【Abstract】In the past several decades， juvenile delinquency is a problematic issue that is having a significant impact in society. Many theories， like Social Learning， Control， and Strain theory are developed to explain why juveniles are prone to be delinquent. This study focuses on analyzing the relationship between peers and juvenile delinquency. Juveniles who have more friends are delinquent are hypothesized that are more likely to be delinquent. The data used secondary dataset： a larger evaluation of the Gang Resistance Education Training （G.R.E.A.T.） program. The study examine whether being a gang member or not differs based on the number of friends they have are good students. In addition， this article analyze whether： there is correlation between the times of being delinquent and the numbers of their friends in gangs. The results suggest that the times of being delinquent are positive correlated to the numbers of close friends in gangs.

【Key words】juvenile； delinquency； peer； gangs

【作者簡(jiǎn)介】王茗涵，江西中醫(yī)藥大學(xué)人文學(xué)院。

Introduction

Juvenile have been overrepresented in both property and violent crimes （Blumstein， 1995）. Social Learning， Control， and Strain theory are developed to explain why juveniles are prone to be delinquent （Akers R. L.， 1994）. Social Learning theory suggests that juvenile delinquency result from the interaction between young people and older people who engage in deviant behavior （Akers， Krohn， Lanza-kaduce， Radosevich， 1979）. Akers explicitly argued that his learning theory contains an important element of rationality and deliberation， indicating that being delinquency is affected by current norms and values one holds. He also indicates that being delinquency or not is also based on the anticipated future rewards and punishments result from different action choices that stem from their direct or friends’ experiences.

This study focuses on analyzing the relationship between peers and juvenile delinquency. Juveniles who have more friends are delinquent are hypothesized that are more likely to be delinquent.

Method

The data used secondary dataset： a larger evaluation of the Gang Resistance Education Training （G.R.E.A.T.） program， a prevention program aimed at middle school students （Esbensen， Evaluation of the Gang Resistance Education Training （GREAT） Program in the United States， 2003）. Cross-sectional study is used in this survey， which selected 5935 students from different schools （Esbensen Osgood， Research in Brief： National Evaluation of G.R.E.A.T ， 1997）

Variables

Dependent variable： variable “Times being delinquent” is an interval variable indicating how many times a student has ever been delinquent. This study measured how often the respondents engaged in 14 different deviant behaviors， ranging from minor acts such as shoplifting to more serious acts such as robbery. This variable also used as the dependent variable in the linear regression analysis. Additionally， independent variable： 1） variable “Friends are good students” is an ordinal variable indicating how many of the students’ current friends have been thought of as good students 2） variable “Ever been gang members” is a dummy variable indicating whether the student has ever been a gang member or not 3） variable “Close friends in gangs” is an interval variable indicating how many close friends of those students are in gangs. This variable also used as the independent variable in the linear regression analysis. Last， the control variables are gender， race and age in the linear regression analysis. Gender has been recoded into a dummy variable： 1 indicates male and 0 indicates female. Race has also been recoded into a dummy variable： 1 indicates black people and 0 indicates non-black people. Age is a scale variable in this study. The mean of age in this sample is 13.79 since the survey used in this study is only including eight-grade students in order to analyze juvenile delinquency. The gender distribution is almost half and half， which is similar to the population of juveniles.

Moreover， it is about 24.9% of participants in this study are black people. And half of the students participated in the study have more good student friends than bad students friends. A large proportion （86.8%） of the students have never been in gang.

The mean of times being delinquent is 38.85. However， the standard deviation is pretty large （298.8）， which means the data is not concentrated around the mean. It is great possible that the outliers are existed in the data. The mean of number of close friends in gangs is 3.41. The standard deviation is also not small， which means it is also possible that the outliers are existed.

Result

First， chi-square analysis is used to find out whether being a gang member or not based on the number of friends they have are good students. Then， correlation analysis is used to determine the correlation between the times of being delinquent and the numbers of their friends in gangs. Last， linear regression analysis is adopted in this study to test how much the numbers of students’ friends in gangs could affect the times of being delinquent after controlling gender， race， and age.

Comparing the obtained chi-square value with the appropriate chi-square value， the table shows that X2=180.79 Obtain X2 > Table X2 （5.99）. Therefore， the obtained chi-square value above the tabled critical value of 5.99. We are able to reject the 1 hypothesis. We found statistically significant evidence that being a gang member or not differs based on the number of friends they have are good students.

The most obvious difference between the observed value and expected value of the variable of Friends Good Student focused on 1） for the students who have been gang members， the expected number of “none or few of their friends are good students” is less than the observed number 2） for the students who have been gang members， the expected number of “all or most of their friends are good students” is more than the observed number.

The students who have been delinquent for more than 2000 times in the past 12 months and the students who have more than 200 close friends in gangs may answer the questionnaires dishonestly. Thus， the cases that either include more than 2000 times being delinquent or more than 200 close friends in gangs would be excluded in the analysis.

Discussion and Conclusions

We found statistically significant evidence that being a gang member or not differs based on the number of friends they have are good students. Moreover， the correlation analysis suggests that the times of being delinquent are positive correlated to the numbers of close friends in gangs. It means the more numbers of close friends in gangs， the more times a student will be delinquent. Also， the more times a student has been delinquent， the more numbers of close friends in gangs they have. However， there is a weak relationship between the times of being delinquent are positive correlated to the numbers of close friends in gangs.

According to linear regression analysis， three independent variable： “close friends in gangs”， “male” and “age” could statistically significantly predicts the dependent variable-the numbers of close friends in gangs. Nevertheless， “race” could not statistically significantly predict the dependent variable-the numbers of close friends in gangs. Besides， the independent variable “ close friends in gangs” is the strongest predictor in this linear regression model to predict the dependent variable-the numbers of close friends in gangs.

This study has a lot of limitations. There are a lot of variables are related to Juvenile Delinquency， like Parental attachment and Social Economic Status. However， this study only includes their friend network in analysis model. It is better for the future study to randomly assign control group in order to control other related variables that affect juvenile delinquency.

References：

[1]Akers，R.L.（1994）.Criminological Theories.Los Angeles：Roxbury.

[2]Akers，R.，Krohn，M.D.，Lanza-kaduce，L.，Radosevich，M.（1979）.Social Learining and Deviant Behavior：A Specific Test of a General Theory.American Sociological Review（44），636-655.

[3]Blumstein，A.（1995）.Violence by Young People：Why the deadly Nexus？.National Institute of Justice（229），2-9.

[4]Esbensen，F(xiàn).-A.（2003）.Evaluation of the Gang Resistance Education Training（GREAT）Program in the United States.Omaha：University of Nebraska at Omaha.