Aug
18

Individual-Level Studies of the Effects of Violence on the Academic Performance of Students

The majority of the current literature about the effects of violence exposure on school-age children has been research looking at individuals as the unit of analysis. These studies often use survey data, interviewing children or their parents about their exposure to violence and then using various individual-level measures to estimate its effects. My project differs from these studies in that I will be examining proximity to violence and the outcome of academic performance at the school level. Nevertheless, these individual-level studies provide important groundwork for my research because they can help illuminate the individual psychological mechanisms that help explain why violence might be associated with student performance. I’ve read a fair number of these studies over the summer, and so I thought I’d summarize a few of them here.

Using a student-level survey, Henrich et al (2004) investigate the long-term effects of exposure to violence on middle school students. The study finds that students who witnessed violence felt less safe at school and were at greater risk for doing worse academically over the 2-year duration of the study. Interestingly, being a victim of violence (as opposed to merely witnessing it) did not exacerbate these negative effects. When discussing the implications of their findings, the authors cite the importance of an “ecological approach” to bolster the academic performance of kids in inner city schools, meaning that those working to improve urban school performance should take into account the many factors at work in the neighborhoods where children live and go to school.

In Bowen and Bowen’s work (1999), an individual-level survey of students revealed that school and neighborhood danger had statistically significant and negative effects on attendance, trouble avoidance, and grades, all indicators of school performance. Interestingly, this study also looked into which types of students were most likely to be exposed to community violence, finding that males, African American students, and urban students were most likely to report exposure to danger. The findings also differentiated between middle and high school students; middle schoolers were more likely to report in-school personal threats, while high schoolers were more likely to report various indicators of neighborhood danger.

Overstreet and Braun (1999) use a survey method to look into the effects of community violence on academic performance, specifically examining whether exposure to violence affects academic functioning independent of emotional distress. The study’s tentative conclusions suggest that exposure to community violence only affects school performance under certain circumstances, when combined with other risk factors such as emotional distress. The results of the study suggested a weak relationship between community violence and academic functioning in general, but the relationship strengthened under certain circumstances (such as when the children exhibited signs of emotional distress or when families had a strong moral-religious emphasis, which appeared to be an insufficient coping mechanism in the face of these problems).

In another individual-level survey, Patchin et al (2006) used a similar method of youth interviews to look into the effects of exposure to community violence on childhood delinquency – specifically, on weapon carrying and assaultive behavior. This study found that individuals who reported higher levels of exposure to community violence were significantly more likely to report possessing a weapon (such as a gun or knife) or engaging in personal assault. In the study, broader indicators of neighborhood disadvantage didn’t have a significant effect on delinquency, while exposure to community violence did. While this study doesn’t deal directly with school performance, I think it’s important to note this evidence of the social costs of community violence. Also, it’s reasonable to surmise that weapon carrying and assaultive behavior may have a detrimental effect on academic efforts.

I also read one other study that was less conclusive about the effects of community violence on student outcomes. Aizer (2008) focused on isolating neighborhood violence from other measures of neighborhood and family disadvantage. After controlling for child characteristics, family characteristics, and neighborhood characteristics, Aizer found that living in violent neighborhoods, knowing gang members, and witnessing violence were not statistically significant in predicting a child’s test scores or internalizing behavioral problems. These factors did, however, have significant positive associations with externalizing behavioral problems, which are problematic behaviors that are directed outward, such as aggression, delinquency, and hyperactivity. This study also conducted an analysis using family fixed effects, looking at siblings from the same family that had experienced different levels of violence exposure. This technique is interesting because it provides a natural control for all family measures of disadvantage (both observable and unobservable). The family fixed effects analysis revealed a borderline significant association between exposure to violence and internalizing behavioral problems, which are problematic behaviors that are directed inwardly (such as depression and anxiety). There was some evidence that knowing violent peers was associated with lower test scores.

Ultimately, these results suggest that “care should be taken in interpreting estimates of the impact of exposure to neighborhood violence on child outcomes” (Aizer 29). Noted! Aizer’s study underscores the importance of including adequate controls to make sure that what may appear to be the effects of violence are not actually reflecting the effects of some other type of disadvantage. I will need to include similar controls at the school level when I design my analysis. Since these studies reveal somewhat mixed results at the individual level, it will be interesting to see what my school-level study reveals.

 

Henrich, Christopher C. et al. “The Association of Community Violence Exposure with Middle School Achievement: A Prospective Study.” Journal of Applied Developmental Psychology 25.3 (2004): 327-248.

Bowen, Natasha K. and Gary L. Bowen. “Effects of Crime and Violence in Neighborhoods and Schools on the School Behavior and Performance of Adolescents.” Journal of Adolescent Research 14.3 (1999): 319-342.

Overstreet, Stacey and Shawnee Braun. “A preliminary examination of the relationship between exposure to community violence and academic functioning.” School Psychology Quarterly 14.4 (1999): 380–396.

Patchin, Justin W. et al. “Exposure to Community Violence and Childhood Delinquency.” Crime & Delinquency 52 (2006): 307-332.

Aizer, Anna. “Neighborhood Violence and Urban Youth.” National Bureau of Economic Research Working Paper No. 13773, 2008.

 

Comments

  1. efpelletier says:

    Meg and Natalie — thanks for your thoughtful comments! The issue of control variables is a tricky one, and I have been looking for other studies that look specifically at spatial characteristics of school neighborhoods to see what’s been done before. As for your question, Meg, I think I’ll look at socioeconomic status at the school level, rather than using neighborhood data–which is often approximated by the number of students eligible for free and reduced lunch at a school. This eliminates the problem of trying to determine socioeconomic status for the specific area contained by the school attendance boundary. Natalie–in addition to controls for socioeconomic status and student demographic characteristics, there are also ways to control for school characteristics–school funding, student/teacher ratio, teacher salary, teacher education, and years of teacher experience, to name a few.

  2. mjschwenzfeier says:

    Hey Lizzy,

    All of these readings sound so interesting! I’m assuming that you’re going to be controlling somehow for factors that could be associated with both crime and student achievement (i.e. neighborhood socioeconomic status), but will it be a challenge to overlay all of these different statistics in the context of school boundaries?

    Meg

  3. Hey Lizzy! This project is so interesting – it’s the intersection of so many issues like urban environments, social issues, and – yes – the future of America! While reading this post, I was thinking “ah well this is all cool but how do they correct for many factors that usually come with violent environments”? Lo and behold, Aizer read my mind! I am unfamiliar with individual-level research methods and the methods that you are using – how do you go about correcting for other types of disadvantage? There seems to be a lot of focus on correcting for the students’ backgrounds, but is there any control for school qualities? Thanks for sharing all your work!

Speak Your Mind

*