Student victimization in schools is a major concern of educators, policymakers, administrators, parents, and students. Understanding the scope of the criminal victimization of students, as well as the factors associated with it, is an essential step in developing solutions to address the issues of school crime and violence.
The National Center for Education Statistics (NCES) collects data on student criminal victimization through its sponsorship of the School Crime Supplement (SCS) to the National Crime Victimization Survey (NCVS), administered by the U.S. Department of Justice, Bureau of Justice Statistics (BJS). The SCS survey is designed to assist policymakers, researchers, and practitioners in making informed decisions concerning crime in schools. The purpose of this report is to provide data on student criminal victimization and the characteristics of crime victims and nonvictims from the 2009 SCS data collection.
This report uses data from the 2009 NCVS Basic Screen Questionnaire (NCVS-1), NCVS Crime Incident Report (NCVS-2), and SCS.1 The NCVS is the nation's primary source of information on criminal victimization and the victims of crime. The NCVS-2 collects data on criminal victimizations that occur at school and in locations other than at school. The SCS collects additional national-level information about the school and student characteristics that may be related to school crime by asking students questions about their experiences with and perceptions of crime and violence occurring inside their school, on school grounds, on the school bus, and going to or from school. The SCS contains questions not included in the NCVS, such as student reports of traditional bullying at school and cyber-bullying anywhere; the presence of weapons, gangs, hate-related words, and graffiti in school, as well as the availability of drugs and alcohol in school; and students' attitudes relating to fear of victimization and avoidance behavior at school.
Created as a supplement to the NCVS and codesigned by NCES and BJS, the SCS has been conducted in 1989, 1995, 1999, 2001, 2003, 2005, 2007, 2009, and 2011.
Each month, the U.S. Census Bureau selects households for the NCVS using a rotating panel design (see appendix A for additional information on sample design and data collection). Households within the United States are selected into the sample using a stratified, multistage cluster design, and all age-eligible individuals in the households become part of the panel. Once in the panel, respondents are administered the NCVS every 6 months over 3 years to determine whether they have been victimized during the 6 months preceding the interview.2 The SCS questionnaire is completed after the NCVS by persons in the sample household ages 12 through 18 who are currently enrolled in a primary or secondary education program leading to a high school diploma or who were enrolled sometime during the school year of the interview, and did not exclusively receive their education through homeschooling during the school year.3
Of the 8,986 NCVS household members who were between ages 12 and 18 and eligible for the 2009 SCS, a total of 5,023 students completed the NCVS and SCS surveys, of whom 4,326 met the requirements for inclusion in this analysis. Specifically, this report includes only students ages 12 through 18 who were enrolled in 6th through 12th grade at any time during the 2008–09 school year and who did not receive all or part of their education through homeschooling. The household completion rate was 92 percent and the student completion rate was 56 percent. The overall unweighted SCS unit response rate (calculated by multiplying the household completion rate by the student completion rate) was 51 percent.
NCES requires that any stage of data collection within a survey that has a unit base-weighted response rate of less than 85 percent be evaluated for the potential magnitude of unit nonresponse bias before the data or any analysis using the data may be released (U.S. Department of Education 2003). Due to the low SCS unit response rate, a unit nonresponse bias analysis was performed. Differences were found between the distributions of respondents and nonrespondents across race/ethnicity categories. White students and students of all other races had higher response rates than did Black and Hispanic respondents. The variable was retained for analysis and reporting because student race/ethnicity is a key population characteristic for consideration by readers, although readers should use caution when interpreting the results derived from this variable.
The mean item weighted response rate for the 2009 NCVS/SCS was greater than 97 percent and, therefore, there is little potential for item nonresponse bias for most items in the survey. Household income was the only analysis variable in this report that had a response rate of less than 85 percent (80 percent). When compared across other key population characteristics, it was found that respondents to the household income item differed across race/ethnicity. White students had higher rates of response for the income item than Black and Hispanic students and students of other race/ethnicities; however, when the distributions of respondents to the household income item were compared to the distribution of all those eligible to respond to the household income item, no measurable differences were found. Nonetheless, readers should use caution when interpreting the results derived from this variable. Refer to appendix A for more information on the respondent criteria for inclusion in the report analysis and the bias analyses that were performed.
NCVS and SCS data are also presented by Robers et al. in the 2010 edition of Indicators of School Crime and Safety, a report produced annually by NCES and BJS. That report compiles data from multiple sources, including national surveys of students, teachers, and principals, as well as universe data collections from federal departments and agencies, including BJS, NCES, the Federal Bureau of Investigation, and the Centers for Disease Control and Prevention. Indicators provides a first look at the SCS data and makes trend comparisons of SCS findings.
This report supplements Indicators by detailing characteristics of school crime, victims of crime, and the relationship between criminal victimization and bullying.
In this report, the definition of criminal victimization4 is derived from the NCVS "type of crime" variable. Criminal victimizations are categorized as "serious violent," "violent," or "theft." Serious violent victimization includes rape, sexual assault, robbery, and aggravated assault and is a subset of violent crimes. Violent victimization includes all serious violent crimes and simple assault. Theft includes attempted and completed purse snatching, completed pickpocketing, and all attempted and completed thefts, excluding motor vehicle theft. Theft does not include robbery, in which the threat or use of force is involved. Victims of "any" crime reported at least one of the victimizations above. Nonvictims of any crime reported none of the victimizations above.
Readers should note that the NCVS counts each incident of crime against an individual as a criminal victimization. However, the estimates in this report are based on the prevalence, or percentage, of students who experience victimizations. For example, if a respondent reports two unique victimizations, such as an assault and a theft, during the previous 6 months, this student would be counted once in the overall prevalence (any) estimate, because any victimization constitutes at least one violent victimization or theft. For many of the findings discussed in this report, the baseline comparison is that of victims of specific crimes to that of nonvictims. Measuring student victimization in this way provides the percentages of students who are directly affected by victimization, rather than the number of victimizations that occur at school. Estimates for serious violent victimization are only provided in detail in table 1. Because the percentage of students who experienced this type of victimization was not large enough to present meaningful cross-tabulations, tables 2 through 7 include estimates for serious violent victimization in the estimates for violent victimization.
For the purposes of this report, victimization at school refers to incidents that occurred inside the school building, on school property, on the school bus, or on the way to or from school. Some characteristics (such as school sector, security measures, and grade level) are drawn from student responses to the 2009 SCS, while others (such as sex, race/ethnicity, and household income) are drawn from NCVS-1 variables appended to the SCS data file. Estimates of victimizations that occurred inside the school building, on school property, on the school bus, or on the way to or from school are obtained from the NCVS-2. See appendixes C and D for selected questions from the NCVS-1 and NCVS-2 instruments and appendix E for the SCS instrument.
Readers should note that limitations inherent to victimization surveys such as the SCS and NCVS could have some effect on the estimates of victimization reported here (see Cantor and Lynch 2000). First, 15 percent of SCS interviews were new to the NCVS panel in 2009. Because there is no prior interview for new respondents to use as a point of reference when reporting victimization, their reports may include victimizations that occurred before the desired reference period. To the extent that these earlier victimizations are included, rates are overreported. Second, respondent recall of a victimization event may be inaccurate. People may forget the event entirely or recall the characteristics of the event inaccurately. This could lead to misclassification of victimizations.
Additional caution should be considered when examining the other variables used in this report. Because all variables of interest in the SCS and NCVS are self-reported, information about the respondent and his or her school may be inaccurate due to errors in recall, falsification, or exaggeration. Finally, readers should be aware of the limitations of the survey design and the analytical approach used here with regard to causality. Conclusions about causality between school or student characteristics and victimization cannot be made due to the cross-sectional, nonexperimental design of the SCS. Furthermore, certain characteristics discussed in this report (e.g., gang presence, security guards, and hallway monitors) may be related to one another, but this analysis does not control for such possible relationships. Therefore, no causal inferences should be made between the variables of interest and victimization when reading these results.
The comparisons in the text have been tested for statistical significance to ensure that the differences are larger than might be expected due to sampling variation. All statements cited in the report are statistically significant at the .05 level. The test procedure used in this report is Student's t statistic, which tests the difference between two sample estimates (see appendix A-10 for a fuller discussion). Multiple comparison adjustments have not been made in the analyses presented in this report, which may cause an increase in the number of significant findings that are reported. For example, when using a .05 alpha level, 5 percent of findings would be expected to be statistically significant by chance. The standard error is calculated for each estimate provided in order to determine the margin of error for the estimates. The standard errors of the estimates for different subpopulations can vary considerably and should be taken into account when making comparisons. It should also be acknowledged that apparently large differences between estimates may not have measurable differences, which may be due to large standard errors.
The results of this report are presented in six sections. The first two sections discuss the prevalence and type of student criminal victimization at school and selected characteristics of victims, including their demographic characteristics and school sector. The third section explores crime victim and nonvictim reports of school conditions, such as the presence of gangs and weapons and the availability of drugs. The fourth section examines criminal victimization and student reports of bullying and cyber-bullying at school. The fifth section examines criminal victimization and student reports of security measures taken at school to secure school buildings and the use of designated personnel and the enforcement of administrative procedures at school to ensure student safety. The sixth section examines fear and avoidance behaviors of crime victims and nonvictims, such as skipping class or avoiding specific places at school.
1 The SCS data are available for download from the Student Surveys link
at the NCES Crime and Safety Surveys portal, located at
2 The NCVS collects data on criminal victimization during the 6 months preceding the interview whereas, since 2007, the SCS has asked students about school characteristics "during this school year." This change in the SCS was made largely based on feedback obtained from students ages 12 through 18 who reviewed the items during cognitive laboratory evaluations conducted by the Census Bureau. These respondents revealed they were not being strict in their interpretation of the 6-month reference period and were responding based on their experiences during the entire school year.
3 Persons who have dropped out of school, have been expelled or suspended from school, or are temporarily absent from school for any other reason, such as illness or vacation, can complete the SCS as long as they have attended school at any time during the school year of the interview. Students who receive all of their education through homeschooling are not included past the screening questions and those who receive part of their education through homeschooling are not included in this report, since many of the questions in the SCS are not relevant to their situation.
4 For ease of presentation, the terms criminal victimization and victimization are used interchangeably throughout this report.