriminant validity. Differential correlations emerged between the forensic and school sample on measures of criminal behavior and substance use. In particular, both proactive and reactive aggression were associated with number of CD symptoms, age of first criminal act, age of first encounter with the law, past physical violence, and more serious delinquency within the forensic sample of female adolescents. However, only proactive aggression was associated with CD and substance use (alcohol, cannabis use) among the forensic sample. In addition, it appears that much of the relationship between aggression and these criterion variables is due to the proactive component based on the partial correlations controlling for the other scale. Conversely, within the school sample, both proactive and reactive aggression were associated with CD, delinquency, and substance use. Thus, in forensic samples of female adolescents, proactive aggression correlates more strongly with other important indicators of antisocial behavior than reactive aggression. These findings lend support to notion that aggression is multifaceted consisting of a reactive and proactive component, given the unique associations with certain criteria.

The results should be considered within the context of some limitations. First, the current study relied almost entirely on self-report, which may have increased the possibility of shared method variance inflating associations between study variables. However, findings from the current study are relatively consistent with theoretical views and past empirical work on the RPQ, including studies that have used multiple reporters. Nevertheless, future research would benefit from utilizing multiple informants, including parent, teacher, and even peer reports. Second, the cross-sectional design and analyses in the current study do not allow us to infer causality in terms of the associations between aggression and other constructs measured. Given prospective research in boys has shown proactive aggression, compared with reactive aggression is associated with more severe long-term outcomes (Fite, Raine, Stouthamer-Loeber, Loeber, & Pardini, 2010), this type of research is especially important to pursue in female populations. In addition, the cross-sectional nature of the current study does not allow for the examination of temporal stability, analysis of test–retest, or associated outcomes for the RPQ. While previous research that included female participants has revealed adequate temporal stability over a 3-year period (e.g., Cima et al., 2013), this study had only a small sample of girls, and thus, future research is warranted to examine these psychometric properties of the RPQ among detained female populations. In addition, the sample on which the current study was based consists only of girls. Directly comparing the utility of the RPQ between girls and boys is a necessary step for further validation of the RPQ.

Understanding how and why girls engage in aggressive behavior has remained somewhat elusive due to a focus on aggression in boys. Yet, we know that individuals who demonstrate high levels of aggression in childhood (both boys and girls) tend to demonstrate stable patterns of aggression (e.g., Bongers, Koot, van der Ende, & Verhulst, 2004; Côté, Vaillancourt, LeBlanc, Nagin, & Tremblay, 2006) and antisocial behavior through adolescence and adulthood (A. K. Andershed & Pepler, 2013; Farrington & Loeber, 2000; Nagin & Tremblay, 1999). Over time, continued engagement in aggressive and antisocial behavior has both individual and societal consequences among female adolescents, including higher mortality rates, less stable employment, psychiatric difficulties, and higher utilization of social service programs (Pajer, 1998). Given the distinct correlates of proactive and reactive aggression, identifying and distinguishing between these subtypes of aggression may provide further insight for treatment or intervention efforts (Merk, et al., 2005).

Footnotes

Acknowledgements

We wish to thank the following Portuguese juvenile detention centers for their collaboration: Bela Vista, Navarro de Paiva, and Santa Clara.

Results regarding alternative models are available upon request from the corresponding author.

This study was conducted at Psychology Research Centre, University of Minho, supported by the Portuguese Foundation for Science and Technology (FCT; Grant SFRH/BPD/86666/2012) with cofinancing of the European Social Fund (POPH/FSE), the Portuguese Ministry of Education and Science, and Fundo Europeu de Desenvolvimento Regional (FEDER) under the PT2020 Partnership Agreement (UID/PSI/01662/2013).

Pedro Pechorro is a postdoctoral researcher at the Center for Research in Psychology at the School of Psychology, University of Minho (Portugal). His research interests include juvenile delinquency, juvenile psychopathy, psychometrics, and developmental criminology.

Rachel E. Kahn is a psychological associate in the research unit at Sand Ridge Secure Treatment Center in Madison, Wisconsin. Her research interests include adult and juvenile psychopathy, empathy development, juvenile delinquency, and sexually violent offending.

James V. Ray is an assistant professor in the Department of Criminal Justice at the University of Central Florida. His research interests include juvenile delinquency and juvenile justice, personality and offending, psychopathy, and developmental criminology.

Adrian Raine is the Richard Perry University Professor of Criminology, Psychiatry, and Psychology at the University of Pennsylvania. His research focuses on the etiology and prevention of antisocial, criminal, and psychopathic behavior in children and adults.

Rui Abrunhosa Gonçalves is an associate professor at the School of Psychology in the University of Minho (Portugal). He has researched and published on the topics of violent and dangerous offenders, namely, psychopaths, sexual abusers, and spouse abusers.

References

American Educational Research Association, American Psychological Association, & National Council for Measurement in Education. (2014). Standards for educational and psychological testing. Washington, DC: American Educational Research Association.

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.

Andershed

A.-K.

Pepler

D. J.

(2013). Life-span continuity in female aggression and violence. In Andershed

A.-K.

(Ed.), Girls at risk: Swedish longitudinal research on adjustment (pp. 105-117). New York, NY: Springer.

Andershed

Kerr

Stattin

Levander

(2002). Psychopathic traits in nonreferred youths: Initial test of a new assessment tool. In Blaauw

Sheridan

(Eds.), Psychopaths: Current international perspectives (pp. 131-158). Haag, The Netherlands: Elsevier.

Ang

R. P.

Huan

V. S.

Florell

(2014). Understanding the relationship between proactive and reactive aggression, and cyberbullying across United States and Singapore adolescent samples. Journal of Interpersonal Violence, 29, 237-254. doi:10.1177/0886260513505149

Ang

R. P.

Huan

V. S.

Chang

W. T.

(2016). Factor structure and invariance of the Reactive and Proactive Aggression Questionnaire in a large sample of young adolescents in Singapore. Child Psychiatry & Human Development, 47, 883-889. doi:10.1007/s10578-015-0619-y

Ang

R. P.

Raine

(2009). Reliability, validity and invariance of the Narcissistic Personality Questionnaire for Children-Revised (NPQC-R). Journal of Psychopathology and Behavioral Assessment, 31, 143-151. doi:10.1007/s10862-008-9112-2

Archer

(2004). Sex differences in aggression in real-world settings: A meta-analytic review. Review of General Psychology, 8, 291-322. doi:10.1037/1089-2680.8.4.291

Baker

L. A.

Raine

Liu

Jacobson

K. C.

(2008). Differential genetic and environmental influences on reactive and proactive aggression in children. Journal of Abnormal Child Psychology, 36, 1265-1278. doi:10.1007/s10802-008-9249-1

10.

Barker

E. D.

Vitaro

Lacourse

Fontaine

N. M.

Carbonneau

Tremblay

R. E.

(2010). Testing the developmental distinctiveness of male proactive and reactive aggression with a nested longitudinal experimental intervention. Aggressive Behavior, 36, 127-140. doi:10.1002/ab.20337

11.

Baş

A. U.

Yurdabakan

(2012). Factor structure of the Reactive-Proactive Aggression Questionnaire in Turkish children and gender, grade-level, and socioeconomic status differences in reactive and proactive aggression. Journal of Psychoeducational Assessment, 30, 284-297. doi:10.1177/0734282911428892

12.

Bass

S. L. S.

Nussbaum

(2010). Decision making and aggression in forensic psychiatric inpatients. Criminal Justice and Behavior, 37, 365-383. doi:10.1177/0093854809360043

13.

Bentler

(2008). EQS for Windows user’s guide. Encino, CA: Multivariate Software.

14.

Bergeron

Schneider

(2005). Explaining cross-national differences in peer-directed aggression: A quantitative synthesis. Aggressive Behavior, 31, 116-137. doi:10.1002/ab.20049

15.

Bongers

I. L.

Koot

H. M.

van der Ende

Verhulst

F. C.

(2004). Developmental trajectories of externalizing behaviors in childhood and adolescence. Child Development, 75, 1523-1537. doi:10.1111/j.1467-8624.2004.00755.x

16.

Borroni

Somma

Andershed

Maffei

Fossati

(2014). Psychopathy dimensions, Big Five traits, and dispositional aggression in adolescence: Issues of gender consistency. Personality and Individual Differences, 66, 199-203. doi:10.1016/j.paid.2014.03.019

17.

Brendgen

Vitaro

Boivin

Dionne

Pérusse

(2006). Examining genetic and environmental effects on reactive versus proactive aggression. Developmental Psychology, 42, 1299-1312. doi:10.1037/0012-1649.42.6.1299

18.

Bushman

B. J.

Anderson

C. A.

(2001). Is it time to pull the plug on hostile versus instrumental aggression dichotomy? Psychological Review, 108, 273-279. doi:10.1037/0033-295X.108.1.273

19.

Byrne

(2006). Structural equation modeling with EQS: Basic concepts, applications, and programming. Mahwah, NJ: Lawrence Erlbaum.

20.

Card

N. A.

Little

T. D.

(2006). Proactive and reactive aggression in childhood and adolescence: A meta-analysis of differential relations with psychosocial adjustment. International Journal of Behavioral Development, 30, 466-480. doi:10.1177/0165025406071904

21.

Chase

K. A.

O’Leary

K. D.

Heyman

R. E.

(2001). Categorizing partner-violent men within the reactive-proactive typology model. Journal of Consulting and Clinical Psychology, 69, 567-572. doi:10.1037/0022-006X.69.3.567

22.

Cima

Raine

(2009). Distinct characteristics of psychopathy relate to different subtypes of aggression. Personality and Individual Differences, 47, 835-840. doi:10.1016/j.paid.2009.06.031

23.

Cima

Raine

Meesters

Popma

(2013). Validation of the Dutch Reactive Proactive Questionnaire (RPQ): Differential correlates of reactive and proactive aggression from childhood to adulthood. Aggressive Behavior, 39, 99-113. doi:10.1002/ab.21458

24.

Clark

Watson

(1995). Constructing validity: Basic issues in objective scale development. Psychological Assessment, 7, 309-319. doi:10.1037/1040-3590.7.3.309

25.

Cooke

Michie

(2001). Refining the construct of psychopathy: Towards a hierarchical model. Psychological Assessment, 13, 171-188. doi:10.1037/1040-3590.13.2.171

26.

Connor

D. F.

Steingard

R. J.

Anderson

J. J.

Melloni

R. H.

Jr. (2003). Gender differences in reactive and proactive aggression. Child Psychiatry and Human Development, 33, 279-294. doi:10.1023/A:1023084112561

27.

Côté

Vaillancourt

LeBlanc

J. C.

Nagin

D. S.

Tremblay

R. E.

(2006). The development of physical aggression from toddlerhood to pre-adolescence: A nation wide longitudinal study of Canadian children. Journal of Abnormal Child Psychology, 34, 68-82. doi:10.1007/s10802-005-9001-z

28.

Crick

N. R.

Dodge

K. A.

(1996). Social information-processing mechanisms in reactive and proactive aggression. Child Development, 67, 993-1002. doi:10.1111/j.1467-8624.1996.tb01778.x

29.

Crocker

Algina

(2008). Introduction to classical and modern test theory. Mason, OH: Cengage Learning.

30.

Dodge

K. A.

(1991). The structure and function of reactive and proactive aggression. In Pepler

D. J.

Rubin

K. H.

(Eds.), The development and treatment of childhood aggression (pp. 201-218). Hillsdale, NJ: Lawrence Erlbaum.

31.

Dodge

K. A.

Coie

J. D.

(1987). Social-information-processing factors in reactive and proactive aggression in children’s peer groups. Journal of Personality and Social Psychology, 53, 1146-1158. doi:10.1037/0022-3514.53.6.1146

32.

Dodge

K. A.

Lochman

J. E.

Harnish

J. D.

Bates

J. E.

Pettit

G. S.

(1997). Reactive and proactive aggression in school children and psychiatrically impaired chronically assaultive youth. Journal of Abnormal Psychology, 106, 37-51. doi:10.1037/0021-843X.106.1.37

33.

Essau

Sasagawa

Frick

(2006). Callous-unemotional traits in community sample of adolescents. Assessment, 13, 454-469. doi:10.1177/1073191106287354

34.

Fanti

K. A.

Frick

P. J.

Georgiou

(2009). Linking callous-unemotional traits to instrumental and non-instrumental forms of aggression. Journal of Psychopathology and Behavioral Assessment, 31, 285-298. doi:10.1007/s10862-008-9111-3

35.

Farrington

D. P.

Loeber

(2000). Epidemiology of juvenile violence. Child and Adolescent Psychiatric Clinics of North America, 9, 733-748.

36.

Fite

P. J.

Raine

Stouthamer-Loeber

Loeber

Pardini

D. A.

(2010). Reactive and proactive aggression in adolescent males. Criminal Justice and Behavior, 37, 141-157. doi:10.1177/0093854809353051

37.

Fite

P. J.

Stoppelbein

Greening

(2009). Proactive and reactive aggression in a child psychiatric inpatient population. Journal of Clinical Child & Adolescent Psychology, 38, 199-205. doi:10.1080/15374410802698461

38.

Fossati

Raine

Borroni

Bizzozero

Volpi

Santalucia

Maffei

(2009). A cross-cultural study of the psychometric properties of the Reactive-Proactive Aggression Questionnaire among Italian nonclinical adolescents. Psychological Assessment, 21, 131-135. doi:10.1037/a0014743

39.

Frick

Barry

Bodin

(2000). Applying the concept of psychopathy to children: Implications for the assessment of antisocial youth. In Gacono

(Ed.), The clinical and forensic assessment of psychopathy: A practitioner’s guide (pp. 1-24). Mahwah, NJ: Lawrence Erlbaum.

40.

Frick

Hare

(2001). The Antisocial Process Screening Device: Technical manual. Toronto, Ontario, Canada: Multi-Health System.

41.

Frick

O’Brien

Wootton

McBurnett

(1994). Psychopathy and conduct problems in children. Journal of Abnormal Psychology, 103, 700-707. doi:10.1037/0021-843X.103.4.700

42.

Frick

P. J.

Ray

J. V.

Thornton

L. C.

Kahn

R. E.

(2014). Can callous-unemotional traits enhance the understanding, diagnosis, and treatment of serious conduct problems in children and adolescents? A comprehensive review. Psychological Bulletin, 140, 1-57. doi:10.1037/a0033076

43.

Fung

A. L.-C.

Raine

Gao

(2009). Cross-cultural generalizability of the reactive-Proactive Aggression Questionnaire (RPQ). Journal of Personality Assessment, 91, 473-479. doi:10.1080/00223890903088420

44.

Gentile

Miller

Hoffman

Reidy

Zeichner

Campbell

(2013). A test of two brief measures of grandiose narcissism: The Narcissistic Personality Inventory-13 and the Narcissistic Personality Inventory-16. Psychological Assessment, 25, 1120-1136. doi:10.1037/a0033192

45.

Hambleton

Merenda

Spielberger

(2005). Adapting educational and psychological tests for cross-cultural assessment. Mahwah, NJ: Lawrence Erlbaum.

46.

Hare

(2003). The Hare psychopathy checklist–revised: Technical manual (2nd ed.). Toronto, Ontario, Canada: Multi-Health Systems.

47.

Hubbard

J. A.

Smithmyer

C. M.

Ramsden

S. R.

Parker

E. H.

Flanagan

K. D.

Dearing

K. F.

. . . Simons

R. F.

(2002). Observational, physiological, and self-report measures of children’s anger: Relations to reactive versus proactive aggression. Child Development, 73, 1101-1118. doi:10.1111/1467-8624.00460

48.

IBM SPSS. (2013). IBM SPSS Statistics Base 22. Chicago, IL: SPSS.

49.

Jolliffe

Farrington

(2006). Development and validation of the Basic Empathy Scale. Journal of Adolescence, 29, 589-611. doi:10.1016/j.adolescence.2005.08.010

50.

Kaplan

Saccuzzo

(2013). Psychological testing: Principles, applications, and issues (8th ed.). Belmont, CA: Wadsworth.

51.

Kempes

Matthys

de Vries

van Engeland

(2005). Reactive and proactive aggression in children: A review of theory, findings and the relevance for child and adolescent psychiatry. European Child & Adolescent Psychiatry, 14, 11-19. doi:10.1007/s00787-005-0432-4

52.

Kempes

Matthys

Maassen

van Goozen

van Engeland

(2006). A parent questionnaire for distinguishing between reactive and proactive aggression in children. European Child & Adolescent Psychiatry, 15, 38-45. doi:10.1007/s00787-006-0502-2

53.

Kimonis

Frick

Skeem

Marsee

Cruise

Munoz

. . . Morris

(2008). Assessing callous-unemotional traits in adolescent offenders: Validation of the Inventory of Callous-Unemotional Traits. International Journal of Law and Psychiatry, 31, 241-252. doi:10.1016/j.ijlp.2008.04.002

54.

La Greca

A. M.

Lopez

(1998). Social anxiety among adolescents: Linkages with peer relations and friendships. Journal of Abnormal Child Psychology, 26, 83-94. doi:10.1023/A:1022684520514

55.

Leech

Barrett

Morgan

(2015). IBM SPSS for intermediate statistics (5th ed.). New York, NY: Routledge.

56.

Levi

M. C.

Nussbaum

D. S.

Rich

J. B.

(2010). Neuropsychological and personality characteristics of predatory, irritable, and nonviolent offenders: Support for a typology of criminal human aggression. Criminal Justice and Behavior, 27, 633-655. doi:10.1177/0093854810362342

57.

Marôco

(2014). Análise de Equações Estruturais: Fundamentos teóricos, software & aplicações [Structural equations analysis: Theoretical foundations, software and applications]. Pero Pinheiro, Portugal: ReportNumber, Ltd.

58.

Marsee

M. A.

Frick

P. J.

(2007). Exploring the cognitive and emotional correlates to proactive and reactive aggression in a sample of detained girls. Journal of Abnormal Child Psychology, 35, 969-981. doi:10.1007/s10802-007-9147-y

59.

Merk

de Castro

B. O.

Koops

Matthys

(2005). The distinction between reactive and proactive aggression: Utility for theory, diagnosis and treatment? European Journal of Developmental Psychology, 2, 197-220. doi:10.1080/17405620444000300

60.

Miller

J. D.

Lynam

D. R.

(2006). Reactive and proactive aggression: Similarities and differences. Personality and Individual Differences, 41, 1469-1480. doi:10.1016/j.paid.2006.06.004

61.

Millsap

Olivera-Aguilar

(2012). Investigating measurement invariance using confirmatory factor analysis. In Hoyle

(Ed.), Handbook of structural equation modeling (pp. 380-392). New York, NY: Guilford Press.

62.

Moffitt

T. E.

Caspi

Rutter

Silva

P. A.

(2001). Sex differences in antisocial behaviour: Conduct disorder, delinquency, and violence in the Dunedin longitudinal study. Cambridge, England: Cambridge University Press.

63.

Moretti

M. M.

Catchpole

R. E.

Odgers

(2005). The dark side of girlhood: Recent trends, risk factors and trajectories to aggression and violence. The Canadian Child and Adolescent Psychiatry Review, 14, 21-25

64.

Muñoz

Frick

(2007). The reliability, stability, and predictive utility of the self-report version of the Antisocial Process Screening Device. Scandinavian Journal of Psychology, 48, 299-312. doi:10.1111/j.1467-9450.2007.00560.x

65.

Nagin

Tremblay

R. E.

(1999). Trajectories of boys’ physical aggression, opposition, and hyperactivity on the path to physically violent and nonviolent juvenile delinquency. Child Development, 70, 1181-1196. doi:10.1111/1467-8624.00086

66.

Nunnally

Bernstein

(1994). Psychometric theory (3rd ed.). New York, NY: McGraw-Hill.

67.

Pajer

K. A.

(1998). What happens to “bad” girls? A review of the adult outcomes of antisocial adolescent girls. American Journal of Psychiatry, 155, 862-870. doi:10.1176/ajp.155.7.862

68.

Patton

Stanford

Barratt

(1995). Factor structure of the Barratt Impulsiveness Scale. Journal of Clinical Psychology, 51, 768–774. doi:10.1002/1097-4679

69.

Pechorro

Andershed

Ray

Marôco

Gonçalves

(2015). Validation of the Youth Psychopathic Traits Inventory and Youth Psychopathic Traits Inventory—Short form among incarcerated juvenile delinquents. Journal of Psychopathology and Behavioral Assessment, 37, 576-586. doi:10.1007/s10862-015-9490-1

70.

Pechorro

Ayala-Nunes

Nunes

Maia

Gonçalves

(2016). The Barratt Impulsiveness Scale—11 among a Portuguese sample of female juvenile delinquents and community youths. International Journal of Forensic Mental Health. Advance online publication. doi:10.1080/14999013.2016.1235627

71.

Pechorro

Ayala-Nunes

Nunes

Marôco

Gonçalves

(2016). The Social Anxiety Scale for Adolescents: Measurement invariance and psychometric properties among a school sample of Portuguese youths. Child Psychiatry & Human Development, 47, 975-984. doi:10.1007/s10578-016-0627-6

72.

Pechorro

Ayala-Nunes

Ray

Nunes

Gonçalves

(2016). Psychometric properties of the Portuguese version of the Barratt Impulsiveness Scale—11 among a school sample of male and female adolescents. Journal of Child and Family Studies, 25, 2753-2764. doi:10.1007/s10826-016-0453-2

73.

Pechorro

Gentile

Ray

Nunes

Gonçalves

(2016). Adaptation of the Narcissistic Personality Inventory among Portuguese juvenile offenders. Psychology, Crime & Law, 22, 495-511. doi:10.1080/1068316X.2016.1168421

74.

Pechorro

Hawes

Gonçalves

Ray

(2016). Psychometric properties of the Inventory of Callous-Unemotional Traits short version (ICU-12) among detained female juvenile offenders and community youths. Advance online publication. Psychology, Crime & Law. Advance online publication. doi:10.1080/1068316X.2016.1239724

75.

Pechorro

Hidalgo

Nunes

Jiménez

(2016). Confirmatory factor analysis of the Antisocial Process Screening Device: Self-Report among incarcerated male juvenile offenders. International Journal of Offender Therapy and Comparative Criminology, 60, 1856-1872. doi:10.1177/0306624X15588903

76.

Pechorro

Marôco

Poiares

Vieira

(2013). Validation of the Portuguese version of the Antisocial Process Screening Device Self-Report with a focus on delinquent behavior and behavior problems. International Journal of Offender Therapy and Comparative Criminology, 57, 112-126. doi:10.1177/0306624X11427174

77.

Pechorro

Marôco

Ray

Gonçalves

(2015). Psychometric properties of the Barratt Impulsiveness Scale version 11 among a Portuguese sample of incarcerated juvenile offenders. Psychology, Crime & Law, 21, 854-870. doi:10.1080/1068316X.2015.1054386

78.

Pechorro

Ray

Barroso

Marôco

Gonçalves

(2016). Validation of the Inventory of Callous-Unemotional Traits among a Portuguese sample of detained juvenile offenders. International Journal of Offender Therapy and Comparative Criminology, 60, 349-365. doi:10.1177/0306624X14551256

79.

Pechorro

Ray

Raine

Marôco

Gonçalves

(2015). The Reactive-Proactive Aggression Questionnaire: Validation among a Portuguese sample of incarcerated juvenile delinquents. Journal of Interpersonal Violence. Advance online publication. doi:10.1177/0886260515590784

80.

Pechorro

Ray

Salas-Wright

Marôco

Gonçalves

(2015). Adaptation of the Basic Empathy Scale among a Portuguese sample of incarcerated juvenile offenders. Psychology, Crime & Law, 21, 699-714. doi:10.1080/1068316X.2015.1028546

81.

Pechorro

Ribeiro

Silva

Andershed

Rijo

Gonçalves

R. A.

(2016). The Youth Psychopathic Traits Inventory: Measurement invariance and psychometric properties among Portuguese youths. International Journal of Environmental Research and Public Health, 13, 852. doi:10.3390/ijerph13090852

82.

Prior

Smart

Sanson

Oberklaid

(1993). Sex differences in psychological adjustment from infancy to 8 years. Journal of the American Academy of Child & Adolescent Psychiatry, 32, 291-305. doi:10.1097/00004583-199303000-00009

83.

Raine

Dodge

Loeber

Gatzke-Kopp

Lynam

Reynolds

. . . Liu

(2006). The Reactive-Proactive Aggression Questionnaire: Differential correlates of reactive and proactive aggression in adolescent boys. Aggressive Behavior, 32, 159-171. doi:10.1002/ab.20115

84.

Raine

Fung

A. L. C.

Portnoy

Choy

Spring

V. L.

(2014). Low heart rate as a risk factor for child and adolescent proactive aggressive and impulsive psychopathic behavior. Aggressive Behavior, 40, 290-299. doi:10.1002/ab.21523

85.

Raskin

Terry

(1988). A principal-components analysis of the Narcissistic Personality Inventory and further evidence of its construct validity. Journal of Personality and Social Psychology, 54, 890-902. doi:10.1037//0022-3514.54.5.890

86.

Rosell

D. R.

Siever

L. J.

(2015). The neurobiology of aggression and violence. CNS Spectrums, 20, 254-279. doi:10.1017/S109285291500019X

87.

Seah

S. L.

Ang

R. P.

(2008). Differential correlates of reactive and proactive aggression in Asian adolescents: Relations to narcissism, anxiety, schizotypal traits, and peer relations. Aggressive Behavior, 34, 553-562. doi:10.1002/ab.20269

88.

Seals

R. W.

Sharp

Michonski

J. D.

(2012). The relationship between the youth psychopathic traits inventory and psychopathology in a U.S. Community sample of male youth. Journal of Personality Assessment, 94, 232-243. doi:10.1080/00223891.2011.650303

89.

Sevecke

Kosson

(2010). Relationships of child and adolescent psychopathy to other forms of psychopathology. In Salekin

Lynam

(Eds.), Handbook of child and adolescent psychopathy (pp. 284-314). New York, NY: Guilford Press.

90.

Skilling

Quinsey

Craig

(2001). Evidence of a taxon underlying serious antisocial behavior in boys. Criminal Justice and Behavior, 28, 450-470. doi:10.1177/009385480102800404

91.

Stanford

Mathias

Dougherty

Lake

Anderson

Patton

(2009). Fifty years of the Barratt Impulsiveness Scale: An update and review. Personality and Individual Differences, 47, 385-395. doi:10.1016/j.paid.2009.04.008

92.

Tuvblad

Raine

Zheng

Baker

L. A.

(2009). Genetic and environmental stability differs in reactive and proactive aggression. Aggressive Behavior, 35, 437-452. doi:10.1002/ab.20319

93.

Urheim

Rypdal

Melkevik

Hoff

H. A.

Mykletun

Palmstierna

(2014). Motivational dimensions of inpatient aggression. Criminal Behaviour and Mental Health, 24, 141-150. doi:10.1002/cbm.1894

94.

Vachon

Lynam

Johnson

(2014). The (non)relation between empathy and aggression: Surprising results from a meta-analysis. Psychological Bulletin, 140, 751-773. doi:10.1037/a0035236

95.

van Baardewijk

Vermeiren

Stegge

Doreleijers

(2011). Self-reported psychopathic traits in children: Their stability and concurrent and prospective association with conduct problems and aggression. Journal of Psychopathology and Behavioral Assessment, 33, 236-245. doi:10.1007/s10862-010-9215-4

96.

van Damme

Colins

O. F.

Vanderplasschen

(2016). The limited prosocial emotions specifier for conduct disorder among detained girls: A multi-informant approach. Criminal Justice and Behavior, 43, 778-792. doi:10.1177/0093854815613104

97.

Vitaro

Barker

Boivin

Brendgen

Tremblay

(2006). Do early difficult temperament and harsh parenting differentially predict reactive and proactive aggression? Journal of Abnormal Child Psychology, 34, 681-691. doi:10.1007/s10802-006-9055-6

98.

Vitaro

Brendgen

Tremblay

R. E.

(2002). Reactively and proactively aggressive children: Antecedent and subsequent characteristics. Journal of Child Psychology and Psychiatry, 43, 495-505. doi:10.1111/1469-7610.00040

99.

West

Taylor

(2012). Model fit and model selection in structural equation modeling. In Hoyle

(Ed.), Handbook of structural equation modeling (pp. 209-231). New York, NY: Guilford Press.

100.

White

Moffitt

Caspi

Jeglum-Bartusch

Needles

Stouthamer-Loeber

(1994). Measuring impulsivity and examining its relation to delinquency. Journal of Abnormal Psychology, 103, 192-205. doi:10.1037/0021-843X.103.2.192