The Tough Truth: What Impacts a Child's Violent Behavior?

Brittany White

 

 

            Aggression and violent tendencies are present in every human being.  In order for a person to act on these instincts, the right circumstances must be present.  For some people, a traumatic or life-altering event needs to occur before they react in a violent way.  For others, however, a slight disagreement or minor upset may invoke these behaviors (Collins, 2009).  Behaving violently toward one's self or others can start at a very young age and most times only gets worse as a child grows up, if the problem is not dealt with appropriately.  Many theorists have studied the effects of growing up in a violent household and what role that plays in a child's development.  Little research has been done, however, that investigates the role that relationships with teachers, parents and peers play in a child's likelihood to act violently. 

            Most researchers would agree that a child's home life is an obvious contributing factor to violence but most focus on the aspect of abuse and not lack of or hindrance in parental involvement (Widom, 1989).  Many parents do not take time to get involved in their child's life or the parent has some psychological issue that is impairing their involvement, but does this cause a child to be more aggressive?  Many children who have anger and aggression issues come from families that would not be considered typical for risk factors for violence.  Another factor that could influence a child's likelihood of being violent is relationships at school, including teachers and peers (O'Donnell, Schwab-Stone, and Muyeed, 2003).

            Parental involvement, teacher's interaction and peer influences play an intricate role in a child's life.  Could these relationships, if hindered in some way, result in a child acting out violently?  This research will use the Kentucky Youth Survey to analyze the correlation between parental involvements and tendencies toward violence.  This paper will examine factors like peer to peer interaction and how this affects a student's view of themselves and others around them.  A correlation between bad or suffering relationships, because of various factors, and violence is the expected outcome. 

 

Literature Review    

            Many researchers have used Hirschi's Bonding Theory to look at different aspects of a person's life and see how they impact their likelihood to be deviant.  The four basic elements of social bond theory are attachment, commitment, involvement in conventional versus deviant or criminal activities, and lastly the common belief system within an individual's society or subgroup (Criminology Wiki 2010). Hirschi's theory points out that attachment is especially important when it come to the person's parental figures.  This theory suggests that conventional figures, such as parents, when bonded make a huge impact in the deterrent of criminal acts (Criminology Wiki 2010).

            One experiment and article, in particular, looked at how parental involvement affected how likely a child is to be a bully in school (Holt, Kaufman, and Finkelhor 2009).  Holt, Kaufman, and Finkelhor (2009) state, 'Bullying perpetration and victimization rates were higher when reported by students than parents, and parents were particularly unaware of their children bullying others (42).'  The authors also added, 'Findings highlight the need to increase parental awareness about bullying and to include parents in school-based bullying prevention programs (43).'  To summarize, home life for both victims of bullying and the bullies themselves, the authors concluded 'Victims' homes were characterized by higher levels of criticism, fewer rules, and more child maltreatment; bullies' homes were characterized by lack of supervision, child maltreatment, and exposure to domestic violence (43).'  Another study confirmed a correlation between maternal behavior and a child's likelihood to be a bully (Georgiou and Fanti 2010).  The study did not, however, discuss the parent/child relationship, only how what a child sees the parent do can send the wrong messages to the child.  The findings also showed that increases in bullying behavior on the part of the child cause a reduction in maternal involvement and monitoring over time.  Finally,  the findings of this study offer support to the hypothesis that parental and child behavior reciprocally influence each other, rather than one shaping the other in a unidirectional fashion (Georgiou and Fanti 2010).

            Maladaptive behaviors by the parents also play a role in a child's delinquency status.  Eiden, Ostrov, Colder, Leonard, Edwards and Orrange-Torchia (2010) examined the association between parents' alcoholism and peer bullying and victimization in middle childhood in community recruited families with and without alcohol problems.  The study states, 'There was a direct association between fathers' alcohol symptoms and bullying of peers...(341)'  Multiple group models indicated that the direct association between parents' alcohol symptoms and bullying was significant for boys but not girls (Eiden, Ostrov, Colder, Leonard, Edwards and Orrange-Torchia 2010).   In most cases, the father was the one to abuse the alcohol and in households with mixed-sex children, the sons would display more violent tendencies than female children.  A related study, done by Jeynes (2008) investigated the relationship between parental involvement and their children being picked on or bullied and being discriminated against based on their race during their elementary and secondary school years.  The influence of parental involvement on academic achievement during that period was also examined (Jeynes 2008).  As stated in the article, 'Specific aspects of parental involvement and an overall parental involvement variable were examined (255).'  The study concluded that it does appear that parental involvement could be related to a lower incidence of children being picked on and discriminated against and also a lower likelihood for a child to be a bully themselves (Jeynes 2008).

            Another important area of a child's life takes place at school, that is where he or she spends most of their waking time.  According to Flaspohler, Elfstrom, Vanderzee, Sink, and Birchmeier (2009) bullying is one of the most prevalent forms of school violence.  'Engagement in bullying has been shown to have adverse effects on perpetrators and victims of bullying (636).'  In this study, the effects of bullying on the well-being of junior high and high school students was examined.  As stated in the article, 'Results suggest that students who bully and/or are bullied experience reduced life satisfaction and support from peers and teachers compared to 'bystanders' (children who are neither victims nor perpetrators of bullying) (637).'  The study also suggests that teacher and peer support might lessen the severity and occurrence of bullying and promotes an intervention of teachers to prevent the harsh effects of bullying (Flaspohler, Elfstrom, Vanderzee, Sink, and Birchmeier 2009).  Another article (Moore 2009) reviews the book, "Children and bullying: How parents and educators can reduce bullying at school", emphasizes the impact parents and educators have on the occurrence of bullying in schools and also provides resources to prevent bullying situations.  Moore (2009) states that parents and teachers do have an impact on the delinquency of children but no research was done on the level of impact.  Finally, a unique study (Allen 2010) was used to look at how classroom management can help to reduce the level of school violence in a given area.  Allen (2010) stated, 'Research from a number of fields suggests that several variables conspire to create environments where bullying is more likely to occur.  These include harsh and punitive discipline methods, lower-quality classroom instruction, disorganized classroom and school settings, and student social structures characterized by antisocial behaviors (1).'  The study suggested that certain measures needed to be taken to ensure that classrooms were adequately able to prevent and respond to bullying, in a timely manner.

            All of the research given had similar results, parental behaviors and parent/child involvement do have an impact on the degree of delinquency a child partakes in but none of them specifically focused on relationships.  For example, a parent can be highly involved with their child but still have a struggling relationship and a parent can not be as involved and have a good relationship with their child.  I am hoping to shed some light on positive and negative relationships between not only parents and children but also peers and teachers.  I am hoping to show a correlation between children behaving violently and their relationships with people closest to them.

 

Data

            The data used to estimate the influence relationships with peers, parents and educators have on a child's likelihood to bully or behave violently toward another or be victimized came from the Kentucky Youth Survey, a project that focused on students in grades 6 through 12.  The survey questioned students about different habits such as smoking, drinking and drug activities and also the student's attitudes toward school, violence, relationships and religious activities.

            The data from the survey was collected in spring of 1996 and taken from several heavily populated counties throughout Kentucky.  Various public schools were targeted and the respondents for the survey represent all students grades 6 through 12 on the day the survey was administered.  The questionnaires were completely anonymous and self-administered by the teachers of each classroom.  The survey took about 45 minutes to complete and parent's and student's consent was obtained before the survey was taken.

            Even though the study was not intended to be sample-based, only 52 percent of all students enrolled in the 22 school completed the survey.  One school alone, however, was the main cause for such low responses, where the response rate was less than 25 percent.  The school was eliminated from the sample analyzed, lowering the total number of student responses from 13,349 to 13,255.  Response rates for the rest of the schools averaged 56 percent and ranged from 25 percent to 84 percent.  Nonresponse in the 21 schools is caused by many factors including absent students, some refusals by students and/or parents and conflicting events.

            The sample does not appear to differ from the entire population of results in the sampled schools, despite the low response rate.  The sample also appears to be unbiased regarding specific school-level demographic characteristics.  For instance, the survey was taken by a mean proportion nonwhite of .34 with a range of .22 and .54.  The correlation between the school-level proportion nonwhite figures is .97.  The Department of Education data show that 51 percent of students in the selected Kentucky schools were male and the students' average age was 14.8.  The survey-based  sample statistic for school-level age is 14.6.

            Almost 1,000 cases were deleted from the 13,255 completed surveys due to missing data on nominal-level measures.  Missing values on ordinal and continuous variables were replaced with school-level mean values.  The total usable sample was 12,343, yet because of space constraints a 50 percent random sub-sample was employed.  Although some cases were dismissed, the final sample of cases is similar to the total number of students surveyed and one would expect the outcomes and correlations to be similar as well.  However, because of software constraints and to ease my analysis, my final sample size is a randomly chosen 1500 cases.  My hypothesis about the findings of my analysis is that the closer a respondent is to their parents and teachers; the less likely they are to behave violently and to use weapons.  On the contrary, the closer a respondent is to their friends, the more likely they are to behave in deviant manners.

 

Measures

Dependent Variables

This study examines two types of outcomes: tendencies to act violently and weapon usage.  The attitude questions all feature Likert-item response categories (strongly disagree, disagree, agree, strongly agree), which I recoded as dichotomous variables where 1=agree or strongly agree and 0=disagree or strongly disagree.  I analyzed respondents' responses to the following 7 statements about acts of violence against another person and 4 statements about weapon usage:

Violent Tendencies

1.      Ever shoved or tripped someone.

2.      Ever sat on someone or pinned someone.

3.      Ever hit, punched, or slapped someone.

4.      Ever hit someone with object.

5.      Ever pulled, twisted, squeezed, or pinched part of other's body.

6.      Ever laid trap for someone so that he/she would get hurt

7.      Ever been in a fist fight.

Weapon Usage

1.      Ever used weapon during a fight.

2.      Ever threatened to hurt someone with a weapon.

3.      Ever been threatened with gun

4.      Ever used a gun to threaten someone.

These variables were then used to create two different scales: a violence scale (using the 7 variables listed above) and a weapons scale (using the last 4 variables).

 

Independent Variables

The three independent variables for this study are relationships with parents, relationships with peers and relationships with teachers.  The attitude questions all feature Likert-item response categories (never, almost never, sometimes, most of the time), which I recoded as dichotomous variables where 1=sometimes or most of the time and 0=never or almost never.  I analyzed respondents' responses to the following 6 statements about their relationships with parents, 5 statements about relationships with peers, and 4 statements about relationships with teachers:

Parent Relationships

1.      How often my parent(s) seem to understand me.

2.      How often my parent(s) make rules that seem fair.

3.      How often my parent(s) know where I am when I'm not at home.

4.      How often my parent(s) know how I am with when I'm not at home.

5.      How often my parent(s) is concerned with how I am doing in school.

6.      How often I share my thoughts and feelings with my parent(s).

Peer Relationships

1.      Would like to be liked by my best friends.

2.      I respect the opinion of my best friends.

3.      I fit in well with my best friends.

4.      My best friends take interest in my problems.

5.      I feel close to my best friends.

Teacher Relationships

1.      Care what teachers think of me.

2.      Most of my teachers are interested in what I say and do.

3.      I would quit school now if I could.

4.      I look forward to coming to school.

These variables were then used to create three different scales: a parents school (using the 6 variables listed above), a peers scale (using the middle 5 variables listed above) and a teachers scale (using the last 4 variables listed above).

 

Individual-Level Control Variables

The first variable I chose to control for was the respondent's race.  Violent tendencies and weapon usage are known to vary by race, so I thought that this variable was definitely important to include in my analysis.  Age is another variable I chose to control for because, it is also known that age greatly impacts suggestibility and this could, in fact, lead to higher tendencies toward violence and weapon usage.  The final variable I chose to control for was gender, because as most leading studies have shown, males are more likely to be violent than females.

Variable	Metric	Mean	S.D.	Range
Violent Tendencies	(0=low violent tendencies', 7=high violent tendencies)	3.66	2.35	0-7
Weapon Usage	(0=no weapon usage', 4=heavy weapon usage)	.52	1.00	0-4
Relationship w/ Teachers	(0=poor relationships',4= good relationships)	1.03	1.06	0-4
Relationship w/ Parents	(0=poor relationship', 6=good relationship)	3.89	2.36	0-6
Relationships w/Peers	(0= poor relationships', 6=good relationships)	4.55	1.75	0-6
Sex	(0=female, 1=male)	.509	.500	0-1
Race	(0=non white, 1=white)	.854	.353	0-1
Age	(number of years)	14.07	1.99	10-20

 

Results

Table 2 reveals that respondents violent behaviors are greatly impacted by relationships with parents, in a negative correlation, meaning that the closer the respondent was to their parent(s) or the more the parent(s) was involved in their child's life, the less likely (-.129) the child was to behave violently toward others.  The table also revealed that the closer a respondent was to their friends; the more likely (.032) they were to act out in a violent manner.  Finally, when it came to violence surprisingly, teachers had a positive influence (.333), on the respondent's behaviors; meaning that the closer the respondent was to their teacher, the more likely they were to act violently toward others.  Relationships with peers positively (.366) affected a respondent's relationship with parents, meaning that the closer the respondent was to their friends the more distant they were to their parents.  Relationships with peers negatively (-.030) affected a respondent's relationship with teachers, meaning that the closer they were to peers, the closer they were to their teachers as well.  Finally, relationships with teachers negatively affected (-.131) their relationships with parents as well.

Table 3, which is my OLS Regression Models, also reveals that parents (-.099) negatively influence violent tendencies, whereas teachers (.678) and peers (.169) positively influence a respondent's likelihood to act violently.  Also, looking at my control variables, males are more likely to behave violently than females and non whites (-.781) are more likely to commit deviant behaviors than whites.

 

	1	2	3	4	5	6	7
1. Weapon Usage
2. Relationships w/ Peers	-.083**
3. Relationships w/ Parents	-.176**	.366**
4. Relationships w/ Teachers	.257**	-.030	-.131**
5. Violent Tendencies	.470**	.032	-.129**	.333**
6. Race	-.120**	.127**	.093**	-.040	-.111*
7. Age	.069**	.104**	.108**	.130**	.055**	.000
8. Sex	.215**	-.180**	-.165**	.106**	.243**	.015	.009

Table 2: Zero-Order Correlations Among Study Variables

	Weapons			Violence
	Coeff.	S.E.		Coeff.	S.E.
Age	.025*	.01		.014*	.03
Sex	.345	.02		1.04	.16
Race	-.282*	.07		-.781*	.16
Peer Relationships	.004*	.02		.169*	.04
Parent Relationships	-.051*	.01		-.099*	.03
Teacher Relationships	.194*	.02		.678*	.05

Table 3: OLS Regression Models Examining the Effects of Social Relationships on Violent Behaviors