Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jun 9.
Published before final editing as: Psychol Violence. 2025 Jun 9:10.1037/vio0000599. doi: 10.1037/vio0000599

Normative Beliefs Approving of Aggression Moderate the Observational Learning of Weapon Use

Meagan Docherty 1,2, Eric F Dubow 1,2, Paul B Boxer 3, L Rowell Huesmann 2
PMCID: PMC12629633  NIHMSID: NIHMS2080953  PMID: 41267754

Abstract

Objective.

A prominent theory accounting for the development and maintenance of aggressive behavior from childhood to adolescence is the social cognitive model, which holds that aggressive behavior is sustained over time through various context-dependent beliefs, biases, and schemas that emerge through repeated observation of aggressive social interactions. In this study, we provide a novel test of whether this model could also account for late adolescent and young adult weapon use.

Method.

We use integrative data analysis to combine information from two longitudinal studies of youth in urban areas (Study 1: N=426, 4 waves over 13 years, from ages 8 to 27; Study 2: N=200, 4 waves over 4 years, from ages 15 to 18; total N=626, 51% female, 56% Black). Data consists of both youth and parent report.

Results.

We show that normative beliefs supporting aggression promote the observational learning of weapon use in late adolescence and early adulthood. In addition, we show that these normative beliefs that promote the observational learning of weapon use in late adolescence and early adulthood are significantly stronger for people who 8-years earlier scored higher on callousness who held normative beliefs approving of retaliatory aggression at that time.

Conclusions.

The results of this study suggest that interventions in childhood and adolescence that counteract normative beliefs approving of aggression and that reduce callousness will lessen the likelihood of the later imitation of observed neighborhood weapon violence by the intervened youth.

Keywords: normative beliefs, weapons, violence exposure, aggression, social cognition

The rate of gun deaths in the U.S. has increased approximately 36% in the last decade, and every day an average of 117 people are killed by guns, and another 210 are shot and injured (Everytown for Gun Safety, n.d.). This gun violence epidemic disproportionately affects marginalized communities, including people of color and those with lower socioeconomic status (Bottiani et al., 2021; Kravitz-Wirtz et al., 2022). Although this is a pressing public health concern that has drawn increased attention in recent years, the application of relevant psychological theory that has helped to explain individual differences in other forms of aggressive and violent behavior has been very limited in studies of gun violence. The current study remedies this by showing how observational learning theory and social cognitive theory applied to longitudinal data from childhood to adulthood combine to explain individual differences in the propensity for weapon use in young adulthood.

Observational Learning of Violence

Substantial research has established that social experiences can shape enduring social beliefs that guide and direct social behavior, including aggressive behavior, over time. In particular, prior to middle childhood and the critical cognitive leaps that occur around 8 years of age (Sameroff & Haith, 1986), social encounters can impact behavior quite directly. Understanding how this observational learning works largely began with Albert Bandura’s classic “Bobo doll” experiments (Bandura et al., 1961), and this work has been replicated and used to form the basis of social cognitive theory, which holds that social behavior is sustained over time principally through various beliefs, biases, and schemas that emerge through social interactions and eventually sustain as cognitive-control mechanisms (Bandura, 1986).

Through additional studies focusing on aggressive behavior, Huesmann and colleagues built on this theorizing to demonstrate that in middle childhood, exposure to violence shapes the development of aggression-supporting beliefs and schemas that endure into later childhood, adolescence, and young adulthood, in turn increasing the likelihood of engaging in later aggressive behavior (Anderson & Huesmann, 2003; Guerra et al., 2003; Huesmann et al., 2017; Huesmann et al. 2021). Central to this line of reasoning has been the construct of the normative belief: a self-regulatory belief that provides information about the “acceptability or unacceptability” of potential responses and prescribes the “range of allowable and prohibited behaviors” (Huesmann & Guerra, 1997, p. 409). Normative beliefs are learned through social experiences that provide information about behaviors that are commonplace, generally permissible or impermissible, and lead to positive or negative outcomes for the person. In action, normative beliefs are foundational schemas that filter out many inappropriate potential behaviors and drive appropriate behaviors. A child who witnesses aggressive behavior that achieves desirable outcomes for the perpetrator repeatedly in the social environment is likely to develop normative beliefs that identify aggression as common and acceptable (Guerra et al., 2003). Then, an adolescent who holds these normative beliefs supportive of aggression is more likely to evaluate positively the use of aggressive behavior as a response to stress or social conflict or as a means to achieve desired goals (Huesmann et al., 2017). In addition to early exposure to aggressive behavior, there may be other factors – such as high levels of emotional callousness (e.g., lack of empathy for others’ pain or of guilt or remorse) – in childhood that contribute to the development of these normative beliefs. Exploring and identifying these factors can help to further understanding of the observational learning of aggression.

The Role of Normative Beliefs Approving of Aggression

There is now a wealth of research demonstrating the critical mediating role of aggression-supporting normative beliefs that emerge during middle childhood in children exposed to violence, and that then become more stable and less prone to change by early adolescence, therefore linking children’s early exposure to violence with later aggressive and violent behaviors (Anderson & Huesmann, 2003; Dodge et al., 2022; Huesmann, 1998; Huesmann, 2018). For example, Guerra et al. (2003) found that fourth graders exposed to more neighborhood violence behaved more aggressively by the sixth grade, and the effect was mediated by the normative beliefs about aggression they developed in the fourth grade. Similarly, Huesmann et al. (2017) showed that youth exposed to more war violence between eight and fifteen years of age behaved more aggressively a year after the exposure, and that effect was also mediated by their normative beliefs approving of aggression that emerged while they were exposed to war violence. Interestingly, although males typically report higher levels of aggression, weapon use, and normative beliefs supporting aggression compared to females, evidence is mixed on whether the associations among these constructs differ as a function of biological sex or gender (Guerra et al., 2003; Huesmann et al., 2017). Previous research has found that these normative beliefs are malleable at age 6 but stabilized by age 10, indicating that from age 10 onwards, these beliefs may act as a moderator and alter youth’s evaluation of their own and others’ aggressive behaviors (Huesmann, 1998; Huesmann & Guerra, 1997). This is also consistent with the broader cognitive development literature, which has found that middle childhood is an important period of neuromaturation (e.g., through myelination and synaptic pruning) and the consolidation and stabilization of cognitive abilities and beliefs (Davis-Kean et al., 2008; DelGiudice, 2018; Mah & Ford-Jones, 2012).

Other than research on the effects of exposure to violence as a precursor to the development of normative beliefs, there has been less research into other developmental precursors, such as emotional callousness. Youth who show high levels of callousness are likely also to exhibit elevated and persistent levels of aggressive and violent behavior (Docherty et al., 2018). Although little is known regarding the specific impact of callousness on normative beliefs or other aggression-supporting social cognitions, callousness is an important covariate in any investigation of aggressive or violent behavior (Boxer et al., 2009). In addition, there are theoretical reasons to expect callousness to precede the development of normative beliefs supporting aggression; at lower levels of empathy for others’ pain and lower levels of guilt or remorse for behaviors causing others’ pain, the perceived penalties for aggressive behavior may be much lower, and the perceived rewards (e.g., status, power) may be much higher. There is also less likely to be an aversive emotional reaction to aggression – such as fear or anxiety, which often serves as a deterrent to aggressive behavior – among people with elevated callousness. Thus, youth with higher levels of callousness may be more attuned to positive outcomes related to aggression, and therefore may be more likely to develop beliefs approving of aggression and encode aggressive scripts. This proposition is supported by cross-sectional research which has found that adolescents with elevated callousness are more likely to endorse attitudes in favor of violence (López-Romero et al., 2015) and normative beliefs supporting aggression (Lui et al., 2017), although longitudinal research in this area has been lacking. In terms of sex differences, males tend to exhibit higher levels of callousness compared to females, although there is mixed evidence on whether callousness predicts outcomes differently as a function of biological sex or gender (Carvalho et al., 2018; Pihet et al., 2015).

Although normative beliefs supporting aggression are malleable during early childhood, once formed, from adolescence onward, they likely act as moderators of the effects of the current social context on behavior (Huesmann, 1998; Huesmann & Guerra, 1997). Because of this developmental timeline, those who have developed normative beliefs more supportive of aggression and violence by middle childhood should be not only more likely to engage in those behaviors in response to violent circumstances later on, but they also should be more likely to learn new aggressive behaviors consistent with their normative beliefs from observing violence. Our position is that normative beliefs supporting aggression serve both to mediate the effect of earlier childhood exposures to violence on adolescent and young adult aggression and, once formed and crystallized, to moderate (specifically, to promote) the acquisition of new scripts for aggression from the observation of violence in adolescence and young adulthood. Once they are crystallized in a person’s social-cognitive system, normative beliefs act as filters affecting which observed scripts will be encoded and which will be rejected.

The Present Study

In the present study, we investigate whether, as our theory suggests, adolescents’ and young adults’ (age 15 to 25) normative beliefs supporting aggression act as moderators of the observational learning of scripts for weapon use, promoting their acquisition and therefore later weapon use. According to our reasoning, those who already have stronger normative beliefs supporting aggression by this age should be more likely to encode weapon use scripts from later observations of neighborhood weapon violence because encoding such scripts is consistent with their existing normative beliefs. Having encoded such scripts for weapon use, they should be more likely subsequently to behave aggressively with weapons. We also aim to examine which individual and contextual variables from earlier childhood and adolescence predict the adolescents’ and young adults’ normative beliefs that support aggression and thus that promote their observational learning of weapon use.

To accomplish these goals, we analyze longitudinal data drawn from two urban samples (one from the midwestern US, one from the northeastern US) and combine these data in a manner that allows us to do robust longitudinal “integrative data analysis” (IDA; see Curran & Hussong, 2009). Thus, although there are differences between the two studies, we are able to leverage the full combined sample in our analyses for more valid and reliable inferences. Building on prior findings in this area reviewed above and our theorizing about the role of normative beliefs, we expected that adolescent and young adult exposure to neighborhood weapon violence would increase the risk of subsequent weapon use by them, but this effect will be moderated by the person’s existing normative beliefs supporting aggression. In particular, the relation between observation of neighborhood weapon violence and subsequent weapon use should be stronger for those who already hold normative beliefs supporting aggression and supporting aggression with weapons. We also hypothesized that earlier levels of aggressive behavior, emotional callousness, beliefs in support of retaliation, exposure to parents’ aggression, and exposure to neighborhood gun violence would predict subsequent normative beliefs supporting aggression, and thus help to explain observational learning of weapon use.

Summary of Hypotheses

Hypothesis 1:

Normative beliefs supporting aggressive behavior will moderate the link between exposure to neighborhood weapon violence and subsequent use. Specifically, exposure to neighborhood weapon violence in adolescence and young adulthood will predict weapon use three years later among youth who endorse beliefs supporting aggressive behavior (Hypothesis 1a), but this association will not be significant among youth who do not endorse these beliefs (Hypothesis 1b).

Hypothesis 2:

Earlier levels of individual and contextual risk factors will predict normative beliefs supporting aggression eight years later. Specifically, we hypothesize that earlier levels of aggressive behavior (Hypothesis 2a), emotional callousness (Hypothesis 2b), beliefs in support of retaliation (Hypothesis 2c), exposure to parents’ aggression (Hypothesis 2d), and exposure to neighborhood gun violence (Hypothesis 2e) will predict later normative beliefs supporting aggression when tested within the same model.

Method

Participants and Procedures

This project incorporates data from two samples studied longitudinally: a cohort- sequential sample of youth from Flint, MI, who were assessed five times over the course of about 13 years, and a longitudinal sample of youth from Jersey City, NJ, which involved four annual assessments. Because the same research team was involved in both studies, many of the procedures and measures were very similar across the two studies. More information on each sample is provided below. All procedures and measures described below were approved by the Institutional Review Boards of the authors’ universities and meet ethical guidelines for human subjects research.

Flint Sample

Youth in the Flint, MI, sample (N=426) were in the 2nd (n=126), 4th (n=173), or 9th (n=127) grade at the initial assessment during the 2006–2007 school year. The sample was roughly split equally by sex (52.6% male) and mostly Black (75%), reflecting Flint’s ethnic/racial make-up. At the first assessment, most of the families (78%) had household incomes of less than $30,000/year, and 28% of the parents were married and living with a spouse. Youth and their parents were interviewed each year for three years, and again 10 and 13 years after the initial assessment, for five interviews over 13 years.

For the first three assessment waves, parental consent to interview youth was obtained during interviews with parents, and youth provided assent. Youth completed survey interviews in their classrooms, and parents completed interviews via mail, over the phone, or in person. For the fourth and fifth assessments, the participants were 18 or older, and provided informed consent. They were interviewed in person for approximately 60 minutes and their parents were interviewed by phone or in person for approximately 45 minutes. Both youth and parents received financial incentives for participating in each assessment.

Jersey City Sample

In Jersey City, NJ, youth were recruited from five public regular and charter high schools in 2016 when they were in the 10th grade. Participants were interviewed at four annual assessments, and their parents were also interviewed for the first three assessments. The initial sample consisted of 200 participants (57% female; 22% Black/African-American, 23.5% Latino/a, 25.5% Asian or Pacific Islander, 9.5% White, 10.5% other and 9% multiracial or mixed race), reflecting Jersey City’s ethnic distribution. At the first assessment, most families (65%) had household incomes of less than $50,000/year, and 58% of the parents were married and living with a spouse.

The first three assessments consisted of in-person interviews, and the final assessment was completed via phone. Participants and parents both provided written consent/assent and were compensated at each wave for their participation.

Measures

For the analyses in this article, we use Integrative Data Analysis (IDA), which represents an analytic method that permits more robust inference through the simultaneous analysis of multiple data sets. IDA confers several important advantages, including having higher frequencies of low-base-rate behaviors (e.g., serious violence) in the integrated data set than in the individual data sets, having greater statistical power, and having the ability to cover an expanded range of development when longitudinal cohort studies are combined (Curran & Hussong, 2009). In this case, to provide an integrated data set allowing robust tests of our hypotheses, we aligned data from the two data sets at three different time points in the following ways, which are illustrated in Table 1.

Table 1.

Modal Ages of Participants in Each Sample (Flint and Jersey City (JC)) at Each Time in the Integrative Data Analysis

T1 scale scores
Flint = Mean of Wave 1, 2, 3 JC = no T1 scores		 T2 scale scores T3 scale scores
Sample and cohort
(grade at first assessment)		 Flint: W1 Flint: W2 Flint: W3 Flint: W4 JC: W1 Flint: W5 JC: W4
Flint 2nd Grade 8 9 10 18 21
Flint 4th Grade 10 11 12 20 23
Flint 9th Grade 15 16 17 25 28
JC 10th Grade 15 18
Open in a new tab

Data at T1 in the integrated data set were created by averaging data from the first three assessments in the Flint sample, when the modal ages of youth in the three cohorts were approximately 8–10, 10–12, and 15–17, respectively. For example, we created an overall cumulative exposure to gun violence score by averaging items indexing exposure to gun violence at each wave (see more information below), and then averaged these three wave- specific scores to create the total score incorporated into analyses. Data at T2 come from the fourth assessment in the Flint sample (modal ages approximately 18, 20, and 25) and the first assessment in the Jersey City sample (modal age approximately 15); and data at T3 come from the fifth assessment in the Flint sample (modal ages approximately 21, 23, and 28) and the fourth assessment in the Jersey City sample (modal age approximately 18). We elected to align the data this way to keep the amount of time between T2 and T3 consistent (about three years) across both samples.

Demographic Variables and Covariates

Youth reported on their sex assigned at birth (male or female) and race/ethnicity (Black, White, Latino/a, Asian American, Other Race, Multiracial) at the first assessment. From the race/ethnicity variable, we created one binary variable: Black (=2) versus not-Black (=1), given that 55.5% of the youth sample was Black, and non-Black participants were distributed over a large range of ethnicities with small group sizes for any particular racial/ethnic identity. The parents who were interviewed were mostly mothers (77% of the 352 interviewed parents in Flint and 74% of the 167 interviewed parents in Jersey City).

Key Outcome Measure: Weapon Use in Late Adolescence/Early Adulthood (ages 18 to 28) at T3

For the measure of weapon use in late adolescence/early adulthood, we used 7 items taken from several sources (CDC, 2014, Youth Risk Behavior Surveillance Survey; Cook & Ludwig, 2004, from the National Survey of Adolescent Males; Huesmann et al., 1984; NIJ, 2007) that assess participants’ weapon carrying, use, and threatening to use weapons in the past year. This approach of compiling information from different sources to create one outcome measure is consistent with some past work examining weapon use (Huesmann et al, 2021) and other forms of antisocial behavior (e.g., gang involvement; Boxer et al., 2015). The items were dichotomized into 0 = no or never, and 1 = yes or at least once in the past year. Sample items include “In the past year, how often have you… threatened or actually shot another person with a gun?”, and “carried some other weapon such as a knife?”. We then created a dichotomous scale score indicating any weapon use in the past year: 0 for no weapon use, and 1 if any “yes” response was endorsed. The choice to dichotomize the weapon use outcome variable is also consistent with prior work (Farrington & Loeber, 2000; Haegerich et al., 2014) and allows us to examine differences between youth who engage in any form of weapon use and those who report no weapon use whatsoever. Although we acknowledge the limitations of dichotomization (Iselin et al., 2013), the outcome variable was heavily skewed and results using a count or continuous version may be biased. In the current study, 23% of participants (n=71) endorsed any weapon use. Among that group, 56% (n=40) reported engaging in one weapon-related behavior, 34% (n=24) reported two behaviors, and 10% (n=7) reported between three and five behaviors.

Hypothesized T2 Influences on Weapon Use at T3

Youth’s Aggressiveness 3 Years Earlier at T2.

We employed the 9-item physical aggression scale from the Buss-Perry Aggression Questionnaire (1992). Respondents indicated how characteristic each statement is of them (0 = not at all true of me to 4 = definitely true of me). Sample items are: “I can think of no good reason for ever hitting a person” (reverse coded), and “There are people who pushed me so far that we came to blows.” We computed an average score for the 9 items for each respondent (Cronbach’s α = .72).

Interviewed Parent’s Aggressiveness 3 Years Earlier at T2.

To assess each interviewed parent’s aggressiveness at T2, we used the same measure as with the youth – the 9-item Buss-Perry Aggression Questionnaire (1992) – and computed an average score for the 9 items for each respondent (α = .76).

Youths’ Exposure to Neighborhood Weapon Violence at T2.

We assessed exposure to neighborhood weapon violence in adolescence and early adulthood (ages 15 to 25) at T2 using 9 items from the Survey of Exposure to Community Violence Scale (Richters & Martinez, 1993) that specifically assess witnessing and being victimized by weapon violence – including guns as well as other weapons such as knives – in the neighborhood, in school, and outside of school. Sample items include: “In the last year, how many times… have you yourself seen someone carrying or holding a gun?”, “have you seen someone carrying or holding another weapon like a knife”, and “have you been shot at with a gun?”. We computed an average across the 9 items (α =.78).

Youths’ Normative Beliefs about Aggression and Gun Use at T2, as a Moderator of the Effect of Exposure to Weapon Violence at T2 on Weapon Use at T3.

We measured youths’ normative beliefs about aggression and gun use at T2 with 8 items from the General Normative Beliefs about Aggression Scale (Huesmann & Guerra, 1997) and 4 items reflecting normative beliefs supporting gun use specifically created for this study. Participants indicated how “okay” they believed each behavior was on a 4-point scale (1 = It’s really wrong to 4 = It’s perfectly OK). Sample items include “In general, is it OK to… hit other people?” and “…use or threaten to use a gun against another person if they have insulted you or disrespected you?”. We computed an average across the 12 items to assess normative beliefs supporting both aggression and gun use (α = .90).

Hypothesized T1 Predictors of this T2 Normative Beliefs Moderator

We also measured hypothesized personal and contextual factors at T1 (Waves 1 to 3, ages 8 to 17, 8–10 years prior to the T2 assessments) that would influence individual differences in the youths’ T2 normative beliefs supporting aggression and weapon use. As only the Flint sample was interviewed during T1, we could only measure these T1 predictors for the Flint sample.

Youths’ Normative Beliefs Supporting Aggressive Retaliation at T1 (ages 8 to 17).

The youth in the Flint sample completed at T1 an 8-item measure of beliefs supporting retaliatory aggression across that (Huesmann & Guerra, 1997) found was reliable and valid for that age probably because it asked about very specific situations, e.g., “Suppose a boy says something bad to a girl. Do you think it’s ok for the girl to hit him?” The response options ranged from 1=“It’s perfectly OK” to 4=“It’s really wrong.” Scores were then averaged across the three waves to yield one scale score for T1 (α = .85–.86 across waves).

Parental Aggressiveness at T1.

We assessed the interviewed parent’s aggressiveness at T1 (when their child was 8 to 17 years old) with the same the 9-item physical aggression scale from the Buss-Perry Aggression Questionnaire (1992) that we used 8 years later for assessing parent and youth aggressiveness during the T2 interviews. Parents indicated how true each statement was for them (e.g., “I get into fights a little more than the average person”; response options ranged from 0=“not at all true of me” to 4=”definitely true of me”). Parents from the Flint sample completed this measure across the three initial waves of the study comprising T1. Scores were averaged across the three waves to yield one T1 scale score (α = .65–.73 across waves).

Youths’ Aggressiveness at T1.

In this study during the 3 waves of the T1 interview period, we were only able to assess the aggressiveness of the oldest cohort of the Flint sample (age 15 at Wave 1 to age 17 at Wave 3; n=127), using the same 9-item Buss-Perry Aggression Questionnaire (1992) that we used for parents (α = .67–.71 across waves).

Youth Callousness at T1.

Youth in the oldest cohort of the Flint sample completed the 9-item callousness scale of the Inventory of Callous-Unemotional Traits (ICU; Frick, 2004; Kimonis et al., 2008) at each of the initial three T1 waves.1 Youth indicated how true each statement was for them (e.g., “I do not feel remorseful when I do something wrong.” Response options ranged from 0=“not at all true of me” to 3=“definitely true of me”). The scale score for youth callousness at T1 was the average of the scores for the three waves (α = .72–.78 across waves).

Youths’ Exposure to Neighborhood Gun Violence at T1.

Early exposure to neighborhood gun violence at T1 (ages 8 to 17) was assessed in the Flint sample in each wave comprising T1 using three “yes - no” questions (yes =1, no =0) about witnessing gun violence in the last year (Attar et al., 1994). Two specifically asked about seeing people shooting guns, e.g., “in your neighborhood, have you seen anyone get hit, shot or really hurt by someone else,” and one asked, “Have you had to hide someplace because of shootings in your neighborhood?” A mean score was calculated at each wave, and we then averaged these scores across the three waves comprising T1 to compute an early exposure to gun violence composite score for T1.

Missing Data

For the Flint sample, retention rates at each of the four waves of follow-up interviews were 78.88%, 68.78%, 52.35%, and 56.10%. For this sample, attrition was mostly due to difficulty relocating participants. For the purposes of our analyses, 33% of this sample had completely observed data for our first model, 17% were missing one variable (most commonly parent aggression at T2), and 35% only had data for race and sex. Attrition was unrelated to sex, cohort, exposure to neighborhood gun violence at T1, normative beliefs related to retaliation at T1, parent aggression at T1, parent-reported youth aggression at T1, or (for the 9th grade cohort) self-reported youth aggression at T1. However, Black youth were more likely to have completely observed data (38%) compared to non-Black youth (25%; p=.021). For our second model, 49% of the sample had completely observed data and 42% were only missing on the outcome, normative beliefs about aggression and gun use at T2. For the Jersey City sample, retention rates at each of the three follow-up waves of interviews were 90.5%, 84%, and 62.5%. Out of 200 participants, 54% had completely observed data for our first model and 35% were only missing one variable (most commonly weapon use at T3). For this sample, attrition and missingness were not related to demographic or major study variables.

For both samples, we assume that data are missing at random (MAR) after accounting for earlier youth and parent aggression and sociodemographic factors because they are likely not missing completely at random (MCAR), and we have no reason to believe that data are missing not at random (MNAR). Because early childhood participants could only be recruited and interviewed in Flint, data on the T1 variables is missing for the Jersey City sample in the integrated data set. Although this is a limitation of the study design, we use a contemporary and best-practice method of handling missing data – full information maximum likelihood (FIML) – which is able to use all available data when estimating model parameters (Enders & Bandalos, 2001). In addition, data from T1 is used in testing hypothesis 2, but not hypothesis 1. Thus, the Jersey City sample is able to contribute to models testing both hypotheses, but notably with less data available than the Flint sample for the model testing hypothesis 2.

Results of models using FIML are more robust and less biased than they would be using listwise deletion, even if data were MNAR (Enders, 2008). FIML is also robust to the presence of some MNAR data (Gomer, 2019). In addition, incorporating auxiliary variables associated with missingness and/or the outcome – as we do in our current model, with sex, race, and earlier levels of youth and parent aggression – in FIML models can help to reduce bias and strengthen the MAR assumption (Enders, 2008; Raykov & West, 2016). However, because observed data are required for the grouping variable in a multiple group model, our analytic sample drops to N=423 (the participants who have observed data on normative beliefs supporting aggression at T2) for the key model in Figure 1 predicting adolescent and young adults weapon use at T3. Nevertheless, with the FIML approach to analyzing this model, we should retain sufficient power to detect effects of expected size (see power analysis below). All analyses were conducted in SPSS version 29 (for descriptive statistics and correlations; IBM Corp, 2023) and Amos version 26 (for path analysis; Arbuckle, 2019).

Fig. 1.

Fig. 1

Open in a new tab

Moderating Effect of Normative Beliefs Approving of Aggression in the Relation Between Exposure to Neighborhood Weapon and Subsequent Weapon Use

Note. This figure shows the results of the structural model predicting weapon use at Time 3 (ages 18 to 28) from exposure to neighborhood weapon violence 3 years earlier at Time 2 (ages 15 to 25) for those scoring high on normative beliefs approving of aggression at Time 2 and those scoring low on those normative beliefs at Time 2, while controlling for the effects of race, sex, the parent’s aggression, and the child’s aggression at Time 2. The model fit is as follows: χ2 (2) = 4.03, p = .134, CFI=. 99, RMSEA=.049, AIC = 108.0. N=423 participants with observed data for the grouping variable, normative beliefs supporting aggression at T2. ***p < .001. **p < .01. *p < .05. +p < .10.

Power Analysis

An a-priori power analysis was conducted as part of the original study to ensure adequate sample size to find statistically significant effects and test study aims (Lakens, 2022). For the current paper, a formal post-hoc power analysis indicates that our sample size is sufficient to adequately test our hypotheses. For hypothesis 1, we calculated a power of 94% using an alpha of .05, correlations among the predictor (continuous), moderator (binary), interaction, and outcome (binary) of .2, reliability estimates of .8 for the predictor, moderator, and outcome, and N=423. For hypothesis 2, we estimated that we had 90% power to detect an effect size as small as f2 = 0.02 – a very small effect – for any individual regression slope in a model with N=626, 7 predictors, and alpha of .05.

Results

Descriptive statistics for the key study variables are presented in Table 2 and the correlations among them and with the race and sex of the participants are presented in Table 3. Race was moderately correlated with T1 aggression, T1 normative beliefs supporting retaliation, T2 parent’s aggression, T2 neighborhood weapon violence exposure, and T3 weapon use (r = .23–.36). Sex had only small correlations with T2 neighborhood weapon violence exposure, T2 normative beliefs supporting aggression, and T3 weapon use (r = .11–.14).

Table 2.

Descriptive Statistics for the Key Variables in the Study

Variable and time assessed N Min Max Mean SD % at Max
S’s Aggressiveness at T1 327 0.22 3.83 1.78 0.65
S’s Interviewed Parent’s Aggressiveness at T1 391 0.00 3.39 1.09 0.67
S’s Callousness at T1 327 0.00 2.56 0.62 0.39
S’s Exposure to Neighborhood Gun Violence at T1 426 0.00 1.00 0.36 0.29 35.7%
S’s Normative Belief Supporting Aggressive Retaliation at T1 424 0.00 3.00 0.96 0.66
S’s Aggressiveness at T2 449 0.00 3.78 1.63 0.71
S’s Interviewed Parent’s Aggressiveness at T2 357 0.00 3.56 1.06 0.72
S’s Exposure to Neighborhood Weapon Violence at T2 423 0.00 2.33 0.50 0.42
S’s Normative Beliefs Supporting Aggression at T2 423 1.00 4.00 1.36 0.48
S’s Group on Normative Supporting Aggression at T2 423 0 (Lo) 1 (Hi) 0.69 0.46 69%
S’s Weapon Use at T3 364 0 1 0.23 0.42 23%
Sex: 0 = female, 1 = male 626 0 1 0.49 0.50 50.0%
Race: 2 = Black, 1 = not-Black 607 1 2 1.56 0.497 49.7%
Open in a new tab

Table 3.

Intercorrelations of the Key Variables in the Study

1 2 3 4 5 6 7 8 9 10 11 12
1 S’s Agg. at T1 1.00
2 S’s Interviewed Parent’s Agg.
at T1		 .10 1.00
3 S’s Callousness at T1 .30*** .15+ 1.00
4 S’s Exposure to Neigh. Gun
Violence at T1		 .37*** .04 .04 1.00
5 S’s Normative Belief Agg.
Retaliation OK at T1		 .34*** .06 .20* .25*** 1.00
6 S’s Agg. at T2 .71*** .17** .36*** .06 .27*** 1.00
7 S’s Interviewed Parent’s Agg.
at T2		 .32*** .31*** .10 −.01 .04 .16** 1.00
8 S’s Exposure to Neigh.
Weapon Violence at T2		 .52*** .01 .17** .20** .23*** .41*** .18*** 1.00
9 S’s Normative Belief Agg. OK
at T2		 .29*** .12+ .46*** .01 .22*** .45*** .08 .22*** 1.00
10 S’s Weapon Use at T3 .21** .12+ .13+ .08 .01 .21*** .19*** .27*** .12* 1.00
11 Sex: 0 = female, 1 = male .06 .01 .08 .03 −.01 .09+ .03 .11* .14** .11* 1.00
12 Race: 2 = Black, 1 = not-Black .26*** −.03 .03 .13** .28*** .06 .29*** .36*** −.03 .23*** .04 1.00
Open in a new tab
***

p < .001.

**

p < .01.

*

p < .05.

+

p < .10.

In Figure 1 we present a two-group (high or low on normative beliefs approving of aggression) structural model analysis that we computed to test our major hypothesis (Hypothesis 1) that young adults’ (between age 18 and 28) weapon use would be predicted by the extent of their exposure to neighborhood weapon violence in the recent past (from ages 15 to 25), but only for youth who held normative beliefs approving of aggression at the same time as they were exposed to weapon violence. Specifically, the structural model predicts a dichotomous weapon use measure from youths’ exposure to neighborhood weapon violence 3 years previously moderated by the normative beliefs approving of aggression (with or without weapons) that they held in the same period as they observed the weapons violence. We assigned to the low approval group all the participants who at T2 scored at the minimum on our scale assessing normative beliefs approving of aggression -- meaning they did not approve of a single aggressive act on the scale. The rest of the participants who scored above the minimum were denoted as the high approval group. As Figure 1 shows, the two-group model appeared to fit the data very well, as indicated by a non-significant chi-squared test (χ2(2)=4.03, p=.134), and a CFI = .99, and a RMSEA = .049. For the low group on normative beliefs supporting aggression (n=132), only the participant’s sex predicted weapon use, such that males with low approval of aggression were more likely to use weapons compared to females with low approval of aggression (β=0.19, p=.045). For the low approval group, exposure to neighborhood weapon violence 3 years previously had no significant effect on weapon use in young adulthood, providing support for hypothesis 1b. However, for the high group on normative beliefs approving of aggression (n=261), exposure to neighborhood weapon violence 3 years earlier significantly predicted weapon use in young adulthood (β=0.19, p=.013), providing evidence in support of hypothesis 1a. This was true even though the model controlled for the youth’s race and sex, the youth’s prior (3 years earlier) aggressiveness, and the youth’s parent’s prior aggressiveness. Importantly, when we recomputed this two-group model with all the parameters constrained to be equal between the two groups, the model fit significantly worse (χ2(26)=64.69, p<.001).

To graphically illustrate the moderation effect of normative beliefs on the relation between exposure to neighborhood gun violence and later gun use, we divided the participants into high and low groups on exposure to neighborhood weapon violence at T2. We assigned the one-third of the participants who scored highest on exposure to neighborhood weapon violence at T2 to the high group and the rest to the low group so the minimum exposure for anyone in the high group (0.67) was greater than the maximum exposure for anyone in the low group (0.56). Then we examined the mean scores on weapon use at T3 when we crossed these high and low groups on exposure to neighborhood gun violence with the high and low groups on normative beliefs approving of aggression. In Figure 2 we show the pattern of mean scores obtained. The mean weapon use score at T3 for the “high exposure with high normative beliefs” group is much greater than the mean weapon use score in any other group, and this interactive effect is statistically significant (F(1,318) = 5.11, p = .025). In other words, either being exposed to violence or approving of aggression was not enough on its own to predict weapon use, but youth with both of these risk factors had more than twice the rate of weapon use (44%) compared to any other group (12–19%). This general pattern – with weapon use highest for youth who have been exposed to violence and who approve of aggression – held for both males and females in this sample, albeit more robustly for females.

Fig. 2.

Fig. 2

Open in a new tab

Mean Scores on Weapon Use in Young Adulthood (Ages 18 – 28) as a Function of a Youth’s Normative Beliefs (NB) Approving of Aggression and Exposure to Neighborhood Weapon Violence Assessed 3 Years Previously

Having shown that normative beliefs supporting aggression promote the observational learning of weapon use from neighborhood weapon violence in late adolescence and early adulthood, we now turn to examining what personal and contextual factors early in life that promote holding such normative beliefs supporting aggression at that stage of life (Hypothesis 2). In Figure 3, we display a structural model that shows what childhood and early adolescent measures in our study predicted the extent of a youth’s approval of aggression (with or without weapons) in late adolescence and early adulthood. The model, which is essentially a multiple regression model, shows that youth’s previous normative beliefs approving of retaliatory aggression at T1 and the youth’s callous-unemotionality at T1 are both significant predictors of the youth holding normative beliefs 8 years later that it is generally OK to behave aggressively with or without weapons. Interestingly, these effects are present even though the analysis controls for the youth’s own earlier aggression and exposure to gun violence and the parent’s earlier aggression, none of which has a significant effect on the development of the youth’s normative beliefs that it is OK to behave aggressively. Race and sex each had small but significant coefficients predicting normative beliefs supporting aggression (r = −.12 and .13, respectively). Again, the model fits well with x2(1)=0,82, p=0.37, RMSEA=0.0, and R2=.318. Thus, our results are consistent with hypotheses 2b and 2c, but not with hypotheses 2a, 2d, or 2e, providing partial support for Hypothesis 2.

Fig. 3.

Fig. 3

Open in a new tab

Early Personal and Contextual Variables that Predict Subsequent Normative Beliefs Approving of Aggression

Note. This figure shows the results of the structural model predicting a youth’s normative beliefs approving of aggression at Time 2 (ages 15 to 25) from their normative beliefs approving of aggression their callous-unemotionality, their parents’ aggressiveness, and their exposure to neighborhood gun violence 8 years earlier (at ages 7 to 17). All inter-correlations between the predictor variables and correlations of them with race and sex are included in the model. The model fit is as follows: X2(1) = 0.82, p = .37; RMSEA = 0.0, R2= .318, N = 626. ***p < .001. **p < .01. *p < .05. +p < .10.

Discussion

In this study we examined longitudinal relations between exposure to weapon-related violence in the community and later weapon use in an integrated dataset comprised of two urban samples of adolescents and young adults. Specifically, we tested a hypothesized model in which weapon violence exposure predicted weapon use three years later, but only for those participants who endorsed normative beliefs that aggression is an acceptable behavior. Our results supported this hypothesis. That is, for youth who approved of aggression, encounters with weapon violence spurred greater engagement with weapons. For youth who did not approve of aggression, encounters with weapon violence had no impact on later weapon carrying or use. In line with relevant theorizing (e.g., Huesmann, 1998, 2018), these findings support the view of normative beliefs about aggression during adolescence as key schemas or filters through which experiences with violence in the social ecology impact behavioral choices. We also observed that normative beliefs about aggression during adolescence were predicted significantly by childhood risk factors for aggression, namely sex (male identified), race (non- Black), earlier normative beliefs supporting retaliation, and callousness. These results underscore the continuity of normative beliefs and highlight the importance of emotional response styles in the development of social-cognitive behavioral schemas (e.g., Boxer & Sloan-Power, 2013; Lemerise & Arsenio, 2000). Taken in sum, our findings have important implications for ongoing refinement of social-cognitive-ecological theory on the development of aggression and violence, as well as for interventions targeting the reduction of these behaviors in higher-risk settings.

Our study links normative beliefs as moderators of the impact of exposure to community violence to the use and/or carrying of weapons, and thus provides a key extension of social- cognitive theory to weapon use. Normative beliefs (and their analogs with different names) have long been recognized as targets for psychoeducational prevention and intervention programs (Boxer & Dubow, 2001; Metropolitan Area Child Study Research Group, 2007). Our findings underscore the importance of normative beliefs as potential filters of risk-promoting experiences. We know that youth growing up in communities with high and persistent levels of violent crime are likely to exhibit aggression as the result, in line with normative observational learning. Yet not all youth do so, and this is certainly the result of variation in youths’ other learning experiences – perhaps even more than their genetic dispositions (Sypher et al., 2019). For example, studies have shown that caregivers in higher-violence communities are likely to engage in a greater degree of close monitoring of their children’s behavior (e.g., Jones et al., 2003), and one aspect of this might be more concerted efforts by caregivers to teach their children that aggression and violence are wrong and to be avoided as a style of responding. Still, these efforts might not override the effect of violent communities on youth aggression (Skinner et al., 2014) – we observed no relation from parents’ own aggressiveness to youths’ normative beliefs.

Given the role that normative beliefs supporting aggression play in the observational learning of violence and weapon use, it is important to consider earlier precursors to these normative beliefs and identify possible strategies for intervention efforts. In the present study, we observed that early callousness was a significant predictor of normative beliefs supporting aggression eight years later, even after controlling for children’s own aggression, their parent’s aggression, and their earlier beliefs supporting retaliation. These findings are novel and suggest that children and adolescents who present as callous may be primed for social learning of aggression through the development of social cognitions, such as normative beliefs, that support the use of aggression. This link between callousness and greater support for aggressive responding is consistent with integrated theory linking emotion-reactivity processes to social- cognitive processes in aggression (Boxer & Sloan-Power, 2013; Huesmann et al. 2017). The less one cares emotionally about the harm one does to others, the more acceptable one should find it to behave aggressively as a form of retaliation or as a general strategy for solving social conflicts. Callousness also emerges through exposure to extreme violence (e.g., Docherty et al., 2023).

In this study, we found no direct association between race and weapon use in the full model, although race was related to neighborhood weapon violence exposure, which then predicted weapon use for youth with normative beliefs supporting aggression. This is in line with previous studies which have found that apparent differences in weapon use across racial groups diminish or disappear when other factors are included in the model, highlighting the role of racial disparities in contributing to unequal outcomes (Beardslee et al., 2018; Docherty et al., 2019). In addition, we also found that non-Black youth were more likely to approve of aggression compared to Black youth, although this effect was small in size. For sex, we found that males were more likely than females to endorse normative beliefs supporting aggression, but we found similar rates of exposure to neighborhood weapon violence for both males and females, as well as similar levels of later weapon use for youth of both sexes who endorsed normative beliefs supporting aggression. Thus, we only found a sex difference among youth who do not approve of aggression, such that in this group males were more likely to use a weapon than females.

Limitations

Though our study has a number of strengths, including a relatively large integrated sample (N = 626), prospective longitudinal design that allows us to study developmental processes from ages 8 to 28, and the use of multi-informant constructs, there are also some limitations worth noting. Although with integrative data analysis and full information maximum likelihood we were able to leverage information from both longitudinal samples, it would have been preferable to collect data for both samples at similar time points. In addition, because there are four different cohorts represented in these samples, but each is too small to reliably test for cohort effects, we are unable to more precisely pinpoint the ages at which these processes are occurring. Still, it is worth noting that we were able to align the data by broad developmental periods and thus results are informative for developmental processes. Although we did include parent-report measures in our analyses, most of the key variables were assessed via self-report, increasing the risk for common method bias which can lead to inflated Type I error. Finally, although we used samples well-suited to test our hypotheses – mostly Black youth from urban environments characterized by high levels of disadvantage and violence in the Midwest and Northeast regions of the U.S. – generalizability of findings to other samples or populations (e.g., to rural youth, higher-SES youth, youth from other regions of the U.S., etc.) may be limited.

Future Research Directions

Although this study used a prospective longitudinal design spanning 20 years of development and thus provided novel information about relations among exposure to violence, normative beliefs, and weapon use across time, we were unable to capture more proximal processes occurring in the short term. Future research could use longitudinal designs with more fine-grained temporal assessments – for example, through the use of ecological momentary assessment – to determine how daily levels of exposure to violence and normative beliefs supporting aggression contribute to immediate increases in risk for weapon use, as well as other personal, contextual, and situational factors that are relevant for these processes. Such designs could provide a more precise test of the hypotheses derived from social-cognitive- ecological theory and inform future intervention work to reduce engagement in and victimization from weapon violence, thereby improving public health and safety.

Clinical Implications

This study represents a key test of social-cognitive-ecological theory and points to normative beliefs supporting aggression as playing a pivotal role in influencing weapon involvement after exposure to weapon violence. Although interventions have already been designed to target normative beliefs, it is possible that these interventions could be more specifically designed to reduce risk for weapon use among youth who have been exposed to violence. For example, it is possible that normative beliefs approving of weapon use specifically may be especially important to target. In addition, because callousness and beliefs in support of retaliation were earlier precursors to normative beliefs supporting aggression, interventions that focus on children and early adolescents may benefit from focusing on these risk factors. Although youth with callousness were historically deemed difficult to treat (Frick et al., 2014; Hawes & Dadds, 2005), recent research suggests that certain types of treatments, especially those that involve parents and are tailored to the specific needs of youth with elevated callousness, have been effective in reducing both callousness and problem behaviors for these youth (Hawes et al., 2014). Thus, it is possible these interventions may also help to deter youth from developing normative beliefs supporting aggression, and in turn reduce the risk for observational learning of weapon use.

Public Significance Statement.

The present study indicates that exposure to violence in late adolescence and early adulthood increases later risk for using weapons, primarily among people who already believe that aggression is an acceptable behavior. We found that such beliefs were predicted by much earlier levels of callousness and normative beliefs approving of aggressive retaliation. Violence prevention work that focuses on reducing callousness, exposure to violence, and beliefs approving of aggression in childhood and adolescence may be particularly useful for reducing later weapon violence.

Acknowledgments

This research was supported by grants HD49837 (L. Rowell Huesmann, PI) and HD84652 (Eric F. Dubow and L. Rowell Huesmann, MPIs) from the Eunice Kennedy Shriver National Institute of Child Health and Human Development and grant CE003302 (Eric F. Dubow, L. Rowell Huesmann, and Paul Boxer, MPIs) from the Centers for Disease Control and Prevention. All procedures involving human subjects were approved by the University of Michigan's IRB. The authors have no conflicts of interest with this research.

The authors wish to acknowledge the contributions of Maureen O’Brien, who supervised the collection of the Flint data in the early waves, and of Cathy Smith and Matt Morley, who assisted in the analyses of the data. The authors also wish to acknowledge the contributions of Brad Bushman, Craig Anderson, and Doug Gentile, who contributed to the theoretical conception of the early part of the project and development of the measures.

Footnotes

1

We chose to use the Callousness scale because it is internally consistent and is more strongly related to aggressive and violent behavior than the other scales in previous research (e.g., Kimonis et al., 2008). However, results using the Total score based on 22 items (excluding items 2 and 10; Kimonis et al., 2008) were not appreciably different from those obtained when using the Callousness score, suggesting robustness of results to this decision. Results with the Total 22-item score are available upon request from the corresponding author.

References

  1. Anderson CA, & Huesmann LR (2003). Human aggression: A social-cognitive view. In Hoag MA, & Cooper J (Eds.), The Sage handbook of social psychology (pp. 296–323). Thousand Oaks, CA: Sage. [Google Scholar]
  2. Arbuckle JL (2019). Amos (Version 26.0) [Computer Program]. Chicago: IBM SPSS. [Google Scholar]
  3. Attar BK, Guerra NG, & Tolan PH (1994). Neighborhood disadvantage, stressful life events, and adjustment in urban elementary-school children. Journal of Clinical Child Psychology, 23(4), 391–400. 10.1207/s15374424jccp2304_5 [DOI] [Google Scholar]
  4. Bandura A (1986). Social foundations of thought and action: A social-cognitive theory. Englewood Cliffs, NJ: Prentice-Hall. [Google Scholar]
  5. Bandura A, Ross D, & Ross SA (1961). Transmission of aggression through imitation of aggressive models. The Journal of Abnormal and Social Psychology, 63(3), 575–582. 10.1037/h0045925 [DOI] [PubMed] [Google Scholar]
  6. Beardslee J, Docherty M, Mulvey E, Schubert C, & Pardini D (2018). Childhood risk factors associated with adolescent gun carrying among Black and White males: An examination of self-protection, social influence, and antisocial propensity explanations. Law and Human Behavior, 42(2), 110–118. 10.1037/lhb0000270 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Beardslee J, Kan E, Simmons C, Pardini D, Peniche M, Frick PJ, Steinberg L, & Cauffman E (2021). A within-individual examination of the predictors of gun carrying during adolescence and young adulthood among young men. Journal of Youth and Adolescence, 50(10), 1952–1969. 10.1007/s10964-021-01464-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bottiani JH, Camacho DA, Lindstrom Johnson S, & Bradshaw CP (2021). Annual Research Review: Youth firearm violence disparities in the United States and implications for prevention. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 62(5), 563–579. 10.1111/jcpp.13392 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Boxer P, & Dubow EF (2001). A social-cognitive information-processing model for school- based aggression reduction and prevention programs: Issues for research and practice. Applied & Preventive Psychology, 10(3), 177–192. 10.1016/S0962-1849(01)80013-5 [DOI] [Google Scholar]
  10. Boxer P, Huesmann LR, Bushman BJ, O’Brien M, & Moceri D (2009). The role of violent media preference in cumulative developmental risk for violence and general aggression. Journal of Youth and Adolescence, 38(3), 417–428. 10.1007/s10964-008-9335-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Boxer P, & Sloan-Power E (2013). Coping with violence: A comprehensive framework and implications for understanding resilience. Trauma, Violence & Abuse, 14(3), 209–221. 10.1177/1524838013487806 [DOI] [Google Scholar]
  12. Boxer P, Veysey B, Ostermann M, & Kubik J (2015). Measuring gang involvement in a justice-referred sample of youth in treatment. Youth Violence and Juvenile Justice, 13(1), 41–59. 10.1177/1541204013519828 [DOI] [Google Scholar]
  13. Buss AH, & Perry M (1992). The Aggression Questionnaire. Journal of Personality and Social Psychology, 63(3), 452–459. 10.1037/0022-3514.63.3.452 [DOI] [PubMed] [Google Scholar]
  14. Carvalho M, Faria M, Conceição A, Gaspar de Matos M, & Essau CA (2018). Callous- unemotional traits in children and adolescents: Psychometric properties of the Portuguese version of the Inventory of Callous-Unemotional Traits. European Journal of Psychological Assessment, 34(2), 101–110. 10.1027/1015-5759/a000449 [DOI] [Google Scholar]
  15. Centers for Disease Control and Prevention. (2014). Youth Risk Behavior Surveillance Survey- United States, 2013. Morbidity and Mortality Weekly Report, Surveillance Summaries, 63 (4). [Google Scholar]
  16. Cook PJ, & Ludwig J (2004). Does gun prevalence affect teen gun carrying after all? Criminology, 42, 27–54. [Google Scholar]
  17. Curran PJ, & Hussong AM (2009). Integrative data analysis: The simultaneous analysis of multiple data sets. Psychological Methods, 14(2), 81–100. 10.1037/a0015914 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Davis-Kean PE, Huesmann LR, Jager J, Collins WA, Bates JE, & Lansford JE (2008). Changes in the relation of self-efficacy beliefs and behaviors across development. Child Development, 79(5), 1257–1269. 10.1111/j.1467-8624.2008.01187.x [DOI] [PubMed] [Google Scholar]
  19. DelGiudice M (2018). Middle childhood: An evolutionary-developmental synthesis. In Halfon N, Forrest CB, Lerner RM, and Faustman EM (Eds.), Handbook of life course health development (pp. 95–108). Springer. [Google Scholar]
  20. Docherty M, Beardslee J, Grimm KJ, & Pardini D (2019). Distinguishing between-individual from within-individual predictors of gun carrying among Black and White males across adolescence. Law and Human Behavior, 43(2), 144–155. 10.1037/lhb0000320 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Docherty M, Decrop R, McManamon B, Boxer P, Dubow EF, & Huesmann LR (2023). Exposure to violence predicts callous-unemotional traits and aggression in adolescence in the context of persistent ethnic-political conflict and violence. Aggressive Behavior, 49, 655–668. 10.1002/ab.22103 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Docherty M, Kubik J, Herrera CM, & Boxer P (2018). Early maltreatment is associated with greater risk of conduct problems and lack of guilt in adolescence. Child Abuse & Neglect, 79, 173–182. 10.1016/j.chiabu.2018.01.032 [DOI] [PubMed] [Google Scholar]
  23. Dodge KA, Bai Y, Godwin J, Lansford JE, Bates JE, Pettit GS, & Jones D (2022). A defensive mindset: A pattern of social information processing that develops early and predicts life course outcomes. Child Development, 93(4), e357–e378. 10.1111/cdev.13751 [DOI] [PubMed] [Google Scholar]
  24. Enders CK (2008). A note on the use of missing auxiliary variables in full information maximum likelihood-based structural equation models. Structural Equation Modeling: A Multidisciplinary Journal, 15(3), 434–448. 10.1080/10705510802154307 [DOI] [Google Scholar]
  25. Enders CK, & Bandalos DL (2001). The relative performance of full information maximum likelihood estimation for missing data in structural equation models. Structural Equation Modeling: A Multidisciplinary Journal, 8(3), 430–457. 10.1207/S15328007SEM0803_5 [DOI] [Google Scholar]
  26. Everytown for Gun Safety (n.d.). How does gun violence impact the communities you care about? EveryStat. Retrieved May 16, 2024, from https://everystat.org/ [Google Scholar]
  27. Farrington DP. & Loeber R (2000). Some benefits of dichotomization in psychiatric and criminological literature. Criminal Behavior and Mental Health, 10, 100–122. 10.1002/cbm.349 [DOI] [Google Scholar]
  28. Frick PJ (2004). Inventory of Callous–Unemotional Traits. 10.1037/t62639-000 [DOI] [Google Scholar]
  29. Frick PJ, Ray JV, Thornton LC, & Kahn RE (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), 1–57. 10.1037/a0033076 [DOI] [PubMed] [Google Scholar]
  30. Gomer B (2019). MCAR, MAR, and MNAR values in the same dataset: A realistic evaluation of methods for handling missing data. Multivariate Behavioral Research, 54(1), 153. 10.1080/00273171.2018.1557033 [DOI] [PubMed] [Google Scholar]
  31. Guerra NG, Huesmann LR, & Spindler A (2003). Community violence exposure, social cognition, and aggression among urban elementary school children. Child Development, 74(5), 1561–1576. 10.1111/1467-8624.00623 [DOI] [PubMed] [Google Scholar]
  32. Haegerich TM, Oman RF, Vesely SK, Aspy CB, & Tolma EL (2014). The predictive influence of family and neighborhood assets on fighting and weapon carrying from mid-to late adolescence. Prevention Science, 15, 473–484. 10.1007/s11121-013-0400-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Hawes DJ, & Dadds MR (2005). The treatment of conduct problems in children with callous-unemotional traits. Journal of Consulting and Clinical Psychology, 73(4), 737–741. 10.1037/0022-006X.73.4.737 [DOI] [PubMed] [Google Scholar]
  34. Hawes DJ, Price MJ, & Dadds MR (2014). Callous-unemotional traits and the treatment of conduct problems in childhood and adolescence: A comprehensive review. 17, 248–267. 10.1007/s10567-014-0167-1 [DOI] [Google Scholar]
  35. Huesmann LR (1998). The role of social information processing and cognitive schema in the acquisition and maintenance of habitual aggressive behavior. In Geen RG & Donnerstein E (Eds.), Human aggression: Theories, research, and implications for social policy (pp. 73–109). Academic Press. 10.1016/B978-012278805-5/50005-5 [DOI] [Google Scholar]
  36. Huesmann LR (2018). An integrative theoretical understanding of aggression: A brief exposition. Current Opinion in Psychology, 19, 119–124. 10.1016/j.copsyc.2017.04.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Huesmann LR, Dubow EF, B Boxer P, Bushman BJ, S Smith C, A Docherty M, & J O’Brien M (2021). Longitudinal predictions of young adults’ weapons use and criminal behavior from their childhood exposure to violence. Aggressive Behavior, 47(6), 621–634. 10.1002/ab.21984 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Huesmann LR, Dubow EF, Boxer P, Landau SF, Gvirsman SD, & Shikaki K (2017). Children’s exposure to violent political conflict stimulates aggression at peers by increasing emotional distress, aggressive script rehearsal, and normative beliefs favoring aggression. Development and Psychopathology, 29(1), 39–50. 10.1017/S0954579416001115 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Huesmann LR, Eron LD, Lefkowitz MM, & Walder LO (1984). Stability of aggression over time and generations. Developmental Psychology, 20(6), 1120–1134. 10.1037/0012-1649.20.6.1120 [DOI] [Google Scholar]
  40. Huesmann LR, & Guerra NG (1997). Children’s normative beliefs about aggression and aggressive behavior. Journal of Personality and Social Psychology, 72(2), 408–419. 10.1037//0022-3514.72.2.408 [DOI] [PubMed] [Google Scholar]
  41. IBM Corp. Released 2023. IBM SPSS Statistics for Windows, Version 29.0. Armonk, NY: IBM Corp. [Google Scholar]
  42. Iselin A-MR, Gallucci M, & DeCoster J (2013). Reconciling questions about dichotomizing variables in criminal justice research. Journal of Criminal Justice, 41(6), 386–394. 10.1016/j.jcrimjus.2013.07.002 [DOI] [Google Scholar]
  43. Jones DJ, Forehand R, Brody G, & Armistead L (2003). Parental monitoring in African American, single mother-headed families. An Ecological approach to the identification of predictors. Behavior Modification, 27(4), 435–457. 10.1177/0145445503255432 [DOI] [PubMed] [Google Scholar]
  44. Kimonis ER, Frick PJ, Skeem JL, Marsee MA, Cruise K, Munoz LC, Aucoin KJ, & Morris AS (2008). Assessing callous-unemotional traits in adolescent offenders: Validation of the Inventory of Callous-Unemotional Traits. International Journal of Law and Psychiatry, 31(3), 241–252. 10.1016/j.ijlp.2008.04.002 [DOI] [PubMed] [Google Scholar]
  45. Kravitz-Wirtz N, Bruns A, Aubel AJ, Zhang X, & Buggs SA (2022). Inequities in community exposure to deadly gun violence by race/ethnicity, poverty, and neighborhood disadvantage among youth in large US cities. Journal of Urban Health, 99(4), 610–625. 10.1007/s11524-022-00656-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Lakens D (2022). Sample size justification. Collabra: Psychology, 8(1), 33267. 10.1525/collabra.33267 [DOI] [Google Scholar]
  47. Lemerise EA, & Arsenio WF (2000). An integrated model of emotion processes and cognition in social information processing. Child Development, 71(1), 107–118. 10.1111/1467-8624.00124 [DOI] [PubMed] [Google Scholar]
  48. Lui JHL, Barry CT, & Schoessler M (2017). The indirect effects of adolescent psychopathic traits on aggression through social-cognitive factors. Journal of Child and Family Studies, 26, 1298–1309. 10.1007/s10826-017-0667-y [DOI] [Google Scholar]
  49. López-Romero L, Gómez-Fraguela JA & Romero E (2015). Assessing callous-unemotional traits in a Spanish sample of institutionalized youths: The Inventory of Callous- Unemotional Traits. Journal of Psychopathology and Behavioral Assessment, 37, 392–406. 10.1007/s10862-014-9469-3 [DOI] [Google Scholar]
  50. Mah VK, & Ford-Jones EL (2012). Spotlight on middle childhood: Rejuvenating the ‘forgotten years’. Paediatrics & Child Health, 17(2), 81–83. 10.1093/pch/17.2.81 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. The Metropolitan Area Child Study Research Group. (2007). Changing the way children “think” about aggression: Social-cognitive effects of a preventive intervention. Journal of Consulting and Clinical Psychology, 75(1), 160–167. 10.1037/0022-006X.75.1.160 [DOI] [PubMed] [Google Scholar]
  52. National Institute of Justice. (2007). Adolescents, neighborhoods and violence: Recent findings from the Project on human development in Chicago neighborhoods. Washington, D.C.: U.S. Department of Justice, Office of Justice Programs. [Google Scholar]
  53. Pihet S, Etter S, Schmid M, & Kimonis ER (2015). Assessing callous-unemotional traits in adolescents: Validity of the Inventory of Callous-Unemotional Traits across gender, age, and community/institutionalized status. Journal of Psychopathology and Behavioral Assessment, 37, 407–421. 10.1007/s10862-014-9472-8 [DOI] [Google Scholar]
  54. Raykov T, & West BT (2016). On enhancing plausibility of the missing at random assumption in incomplete data analyses via evaluation of response-auxiliary variable correlations. Structural Equation Modeling: A Multidisciplinary Journal, 23(1), 45–53. 10.1080/10705511.2014.937848 [DOI] [Google Scholar]
  55. Richters JE, & Martinez P (1993). Children as victims of and witnesses to violence in a Washington, D.C. neighborhood. In Leavitt LA & Fox NA (Eds.), The psychological effects of war and violence on children (pp. 243–278). Lawrence Erlbaum Associates, Inc. [Google Scholar]
  56. Sameroff AJ, & Haith MM (1986). The five to seven year shift. Chicago: University of Chicago Press. [Google Scholar]
  57. Skinner AT, Bacchini D, Lansford JE, Godwin JW, Sorbring E, Tapanya S, Tirado LMU, Zelli A, Alampay LP, Al-Hassan SM, et al. (2014). Neighborhood danger, parental monitoring, harsh parenting, and child aggression in nine countries. Societies, 4, 45–67. 10.3390/soc4010045 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Sypher I, Hyde LW, Peckins MK, Waller R, Klump K, & Alexandra Burt S (2019). Effects of parenting and community violence on aggression-related social goals: A monozygotic twin differences study. Journal of Abnormal Child Psychology, 47(6), 1001–1012. 10.1007/s10802-018-0506-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES