read

References

Citekey: @KerresMalecki2003

Kerres Malecki, C., & Kilpatrick Demaray, M. (2003). Carrying a weapon to school and perceptions of social support in an urban middle school. Journal of Emotional and Behavioral Disorders, 11(3), 169–178. http://doi.org/10.1177/10634266030110030401

Notes

Notes to me as an instructor:

  • Stats involved in this paper
    • descriptive analysis
    • ANOVA
    • MANOVA
    • logistic regression
    • internal consistency as measured by alpha coefficient
  • Discussion points
    • roles of quantitative analyses
    • limitations of this study
    • next steps of research you can think of?

Highlights

Students who reported carrying weapons to school reported less overall or total perceived social support than did their peers who did not carry weapons. In addition, the former reported significantly less perceived social support from all sources—parents, teachers, classmates, close friends, and the school. Another finding was that in addition to other commonly cited and researched risk factors (e.g., drug use, alcohol use), individual social support from parents, teachers, classmates, and close friends also was a significant predictor of weapon carrying in school. (p. 169)

RISK FACTORS (p. 170)

Violence (p. 170)

providing more evidence that adult support is negatively related to violent behavior. (p. 170)

Although these common characteristics have been identified, a clear “profile” of a student who may become lethally violent in school does not exist. Rather, researchers have identified an interaction among many factors. (p. 170)

These three studies thus suggest that higher peer support is related to higher rates of violence, whereas lower peer support is related to higher rates of being victimized. (p. 170)

• individualfactors(e.g.,bringing weapons to school, discipline problems, aggressive behavior); • familyfactors(e.g.,lackoffamily support); • schoolorpeerfactors(e.g.,gang membership, being bullied at school, low peer support); • societalfactors(e.g.,accessto weapons); and • situational factors (e.g., stressful events; Verlinden et al., 2000). (p. 170)

few empirical studies have specifically examined the relationships among social support variables and school violence or weapons possession. (p. 170)

Specifically, social support was defined as an individual’s perceptions of general support or specific supportive behaviors from people in his or her social network, which enhances the person’s functioning and/or may buffer him or her from adverse out- (p. 170)

Weapons Carrying in Schools (p. 170)

comes (Malecki & Demaray, 2002). We measured social support with The Child and Adolescent Social Support Scale– Revised Edition (CASSS; Malecki, Demaray, & Elliott, 2000), which was developed using Tardy’s (1985) model of social support (p. 171)

Tardy described several elements in measuring social support, including whether support is available or actually enacted, whether support is described (frequency) or evaluated (importance of support perceived), and whether support is given or received. In addition, Tardy noted that there are several types of social support (emotional, informational, instrumental, and appraisal ) and that support comes from many sources. (p. 171)

Materials (p. 171)

TheCASSS. Thisinstrumentassesses perceived social support via students’ selfreports of support from five sources: parents, teachers, classmates, a close friend, and the school. Four types of support (emotional, informational, instrumental, and appraisal) within each source of support are assessed. Thus, three items for each of the four types of support make up each of the source subscales (12 items for each subscale), resulting in 60 total items. (p. 171)

RESEARCH PREDICTIONS (p. 171)

METHOD (p. 171)

Each item is rated by students on the perceived frequency of a supportive behavior (1 = Never to 6 = Always) and on the importance that they attribute to that behavior (1 = Not Important to 3 = Very Important ). (p. 171)

Health Questionnaire. The questions used in the current study were taken from a larger questionnaire originally developed by an outside community agency. (p. 172)

Risk Code. Twelve of the questions on the survey were summed to create a risk code. These 12 items addressed bullying, being bullied, delinquent behavior, and drug and alcohol use. (p. 172)

The three questions regarding bullying asked “How often have you done these things to someone else at school in the last year?” and included “attacked someone,” “threatened someone with a weapon,” and “used a weapon to hurt someone.” Responders rated each item on a 5-point scale (0 = Never, 1 = 1 to2times,2=3to5times,3=6to9times, and 4 = 10+ times). (p. 172)

A total of 14 items from the questionnaire were used in the current study (3 to assess weapons possession and 12 to develop a “risk code” score) and are described next. (p. 172)

Weapons Possession. (p. 172)

The internal consistency of the frequency scores has been demonstrated with an alpha coefficient of .97 for the overall support score (n = 192) and strong coefficients on the subscales (rs = .92–.95; ns = 193–263). Test–retest reliability evidence has been demonstrated for the frequency scores with 8to 10-week coefficients of .85 on the overall support score and .47 to .83 on the subscales (p. 172)

Students were classified as having carried a weapon to school by analyzing the following three items from the health questionnaire: 1. “How many times in the last year have you carried a knife to school?” 2. “How many times in the last year have you carried a gun to school?” 3. “How many times in the last year have you carried another weapon to school?” (p. 172)

In regards to assessing drug and alcohol use, students were asked “How many times in the last year have you been drunk or high at school?” and “How many times in the last 30 days did you use marijuana?” (p. 172)

RESULTS (p. 173)

Preliminary Analyses (p. 173)

Descriptive statistics were compiled on the number of students in the sample who reported carrying a knife, gun, or other weapon to school in the last year (see Table 2). Descriptive statistics for the 42 students who reported carrying a weapon to school at least once are presented in Table 3. (p. 173)

In order to investigate gender and grade differences in the main variable of interest, carrying a weapon to school, an analysis of variance (ANOVA) was conducted on weapons-carrying status (0 = have not carried a weapon, 1 = carried a weapon at least once) by gender and grade (sixth, seventh, and eighth). Results indicated a main effect for gender, F(1, 446) = 13.77, p < .001, with girls (M = .05, SD = .21) reporting less carrying of weapons than boys (M = .14, SD = .35), and a main effect for grade level (p. 173)

Post hoc Sheffé analyses revealed that sixth graders (M = .04, SD = .19) reported carrying weapons less often than both seventh-grade students (M = .12, SD = .33) and eighth-grade students (M = .12, SD = .33). No significant gender by grade level interaction was found (p. 174)

Main Analyses (p. 174)

conducting an ANOVA on the CASSS total support score by weapons-carrying status. Students who reported carrying weapons obtained significantly lower total support scores than students who did not report carrying weapons, F(1, 459) = 14.26, p < .001. (p. 174)

Differences in Perceived Social Support. Means and standard deviations for social support scores by weaponscarrying status are reported in Table 4. (p. 174)

conducting a MANOVA on the CASSS subscale scores by weaponscarrying status. The MANOVA was significant, Wilks’s lambda = .95, F(5, 455) = 4.96, p < .001. Results of followup univariate analyses indicated that students who self-reported carrying weapons scored lower on all subscales (Parent, Teacher, Classmate, Close Friend, and School) than did students who did not report carrying weapons. (p. 175)

PredictingWeaponCarriers. Alogistical regression analysis was conducted to investigate the predictive power of several risk factors, demographic variables, and social support in predicting weapons carriers. (p. 175)

The logistical regression was significant, χ2(4, N = 461) = 65.53, p < .0001. See Tables 5 and 6 for results. (p. 175)

DISCUSSION (p. 175)

Another finding was that in addition to other commonly cited and researched risk factors (e.g., drug use, alcohol use, victimization, bullying, delinquent behavior), overall social support from parents, teachers, classmates, and close friends is a significant individual predictor. (p. 176)

As predicted, students who reported carrying weapons to school reported less overall or total perceived social support than did their peers who did not carry weapons. In addition, they reported significantly less perceived social support from all sources: parents, teachers, classmates, close friends, and people in their school. (p. 176)

Thus, researchers and practitioners should be aware that measuring students’ levels of perceived social support can significantly improve risk assessments concerning carrying a weapon to school. (p. 176)

The current study provides empirical evidence that students who report carrying weapons to school also report significantly less perceived social support than their peers who do not carry weapons. (p. 176)

This study has several limitations. First, as with the majority of the research in this area, all of the data were gathered via self-report from students. (p. 176)

Blog Logo

Bodong Chen


Published

Image

Crisscross Landscapes

Bodong Chen, University of Minnesota

Back to Home