DIFFERENCES IN EXECUTIVE FUNCTIONING i

Title Page

Differences in Executive Functioning Between Current and Former Gang Members

Dissertation

Presented to the

Graduate Faculty of the

Clinical Forensic Psychology Program

California School of Forensic Studies, Irvine

Alliant International University

In Partial Fulfillment

of the Requirements for the Degree of

DOCTOR OF PSYCHOLOGY

By

Heather Farris

January 2013
 UMI Number: 3557654

All rights reserved

INFORMATION TO ALL USERS

The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.

UMI 3557654
Published by ProQuest LLC (2013). Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code

ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106 - 1346
DIFFERENCES IN EXECUTIVE FUNCTIONING ii

Copyright page
DIFFERENCES IN EXECUTIVE FUNCTIONING iii

Signature page

Differences in Executive Functioning Between Current and Former Gang Members

ALLIANT INTERNATIONAL UNIVERSITY, Irvine

The dissertation of Heather Farris, directed and approved
by the candidate’s Committee, has been accepted by the
Faculty of the California School of Forensic Studies
In partial fulfillment of the requirement for the Degree of
DOCTOR OF PSYCHOLOGY

________________________
DATE

Dissertation Committee:

________________________
Noor Damavandi, Ph.D., Chairperson
Alliant International University

__________________
Daniel Levinson, Ph.D., Committee Member
Alliant International University
DIFFERENCES IN EXECUTIVE FUNCTIONING iv

Acknowledgements

First, I would like to express my gratitude to the volunteers at Homeboy

Industries for their participation in this research. You are all an inspiration. Without

your contribution, my research efforts would not have been possible. I would like to

expressly thank James Horton at Homeboy Industries for his diligence and hard work.

Without your assistance, this most certainly would not have been possible. Homeboy

Industries does amazing work, and I hope my journey brings me back there one day.

To Jami, few words can ever express what your friendship means to me. I am so

grateful for your honesty and willingness to tell me what I needed to hear no matter what.

I wish more people were able to experience what it is like to have a truly unconditional

friendship. Your words stuck with me even when our schedules and distance kept us

from catching up for periods of time. Your wisdom and strength carried me through so

much of this process. For that, and so much more, I am forever grateful. You are my

ride-or-die friend. I love you.

To Rose, Robyn, and Alana, I could not have gotten through this process without

your friendship. I will always hold close our conversations, whether laughing or crying.

I am fortunate to have gained three new lifelong friends. We did it!

I would like to thank my brother, Jason, and sister-in-law, Denise, for your love

and friendship. Life keeps us busy, but I am grateful to you for always keeping me close

in your hearts and extending invites for get-togethers and getaways. They are just what I

needed.
DIFFERENCES IN EXECUTIVE FUNCTIONING v

To Mom and Dad, thank you for being a part of my dream. I attribute my

successes to your unwavering support. You have always been there for me, guiding me,

but letting me find my own way. You are the epitome of devotion. I hope to one day

devote my life to my child’s in the same ways you have.

Finally, to my new niece Kamila, the sky is the limit, sweet girl. Anything is

possible. Dare greatly. Do not ever limit yourself.

DIFFERENCES IN EXECUTIVE FUNCTIONING vi

Table of Contents

Table of Contents

Title Page ............................................................................................................................. i

Copyright page .................................................................................................................... ii
Signature page .................................................................................................................... iii
Acknowledgements ............................................................................................................ iv
Table of Contents ............................................................................................................... vi
Abstract .............................................................................................................................. ix
CHAPTER I ........................................................................................................................ 1
Overview ............................................................................................................................. 1
Introduction and Background .......................................................................................... 1
Significance of the Study ................................................................................................ 3
Nature of the Study ......................................................................................................... 3
Research Questions ......................................................................................................... 3
Conceptual/Theoretical Framework ................................................................................ 4
Assumptions .................................................................................................................... 4
Scope and Delimitations.................................................................................................. 4
Limitations ...................................................................................................................... 5
Chapter Summary............................................................................................................ 6
CHAPTER II ....................................................................................................................... 7
Literature Review................................................................................................................ 7
Gangs and Delinquent Behavior ..................................................................................... 7
Predictors of Gang Involvement ..................................................................................... 9
Family, Neighborhood, and School Characteristics .................................................... 9
Peer Influences .......................................................................................................... 17
External Factors to Gang Involvement.......................................................................... 19
DIFFERENCES IN EXECUTIVE FUNCTIONING vii

The Neighborhood’s Role ......................................................................................... 19

Socio-Cultural Perspective ........................................................................................ 25
Self-Control ................................................................................................................... 27
Heritability Factors ........................................................................................................ 35
Gene-Environment Interactions ................................................................................. 35
Biology, Self-control, and Delinquency .................................................................... 36
Delinquency Trajectory ................................................................................................. 43
Executive Functioning................................................................................................... 60
Executive Function and Biology ............................................................................... 60
Executive Function and Antisocial Behavior ............................................................ 62
Summary ....................................................................................................................... 81
CHAPTER III ................................................................................................................... 83
Methods............................................................................................................................. 83
Introduction ................................................................................................................... 83
Research Design ............................................................................................................ 83
Appropriateness of Design ............................................................................................ 83
Research Questions and Hypotheses ............................................................................. 84
Population / Participants................................................................................................ 84
Informed Consent .......................................................................................................... 85
Sampling Frame ............................................................................................................ 85
Confidentiality............................................................................................................... 86
Geographic Location ..................................................................................................... 86
Instrumentation / Validity and Reliability..................................................................... 86
Behavior Rating Inventory of Executive Function–Adult Version (BRIEF-A) ........ 86
Wisconsin Card Sorting Test (WCST) ...................................................................... 88
Demographic Questionnaire ...................................................................................... 90
Data Collection .............................................................................................................. 90
Data Analysis ................................................................................................................ 91
Summary ....................................................................................................................... 92
CHAPTER IV ................................................................................................................... 93
DIFFERENCES IN EXECUTIVE FUNCTIONING viii

Results ............................................................................................................................... 93
CHAPTER V .................................................................................................................. 101
Discussion ....................................................................................................................... 101
Limitations .................................................................................................................. 105
Internal Validity....................................................................................................... 105
External Validity...................................................................................................... 107
References ....................................................................................................................... 108
Appendix A ..................................................................................................................... 112
Table 1 ......................................................................................................................... 112
Table 2 ......................................................................................................................... 112
Table 3 ......................................................................................................................... 112
Appendix B ..................................................................................................................... 113
Table 4 ......................................................................................................................... 113
Table 5 ......................................................................................................................... 113
Appendix C ..................................................................................................................... 114
Demographic Questionnaire ........................................................................................ 114
Appendix D ..................................................................................................................... 115
Demographic Table ..................................................................................................... 115
Appendix E ..................................................................................................................... 117
Informed Consent ........................................................................................................ 117
DIFFERENCES IN EXECUTIVE FUNCTIONING ix

Abstract

Executive functioning entails reasoning, self-regulation, and goal-directed

behavior. Deficits in executive functioning may be manifested by impulsivity and

aggression (Blair, 2007; Giancola, Roth, and Parrott, 2006; Dolan & Anderson, 2002). It

is proposed that delinquency in adolescence is a normal activity, understandable from the

perspective of like-minded teens (Moffitt, 1993). Within this theory, it is believed that

delinquency shifts from individual psychopathology in childhood to a normative social

experience in adolescence, and then reverts to a psychopathological state in adulthood.

Most delinquency remits after adolescence; however, a small percentage persists into

adulthood, and some continues for a lifetime. The antisocial behavior of gang members

can be thought of as stable and persistent, and the gang culture breeds an enmeshment

that is difficult to break from; hence, the cycle of antisocial behavior persists. However,

some gang members make a choice to break from the gang, and live a more prosocial life.

This decision and planning can be considered an executive function. The outcomes of

antisocial behavior in midlife have seldom been examined. This research will examine

executive functioning differences between current and former gang members, and what

role it might play in a gang member’s ability to leave the gang. This researcher proposes

there will be significant differences in executive functioning between current and former

gang members, with former gang members exhibiting greater executive functioning

abilities.
DIFFERENCES IN EXECUTIVE FUNCTIONING x
DIFFERENCES IN EXECUTIVE FUNCTIONING 1

CHAPTER I

Overview

Introduction and Background

Executive function (EF), housed in the frontal lobes of the brain, includes the

capacity for the most complex brain functions. In a broad sense, EF is linked to

ambition, purposefulness, foresight, insight, and complex decision making (Goldberg,

2009). More specifically, EF drives reasoning, problem solving, and organization. It

allows a person to process and respond to social cues via self-regulation. Subtle damage

to the frontal lobe can cause apathy and indifference (Goldberg, 2009). More important

to society, EF damage has been linked to aggressive behavior, difficult temperament, and

impulsivity (Blair, 2007; Dolan & Anderson, 2002; Giancola, Roth, and Parrott, 2006).

Gangs pose a different construct from mainstream society for the individuals

drawn to them. When complicated external factors, such as socio-economic difficulties,

discrimination, and community violence are normal daily concerns for individuals,

changes in cognitive functioning are likely to occur in order to function in this

environment. An important question is whether changes in cognitive functioning occur

prior to an individual joining a gang, which enables him to transition into antisocial

behavior, or whether these changes occur after joining the gang, as a result of antisocial

activity. Additionally, what bearing does the timing of these cognitive deficits have on

potential for rehabilitation?

As a young adult gang member begins to question his or her place in the world,

the jump from contemplating criminal acts versus prosocial activities (e.g., job,
DIFFERENCES IN EXECUTIVE FUNCTIONING 2

responsibilities), entail planning and a degree of cognitive flexibility. The executive

functioning to entertain these prosocial ideas must originate somewhere. Theories

abound regarding delinquency that extinguishes in adolescence and delinquency that

persists throughout the life course (Moffit, 1993; Piquero, 2001; Raskin, White, Bates, &

Buyske, 2001). Specific to gangs, this study will examine differences in executive

functioning between former and current gang members, and determine whether they can

fit into the delinquency trajectory theory.

Problem Statement

Among the complex factors that contribute to gang membership, executive

functioning may be a critical factor in gang initiation, as well as gang membership

continuity. Delinquency trajectory theory posits that the majority of adolescents who

engage in delinquent behavior will outgrow this behavior by late adolescence or early

adulthood (Moffitt, 1993); however, it is still a decision, which falls under the umbrella

of executive function. Factors that lead to a gang member’s decision to leave should be

further researched so that executive function can be better understood in its role in

delinquent behavior in adolescence and early adulthood.

Purpose Statement

The discontinuance of antisocial behavior in midlife has seldom been examined.

In addition, studies on antisocial behavior have not been specific to gangs or specific

factors that contribute to a young male exiting this lifestyle. The purpose of this study is

to examine executive functioning differences between current and former gang members

in the crucial developmental stage of late adolescence/early adulthood, when the frontal

lobe reaches full development. This researcher hopes this study will contribute additional
DIFFERENCES IN EXECUTIVE FUNCTIONING 3

data to the understanding of executive functioning and its role in antisocial behavior in

late adolescence and early adulthood, as well as the causes that may lead this behavior to

remit or not remit, during this developmental stage.

Significance of the Study

Differences in executive functioning between current and former gang members

may have a more critical role than supporting the delinquency trajectory theory of falling

into either adolescence-limited or life-course persistent delinquency. If it is determined

that there are significant differences in executive functioning, it may support the idea that

improving or rehabilitating executive function may decrease the likelihood of antisocial

behavior (Giancola et. al., 2006). This may lead to a paradigm shift in the treatment of

antisocial behavior throughout the life span in gang members and other delinquents.

Nature of the Study

This study is a quasi-experimental design comparing the performance on

executive function measures among current gang members and former gang members.

Research Questions

Current literature suggests executive dysfunction contributes to impulsivity,

aggressiveness, and delinquency (Blair, 2007; Dolan & Anderson, 2002; Giancola et. al.,

2006). Once engaged in this line of behavior, what factors may exist that allow a person

to make a change in their life course trajectory? The research question is as follows:

1. Are there significant differences in executive functioning between current

gang members and former gang members?

DIFFERENCES IN EXECUTIVE FUNCTIONING 4

Conceptual/Theoretical Framework

Prior research indicates a relationship between executive functioning and

antisocial behavior, specifically aggression and impulsivity (Blair, 2007; Dolan &

Anderson, 2002; Giancola et. al., 2006). This researcher theorizes there will be

significant differences in executive functioning, with the former gang members having

greater executive functioning abilities than the current gang members. This will confirm

prior studies that adolescence-limited delinquency has less executive functioning deficits

than life-course persistence delinquency, significant differences among adolescent-

limited delinquents and life-course persistent delinquents, and executive function

(Donnellan, Ge, & Wenk, 2000; Huesmann, Dubow, & Boxer, 2009; Raine, Moffit,

Caspi, Loeber, Stouthamer-Loeber, and Lynam, 2005).

Assumptions

This study has the following assumptions:

1. The BRIEF-A and Wisconsin Card Sorting Test (WCST) are valid

and reliable measures of executive functioning.

2. All participants will have current or former involvement in a gang.

3. All participants will make an effort to respond to all test items to

the best of their ability.

Scope and Delimitations

The scope of this study encompasses executive functioning levels of current gang

members compared with former gang members. Delimitations which will narrow the

scope of the study include the self-report measures used to capture gang membership as
DIFFERENCES IN EXECUTIVE FUNCTIONING 5

opposed to recorded data of gang affiliation, as well as the limited scope of

neurocognitive measures being used. In addition, the participants recruited will be from

one non-profit organization that assists with gang intervention and prevention; in essence,

the participants will be in a frame of mind that is open to assistance. Although some may

still be in the gang, they are seeking support, therefore, the scope of current gang

affiliation will be narrower than someone enmeshed in the gang and not seeking help. In

addition, many life-course persistent offenders may not be available because they are

incarcerated or deceased.

Limitations

Limitations of the study include:

1. The arena of available participants is limited. One non-profit

organization will be contacted for voluntary participation by its

affiliates/volunteers/employees. For safety precautions, true current gang

members may not be able to be assessed.

2. Executive functioning will be assessed with participants in adulthood;

therefore, assumptions will be limited as to executive dysfunction that

existed prior to gang affiliation.

3. Substance and alcohol abuse will be limitations in that the use of these

substances may have contributed to prefrontal cortex damage over time.

4. Head trauma may also be a limitation in that the lifestyle of gang

members may invite increased risk for altercations and possible head

trauma, which may impact frontal lobe and executive functioning.

DIFFERENCES IN EXECUTIVE FUNCTIONING 6

Chapter Summary

Executive functions include the capacity to plan, reason, organize thoughts, and

control behavior. Deficits in executive function may contribute to impulsivity and

aggression (Blair, 2007; Dolan & Anderson, 2002; Giancola et. al., 2006). These deficits,

as well as psychosocial factors, heritability factors, and delinquency trajectories, all play

an intricate role in the antisocial behavior associated with gangs as suggested in the

literature to follow. This study will examine whether differences in executive function

exist between former gang members and current gang members.

DIFFERENCES IN EXECUTIVE FUNCTIONING 7

CHAPTER II

Literature Review

Gangs and Delinquent Behavior

The normative view of gangs tends to follow the idea that they are outlaw

organizations steeped in a culture of deviance that condones and promotes antisocial

behavior (Gibbs, 2000). Gang members can generally be diagnostically classified as

having conduct disorder, and having antisocial personality disorder as their behavior

extends into adulthood. Conduct disorder is defined as a repetitive and persistent pattern

of behavior in which the basic rights of others or societal norms or rules are violated

(APA, 2000). These behaviors include aggression towards people or animals, destruction

of property, deceitfulness and theft, and a serious violation of rules that may include

truancy, running away from home, or staying out past curfew.

Gang members can be attributed to having the majority of these behaviors, in

addition to risk factors identified by Maxmen, Ward, and Kilgus (2009), as always being

in trouble, poor frustration tolerance, irritability, frequent temper outbursts, as well as

having inadequate supervision, inconsistent discipline, and having an environment that

offers more temptations. As a gang member grows into adulthood and continues to be

involved in the gang, a diagnosis of Antisocial Personality Disorder would include a

pervasive pattern a disregard for the rights of others as well as a combination of

impulsivity, lack of remorse, a failure to conform to social norms, and consistent

irresponsibility (APA, 2000). According to Maxmen et al. (2009), estimates place the
DIFFERENCES IN EXECUTIVE FUNCTIONING 8

likelihood that those with conduct disorder will develop Antisocial Personality Disorder

at around fifty percent.

Some gang members enter the gangs with these characteristics; others develop

them by association. Gangs are known to participate in activities including stealing,

lying, truancy, participating in graffiti, selling drugs, robbery, auto theft, and assault. In

their annual gang threat assessment, the National Gang Intelligence Center (NGIC, 2009)

identified drug trafficking as the most frequent gang-related activity, but noted human

and weapons trafficking as activities that are increasing in frequency. As of September

2008, approximately one million gang members belonging to more than 20,000 gangs

nationwide were criminally active (NGIC, 2009).

Some of the current literature (e.g., Craig, Vitaro, Gagnon, & Tremblay, 2002;

Hill, Howell, Hawkins, & Battin-Pearson, 1999; Lahey, Gordon, Loeber, Stouthamer-

Loeber, & Farrington, 1999) measures factors that predict gang involvement as well as

differences between gang members and non-gang members; however, it is possible that

some of these differences may arise after a youth has entered gangs and may not be a

contributing factor in their decision to join (Lahey et al., 1999). Gangs and gang

members cannot simply be categorized as conduct disordered and reactively addressed in

the criminal justice system and society. There are other factors that must be looked at

holistically; it is important to question other influences, both psychosocial and biological,

that lead youth to joining gangs and participating in antisocial gang behavior.

When an adolescent joins a gang, they acquire an identity based on their

involvement in street life and delinquent behavior (Valdez, Kaplan, & Codina, 2000).

Adolescent gang identities are complex and range from normal to extreme antisocial
DIFFERENCES IN EXECUTIVE FUNCTIONING 9

behavior. There are many reasons non-antisocial youths may join gangs, which may

include self-esteem, power, and protection, but then are encouraged to participate in

antisocial behavior by the group after joining (Lahey et al., 1999). There are some gang

members with normal psychological characteristics who experience the gang as an

extension of neighborhood based peer groups, and participate in deviant behaviors as a

means of maintaining their gang status (Valdez et al., 2000). Children growing up in

these communities come to know the world as dangerous and view aggression and deceit

as the safest and most effective response to situations. There are youths on the other end

of the spectrum who are predisposed to violence, and whose behavior is rewarded within

the gang hierarchy (Valdez et al., 2000).

The causal mechanisms of gang membership are multiple and complex. Some

theories are more developed than others, and each have a psychosocial or biological

component. It is likely a multitude of factors play a role in a person’s draw towards

gangs. The question remains why some youth continue with the gang life and why some

leave. More importantly is how these individuals come to the decision to leave the gang

life, and what cognitive processes take place that might differentiate them from those that

are unable to or do not desire to leave.

Predictors of Gang Involvement

Family, Neighborhood, and School Characteristics

Considering development from a contextual perspective, some of the literature

examines the microsystems a child experiences in everyday life and how they might

influence gang involvement. Hill, Howell, Hawkins, and Battin-Pearson (1999)

postulated that delinquency, violence, and substance abuse are not often found to be
DIFFERENCES IN EXECUTIVE FUNCTIONING 10

synonymous with gang membership; however, predictors of these behaviors provide a

starting point for examining the predictors.

Using multiple self-report measures completed by the youth, their parents, peers,

and teachers, Craig, Vitaro, Gagnon, and Tremblay (2002) addressed multiple objectives

in their study. In particular, they examined the stability of gang membership between the

ages of 10 to 14; compared the family, behavioral, peer and school profiles among stable,

unstable, and non-gang members; and identified grade achievement among the

participants.

The subjects of the Craig et al. (2002) study were originally part of a longitudinal

study in a low socioeconomic area in Quebec, Canada that included 1,034 boys that were

in kindergarten in 1984. To control for culture, both parents of the boys had to be born in

Canada and their first language be French. A subsample of 142 boys was selected for

this study. The boys were divided into three groups: stable gang members, identified as

children who belonged to a gang at ages 13 and 14; unstable gang members, identified as

children who belonged to a gang at either age 13 or 14; and non-gang members. There

were a disproportionate number of non-gang members, so a random sample was selected

for this comparison group, to provide a total of 66 non-gang members, 25 stable gang

members, and 51 unstable gang members.

Measuring the stability of gang membership year over year, Craig et al. (2002)

employed four loglinear models using the self-report data. Each of the models between

each age was found to be non-significant, with the exception of ages 13 to 14, which

were found to have a significant association. They determined the probability for gang

members at age 13 to belong to a gang at age 14 was .49, and were able to define stable
DIFFERENCES IN EXECUTIVE FUNCTIONING 11

gang members as belonging to a gang at ages 13 and 14. These results were the rationale

for the identification of each participant group. This prompted the researchers to question

what factors may be at play at 13 and 14 that cause this shift from inconsistent

membership and occasional association with the gang to complete immersion.

With their self-report measures, Craig et al. (2002) found that at ages 10 and 11

there were no significant differences between the behavior of the unstable gang members

and non-gang members in their sample. Their research suggested that as adolescents

engage themselves in gang activity at the ages of 10 and 11, they are not exhibiting

significant antisocial behavior different from those of non-gang members; however, their

levels of deviant behavior increase as membership in the gang becomes more stable.

They found that gang members are not uniformly deviant, but their delinquent behavior

escalates as they engage further with the gang. This is important for future research to

determine what factors at this age lead a person who does not exhibit more antisocial

behavior than a non-delinquent, into a gang.

In their research, Craig et al. (2002) also examined the difference between stable

gang members and non-stable gang members, and differences in their behavior profiles.

They hypothesized that stable gang members have more behavioral problems than the

non-stable gang member. At the ages of 12, 13, and 14, the stable gang members scored

higher than unstable and non-gang members, on fighting, stealing, and vandalism. From

ages 11 to 14, the stable gang members reported more school deviance than the unstable

gang members, and stable gang members reported more friend delinquency than unstable

gang members and non-gang members. This suggests the more behavioral problems an

individual has going into a gang, the more likely they are to become a stable gang
DIFFERENCES IN EXECUTIVE FUNCTIONING 12

member. Craig et al. (2002) determined that as stable and unstable gang members age,

they engage in wider range delinquent behavior than non-gang members, and the

frequency of their delinquent acts increase with age.

Another antecedent that leads to gang membership is difficulty in school, and

later dropping out completely. Craig et al. (2002) performed a chi-square analysis to

examine grade attainment among distributions of the three groups. They found a

significant relationship between gang membership and grade achievement at age 14; p =

.03. 59% of stable gang members were behind in school, versus 47% of unstable gang

members and 41% of non-gang members. Craig et al. (2002) suggested that high dropout

rates among gang members is due to the fact that school is an arena in which individual

deficits may become apparent, and dropping out to hang with a peer group where these

deficits can be hidden, suggests an early sign of alienation.

Craig et al. (2002) suggested that there is both continuation and accumulation of

deviant behavior problems with development. The pathway to stable gang membership

may include antecedent behavior problems, as well as low levels of anxiety and failure to

learn prosocial alternatives to antisocial behaviors. Craig et al. (2002) found that both

stable and unstable gang members had antecedent behavior problems in a variety of

contexts as rated by self, parents, and teachers; however, unstable gang members

exhibited fewer problematic behaviors than stable gang members, which suggested that

these two groups may have different developmental trajectories. These differing

trajectories may be due to unstable gang members being on the outside of the antisocial

peer group of gang members. They are not “street” enough, but have to portray this

image to survive in their neighborhood (Anderson, 1999). They may have a tendency
DIFFERENCES IN EXECUTIVE FUNCTIONING 13

towards more prosocial behaviors and peer groups, but view the gang as a way to

maintain friends and status in their neighborhoods. Their activities are for appearances,

not to fulfill violent tendencies (Gibbs, 2000).

While Craig et al. (2002) examined factors that might lead to stability of gang

membership for younger youth, Hill, Howell, Hawkins, and Battin-Pearson (1999) took

self-reports of 10 to 12 year olds and measured the risk factors of them joining a gang

between 13 and 18 years old. Data was collected annually from ages 13 to 18. Some of

the demographic information to note is that 77% of the 10 year olds were in schools

located in high crime areas; 46% of parents reported an annual income of less than

$20,000; and more than half (52%) of the student sample participated in the school lunch

program at some point during the study, indicating they were from families living in

poverty. The sample was multiethnic with 46% European American, 24% African

American, 21% Asian American, 2% Native American, and the remaining 7% of other

ethnic backgrounds, primarily Hispanic. A self-report measure was used and data

obtained from multiple sources, including the youth, their parents or caretakers, teachers,

school records, and court records.

Hill et al. (1999) found that cumulatively, 15.3% of the sample reported ever

belonging to a gang between age 13 and 18, and membership peaked at age 15. The

specific risk factors examined, neighborhood, family, school, peer, and individual, all

were significant in predicting gang involvement. Using prospective logistic regressions,

youth with neighborhood risk factors demonstrated the highest odds ratio (OR) for

joining gangs. Specifically, youth that had availability of marijuana in their

neighborhood had more than three times greater odds of joining a gang than youth that
DIFFERENCES IN EXECUTIVE FUNCTIONING 14

did not. Youth with neighborhood peers that were in trouble had three times greater odds

of joining a gang than those youth who were not exposed to other troublesome youth.

Taking into account family structure, Hill et al. (1999) found that youth living

with one parent, with one parent and other adults, and youth with no parents in the home,

when compared to youth living with two parents (biological or adoptive), had more than

two times greater odds of joining a gang. In addition, poverty and parental pro-violent

attitudes contributed to having more than two times greater odds for joining a gang.

School level predictors of joining gangs were also found to be significant with

low academic achievement in elementary schools and identified as learning disabled at

more than three times greater odds of joining a gang than those with higher academic

achievement and no learning disabilities. Hill et al. (1999) stressed the importance of

elementary school experiences as predictors in later gang membership, and an area that

needs more focus.

Individual predictors as reported at ages 10 to 12 that had the most significant

indicators of gang membership were marijuana experimentation and violence, which was

reported as fighting, throwing objects, and hitting a teacher, at more than three times

greater odds than those who do not engage in this behavior. Other individual factors

included antisocial beliefs (OR = 2.0), externalizing behavior (OR = 2.6), hyperactivity

(OR = 1.7), and poor refusal skills (OR = 1.8).

Overall, Hill et al. (1999) demonstrated that neighborhood, family, and school

influences all have an independent impact on the likelihood a youth will join a gang.

Taking into consideration the impact of multiple risk factors, and creating an index of

four quartiles to group risk factors from no risk to high risk, Hill et al. (1999) determined
DIFFERENCES IN EXECUTIVE FUNCTIONING 15

that with each increase of risk factors, the odds of joining a gang approximately doubled,

and youth with exposure to seven or more risk factors at ages 10 to 12 had more than 13

times greater odds of joining a gang than if exposed to 0 to 1 risk. Collectively, the more

risk factors a youth is exposed to, the more likely they are to join a gang.

Lahey, Gordon, Loeber, Stouthamer-Loeber, and Farrington (1999) hypothesized

that the factors that predict gang entry would include both antisocial behaviors prior to

joining the gang as well as family and neighborhood characteristics. They controlled for

antisocial behavior prior to gang entry to determine whether any of the contextual factors

of peer, family, and neighborhood are independent predictors.

Lahey et al. (1999) used data from the second cohort of the Pittsburgh Youth

Study with a sample of 347 boys being in the seventh grade in 1987-1988 to measure

serious antisocial behavior. Data was collected in waves over six and a half years,

varying between six and 12 month intervals; however, two of the waves were not used in

order to analyze the data on an annual basis. They identified two definitions of gang

membership: membership in any gang, regardless of engagement in antisocial behavior,

and serious gang membership, whose members engaged in at least one antisocial

behavior (fighting, selling drugs, robbery, stealing cars, or homicide). Participants were

classified as either high risk or low risk, with high risk scoring above the 67th percentile.

Of the 347 boys, 181 were classified as high risk and 166 were placed in the low risk

group. The boys’ family median income was $16,579 (in 1996), and 49% lived with a

single parent.

Of the 347 participants, 95 (27%) boys reported ever entering a gang. When

looking specifically at serious gang membership, 62 boys (18%) reported entering a gang.
DIFFERENCES IN EXECUTIVE FUNCTIONING 16

Using only African American participants in a multivariate analysis, Lahey et al. (1999)

found that both higher antisocial behavior at the beginning of the study and increases in

this conduct disorder over time, independently predicted future gang involvement. In an

analysis of three types of self-reported delinquency (crimes against persons, crimes

against property, and drug sales), Lahey et al. (1999) suggested that only crimes against

persons independently predicted adolescents joining gangs after other types were

controlled.

With regards to peer delinquency, Lahey et al. (1999) identified four types that

were reported: school related delinquent acts; covert delinquent acts; aggressive

delinquent acts; and drug use, and found that only the school-related delinquent acts were

significantly associated with gang involvement when the others were controlled. The

measures of family and neighborhood characteristics that were considered as predictors

of gang entry were household structure, family income, parental supervision, and

neighborhood crime. In contradiction to Hill et al. (1999), family and neighborhood

contextual factors were not found to be independently related to gang involvement.

However, upon combining prior antisocial behavior and the contextual factors,

using entry into any gang as the dependent variable, Lahey et al. (1999) did experience

some significant findings. They found that boys who committed crimes against persons

and living in a high-crime neighborhood, had higher rates of joining gangs.

Although Lahey et al. (1999) did find conduct disorder behaviors in their first

wave to predict serious gang entry, they also found a high degree of overlap in the

conduct disorder behaviors of future gang members and boys who did not join a serious

gang. Some boys with low baseline conduct disorder behavior entered gangs, and others
DIFFERENCES IN EXECUTIVE FUNCTIONING 17

with high conduct disorder behavior had not entered a serious gang by the end of their

study. This suggests the complexities that are involved in gang membership; it is not

solely due to socio-cultural reasons, nor solely due to high conduct disorder delinquent

behavior.

When examining various interactions with age, comparing early adolescence with

late adolescence, Lahey et al. (1999) also discovered significant findings. Peers that

engaged in aggressive delinquent acts significantly influenced gang entry only in early

adolescence. In addition, the higher the family income and the less parental supervision

available, the higher the risk for serious gang entry, but only in early adolescence. The

association between an adolescent’s initial conduct disorder behavior and the likelihood

of entering a serious gang weakens as age increases. The study suggested that boys who

exhibited antisocial behavior were already on this path of gang entry. As Lahey et al.

(1999) posited, “…gang entry might be thought of as the next developmental step in the

escalating antisocial behavior of some boys.”

Peer Influences

Kupersmidt, Coie, and Howell (2004) postulated that peer groups have the most

profound influence over whether or not an adolescent participates in a behavior outside

the norm. A question was posed whether problems among aggressive youths are

associated with friendships, and Kupersmidt et al. (2004) answered with a definitive, yes.

They reported that antisocial youths confirm that their antisocial friends negatively

influence them and even teach them how to do things that get them into trouble. They

called it “deviancy training” and suggested aggressive friends support “forbidden”

DIFFERENCES IN EXECUTIVE FUNCTIONING 18

behavior. Kupersmidt et al. (2004) suggested the proliferation of gang activity in the

United States reflects the tendency of antisocial youths to reinforce other’s antisocial

behavior.

In their study, Craig et al. (2002) examined peer relations as an antecedent, and

compared the friendships of boys who belong to gangs to those who do not belong to a

gang. They allude to the idea that unstable gang members may spend less time with their

delinquent peers and have less knowledge about their friends’ delinquent activities than

do stable gang members. This raises a question of whether it is pressure from the gang

that contributes to these youth participating in the deviant behaviors associated with

gangs or some other influence, external or internal, that allows them to participate in this

behavior.

Hill et al. (1999) found that associating with delinquent friends was a potent

factor for predicting gang membership, and having a strong attachment to conventional

peers decreased the probability of joining gangs. Having peers that engaged in problem

behaviors, assessed by the number of friends that got them into trouble with the teacher

or tried alcohol without parents’ knowledge, accounted for a twofold increase in the

likelihood of joining a gang.

These studies reveal the complexities of gang involvement. While some found

that neighborhood, family, school, and peer influences affect gang entry; others found

that these factors only had influence when there was early antisocial behavior. Craig et

al. (2002) pointed out that understanding the mechanism of the catalyst for the initiation,

continuation or even discontinuation of gang membership, is an important issue.

DIFFERENCES IN EXECUTIVE FUNCTIONING 19

External Factors to Gang Involvement

The Neighborhood’s Role

Opportunities to join gangs abound and attitudes towards aggression appear to be

ambivalent in the disadvantaged communities gangs tend to form in; however, not all

youth respond in the same manner to these opportunities. As found in prior studies,

neighborhood context has a fairly significant influence on the youth. Beyers, Bates,

Pettit, and Dodge (2003) suggested that neighborhoods are characterized by their social

structure, that is, the aspects of the individuals who compromise them, as well as by the

nature of the way individuals relate to one another. Neighborhoods where parents talk to

each other about their children’s activities or about parenting strategies have more social

capital than neighborhoods where parents do not communicate. Similarly, in

neighborhoods where adults intervene if they see a group of teenagers acting out, they

have more collective efficacy and improve children’s opportunities than neighborhoods

where adults ignore these youth transgressions (Beyers et al., 2003).

Dupere, Lacourse, Willms, Vitaro, and Tremblay (2007) examined the interaction

between the facilitating neighborhood characteristics and individual propensity to join

gangs. They expected that children with psychopathic tendencies, that is, high

hyperactivity, low anxiety, and low prosociality, would be more at risk to affiliate with

youth gangs, especially in a neighborhood environment prone to gangs.

The sample used was from the National Longitudinal Survey of Children and

Youth (NLSCY), a study based on a nationally representative sample of Canadian

children and adolescents throughout the ten provinces. For this study, the sample was

reduced to n=3,522 based on specific participant requirement. Gang membership was

DIFFERENCES IN EXECUTIVE FUNCTIONING 20

measured in mid-adolescence, when respondents were 14 to 15 years old, incorporating

family SES, residential instability, and childhood affiliation to deviant peer groups. Non-

intact family variables were also considered, and neighborhood characteristics were

measured using the 2001 Census of Canada. Three parent-reported scales were used to

assess late childhood psychopathic tendencies, and approximately 10.7% of the sample

presented the psychopathic tendency profile. Six percent of the participants reported on a

single self-report item that within the last 12 months they had been “part of a gang that

broke the law by stealing, hurting people, damaging property, etc.” (Dupere et al., 2007).

Dupere et al. (2007) found adolescents with psychopathic tendencies in late

childhood were approximately 1.6 times more likely to have a gang affiliation in mid-

adolescence than their peers without these tendencies. They found neighborhood

economic disadvantage had 1.6 times greater odds of youth joining gangs than those who

live in neighborhoods with more economic resources. In addition, residential instability

was found to have 2.7 higher odds of youth gang affiliation than adolescents living in

average neighborhoods.

Performing multivariate logistic regressions, Dupere et al. (2007) determined that

once family background contextual factors were controlled for, psychopathic tendencies

and residential instability were found to be independently associated with youth gang

membership, and neighborhood economic disadvantage was not found to be a significant

independent factor. They also found independent significant associations with gang

membership and the various family background variables such as non-intact families,

family SES, family residential instability, and previous deviant peer group affiliation.
DIFFERENCES IN EXECUTIVE FUNCTIONING 21

Upon adding interaction terms to the model, Dupere et al. (2007) found that youth

living in unstable residential neighborhoods were only influenced towards gang

membership if the youth also had the psychopathic tendencies (higher hyperactivity, low

anxiety, and low prosociality). Youth living in residential instability communities have a

five times higher probability (5.2% vs. 26.5%) of joining a gang if they also demonstrate

the psychopathic profile. There was no significant interaction found between the

psychopathic tendencies and economic disadvantage.

Beyers et al. (2003) determined that increased neighborhood disadvantage and

residential instability were associated with less parental monitoring, and less parental

monitoring predicted increased externalizing behavior and less supervision among

adolescents. According to Dupere et al. (2007), neighborhood residential instability is

likely to drastically reduce collective capacity to supervise youth groups within the

neighborhood. They suggested that single parent households, which are high in unstable

neighborhoods, reduce parent-to-child ratio, and fewer guardians are available to oversee

youth activities. Beyers et al. (2003) concluded that it is important to address family

functioning within the neighborhood context when considering the development of youth

conduct problems.

McMahon, Felix, Halpert, and Petropoulos (2009) examined the relationship

between exposure to community violence and specifically aggressive behavior. They

hypothesized that the more exposure one has to community violence would lead to

increased beliefs that support retaliatory aggression, which in turn would lead to lower

self-efficacy to control aggressive behavior, and eventually lead to actual increased

aggressive behavior.
DIFFERENCES IN EXECUTIVE FUNCTIONING 22

McMahon et al. (2009) examined two samples: a cross-sectional model of two

elementary schools of four classes of sixth through eighth graders, and a longitudinal

model of a third elementary school with five classes of sixth through eighth graders,

where Time 1 (T1), exposure to community violence and retaliatory beliefs about

aggression lead to Time 2 (T2), self-efficacy and aggressive behavior. There were a few

fifth graders in their study as a result of a split fifth/sixth grade class.

The cross-sectional sample included 126 youth (43 boys, 83 girls), ages 10 to 15,

and was conducted in the fall of the academic year. The longitudinal sample included 81

students (35 boys, 46 girls). There were 129 students at T1 and 106 students at T2. The

longitudinal study was conducted during the fall and spring of the following academic

year.

All three schools served urban African American youth who lived in a Chicago

public housing development community, of which 96-100% were low income, which was

defined as families that receive public aid, are supported in foster homes with public

funds, or are eligible to receive free or reduced priced lunches. The schools in this study

also had chronic truancy rates (8.1 – 14.2%) that differed dramatically from those of the

state of Illinois (2%), and were of particular interest to this study (McMahon et al., 2009).

McMahon et al. (2009) measured exposure to violence, normative beliefs about

aggression, self-efficacy, and aggressive behavior for each sample of students. There

was a slight difference in the assessment of exposure to violence between the two groups.

The cross-sectional sample was administered the 51-item Screening Survey of Exposure

to Community Violence, which assesses various types of violence including assault,

physical threats, use of weapons, burglary, and gang and drug activity. The longitudinal
DIFFERENCES IN EXECUTIVE FUNCTIONING 23

sample was administered the 12-item Children’s Exposure to Violence Scale due to a

need for a shorter assessment tool. McMahon et al. (2009) thought the self-report

measures were a better assessment as the self-perception of violence exposure may be

more related to future behavior than other measures.

Other measures included the Normative Beliefs about Aggression Scale, which

measures retaliatory beliefs through two subscales: one that presents situations, and one

that poses more general questions; the Self-Efficacy to Control Aggression, a 5-item

scale from the Teen Conflict Survey and assesses a person’s confidence in controlling

their anger; and the Aggressive Behavior Scale, which measures the frequency of self-

reported aggressive behaviors.

McMahon et al. (2009) hypothesized that the more exposure one has to

community violence would lead to increased beliefs that support retaliatory aggression,

which in turn would lead to lower self-efficacy to control aggressive behavior, and

eventually lead to actual increased aggressive behavior. This model was tested using

structural equation modeling. In addition to this model, McMahon et al. (2009) proposed

there would be a direct link between violence exposure and aggressive behavior.

The cross-sectional data from the first sample was tested first. It was found that

the effect sizes of each of the model pathways were significant and their hypotheses were

confirmed. Exposure to violence was significantly related to retaliatory beliefs (p < .05),

retaliatory beliefs was significantly related to self-efficacy (p < .01), and self-efficacy

was significantly related to aggressive behavior (p < .01). McMahon et al. (2009) also

found a significant relation between exposure to violence and aggressive behavior (p <

.01). In essence, a youth living in a community where they are exposed to violence learn
DIFFERENCES IN EXECUTIVE FUNCTIONING 24

to develop beliefs that retaliatory behavior is the norm, and this in turn leads to beliefs of

inadequacy in the ability to control their aggression, and these negative views on

effectiveness of self-control lead to aggressive behavior.

McMahon et al. (2009) then tested the model with the longitudinal sample where

exposure to community violence and retaliatory beliefs about aggression were introduced

to the model at T1, and self-efficacy and aggressive behaviors were included in the model

at T2. Results from this longitudinal sample also confirmed the model. Exposure to

violence was significantly related to retaliatory beliefs (p < .05), retaliatory beliefs was

significantly related to self-efficacy (p < .05), and self-efficacy was significantly related

to aggressive behavior (p < .01). Exposure to violence was also found to be significantly

related to aggressive behavior (p < .05).

McMahon et al. (2009) also examined the correlations of each variable across

time from T1 to T2, and found a substantial amount of continuity across time: violence

exposure, r=.63; retaliatory beliefs, r=.63; self-efficacy to control aggression, r=.58;

aggressive behavior, r=.54. This stability suggested that this process that leads to

aggression is generally maintained over time unless there is an intervention.

There were limitations to the study that included small sample size with limited

power, assessments were limited to self-reports, and the longitudinal study was only over

two time points in the course of one year. Despite these limitations, within this specific

population, McMahon et al.’s (2009) study discovered that exposure to community

violence mediates a number of cognitions with potential long term negative effects,

specifically, aggressive behavior.

DIFFERENCES IN EXECUTIVE FUNCTIONING 25

“The colossal individual and social costs of youth gang membership calls for a

better understanding of its antecedents and the process leading up to it” (Dupere et al.,

2007). Additional research on diverse populations, where the majority of gangs exist, can

help answer questions regarding whether psychopathology is a secondary occurrence to

their circumstances or has causal factors for explaining violent behavior.

Socio-Cultural Perspective

From a socio-cultural perspective, Gibbs (2000), using a comparative study,

looked at African American youth in Los Angeles, California, as well as London,

between the ages of 15 to 30 years. She gathered self-reported information with semi-

structured interview instruments in focus groups as well as individual interviews on

attitudes, experiences, and behaviors of these participants who were involved or

influenced by gangs. The data reported is from a larger study, conducted from 1992-

1994, of the socioeconomic status of Black youth in Canada, Great Britain, and the

United States.

Gangs and neighborhood violence, drugs, lack of community resources, poor

schools, and lack of jobs, were among the five major issues identified as impacting these

survey participants. Drugs and gang/violence were the top two concerns. Eighty percent

of the participants in Los Angeles and 55% of the participants in London either

personally knew a gang member, lived in communities claimed by specific gangs, or had

formerly been a gang member.

Gibbs (2000) was able to identify multiple functions and reasons these

individuals joined gangs, including safety, friendship, surrogate family, social status,
DIFFERENCES IN EXECUTIVE FUNCTIONING 26

respect, excitement, income, power, and to alleviate boredom. The individuals viewed

any engagement in violence as instrumental in achieving these listed goals, rather than an

end in itself. Their sole intent was not to inflict harm on people; it was a means to an

end. The gang members in the study held the perspective that the gang has its own norms

and internal structures, and serves functions other than perpetuating deviant behaviors.

Valdez et al. (2000) postulated that adolescents that grow up in these communities

may not internalize so called mainstream orientations to norms, values, expectations and

behavior. In essence, they adapt to the social structures and conditions they are

surrounded by. According to Anderson (1999), families from the “street” are more

invested in the unwritten “code of the street” and may aggressively socialize their

children into the “code” in a normative way. A community can become so alienated that

an oppositional culture, like gangs, can develop and flourish. A number of community

residents learn not to conform to the dominant society in order to survive the street

(Anderson, 1999). The DSM-IV-TR (APA, 2000) notes that concerns have been raised

that a conduct disorder diagnosis may at times be misapplied to individuals in settings

where patterns of undesirable behavior are sometimes viewed as protective, such as

living in a threatening or high crime area. The issue of gangs is more complex than a

conduct disorder diagnosis.

Many gang members agreed that the gang provided a more stable social structure

with clear rules, discipline, and norms, in contrast to their home environments (Gibbs,

2000). Some participants stated they derived greater social status from the gang members

than from their family, particularly when they were successful in generating income.

Some rationalized that the income gained from illegal activities is a form of reparations
DIFFERENCES IN EXECUTIVE FUNCTIONING 27

from the dominant society that has oppressed them.

Gang membership, according to the subjects in the study, serves as a transitional

structure through which the members can meet a number of psychological, social and

economic needs in a society in which they feel excluded Gibbs (2000). Many find

themselves adrift between the street and more conventional society (Anderson, 1999).

Gibbs (2000) also found that many of the gang members age out and tire of the risk

taking behavior, usually around age 30. But as Vigil (2003) points out, many gang

members, who usually have been raised in marginalized, highly stressful families, have

their social development arrested and remain peer-dependent into their thirties and

forties.

Gibbs (2000) identified some of the socio-cultural factors that may influence a

young personal to join a gang. However, gang membership is not as simple as labeling it

a socio-cultural issue. There has to be individual accountability as well. An interest lies

in determining individual factors that contribute to this phenomenon.

Self-Control

In light of McMahon et al.’s (2009) study on the role that exposure to violent

communities plays in one’s ability to control aggressive behavior, it is important to

briefly review literature on the role family and community factors play in self-control.

Pratt, Turner, and Piquero (2004) highlight collective socialization as an important

contributing factor to the development of self-control. Expanding on Gottfredson and

Hirschi’s (1990) theory of parental social control, collective socialization includes

monitoring and supervising, as well as role-modeling within the surrounding

DIFFERENCES IN EXECUTIVE FUNCTIONING 28

neighborhood and community. In their study, Pratt et al. (2004) examined the

relationship between collective socialization and self-control. Collective socialization

included the frameworks of parental socialization as well as adverse neighborhood

conditions. They also examined race as a variable. Pratt et al. (2004) used data from the

National Longitudinal Survey of Youth (NLSY). For purposes of their study, they

limited their sample to participants who were age 10 in 1992 (n = 463), and examined

measures cross-sectional as well as longitudinally when the participants were 12. From a

parental socialization perspective, Pratt et al. (2004) measured parental supervision and

parental monitoring/discipline. They incorporated adverse neighborhood conditions into

their measures as well as self-control. All four of these factors were measured using self-

report questionnaires.

Using ordinary least squares regression equations, Pratt et al. (2004) first

examined predictors of the two measures of parental socialization. They found parental

supervision to be predicted by gender, non-White, and adverse neighborhood conditions,

suggesting that parents, who were male, non-White, and exposed to adverse

neighborhood conditions, were less likely to supervise their children. With regards to

parental monitoring/discipline, Pratt et al. (2004) found being non-White and exposed to

adverse neighborhood conditions is related to parents being more likely to discipline

children who receive low grades in school.

Pratt et al. (2004) then examined predictors of self-control. Again using ordinary

least squares regression equations, they found poor parental supervision and higher levels

of parental discipline to be significantly associated with lower self-control. They found

that being non-White is significantly associated with better self-control. In measuring the
DIFFERENCES IN EXECUTIVE FUNCTIONING 29

relationship between adverse neighborhood conditions and self-control, it was found that

adverse neighborhood conditions are significantly positively related to self-control. This

indicates that in the neighborhoods that were reported to exhibit poorer conditions, the

children are more likely to exhibit lower self-control. In the longitudinal analysis of the

variables at age 10 measured against self-control at age 12, the findings were replicated.

Pratt et al. (2004) then looked at predictors of self-control by race. At age 10,

Whites and non-Whites exhibited similar results: both were more likely to experience

lower levels of self-control as a result of lower levels of parental supervision, higher

levels of discipline, and living in neighborhoods with adverse conditions. The only

difference between these two groups was the effect of race, and it was only found to be

marginally significant for Whites (p<.10). Looking at the longitudinal analysis, among

Whites, poor parental supervision and high parental discipline are significantly associated

with lower self-control; whereas with non-Whites, poor parental supervision, high

parental discipline, as well as adverse neighborhood conditions, are all significantly

associated with self-control at age 12.

Pratt et al. (2004) also examined the total effect and fit of the models with

structural equation modeling. Overall, the model provided a good fit to the data and the

results from the previous analyses were replicated. Self-control was found to be

significantly predicted by parental supervision, parental discipline, and adverse

neighborhood conditions, while being non-White only moderately predicted self-control

(p<.10). Adverse neighborhood conditions were found to have significant effects on

parental discipline and self-control.

DIFFERENCES IN EXECUTIVE FUNCTIONING 30

In examining the same structural equation model, this time split by race, Pratt et

al. (2004) found that parental supervision, parental discipline, and adverse neighborhood

conditions, all predict self-control for both Whites and non-Whites. However, an

important relationship between adverse neighborhood conditions and parental supervision

was found. Adverse neighborhood conditions was found to have a significant effect on

parental supervision for non-Whites but not for Whites, indicating that in neighborhoods

with adverse conditions, non-White children are less likely to be supervised by their

parents. Beaver, Wright, and Delisi (2007) examined the possibility of self-control as an

executive function. Both of these concepts incorporate the ability to regulate emotions

and sustain attention. They argue that variability in self-control is likely due to

variability in the function of the prefrontal cortex, or executive function area of the brain.

They examined whether measures of neuropsychological deficits predicted levels of self-

control.

The data for their study came from the Early Childhood Longitudinal Study,

Kindergarten Class of 1998-1999 (ECLS-K), which is the largest nationally

representative sample of American children and thus far has had six waves of data

collected. For purposes of this study, Beaver et al. (2007) limited their data to the fall

and spring of kindergarten (Waves 1 and 2) and the spring of first grade (Wave 4). Due

to the large sample size, 15% of the sample was randomly selected, and a final sample of

almost 3,000 male and female children was obtained. There were three measures Beaver

et al. (2007) examined: low self-control, neuropsychological deficits, and socialization.

Low Self-Control was measured by developing a composite Low Self-Control scale

based on parent and teacher responses from the Social Skills Rating Scale (SSRS) at each
DIFFERENCES IN EXECUTIVE FUNCTIONING 31

wave. Neuropsychological deficits were measured using the Early Screening Inventory-

Revised (ESI-R) administered at Wave 1. A Fine Motor Skills scale and Gross Motor

Skills scale were used from the ESI-R. Six socialization measures were developed which

included parental involvement, parental withdrawal, parental affection, family rules,

physical punishment, and neighborhood disadvantage. They used gender and race as

controls.

First neuropsychological measures were examined to determine if they are

predictive of self-control. Using ordinary least squares regression equations, Beaver et

al. (2007) found that both fine and gross motor skills are predictive of low self-control for

males, suggesting that higher motor skills are associated with lower self-control.

However, when prior self-control was introduced, gross motor skills were no longer

significant. Four of the socialization variables were found to be significantly predictive

of low self-control: parental involvement, parental withdrawal, parental affection, and

physical punishment. When the measure of prior self–control was introduced, only

parental withdrawal and physical punishment had a significant effect on low self-control.

Beaver et al. (2007) next tested whether the neuropsychological measures have

the ability to predict low self-control later in childhood from the scale measured in Wave

4. For the male sample, both neuropsychological measures were significantly related to

low self-control in both models, with and without prior self-control. Three of the

socialization variables were found to be significantly predictive of low self-control:

parental involvement, parental withdrawal, and physical punishment. When prior self–

control was incorporated, only parental withdrawal had a significant effect on low self-

control.
DIFFERENCES IN EXECUTIVE FUNCTIONING 32

These findings suggest that neuropsychological measures have consistent

significant effects on self-control and should be viewed as an executive function.

Parenting and other social measures had some effect on self-control, but results were

small and inconsistent. One of the limitations is the sample age, and Beaver et al. (2007)

were unable to determine whether neuropsychological deficits could be related to levels

of self-control in adolescents and young adults. Another limitation is that only

neuropsychological tests that measure motor skills were included. Other measures might

be used that specifically test for executive functioning. In addition, the motor skills

scales may have identified dysfunctions that are not related to self-control.

Ratchford and Beaver (2009) examined the relationship between social and

biological risk factors, self-control, and delinquency in a three step study, using ordinary

least squares regression equations, which first looked at whether social factors (family

and community) and neuropsychological deficits, among other biological risk factors, are

associated with self-control. They then examined neuropsychological deficits as the

dependent variable to determine predictor variables that might be associated with these

deficits. Finally, they measured delinquency to determine whether socialization

variables, neuropsychological variables, or self-control, might be considered predictor

variables in delinquency.

Ratchford and Beaver’s (2009) data was drawn from the National Survey of

Children (NSC), a longitudinal survey that includes three waves of data collected over 10

years. Ratchford and Beaver (2009) only used data from Waves 1 and 2 for purposes of

their study. Wave 1 was collected in 1976 when the children were between the ages of

six and 12, for a total participant pool of 2,301. Wave 2 data was gathered five years
DIFFERENCES IN EXECUTIVE FUNCTIONING 33

later in 1981, when the participants were between 12 and 17 years with a total of 1,423

respondents.

For the self-control measure, Ratchford and Beaver (2009) developed their own

composite self-control scales based on information gathered from parents and teachers

from the NSC. Biological risk was assessed by measuring neuropsychological deficits as

well as birth complications and low birth weight as reported by the mothers in Wave 1.

The neuropsychological deficits were measured with the Peabody Picture Vocabulary

Test-Revised (PPVT-R). Socialization risk was assessed by measuring family rules,

parental punishment, and neighborhood disadvantage. Delinquency was measured in

Wave 2 via a self-report survey on misconduct and involvement in delinquency. Race,

gender, and age were used as control variables.

In the first step, measuring self-control with any social or biological predictor

variables, Ratchford and Beaver (2009) found that neuropsychological deficits are a

significant predictor of low self-control. Among the socialization variables, parental

punishment and neighborhood disadvantage had significant effects on the parental low

self-control scale as well as the composite scale. The teacher low self-control scale was

predictive of only parental punishment. In other words, the more parental punishment and

neighborhood disadvantage was indicated, the lower the self-control. Gender and race

were also found to have significant associations with low self-control, indicating males

and non-White participants were more likely to have low self-control.

Ratchford and Beaver (2009) found that self-control is influenced by both social

and biological factors; however, the neuropsychological deficits had the most consistent

effects on self-control. With this in mind, they measured predictors of

DIFFERENCES IN EXECUTIVE FUNCTIONING 34

neuropsychological deficits, only considering social risk factors and the other two

biological risk factors, as well as gender and race. They found that birth complications,

low birth weight, gender, and race all had significant effects on neurological deficits. In

other words, male, non-White children who had suffered birth complications and were

born with a low birth weight, had more neurological deficits than children who did not

experience these variables.

Finally, Ratchford and Beaver (2009) considered delinquency and its predictors,

including the parent and teacher self-control variables as well as the socialization and

biological variables. Parental punishment was the most consistently significant of the

social predictors, indicating that the more parental punishment, the more likely the child

will be delinquent. None of the biological measures had an effect on delinquency. Each

of the low self-control scales (parent, teacher, composite) in both waves had significant

effects on delinquency, suggesting that the less self-control a child has; the more likely

they are to be delinquent.

Although biological factors were not predictive of delinquency in this study, they

were significantly related to self-control; self-control was, however, significantly related

to delinquency, although it was self-reported. Beaver, Wright, and Delisi (2007)

determined self-control to be an executive function. Ratchford and Beaver (2009), as in

the prior studies, also found collective socialization to be predictive of self-control and

delinquency. These suggest that a biosocial approach, incorporating both social and

biological factors, may be best in studying antisocial behavior.

DIFFERENCES IN EXECUTIVE FUNCTIONING 35

Heritability Factors

Gene-Environment Interactions

Tuvblad, Grann, and Lichtenstein (2006) examined heritability of antisocial

behavior, and how it differs among varied socioeconomic status, or what they referred to

as gene-environment interaction. They hypothesized that genetic factors for antisocial

behavior would be expressed more among adolescents in socioeconomically advantaged

environments, and environmental factors on antisocial behavior would be more

pronounced among adolescents in socioeconomic disadvantaged environments.

Their sample included 2,133 16 to17 year old twins and 1,067 parents from the

third wave of the Twin study of CHild and Adolescent Development (TCHAD)

conducted in Sweden. The twins were measured on a 32-item self-report of antisocial

behavior derived from a questionnaire developed by the Department of Criminology at

Stockholm University. The items were found to have high validity (97.5%) and

reliability (90%), as well as high internal consistency with a Cronbach’s alpha of .92.

Family socioeconomic status was measured with self-reported data from the parents of

the twins on occupational status and education level. Neighborhood socioeconomic

conditions were measured using five variables: ethnic diversity, basic education level,

unemployment level, buying power, and crime rate.

In order to understand the relative importance of any genetic and environmental

components (shared environment and non-shared environment) of variance in antisocial

behavior, Tuvblad et al. (2006) performed structural equation modeling. These

phenotypic variance factors are also portioned into interaction effects with socioeconomic

status, as well as a main effect of socioeconomic status on antisocial behavior. They also
DIFFERENCES IN EXECUTIVE FUNCTIONING 36

examined whether genetic and environmental influences on antisocial behavior varied

between socioeconomic conditions.

Tuvblad et al. (2006) found modest phenotypic correlations between antisocial

behavior and genetic, shared environment, and non-shared environment factors. Testing

on the interaction between antisocial behavior and socioeconomic status indicated that

adolescent boys’ sensitivity to display antisocial behaviors differ based on their

surrounding environment. They found that genetic factors are more expressed in a

socioeconomically advantaged environment (37% compared to 1%), where an antisocial

adolescent lacks the environmental risk factors that push or predispose him/her to behave

antisocially. Likewise, they found genetic factors to be weaker and the shared

environment more important (69% compared to 13%) in adolescents from disadvantaged

backgrounds because the environment in these areas will conceal the genetic contribution

towards antisocial behavior. In summary, this study provides evidence that

socioeconomic status may be a mitigating influence over genetic and environmental

effects on antisocial behavior.

Biology, Self-control, and Delinquency

As demonstrated in previously cited studies, gang involvement has been found to

be associated with low self-control, associating with delinquent peers, as well as the

environment one is brought up in conjunction with biological influences. The goal of

Beaver, Shutt, Boutwell, Ratchford, Roberts, and Barnes’ (2009) study was to expand on

many criminological studies and their ideas that low self-control among delinquents is

due to social learning theory. They posited that biology and genetics should also be
DIFFERENCES IN EXECUTIVE FUNCTIONING 37

considered in identifying causal influences on low self-control as well as delinquent peer

associations. Beaver, Shutt et al. (2009) specifically wanted to determine any effects of

genetic and environmental factors, both shared and non-shared, in the development of

self-control and delinquent peer associations.

Data was from the National Longitudinal Study of Adolescent Health (Add

Health), which used a nationally representative sample of adolescents in seventh through

twelfth grade among 80 high schools and 52 middle schools, who were administered a

self-report survey in 1994. There were three waves of data collected over seven years,

with the final wave occurring with the participants ranging between 18 and 26 years.

Within the sample of participants was a subsample of siblings, who were asked to

indicate whether they were part of a twin pair. Beaver, Shutt et al. (2009) only included

monozygotic (MZ) twin pairs and same-sex dizygotic (DZ) twin pairs. For purposes of

their study, they only used data from the first two waves; their final analytical sample

ranged between 662 and 914 twins.

Beaver, Shutt et al. (2009) used low self-control and drug-using peers as the

dependent variables and parental influences as social control variables, specifically

maternal disengagement, maternal attachment, maternal involvement, and parental

permissiveness. To estimate genetic and environmental effects on low self-control and

drug-using peers, they used the DeFries-Fulker (DF) augmented analysis, a regression-

based approach that allows for analysis of data with sibling pairs. Beaver, Shutt et al.

(2009) used double entry of the twins to allow for each twin to be used as the dependent

variable and then as the independent variable.

DIFFERENCES IN EXECUTIVE FUNCTIONING 38

Their results were broken into four models with Wave 1 represented by Models 1

and 2, and Wave 2 represented by Models 3 and 4. Models 1 and 3 accounted for

genetic, shared, and non-shared environmental influences, and the four specific non-

shared environmental variables were introduced in Models 2 and 4. In looking at low

self-control, Beaver, Shutt et al. (2009) found that shared environment did not account for

any of the variance in low self-control. In Wave 1, 56% of the variance was accounted

for by genetic influences and the non-shared environment accounted for the remaining

44% of variance. When the four parental non-shared environmental factors were

introduced, genetic influences still accounted for more than half (52%) of the variance;

however, of the four measures, only maternal disengagement was statistically significant

in effecting low self-control.

Among Wave 2 participants, genetics accounted for 40% of the low self-control

variance, and the non-shared environment accounted for 60% of the remaining variance.

When the four non-shared environmental variables were introduced, genetic variance

increased to 44% and non-shared variance reduced to 56%. Two of the parental variables

were statistically significant in effecting low self-control: maternal disengagement and

parental permissiveness.

Using the same four model structure, when looking at drug-using peers, Beaver,

Shutt et al. (2009) found some slightly different results. Shared environment was found

to have an effect on Wave 1, but not on Wave 2. In Wave 1, Beaver, Shutt et al. (2009)

found that genetics accounted for 37% of variance, the shared environment for 27% of

variance, and the non-shared environment and measurement error accounted for 36% of

variance. When the four non-shared variables were added, the variance of genetic effects
DIFFERENCES IN EXECUTIVE FUNCTIONING 39

increased to 40% and the shared environment variance decreased to 22%. There was one

non-shared environmental variable significantly associated with drug-using peers, that of

maternal attachment.

In Wave 2 of assessing for drug-using peers, 53% of the variance was found to be

accounted for by genetics, a large jump from Wave 1. There was no shared environment;

hence, the remaining 47% of variance was accounted for by the non-shared environment

and measurement error. Lastly, when the four non-shared environment variables were

added, the genetic influence increased to 62% variance, and the non-shared environment

accounted for 38% of the variance. None of the four variables were found to be

significant in accounting for drug-using peers.

In short, Beaver, Shutt et al. (2009) found that both genetic and non-shared

environmental factors play a role in low self-control, as well as engaging with drug-using

peers. Although non-shared environment accounted for a large proportion of variance in

this study, the specific four measures were limited and had small, inconsistent effects.

Another item to note is that this analysis was based on a sample of MZ and DZ twins, and

may not generalize to the larger population. Future research might concentrate on

specific genetic causes to low self-control as well as the complex ways the non-shared

environment may lead to maladaptive behavior.

Beaver, DeLisi, Vaughn, and Wright (2010) examined the existence of overlap

among low self-control, monoamine oxidase A (MAOA) genotype, neuropsychological

deficits, and delinquency. They hypothesized a few interactions. One,

neuropsychological deficits will contribute to differences in self-control and delinquency.

Second, MAOA will not have any effects on low self-control or delinquency; however,
DIFFERENCES IN EXECUTIVE FUNCTIONING 40

MAOA will interact with neuropsychological deficits to predict significant differences in

self-control and delinquency.

Beaver, DeLisi et al. (2010) used data from the National Longitudinal Study of

Adolescent Health (Add Health), which included in Wave 1 more than 90,000 students in

seventh through twelfth grade in 1994-1995 from 132 middle schools and high schools.

Three waves of in-home interviews occurred over six years. In Wave 3, a subsample of

participants who had a sibling or co-twin who was also participating in the study was

asked to submit buccal cells for genotyping. This resulted in 2,574 respondents being

genotyped. Beaver, DeLisi et al. (2010) limited their study to white males as they did not

have sufficient statistical power to examine other groups. After removing one

monozygotic twin from each monozygotic twin pair, they had a final sample of 767 white

males.

The MAOA was measured by dividing the gene into two groups: low MAOA

activity and high MAOA activity. The low MAOA activity group comprised the 2- and

3- repeat alleles, and the high MAOA activity group included the 3.5-, 4-, and 5-repeat

alleles (Beaver, DeLisi et al., 2010). The Peabody Picture Vocabulary Test (PPVT)

measured neuropsychological deficits. Low self-control was measured by the summed

score of five questions that were confirmed via factor analysis to be significant indicators

of a latent construct. Delinquency was measured via summed responses that created a 15-

item Wave 1 delinquency scale and 14-item Wave 2 delinquency scale.

Beaver, DeLisi et al. (2010) focused their analyses on Waves 1 and 2 to measure

the delinquency scale as the dependent variable. Using a negative binomial regression,

they ran three models. First MAOA, neuropsychological deficits, age, and age squared
DIFFERENCES IN EXECUTIVE FUNCTIONING 41

were examined as predictor variables of delinquency; only age and age squared were

found to have a significant effect on delinquency. Next, the multiplicative interaction

between MAOA x neuropsychological deficits was added. This interaction was found to

have a significant positive effect on delinquency, as were neuropsychological deficits by

themselves. In the third model, the Wave 1 low self-control scale was introduced and

found to be significant and have the strongest predictor effect on delinquency. The

addition of low self-control also weakened the effect of the MAOA x neuropsychological

deficits interaction.

In examining Wave 2, Beaver, DeLisi et al. (2010) found similar results as in

Wave 1. Only age and age squared had a significant effect on delinquency. The

interaction between MAOA x neuropsychological deficits when added was again found

to have a significant positive effect on delinquency, and remained at the same

significance level after adding the low self-control variable. Neuropsychological deficits

were no longer significant in this wave after adding the MAOA x neuropsychological

interaction. Low self-control was also found to have a significant positive effect on

delinquency in Wave 2 and was the strongest predictor of delinquency.

Beaver, DeLisi et al. (2010) furthered their analysis by examining violent

delinquency scales in Waves 1 and 2. In the first model, only neuropsychological deficits

were found to have a significant positive effect on violent delinquency in Waves 1 and 2.

MAOA was not found to have a significant main effect. When the MAOA x

neuropsychological deficits interaction was added, it also had a significant positive effect

on violent delinquency in both Waves 1 and 2; however, when low self-control was

introduced in the third equation, this interaction no longer had a significant main effect
DIFFERENCES IN EXECUTIVE FUNCTIONING 42

on violent delinquency. Low self-control, however, had a significant effect on violent

delinquency. As in the prior analysis, low self-control was the strongest predictor of

violent delinquency.

Lastly, Beaver, DeLisi et al. (2010) used the low self-control scale as the

dependent variable to examine any effects MAOA and neuropsychological deficits may

have on low self-control. They found that neither MAOA or neuropsychological deficits

on their own have a significant main effect on low self-control; however, when the

MAOA x neuropsychological deficits interaction was added, it was found to have a

significant positive effect on violent delinquency in Wave 1, as well as a significant

positive relation between neuropsychological deficits and low self-control. In Wave 2,

findings were similar; however, the addition of the MAOA x neuropsychological deficits

interaction had a significant positive effect on the interaction only.

Beaver, DeLisi et al. (2010) confirmed most of their hypotheses, and found that

neuropsychological deficits contribute to differences in delinquency, and MAOA did not

have any effects on low self-control or delinquency independently, but did interact with

neuropsychological deficits in the prediction of delinquency and self-control.

Some limitations include the small sample of respondents that were genotyped

and the affect this may have on generalizability, as well as the use of a delinquency scale

as opposed to a direct measure of life-course persistent (LCP) offending. In addition, the

Add Health sample was a community sample and more than likely did not include an

adequate number of LCP offenders. Lastly, the sample only included white males. This

affects generalizability by only including one ethnic group and not a diverse group of

participants.
DIFFERENCES IN EXECUTIVE FUNCTIONING 43

Delinquency Trajectory

In 1993, Terrie E. Moffitt proposed a dual taxonomy of delinquent behavior,

hypothesizing that temporary and persistent antisocial individuals are comprised of two

distinct types of people. She posited that timing and duration of antisocial behavior are

the defining features of her two proposed types of offenders. A larger group of

individuals are thought to engage in delinquent behavior for a short period of time in

adolescence, which Moffitt (1993) identified as adolescent-limited, and a small group

that engages in some kind of antisocial behavior at every stage in life, she identified as

life-course persistent. She noted a feature of her taxonomy is that knowledge of

preadolescent behavior is required for distinguishing between adolescent-limited and life-

course-persistent.

Moffit (1993) defined adolescent-limited delinquency as teenage delinquents who

have no notable antisocial behavior in childhood and little chance of such behavior in

adulthood. Their delinquency is often abrupt, particularly during the onset and

desistance, and their behavior is inconsistent. They may ditch school but abstain from

using drugs. A social learning theory is predominant in adolescent-limited delinquency

in that participation in antisocial behavior that reaps rewards is maintained and antisocial

behavior that has severe consequences is extinguished. Moffit (1993) suggested two

prerequisites for adolescence-limited delinquency: the motivation to close a maturity gap

between childhood and adulthood, and influence of life-course persistent antisocial role

models. She posited the reason adolescence-limited delinquents begin engaging in

antisocial activity is social mimicry, which leads one to believe that delinquency leads to

an end with a desirable resource. This desired resource is postulated to be a mature status
DIFFERENCES IN EXECUTIVE FUNCTIONING 44

which attempts to close the gap between biological age and social age. Once the

developmental process takes hold and a teen gets closer to adulthood with adult roles,

their interest in antisocial behavior diminishes as this maturity gap closes; they no longer

have anything to prove. Moffit (1993) suggested that the antisocial behavior that was

once thought of as rewarding is now punitive, and remits based on them now having

something to lose.

Moffitt (1993) defined the life-course-persistent delinquent as one who has

continuity of antisocial behavior over their life. They bite and hit at age four, shoplift and

are truant at age 10, sell drugs and steal cars at age 16, rob and rape at age 22, and

commit fraud and child abuse at age 30. She suggested that there is no reason to assume

that life-course persistent delinquents assume prosocial tendencies after several decades

of antisocial behavior.

In those whose behavior is stable and persistent, Moffitt (1993) suggested you

have to look for roots in early development. There could be disruption in the

development of the fetal brain by maternal drug abuse, poor prenatal nutrition, and

deprivation of nutrition, stimulation, and affection in early life. Two neuropsychological

deficits empirically associated with antisocial behavior are verbal and executive

functions. Verbal deficits may include difficulties with receptive listening and reading,

problem solving, expressive speech, writing, and memory. Executive functioning deficits

include problems with attention and impulsivity.

Three forms of interaction may play a role in promoting antisocial delinquency as

well as maintain it over a life course. One is evocative interaction, which occurs when a

child’s behavior evokes particular responses from others (Caspi, 1987). The other two
DIFFERENCES IN EXECUTIVE FUNCTIONING 45

interactions are reactive and proactive. Reactive aggression occurs when a person reacts

to an environment based on their style. A person may read into cues as harmful intent

even if there is none, and respond aggressively. Proactive interaction occurs when people

create an environment that supports their style, such as being selective with others they

interact with (Moffitt, 1993).

Moffitt (1993) posited two sources of continuity that she suggested are reasons

why life-course persistent delinquents are impervious to change. One is failing to learn

prosocial alternatives to behavior. She suggested that once the deviant behavior begins, a

lack of recourse to more prosocial alternatives keeps them on the course toward life

persistency. They miss opportunities at every stage of development due to being rejected

by peers and adults. Second is becoming entangled in a deviant life-style due to crime’s

consequences. Poor self-control and impulsivity increase the risk that antisocial youth

will make irreversible decisions that close the doors of opportunity.

Moffit (1993) suggested that the stability of antisocial behavior is linked to its

extremity; however, those that engage in extreme antisocial behavior will not necessarily

have stability over time. Gangs are generally involved in extreme antisocial behavior,

ranging from assault, selling drugs, car theft, and murder. They are a stable phenomenon

in the sense of their extreme crimes, but their stability is not implied and change is

possible depending on their own capacities for change. There have been multiple studies

that examine this proposed developmental taxonomy.

Raskin White, Bates, and Buyske (2001), examined delinquency that persists into

adulthood and hypothesized three trajectories would identify themselves in their study:

non-delinquents, adolescence-limited delinquents, and persistent delinquents that extend

DIFFERENCES IN EXECUTIVE FUNCTIONING 46

their behavior from adolescence into adulthood. They also thought these three groups

would differ on measures of neuropsychological functioning, personality, and

environmental risk.

Raskin White et al. (2001) used data collected as part of a larger longitudinal

study on adolescent development from the Rutgers Health and Human Development

Project (HHDP). Participants of the HHDP study included 698 males and 682 females

tested between 1979 and 1981 (Time 1, T1). Raskin White et al. (2001) only used male

participants for their study due to life-course persistent delinquency usually being

inclusive of men. Three cohorts were tested at T1, with the youngest cohort being age

12, the middle cohort being age 15, and the third cohort being age 18. Participants of

each cohort were tested three more times over fifteen years, with the first retest in 1982-

1984 (T2), the second in 1985-1987 (T3), and the last in 1992-1994 (T4). Of the original

1979-1981 sample, 1,257 returned for the T4 testing, providing a 91% response rate. The

original sample was 90% Caucasian with a median family income between $20,000 and

$29,000 in 1981.

Delinquency and risk factors were measured using a self-report questionnaire.

Delinquency was measured by the number of times a respondent engaged in petty theft,

grand theft, motor vehicle theft, breaking and entering, avoiding payment, vandalism,

assault, and armed robbery over the prior three year period. An aggressive scale was also

assessed by frequency of armed robbery, assault, and fight with a weapon or in a gang.

Risk factors of neuropsychological concerns were measured on the basis of two

birth risk factors: low birth weight and premature birth. At T3, additional

neuropsychological functioning measures were assessed in verbal ability and executive

DIFFERENCES IN EXECUTIVE FUNCTIONING 47

functioning. Verbal ability was measured by the Shipley Institute of Living Scale (SILS)

Vocabulary Test. Executive functioning was assessed by the Trail Making Test - Part B

(TRAIL) and the Booklet Category Test (BCT).

Personality measures employed were the Impulsivity and Harm Avoidance

subscales from the Jackson’s Personality Research Form E, an abbreviated version.

Raskin White et al. (2001) also assessed sensation seeking with the Disinhibition

subscale of Zuckerman’s Sensation Seeking Scale. Disinhibition was assessed at T2 for

the youngest cohort due to lack of reliability prior to age 14. Family adversity was

assessed with three measures: family SES, family structure, and family hostility.

Raskin White et al. (2001) employed a semiparametric, group-based modeling

technique to identify any delinquency trajectories. Seven age scores for delinquency

were used to model the trajectories: 12, 15, 18, 21, 24.5, 28, and 31. Each age group was

added and the analyses repeated to determine the optimal number of groups that best

describes the data. An EM algorithm was used to determine the trajectories and the

membership probabilities. Raskin White et al. (2001) then measured the association

between delinquency trajectory and the risk factors and their interactions with

hierarchical polytomous logistic regressions.

Raskin White et al. (2001) identified a four group trajectory model as best fitting:

non-delinquents, adolescence-limited delinquents, escalating delinquents, and persistent

delinquents. Persistent delinquents were identified as scoring significantly higher on the

delinquency scale than adolescence-limited at all ages except 12; however, the aggressive

scale only had statistically significant differences at ages 18 and 24.5. The escalating

group did not show any statistically significant differences until age 21, compared to the
DIFFERENCES IN EXECUTIVE FUNCTIONING 48

adolescence-limited group, and became significantly more delinquent than the persistent

group at age 31.

Raskin White et al. (2001) then looked at the interactions of neuropsychological,

personality, and environmental risk factors, and whether their influences could

distinguish the delinquency trajectories. They found that the main effects of these

predictors and age cohort were significant, as were the addition of the neuropsychological

-environmental interaction, which improved the model. This indicated that individually,

neuropsychological, personality, and environmental risk factors influence the course of

delinquency, as does the interaction of the neuropsychological and environmental risk

factors. The addition of the neuropsychological-personality interactions and personality-

environment interactions did not significantly improve the model fit, indicating these two

interactions did not influence the delinquency trajectories.

In examining odds ratios (OR), Raskin White et al. (2001) found that delinquents

had significantly greater odds of having higher impulsivity (OR = 1.15), lower harm

avoidance (OR = 1.11), higher disinhibition (OR = 1.41), higher parental hostility (OR =

1.03), and one parent families (OR = 2.16), than non-delinquents. Disinhibition was the

only risk factor found to be significantly different between adolescence-limited and

persistent delinquency, with an OR of 1.19. They also found that persistent delinquents

have significantly greater odds of higher disinhibition (OR = 1.20), lower verbal ability

(OR = 1.06), lower harm avoidance (OR = 1.15), higher parental hostility (OR = 1.03),

and coming from a one-parent family (OR = 3.49), than escalating delinquents.

Raskin White et al. (2001) not only identified the three trajectory models, but a

fourth model was noted, which was unexpected. They were able to identify some
DIFFERENCES IN EXECUTIVE FUNCTIONING 49

differences in the escalating delinquents from the persistent delinquents; however, more

follow up is needed on these two trajectories. Some of the predictor variables did not

produce significant effects, and more research into this area is needed. This may be a

matter of the risk factors being unable to distinguish adolescence-limited from persistent,

or from the escalating group identified, and because other risk factors need to be

researched to better gauge this older population and their trajectories for delinquent

behavior.

Piquero (2001) sought to assess the relationship between neuropsychological risk

and four manifestations of life-course persistent offending: early onset, violent and

nonviolent offending, seriousness of criminal offending, and chronic offending. He

hypothesized that poor performance on the neuropsychological tests would predict the

four manifestations to be in the positive direction toward life-course persistent offending.

Piquero (2001) gathered data from three sources: the National Collaborative

Perinatal Project (NCPP), Philadelphia schools, and Philadelphia Police Department.

The NCPP followed over 56,000 pregnancies from pre-term to age seven. For purposes

of his study, Piquero (2001) identified 207 African–American youth (144 males, 63

females) who had at least one recorded offense by age 18.

The four dependent variables were coded for ease of analysis. Onset of offending

was identified using age 14 as a cut-off; life-course persistent offenders were coded has

having an onset prior to age 14, and adolescent-limited offenders were coded as having

an onset at or after age 14. Offenders were also classified as either being nonviolent or

having offenses that were both nonviolent and violent. Seriousness of criminal offending

was identified via a score from the Sellin and Wolfgang seriousness scale. This scale
DIFFERENCES IN EXECUTIVE FUNCTIONING 50

takes into account the type of criminal acts and the aggravating factors associated with

the acts (Piquero, 2001). Finally, chronic offending was differentiated on the basis of

number of offenses incurred by age 18: four or less offenses were less chronic, and five

or more offenses were more chronic.

Independent variables examined were low birth weight, weak family structure,

biosocial interaction, and neuropsychological tests. Low birth weight was differentiated

by being either above or below six pounds. Weak family structure was measured by

summing the scores of three indicators when the child was age seven: number of changes

in the mother’s marital status, who the child lives with, and whether the father of the

child was present in the household. Biosocial interaction was assessed by examining the

relationship between low birth weight and weak family structure. The

neuropsychological test administered was the short form of the Wechsler Intelligence

Scale for Children (WISC).

Piquero (2001) also included two control variables: gender and socioeconomic

status (SES), which was a composite measure of three indicators collected when the boys

were age seven: education, income, and occupation of the head of household.

In looking at the relation between the WISC and the four manifestations of life-

course persistent offending, Piquero (2001) found that verbal and performance subscales

of the WISC were significantly positively correlated with one another. However, in

looking specifically at each of the subscales, the four manifestation variables were all

found to be negatively correlated with both the verbal and performance subscale of the

WISC; only some were significant. The verbal subscale was found to have the most

significant correlations; it was significantly negatively correlated with involvement in

DIFFERENCES IN EXECUTIVE FUNCTIONING 51

violent and nonviolent offending (r = -.182, p<.05), seriousness of criminal offending (r =

-.201, p<.05), and chronic offending (r = -.193, p<.05). The performance subscale was

only found to be significantly negatively correlated with involvement in violent and

nonviolent offending (r = -.147, p<.05).

Using hierarchical logistic regressions, Piquero (2001) first examined how each of

the five independent variables and two control variables might predict the four

manifestations of life-course persistent offending. In the second model of the regression,

he added the two WISC subscales to measure any effect they may have on the

manifestations of life-course persistence.

In the first regression, Piquero (2001) examined early onset of offending and how

birth weight, family structure, gender, SES, and biosocial interaction might predict its

occurrence. He found that low birth weight, SES, and the biosocial interaction of low

birth weight and family structure, significantly predicted early onset of offending. When

adding the WISC subscales, Piquero (2001) did not find any significant effect on early

onset.

In the results for involvement in violent and nonviolent offending, Piquero (2001)

found only gender to have a significant effect on violent and nonviolent offending in

model 1. In model 2, when the WISC subscales were added, gender remained a

significant effect, but the verbal subscale was also determined to have a significant effect

on involvement in violent and nonviolent offending. Essentially, those who score lower

on the verbal subscales are more likely to engage in both violent and nonviolent offenses.

When examining estimates for the prediction of the Sellin-Wolfgang seriousness

score in seriousness of offending, Piquero (2001) found that in model 1, only gender was
DIFFERENCES IN EXECUTIVE FUNCTIONING 52

found to have a significant effect on seriousness of offending. In adding the WISC

subscales in model 2, gender remained a significant effect, but the verbal subscale was

also found to have a significant effect on seriousness of offending. Males are

significantly more likely than females to be involved in seriousness of offending; those

who score lower on the verbal subscales are more likely to be involved in serious

offending.

Finally, in assessing chronic offending, Piquero (2001) determined that none of

the variables in the first model are significantly indicative of this behavior. When the two

WISC subscales are added in model 2, gender and the verbal subscale of the WISC are

found to have a significant effect on predicting chronic offending. As found in the prior

two predictor regressions, males are significantly more likely than females to be involved

in chronic offending, as well as those who score lower on the verbal subscales.

Piquero (2001) determined that poor verbal ability has a significant relation to

being more likely to engage in both violent and nonviolent offenses, serious offending,

and chronic offending. As Goldberg (2009) suggested, two evolutionary developments

occurred simultaneously: the emergence of language and the explosive rise of executive

functions, or more specifically, the frontal lobes. Although any relationship of these two

developments is not often discussed, Goldberg (2009) posited their emergence at the

same time in evolution is not a coincidence. Language has a generative power to build

new constructs; however, this ability may depend on the frontal lobes to make use of and

manipulate the language constructs into more complex operations. The question remains

whether the poor verbal ability demonstrated in this study is the result of poor cognitive

functioning or the result of a poor education.

DIFFERENCES IN EXECUTIVE FUNCTIONING 53

Although Piquero (2001) found that poor neuropsychological test scores can

predict various manifestations of life-course persistent offending, there were limitations

to his study. First, the generalizability of the study is limited by the class and race of the

sample. Second, the WISC was not able to directly measure neuropsychological risk.

The use of more direct tests may provide stronger results. Lastly, criminal history data

was only available through age 17. The neuropsychological test scores obtained as an

adolescent may affect the ability to predict criminal activity into adulthood, and affect the

categorization of the sample as life-course persistent or adolescent-limited offenders.

Also, the criminal history of the sample is from the 1980s, and criminal offenses

committed then may have limited generalizability to the offenses committed by offenders

today.

Huesmann, Dubow, and Boxer (2009) examined long-term consequences of

aggressive behavior in childhood and adolescence, and how continuity or change in

behavior might predict negative or positive outcomes in middle adulthood. They

analyzed the data from the Columbia County Longitudinal Study, initiated in 1960 with

856 children in third grade, and followed their aggressive and antisocial behavior from

age eight to 30. Huesmann et al. (2009) extended this study to predict criminality

through age 48.

This study was initially broken into three waves; Huesmann et al. (2009) added a

fourth wave. The first wave included 856 children at eight years old. Over 90% were

Caucasian, and 51% were male. The number of ethnic minorities was too small to allow

separate analyses, so they were excluded from this study. Eighty-five percent of the

participants’ mothers and 71% of the fathers were interviewed in this wave.
DIFFERENCES IN EXECUTIVE FUNCTIONING 54

Wave 2 was conducted in 1970 when the participants had a modal age of 19.

Four hundred and twenty-seven of the original 856 children participated, with 211 boys

and 216 girls. They had completed an average of 12.6 years of education. In 1981,

Wave 3 was conducted with 409 of the original participants. The modal age was 30, and

consisted of 198 males and 211 females. The average education level was some college

or a technical school (Huesmann et al., 2009).

Wave 4 was completed between 1999 and 2002. Five hundred and twenty-three

of the original 856 sample completed this wave (61%). This sample consisted of 268

males and 255 females, with an average age of 48.46 years. The average education level

was between some college and a college degree. The sample sizes vary considerably

over all four waves depending on the variables that are studied and what information is

available. For this study and aggression, complete data is available for 285 participants

for the first three waves and 230 participants for all four waves (Huesmann et al., 2009).

Huesmann et al. (2009) incorporated the three specific aggression measures used

in the original study into their fourth wave. Peer-nominated aggression was assessed at

ages eight and 19 using ten peer-nominated aggression items that cover physical, verbal,

acquisitive, and indirect acts of aggression. Severe physical aggression was assessed at

ages 19, 30, and 48, via the participants’ self-reports of how often in the prior year they

had engaged in each of the following four behaviors: choked someone, slapped or kicked

someone, punched or beaten someone, knifed or shot at someone, or threatened to do so.

Aggressive personality was assessed at ages 19, 30, and 48, with the sum of scales 4, 9,

and F of the MMPI.

DIFFERENCES IN EXECUTIVE FUNCTIONING 55

Adult outcomes were identified with 13 variables in the last wave of the original

study, and Huesmann et al. (2009) applied these same items to wave 4. These items,

which assist in measuring the outcome of the participants’ adult life, include self-reports

of criminal behavior, self-reports of traffic violations, antisocial behaviors, aggression,

aggression toward spouse, divorce, depression, self-reported health, problem drinking,

occupational status, educational attainment, verbal achievement, and frequency of

religious attendance.

To gauge the continuity of aggression over multiple waves, Huesmann et al.

(2009) used structural equation modeling first over the first three waves, and then

recomputed the data with their current data over all four waves. The first three waves

produced disattenuated stability coefficients of .50 for males and .35 for females. After

recomputing the model with their data from wave 4, the disattenuated continuity

coefficients were very similar, .50 for males and .42 for females.

In order to assess “who” is contributing to the stability of aggression, the number

of initially high and low participants that stay high or low was compared on all four

waves. Two criteria were used to identify those that scored high on aggression on all

four time points, and who scored low on all four time points. The first criterion was

those that fell above or below the overall median in any wave, and the second was those

that fell above or below the 33rd and 67th percentiles in any wave.

Huesmann et al. (2009) found that just as many initially low-aggressive

participants stay low (37%) through middle adulthood as do initially high-aggressive

participants stay high (35%) when using the median. This suggests the continuity of

aggressiveness is not solely due to the high-aggressive staying high. When the results are
DIFFERENCES IN EXECUTIVE FUNCTIONING 56

broken down by gender, males tend to hold their position in their classification (40%)

more than females (33%). Males stay in the high category (38%) more so than females

(18%); however, on the other end, males and females tend to consistently stay in the low

category (38% and 36% respectively).

Huesmann et al. (2009) used logistic regression to examine what child and family

characteristics can predict life-course persistent aggression. They compared the life-

course persistent aggressive group with the life-course persistent nonaggressive group

and found that being male (OR = 4.2), having parents with lower education (OR = 0.25),

having parents who hit you (OR = 3.5), and having parents who reject you (OR = 1.3), all

have significant effects in increasing the odds a child will have life-course persistent

aggression.

Huesmann et al. (2009) defined five subsets of individuals based on their

composite aggression scores. The subsets are life-course persistent low aggressive, life-

course persistent high aggressive, adolescent-limited aggressive, childhood-limited

aggressive, and late-onset aggressive. In looking at the scores of each of these trajectory

groups on the adult outcome measures, the life-course persistent high aggressive group

had consistently poorer outcomes at age 48 than any other group. For example, 7% of the

life-course persistent aggressive group were arrested between ages 30 and 48, compared

to none of the other groups being arrested during that time. This group scored

significantly higher on general aggression as well as aggression toward spouse. They

were significantly more likely to report problems with alcohol and scored significantly

more depressed.
DIFFERENCES IN EXECUTIVE FUNCTIONING 57

The childhood-limited and adolescent-limited groups exhibited fewer negative

consequences than the life-course persistent low aggressive group. The only differences

reported were more problems with alcohol, and the adolescent-limited group scored

higher on the trait of aggression at age 48 than did the life-course persistent low

aggressive group. The late-onset group showed more negative outcomes than the low-

aggressive group, but not as many as the life-course persistent high aggressive group

(Huesmann et al., 2009).

This study suggests that continuity of aggression can be viewed as a phenomenon

that occurs over entire population levels and is not driven by one trajectory group

maintaining their behavior over a life course (Huesmann et al., 2009). Continuity was

found to be the result of low aggressive remaining low as much as it was high aggressive

remaining high.

Donnellan, Ge, and Wenk (2000) tested a hypothesis derived from Moffitt’s

(1993) developmental taxonomy theory that life-course persistent criminals will perform

poorer on cognitive ability tests than offenders who desist from criminal behavior in

adolescence. They used a unique data set of 20 years of arrest records from a large and

diverse sample who were assessed with a battery of cognitive tests.

The sample was part of a larger study of 4,164 adolescent males committed to the

Reception Guidance Center at the Duel Vocational Institution in Tracy, California,

between January 1964 and December 1965. Using arrest records collected in 1984 and

1985 to classify the participants, 888 in the adolescent-limited category and 774

offenders in the life-course persistent category, Donnellan et al. (2000) identified 1,662

offenders for their final analysis. The ethnic breakdown was 27% African American,
DIFFERENCES IN EXECUTIVE FUNCTIONING 58

52% Caucasian, and 21% Hispanic. The ethnic breakdown of the offender classification

system was notably different. The adolescent-limited category was 17% African

American, 68% Caucasian, and 15% Hispanic, while the ethnic breakdown of the life-

course persistent category was 38% African American, 34% Caucasian, and 28%

Hispanic. The average age of the sample was 19.

Data were collected using self-report questionnaires, a battery of psychological

tests, and official arrest records. The cognitive measures used were the California

Achievement Test (CAT), California Test of Mental Maturity (CTTM), General Aptitude

Test Battery (GATB), and the Raven Test of Progressive Matrices, all administered in

1964 and 1965 upon the youths’ admission to the institution in Tracy.

With such a diverse sample, Donnellan et al. (2000) controlled for any impact of

ethnicity by performing a series of MANOVAs and ANOVAs on each of the cognitive

tests for each ethnic group. Three separate MANOVAs were run for the CAT, CTTM,

and GATB, and an ANOVA was run for the Raven Test.

Caucasians and Hispanics both showed substantial differences in performance

between adolescent-limited and life-course persistent offenders; African Americans did

not. In the Caucasian sample, adolescent-limited performed significantly better on the

cognitive measures than did the life-course persistent offenders in nine of 12 tests. The

non-significant results were evidenced in the spatial and perceptual tests of the GATB

and the Raven Test.

In the Hispanic sample, there were significant differences in six of the 12 tests

between the adolescent-limited and life-course persistent offenders. The non-significant

differences were in the reading and arithmetic tests of the CAT, the language and non-
DIFFERENCES IN EXECUTIVE FUNCTIONING 59

language tests of the CTTM, the spatial test of the GATB, and the Raven test. The

observed effect sizes for significant results in both the Caucasian (.17 to .26) and

Hispanic (.22 to .36) samples were small using Cohen’s designations, which was

expected (Donnellan et al., 2000).

The African American sample did not provide any significant differences between

adolescent-limited and life-course persistent offenders. Although effect sizes were

expected to be small, in this sample, they were particularly weak with only three greater

than .10. In three of the tests, the effect sizes were negative due to the observed

differences being in the opposite direction from expectations (Donnellan et al., 2000).

Hispanics and African Americans had higher numbers of life-course persistent

offenders (66.3% and 61.7%, respectively) than did Caucasians (30.2%). Future work

would need to address the role of ethnicity in criminal behavior and its influence over

life-course criminality. Second, the study did not control for SES and its role in any

relationship between offender classifications (Donnellan et al., 2000).

Donnellan et al. (2000) found that the theory of differences in cognitive abilities

between adolescent-limited and life-course persistent offenders is ethnicity dependent.

One explanation Donnellan et al. (2000) postulates is that cognitive abilities are not as

influential in the contexts in which African Americans live, and we must account for the

importance of a person-context interaction when considering any relationship between

crime and cognitive abilities.

These studies that measure delinquency trajectories suggest that life course

persistent offenders begin their trajectory of crime in childhood, much sooner than

adolescent-limited offenders who tend to begin later and age out in late teens or early
DIFFERENCES IN EXECUTIVE FUNCTIONING 60

twenties. This may suggest a different cognitive development trajectory, and something

that needs further research. In addition, as found by Donnellan et al. (2000), ethnicity

may play a large role in determining the delinquency trajectory, as does environmental

factors as determined by Raskin White et al. (2001).

Executive Functioning

Executive Function and Biology

Goldberg (2009) likened the prefrontal cortex to a conductor in an orchestra or a

CEO of a company; it coordinates and integrates all the other structures of the brain. He

stated that human cognition is forward looking and we require a strong mechanism that

allows us to plan, dream, aspire, and set goals. The prefrontal cortex is the “seat of

intentionality, foresight, and planning” (Goldberg, 2009). It allows us to reason and

regulate emotion. As a process, the prefrontal cortex is otherwise known to facilitate the

executive functions. Once the abilities of the prefrontal cortex and executive functions

are compromised, this results in cognitive inflexibility which may contribute to

impulsivity and the inability to see another’s point of view. Some research suggests that

dysfunctions in executive functioning may contribute to aggressive or antisocial

behavior.

Since Gottfredson and Hirschi (1990) proposed their general theory of crime,

there have been many studies on the antecedents to antisocial behavior. Some studies

support Gottfredson and Hirschi’s (1990) social theory of crime and others support a

more biogenic cause for antisocial behavior. One neurobiological explanation is by van

Goozen, Fairchild, and Harold (2008), who argue that neurobiological deficits link
DIFFERENCES IN EXECUTIVE FUNCTIONING 61

genetic influences and early trauma or adversity to antisocial behavior. They also

contend that low stress-response systems are indicative of antisocial behavior.

From a biological standpoint, van Goozen, Fairchild, and Harold (2008) suggest

that serious stressful events early in life play an important role in “programming” the

stress systems. Those who grow up in adverse conditions and communities, such as gang

members, are commonly in the face of chronic stress in the form of safety concerns,

employment concerns, and where the next meal might come from, among others. This

might result in an adaptive down-regulation of the stress-response system, which would

include avoidance of chronic arousal and energy expenditure that could result in serious

pathophysiological consequences. As a result of this lowered stress-response system, van

Goozen, Fairchild, and Harold (2008) posited that antisocial individuals are less likely to

avoid stressful situations and are likely more fearless and less concerned about possible

negative consequences.

From the perspective of biology and heritability, there is some research that

becoming deviant has some physiological causation. The literature has demonstrated that

there are psychosocial risk factors for antisocial behavior, as well as biological risk

factors; but how do these two risk factors interact? Raine (2002b) suggested that due to

the human’s ability to live in a close social system, mostly as a result of the prefrontal

cortex, it is necessary that we have a system of effective aggression regulation. However,

there are biological processes that can affect this system regulation and cause reasoning

and decision making to go haywire. Raine (2002b) argued that there is a prefrontal

dysfunction theory based on the delayed growth of the prefrontal cortex, which can
DIFFERENCES IN EXECUTIVE FUNCTIONING 62

contribute to the understanding of crime in adolescence. He entertained the notion that

social and executive function demands are being placed on adolescents.

As Raine (2002b) explains, the prefrontal cortex is still developing during

adolescence, which is a time of increased requisite functionality, and continues to

develop into the 20s and even the 30s. There are enormous demands during late

adolescence where efficient regulation and control of a growing sex drive, challenges to

social status, and planning for the future is necessary. Adolescents in a lower SES

community tend to have more complex challenges and competing day to day struggles.

With this magnified cognitive load, Raine (2002b) hypothesized that an overload on the

prefrontal cortex occurs. This overload may cause prefrontal dysfunction and a lack of

inhibitory control over antisocial, violent behavior that peaks during adolescence (Raine,

2002a). This age is the prime time that youths consider gang activity, and a time when

they are becoming more independent of their parents and learning ways to express

themselves. Raine (2002a) summarized that genetic processes need an environment to

become expressed, and those with prefrontal dysfunction and poor executive functions

would be predisposed to antisocial behavior. This behavior would be most pronounced in

those with a less structured, less stable psychosocial environment, which most gang

members tend to experience in both their homes and in their communities.

Executive Function and Antisocial Behavior

Many gang members continue a trajectory of delinquent behavior once the

prefrontal cortex has matured and they have a full capacity to manage and self-regulate
DIFFERENCES IN EXECUTIVE FUNCTIONING 63

behavior. This raises the question of how the prefrontal cortex might be altered for life-

long response patterns due to adolescent exposure and responses to stimuli.

Reactive versus proactive aggression, or as some call it, automatic versus

instrumental aggression, has been discussed in the literature. Specifically, free will,

aggression, and psychopathy were examined from a neurocognitive perspective by Blair

(2007). Reactive aggression occurs as a result of responses to stimuli becoming so over-

learned they occur automatically; there is no thought process that goes into these

responses. Reactive aggression is often the result of a frustrating event and frequently

induces anger. Instrumental aggression incorporates the idea of free will; a person made

one decision when they could have made another. It is goal oriented and involves

executive functioning processes that allow one to come to their desired outcome. Most

antisocial behavior is instrumental, as it usually entails goal oriented acts such as theft,

fraud, or robbery (Blair, 2007). Although some instrumental aggression can become

reactive due to over-learning, particularly due to learned responses to the complex social

milieu encountered in lower SES communities, these behaviors appear to be choices.

People with personality impairments tend to choose an aggressive course of action due to

deficient decision making.

Raine, Moffit, Caspi, Loeber, Stouthamer-Loeber, and Lynam (2005) attempted to

make contributions to five gaps they identified in neurocognitive impairments and

antisocial behavior: whether offenders have spatial as well as verbal impairments;

identify any impairments in memory functions; characterize neuropsychological

impairments in life-course persistent (LCP) and/or adolescent-limited (AL) offenders;

characterize neurocognitive impairments in childhood-limited (CL) offenders; and

DIFFERENCES IN EXECUTIVE FUNCTIONING 64

determine whether impairments can be attributable to ADHD, abuse, social adversity, or

head injury.

Participants for their study came from the youngest cohort of the Pittsburgh Youth

Study, which measured 21 serious antisocial behaviors. The original sample for the

Pittsburgh Youth Study had 503 participants; however, due to attrition and other reasons,

Raine et al. (2005) had a total of 335 participants for their biosocial study of aggression

and violent behavior. These participants had an average age of 16.5 at the time of the

testing, with an ethnic breakdown of 41.2% Caucasian and 58.8% African American.

To measure for delinquency, assessments were gathered by parents, teachers, and

the boys themselves, beginning when the participants were age seven, and taken every six

months from ages seven to eleven, then taken annually up to age 17. Participants were

classified into six levels of delinquency seriousness which were determined by the most

serious reported act.

Raine et al. (2005) also assessed environmental factors and ADHD at age seven,

and assessed neurocognitive functioning and any history of head injury at age 17. In

order to meet the diagnostic criteria of ADHD, participants had to present with at least six

months duration with eight of the 14 behaviors. The rate of ADHD in this sample at age

seven was 14.1%. Raine et al. (2005) also incorporated child abuse data that was

collected from official court records at the Children and Youth Services offices. In

addition, Raine et al. (2005) initially selected 19 measures of psychosocial adversity;

however, after a factor analysis and Oblimin rotation were performed, a screen test

identified three factors of psychosocial adversity: poverty, parental psychopathology, and

parental neglect. All 19 of the initial measures loaded on one of the three factors.
DIFFERENCES IN EXECUTIVE FUNCTIONING 65

The assessment of neurocognitive functioning included various tests that can

measure verbal and spatial IQ, verbal memory, and visual-spatial memory, using subtests

of the WAIS and Wechsler Memory Scale, as well as tests that can measure for any head

injuries in frontal lobe functioning such as the Continuous Performance Task (CPT), the

Wisconsin Card Sorting Task (WCST), and Verbal Dichotic Listening.

In a cluster analysis, Raine et al. (2005) placed significantly varying delinquency

scores into four groups. The groups identified were the control group, who remained low

on antisocial behavior throughout the study. The second group was the adolescent

limited (AL) group, who started off near the control group, but progressed to significant

levels of antisocial behavior by late adolescence. The third group is the childhood

limited (CL) group, who has high antisocial behavior up to about age eleven, then it

drops off. Lastly, the life course persistent (LCP) group starts off with high antisocial

behavior that spikes in late adolescence.

A MANOVA found that overall the groups significantly differed on

neurocognitive functioning. Specifically for intelligence, Raine et al. (2005) found

significant main group effects on verbal and spatial IQ. The LCP group had lower verbal

scores compared to the ALs, as well as lower verbal and total IQ than the control group.

The CL group scored lower than the control in verbal, spatial, and total IQ measures.

Raine et al. (2005) did not find any significant difference in the three IQ measures

between the ALs and the controls.

In measuring memory, there was a significant difference in immediate and

delayed recall on both verbal and visual-spatial memory. The LCP group yielded spatial

but not verbal memory deficits, and the CLs showed deficits in all measures except
DIFFERENCES IN EXECUTIVE FUNCTIONING 66

spatial immediate memory. With regards to frontal functioning, individual ANOVAs

found significant differences on the CPT, with the LCP group and CLs having had

performed worse than the control group. There were no significant effects on the WCST.

Finally, in dichotic listening, a MANOVA indicated significant performance results in

right-ear listening, with the CL group performing significantly worse than both the

control group and AL group.

When measuring the additional variables of psychosocial adversity, abuse, history

of head injury, and ADHD, they were all found to have significant effects on antisocial

grouping, with the exception of head injuries. Raine et al. (2005) viewed head injury as a

potential confound, and ran an analysis with uncorrected post hoc comparisons to assess

whether more head injuries would be found in an LCP group. Out of the six types of

head injuries measured, only number or times knocked unconscious provided evidence

that could indicate an LCP group has a greater history of head injuries than the control as

well as the CL group.

With regards to court-recorded abuse, the LCP group was found to have almost

four times the rate of abuse than the control group, and more than two times the rate of

abuse than the AL group. The CL group had more than two times the rate of abuse than

the control group, but did not differ from the LCP group. ADHD was found to have rates

twice as high in the LCP and CL groups compared to the control and AL groups.

All three measures of psychosocial adversity: poverty, parental psychopathology,

and parental neglect, were found to have a significant influence on antisocial grouping.

The LCP group was found to have higher levels of poverty and parental neglect than the
DIFFERENCES IN EXECUTIVE FUNCTIONING 67

control and AL groups, and the CL group had higher levels of all three measures than the

control group.

Raine et al. (2005) needed to determine whether ADHD, particularly with its high

rate of occurrence in this sample, and psychosocial adversity, were confounds that could

impact the neurocognitive-delinquency relationship. The main group effects remained

significant after these were both added as covariates, and determined that any links in

neurocognitive and antisocial behavior were independent of the influence of both ADHD

and psychosocial adversity.

In contributing to the five gaps in neurocognitive impairments and antisocial

behavior with the above results, Raine et al. (2005) determined that offenders

demonstrate spatial and verbal impairments; they are impaired on memory tasks, life

course persistent offenders are more impaired on neurocognitive and psychosocial factors

than the control, childhood-limited offenders do have cognitive impairments, and finally,

cognitive impairments are not attributable to child abuse, psychosocial adversity, head

injuries, or ADHD. Neurocognitive measures were not assessed prior to the onset of any

antisocial behavior, so a direct causal pathway cannot be confirmed. In addition, results

pertain to those on a path to lifetime offending. Raine et al. (2005) suggested follow up

past age 17 years into adulthood would be needed to truly assess whether these

individuals are truly life course persistent antisocials.

After a review of literature on temperament, aggression, and executive

functioning, Giancola, Roth, and Parrott (2006), proposed to test a hypothesis that

executive function (EF) mediates the relation between difficult temperament (DT) and

aggression. Participants were recruited by newspaper ads in Lexington, Kentucky and

DIFFERENCES IN EXECUTIVE FUNCTIONING 68

included 310 men (n = 152) and women (n = 158) between 21 and 35 years old. There

were 286 Caucasians, 23 African Americans, and one Hispanic. Any reports of drug or

alcohol problems, serious head injuries, learning disabilities, or serious psychiatric

symptomatology, were excluded.

Giancola et al. (2006) took into consideration functional and neuroanatomical

elements to guide selection of executive functioning measures. From a functional

perspective, tests of EF were chosen that encompassed a wide variety of skills such as

attentional control, strategic goal planning, abstract reasoning, cognitive flexibility (set

shifting), hypothesis generation, inhibition, and the ability to organize and utilize

information contained in working memory. From a neuroanatomical perspective, the tests

that were selected are generally accepted measures of functions that originate primarily

from the prefrontal cortex. The cognitive tests administered were the Porteus Maze Test,

Go/No-Go Task, Trails B of the Trail Making Test, Stroop Task, Conditional Associative

Learning Test, Tower of Hanoi, and the Wisconsin Card Sorting Test.

Temperament and aggression were measured using self-report questionnaires.

Temperament was assessed with the Dimensions of Temperament Survey-Revised

(DOTS-R). The DOTS-R is comprised of 10 scales, none of which measure aggression.

The 10 subscales were summed and a difficult temperament index was created. Physical

aggression was measured using the Buss-Perry Aggression Questionnaire (BPAQ). This

self-report has four subscales: Physical Aggression, Verbal Aggression, Anger, and

Hostility. For the purposes of this study, aggression was indexed by the scores on the

Physical Aggression subscale.

DIFFERENCES IN EXECUTIVE FUNCTIONING 69

Giancola et al. (2006) used a linear, 3-step design to measure the mediation of EF

and the relation between DT and aggression. They used regression on: (a) the mediator

(EF) on the independent variable (DT); (b) the dependent variable (aggression) on DT;

and finally (c) aggression on both EF and DT. Mediation is noted as present if: (a) the

relations in equations (a) and (b) are significant; (b) EF is significantly related to

aggression equation (c); and (c) the influence of DT on aggression is reduced following

the inclusion of EF (Giancola, et al., 2006). These tests were performed for men and

women separately.

The design described above was found to be significant for men with: (a) EF onto

DT (B = -.15, p < .05); (b) aggression onto DT (B = -.17, p < .05); and (c) aggression

onto EF and DT (B for DT = -.12, p = ns; B for EF = .31, p < .001). Adding EF to the

equation reduced the relation between DT and aggression by 30%, making it non-

significant, and therefore indicating that EF mediated the relation between DT and

aggression for men (Giancola et al., 2006).

The relation between DT and aggression was not significant for women (B = -.09,

p = ns); therefore, mediation could not be tested. However, Giancola et al. (2006) did

note that the relation between EF and aggression was significant for women (B = .24, p <

.05).

The aggression aspect of this study might have had more valid and reliable

findings if additional measures of aggression had been utilized, like clinical interviews

and collateral reports of the participants’ history of violence. It is also thought that the

generalizability of the results might have been broadened by including individuals with a
DIFFERENCES IN EXECUTIVE FUNCTIONING 70

clinical diagnosis associated with some type of disinhibitory psychopathology (Giancola

et al., 2006).

This study suggests that by enhancing executive function, a difficult temperament

can be mitigated, which in turn can decrease the act of aggression (Giancola et al., 2006).

In essence, increasing cognitive flexibility, attentional skills, planning, and organizing

ability can reduce low adaptability to change, distractibility, and impulsivity, and lead to

reduced acts of aggression toward others.

Dolan and Anderson (2002) looked to examine relationships between aggression,

impulsivity, and neuropsychological functioning. They hypothesized that aggression and

impulsivity will be highly correlated, and those with high aggression/impulsivity scores

would exhibit greater executive and memory deficits than healthy controls as well as

those who assault but are not impulsive.

Dolan and Anderson’s (2002) sample consisted of 60 DSM III-R diagnosed

personality disordered offenders detained in maximum security hospitals, as well as 27

healthy male staff. All participants were Caucasian, between 18 and 50 years old, had no

current or lifetime Axis I diagnosis, and had been medication-free for at least 12 months.

The average length of hospital stay for the offenders was 5.5 years, and all had a history

of convictions for violent or sexually violent offenses.

The measures administered included two intelligence tests: the National Adult

Reading Test (NART) and the Wechsler Adult Intelligence Scale – Revised (WAIS-R).

Executive function tests included the Wisconsin Card Sorting Test (WCST), Stroop

Neuropsychological Screening Test (SNST), the Proverbs Interpretation Test (PT),

Controlled Oral Word Association test of Verbal Fluency (VF), Trail Making Test
DIFFERENCES IN EXECUTIVE FUNCTIONING 71

(TMT), and the Cognitive Estimation task. Memory tasks included the logical memory

(LM) and visual reproduction (VR) tests from the Wechsler Memory Scale – Revised

(WMS-R), and the Recognition Memory Test (RMT).

A composite executive (frontal) function score was derived by summing the

scores of the perseverative error score on the WCST, the B-A trails score, and the

Cognitive Estimation, Proverbs Interpretation, Verbal Fluency, and Stroop (SNST)

scores. A composite memory (temporal) function score was derived from the summed

scores of the Logical Memory – immediate and delayed, Visual Reproduction Memory –

immediate and delayed, and Word Recognition on the RMT.

Participants also completed the Special Hospital Assessment of Personality and

Socialisation (SHAPS), an assessment of personality based on Belligerence and

Withdrawal (Dolan & Anderson, 2002). SHAPS psychopaths were differentiated from

non-psychopaths by high scores on the Belligerence factor of the measure. They

determined there were 51 psychopaths and nine non-psychopaths. The Withdrawal

dimension differentiated primary psychopaths with low scores and secondary

psychopaths with high scores.

Subjects also completed trait impulsivity and aggression assessments. A

composite impulsivity score was derived by the summed scores of the impulsivity

subscales of the Impulsiveness-Venturesome-Empathy inventory (IVE), the Barratt

Impulsivity Scale (BIS), and impulsivity subscale of the SHAPS. A composite

aggression score was derived from the aggression subscales of the Buss-Durkee Hostility

Inventory (BDHI), the Brown and Goodwin lifetime history of Aggression scale (BGA),

and the SHAPS aggression subscale (Dolan & Anderson, 2002).

DIFFERENCES IN EXECUTIVE FUNCTIONING 72

Using an ANOVA, Dolan and Anderson (2002) found that SHAPS psychopaths

as well as non-psychopaths had higher composite aggression scores than controls.

SHAPS psychopaths also had significantly higher composite impulsivity scores than both

the SHAPS non-psychopaths and controls. On the intelligence measures, SHAPS

psychopaths were found to have lower mean WAIS-R full and verbal subscale scores

than both controls and non-psychopaths. The SHAPS psychopaths had lower mean

scores on the WAIS-R performance tests than the non-psychopaths, but did not differ

significantly from the controls on this subscale.

When examining the neuropsychological tests in relation to differences among

psychopaths, non-psychopaths, and controls, there were several significant group

differences. SHAPS psychopaths exhibited poorer performance compared with controls

on a number of executive function tasks: WCST, Verbal Fluency, Logical Memory -

immediate and delayed recall, Visual Reproduction - immediate, Recognition Memory

for faces (RMF), and composite memory score. There were fewer significant group

differences SHAPS psychopaths exhibited poorer performance on, compared with both

controls and non-psychopaths: the Stroop (SNST), Proverbs Interpretation, Cognitive

Estimation, and composite executive score.

To reduce the number of variables and neuropsychological data, Dolan and

Anderson (2002) had all of the executive function and memory tests ran under a principal

components analysis (PCA). They identified four principal components which accounted

for 67.5% of the variance. Factor 1 was identified as frontal/concept formation and

loaded Cognitive Estimation, Proverbs Interpretation, perseverative errors of the WCST,

and Recognition Memory for Faces. Factor 2 was labeled semantic/contextual memory
DIFFERENCES IN EXECUTIVE FUNCTIONING 73

function and loaded Logical Memory – immediate and delayed recall. Factor 3 was

labeled attention/concentration and loaded Verbal Fluency, Trail Making Test, and

Stroop (SNST) test. Factor 4 was labeled unspecified memory and loaded Visual

Reproduction memory and Recognition memory for Words.

An ANOVA indicated significant group differences with the SHAPS psychopaths

performing worse than controls on Factors 1 and 2. Psychopaths also scored significantly

worse than non-psychopaths on Factor 1. When considering the composite scores of

executive function and memory function, Dolan and Anderson (2002) found that SHAPS

psychopaths had significantly lower executive function scores than both the controls and

non-psychopaths, and remained significant after correcting for IQ. There were no

significant differences among the groups on the composite memory scores, even after

correcting for IQ.

As hypothesized, Dolan and Anderson (2002) found the composite impulsivity

and aggression measures to be highly correlated (r = 0.83, p < 0.001). In addition,

correlations between IQ and composite measures of executive (r = 0.75, p < 0.001)

function and memory (r = 0.46, p < 0.001) function were also found to be highly

significant.

The composite impulsivity score negatively correlated with IQ (r = -0.33, p <

0.01), composite executive function score (r = -0.52, p < 0.01), and composite memory

function score (r = -0.34, p < 0.05). The composite aggression score also negatively

correlated with IQ (r = -0.21, p < 0.05), executive function (r = -0.46, p < 0.01), and

memory function (r = -0.29, p < 0.05). Essentially, the higher the impulsivity and

aggressive scores, the lower the IQ score, executive function score, and memory function
DIFFERENCES IN EXECUTIVE FUNCTIONING 74

score. When adjusting for IQ, Dolan and Anderson (2002) determined that correlation

between composite impulsivity and aggression scores and memory function lost their

significance; however, the correlation between composite impulsivity and aggression

scores and executive function remained significant.

Seguin, Pihl, Harden, Tremblay, and Boulerice (1995) examined what types of

cognitive functions might differentiate boys with a history of physical aggression from

those without such a history. Participants were a community sample of White, French-

speaking boys from 53 schools with the lowest SES index of a Montreal School Board.

The boys were assessed annually from ages six to 12, as were their parents and teachers.

Physically aggressive behavior was assessed at ages six, ten, eleven, and 12, with the

fighting subscale of a French Canadian version of the Social Behavior Questionnaire.

Three items made up this scale: (1) fights with other children, (2) kicks, bites and hits

other children, and (3) bullies or intimidates other children.

After removing those boys who either withdrew or had missing values on

aggression at follow up, Seguin et al. (1995) identified 893 boys who had stability and

severity of physical aggression. They defined aggression as stable physical fighting over

a seven-year period from ages six to 12. Nineteen percent of the sample was classified as

stable aggressive boys and fell above the 70th percentile at two or more assessment points

at age six. Nonaggressive boys fell below the 70th percentile at all assessment points, and

comprised 35% of the sample. Participants who did not meet criteria for either of the

prior two classifications were identified as unstable aggressive boys and made up 46% of

the sample.
DIFFERENCES IN EXECUTIVE FUNCTIONING 75

Seguin et al. (1995) identified a subsample of boys at age 13 with overlapping

criteria. After exclusion criteria were applied, a total of 203 boys agreed to laboratory

testing at age 13. Of those, 177 returned for further testing at age 14. The 177 boys who

completed both years of testing, 63 were classified in the stable aggressive group, 59 in

the nonaggressive group, and 55 in the unstable aggressive group. Seguin et al. (1995)

also compiled a familial adversity composite index from variables collected from the

mother of the boys when they were age six, including parental age at first child’s birth,

parental education level, family status, and SES.

The measures used were validated with the task-specific double-dissociation

paradigm and covered a broad range of abilities, including verbal memory and executive

functions, cerebral dominance, and spatial memory. They were Dichhaptic

Lateralization, Digit Span, Nonspatial Conditional Association, Paired Associates, Self-

Ordered Pointing, Spatial Memory, Strategic Problem Solving, Subjective Ordering, and

Verbal Fluency.

Seguin et al. (1995) controlled for age, gender, race, cultural background,

education level, SES, and brain damage. They broke the analysis into two parts: part 1

was the initial study from ages six to 12 where they performed a factor analysis of the test

battery to summarize any pattern of correlations among the large number of variables and

isolate dimensions of cognitive processes; part 2 was the subsample of 13 and 14 year

olds where they used the factor-analyzed dimensions from part 1 in a MANCOVA to

identify which of the three levels of severity and stability of aggression the dimensions of

cognitive processes would differ.

DIFFERENCES IN EXECUTIVE FUNCTIONING 76

In the part 1 factor analysis, Seguin et al. (1995) identified twelve variables of

functioning from the nine psychological tests administered that incorporated executive

functioning, memory, and cerebral dominance. In order to reduce the number of

variables, they performed principal components extraction followed by oblique rotation.

This resulted in four factors, which accounted for 58% of the variance. Factor 1

contained variables that reflect verbal learning. Factor 2 represented Spatial Incidental

Learning and includes variables that assess spatial memory tasks. Factor 3 represented

cerebral dominance in the form of Tactile Laterality, and Factor 4 included variables that

measure executive functioning.

While controlling for social disadvantage and anxiety, a 1 x 3 between-subjects

MANCOVA was performed on each of the four factors in part 2. Seguin et al. (1995)

found the multivariate effect of the family adversity and anxiety covariates to be

significant. In investigating the effect of the covariates on each of the factors with

multiple regressions, they were found to provide significant adjustment only to Factor 1,

verbal learning.

Seguin et al. (1995) entered Spatial Incidental Learning and Tactile Laterality in

the model prior to Verbal Learning in order test the hypothesis that deficits in Verbal

Learning were more associated with physical aggression than deficits in spatial learning

or cerebral dominance. They did the same with executive functioning and entered this

factor into the model last to test the hypothesis that deficits in executive functioning are

associated with physical aggression over and above all other cognitive-

neuropsychological abilities.
DIFFERENCES IN EXECUTIVE FUNCTIONING 77

The overall MANCOVA was found to be significant (p = .020), and significant

main effects were found on the executive functions factor only (p = .005). The stable

aggressive boys were found to perform more poorly on all factors except cerebral

dominance, or Tactile Laterality. When comparing the classification of boys within the

executive function factor, the comparison between the stable aggressive boys and

nonaggressive boys was significant (p < .0003), as was the unstable aggressive boys

compared with the nonaggressive boys (p < .007).

In this multivariate design, executive functioning was the most strongly related to

physically aggressive behavior. When impaired, likely deficits may include an inability

to organize several stimuli at once, anticipate consequences of choices, and reflect

abstractly. One may quickly become overwhelmed with these impairments and unable to

give a more socially appropriate response to a situation, become overwhelmed, and result

in impulsivity (Seguin et al., 1995).

Bergeron and Valliant (2001) examined differences in executive functioning,

personality, and cognitive ability between youth and adult offenders and non-offenders.

They hypothesized that offenders would have lower social competency and ability to

plan, would have elevated scores on personality testing, and that cognition would be an

identifiable difference but less significant.

In this study, Bergeron and Valliant (2001) had 55 male participants, 26 between

the ages of 16 and 18, and 29 between the ages of 20 and 40 years. Thirteen of the youth

offender participants were gathered from a youth incarceration facility and the 13 non-

offenders were gained from a secondary school and had no prior arrests. The adult
DIFFERENCES IN EXECUTIVE FUNCTIONING 78

participant pool consisted of 15 incarcerated offenders, and 14 volunteers from a

university with no prior arrests.

Numerous tests were administered to measure executive functioning, personality,

and cognition. The methods employed to asses for executive functioning (EF) were the

Qualitative Score (Q-score) of the Porteus Maze to identify levels of visual-motor

planning and foresight, the Paragraph Completion Method (PCM) to obtain abstract

reasoning, thinking style and social maturity/competence, and the Wisconsin Card

Sorting Test (WCST) to measure abstract ability and flexibility.

Personality (P) was measured by the Minnesota Multiphasic Personality Inventory

(MMPI), Form 168, an abbreviated version of the test, the Coopersmith Self-Esteem

Inventory (CSEI), the Buss Durkee Hostility Inventory (BDHI) to measure aggression,

and the Carlson Psychological Survey (CPS) which measures attitudes towards crime.

Bergeron and Valliant (2001) also administered the Michigan Alcohol Screening Test

(MAST) and Drug Abuse Screening Test (DAST) to measure the degree of alcohol and

drug abuse among the participants.

To measure cognition (C), Bergeron and Valliant (2001) employed the Test of

Non-verbal Intelligence Form A (TONI) to gauge abstract reasoning, and the Vocabulary

and Block Design subtests of the Wechsler Adult Intelligence Scale Revised (WAIS-R)

which measures general verbal and performance abilities. These cognition measures

were used to gain a baseline of cognitive functioning and then statistically analyzed to

determine their significance in executive functioning and personality (Bergeron &

Valliant, 2001).
DIFFERENCES IN EXECUTIVE FUNCTIONING 79

Each of the multi-scale tests had separate factor analyses ran in order to avoid

confusion in interpretation as well as to reduce the scales to discernible factors. The

MMPI’s ten scales were reduced to three factors: Psychotic, Neurotic, and Androgyny.

The five scales of the CPS identified two factors as Criminality and Validity. The BDHI

identified two factors as well which were denoted as Ruminative Hostility and Overt

Hostility. The WCST generated three factors identified as Perseverative, Test

Performance, and Set Failure. Bergeron and Valliant (2001) then ran four stepwise linear

discriminant function analyses to determine the variables that could differentiate the

groups.

The first discriminant function between general offenders and non-offenders

indicated statistical significance with a canonical correlation at .88, indicating that this

function explained 78% of the variance between groups. This indicated that the offenders

scored lower than non-offenders on executive functioning and cognition measures

(Conceptual Level and Block Design), and higher on personality measures (MAST and

Psychotic Factor).

The second discriminant function compared adult offenders and adult non-

offenders, and also indicated statistical significance. The canonical correlation was .97,

indicating this function explained 94% of the variance between these two groups. This

function determined adult offenders would score lower than adult non-offenders on the

Androgyny/Social Introversion factor (P), the Test Performance factor (EF), and

Conceptual Level (EF); however would score higher on the Q-score (EF) and the MAST

(P).
DIFFERENCES IN EXECUTIVE FUNCTIONING 80

The third discriminant function compared youth offenders and youth non-

offenders, and again indicated statistical significance with the canonical correlation of

.95, indicating this function explained 90% of the variance between these two groups.

This third function indicated young offenders would typically score higher on the

Psychotic factor (P), the Overt Hostility factor (P), the Q-score (EF), and the DAST (P).

Finally, the fourth discriminant function compared young and adult offenders.

This function indicated statistical significance with the canonical correlation of .74

indicating this function explained 54% of the variance between these two groups. This

function indicated adult offenders would score higher than young offenders on the MAST

(P), Psychotic factor (P), the Vocabulary subtest (C), and the Q-score (EF). Adult

offenders would score lower on the TONI (C).

Bergeron and Valliant (2001) ran a series of ANOVAs to determine any

significant effects. ANCOVAs were then executed to determine whether cognition or

age was masking any of the main effects that resulted and could be explained by a

developmental lag. They determined that the various groups could be differentiated by

variables in personality, executive functioning, and cognition, but could not identify any

support for the notion of a developmental lag from age.

Bergeron and Valliant (2001) found that offenders and non-offenders could be

differentiated on the basis of social maturity, measured by the Conceptual Level score of

the PCM (EF), which had the most significant results, accounting for 70% of the variance

between offenders and non-offenders. This score measures foresight, planning, social

maturity, and conceptualization. Offenders tend to have deficiencies in their ability to

interpret social situations and misinterpretation or inability to read these social cues may
DIFFERENCES IN EXECUTIVE FUNCTIONING 81

cause frustration (Bergeron & Valliant, 2001). This frustration coupled with reduced

verbal abilities, and when accompanied by inappropriate attitudes toward crime and

behavior, often makes it difficult for these individuals to engage in the normative

behaviors of society.

Summary

As suggested in the literature, the complexity of antisocial behavior, particularly

in disadvantaged communities where gangs tend to form, is far reaching. The literature

suggests many factors contribute to a person’s developmental trajectory, including

psychosocial and biological variables. Some of the literature confirmed that the

neighborhood one grows up in has a heavy influence on the development of delinquency

(Beyers et al., 2003; Dupere et al., 2007; McMahon et al., 2009). Yet another study

determined that the delinquency trajectory one follows is ethnicity dependent and must be

looked at from a person-context perspective (Donnellan et al., 2000).

If, according to Gibbs (2000), many of the gang members tire of gang life and age

out around 30, can it be hypothesized that most gang members will fall into Moffit’s

(1993) adolescent-limited framework of delinquency? We can question whether the

structure of the gangs and the needs they fulfill tend to draw those looking for somewhere

to fit in during adolescence, and once adulthood is realized, the gang no longer fills a

need either for fulfillment or for delinquency. There will remain those gang members

who persist into adulthood and maintain their gang status, and there may be even more

complex reasons that can be explained by neuropsychology, biology, and genetics.

DIFFERENCES IN EXECUTIVE FUNCTIONING 82

As Raine (2002a, 2002b) suggested, the prefrontal cortex is still growing in

adolescence and is not completed until generally the mid-20s, suggesting a biological

function of delinquency. As such, an overload may contribute to the inability to reason

or make appropriate decisions in risky situations, and likely contributes to the

impulsivity, rule-breaking, and reckless, irresponsible behavior of gang members. Other

literature found a connection between self-control, aggression, and executive functioning

(Beaver et al., 2009; Beaver et al., 2010). Additional literature highlighted the notion that

aggression and delinquent behavior are effected by impairments in executive functions

(Bergeron and Valliant, 2001; Dolan & Anderson, 2002; Giancola et al., 2006; Raine et

al., 2005; Seguin et al., 1995). There remains the possibility of aggression and

delinquency being mitigated by improving executive function (Giancola et al., 2006), or

improving environmental influences (Craig et al., 2002; Hill et al., 1999; Lahey et al.,

1999).

A barrier to future research is that psychosocial research is conducted separately

from biological research. Holistic efforts should be made to examine the complex

problem as a whole, including psychosocial and biological factors. “What many often

perceive as merely a simplistic societal deviance from the mainstream on the part

of…gang members is truly complex” (Arfaniarromo, 2001).

DIFFERENCES IN EXECUTIVE FUNCTIONING 83

CHAPTER III

Methods

Introduction

This study will attempt to measure differences in executive functioning between

current and former gang members. The Behavior Rating Inventory of Executive

Function - Adult Version (BRIEF-A), a test measuring executive function performance,

and the Wisconsin Card Sort, a test measuring abstract reasoning and ability to set shift

cognitive strategies, will be used to measure these differences.

Research Design

This research will be a quantitative quasi-experimental design that will be

analyzed using a Multivariate Analysis of Variance (MANOVA). This study will

examine any significant differences in executive functioning between current and former

gang members as measured by the self-report from The Brief Rating Inventory of

Executive Function (BRIEF-A), and the objective measure of the Wisconsin Card Sorting

Test (WCST).

Appropriateness of Design

Gangs are a complex phenomenon and multivariate in nature, and instances when

a single variable completely explains a phenomenon are rare. This study will examine

two measures of executive function: a self-report measure, and an objective measure, to

determine whether significant differences exist between the two identified samples. In

much of the research cited in the literature review, significant effects of executive

function among sample groups were measured by various analysis of variance

DIFFERENCES IN EXECUTIVE FUNCTIONING 84

techniques, including ANOVAs, MANOVAs, and MANCOVAs (Bergeron & Valliant,

2001; Dolan & Anderson, 2002; Donnellan et al., 2000; Raine et al., 2005; Seguin et al.,

1995). MANOVAs are commonly accepted and utilized by researchers when measuring

significant differences between multiple dependent variables, as suggested in this study.

This design was chosen in an attempt to follow a procedure most consistent with accepted

peer reviewed scientific practices.

Research Questions and Hypotheses

The research questions that will be examined in this study are as follows:

1. Are there significant differences in executive functioning between current

gang members and former gang members?

The researcher hypothesizes that the following will be true:

Ho: Current gang members will not demonstrate a significant difference in

executive functioning than former gang members

H1: Current gang members will demonstrate significant differences in executive

functioning than former gang members

Population / Participants

Forty male adults, ranging from 25 to 50 years, are expected to participate in the

study. Demographic information will be gathered via a questionnaire; however,

demographics of the participants are expected to be of mostly Hispanic or African-

American descent, and of lower SES backgrounds. Participants will be recruited from

one non-profit organization: Homeboy Industries. Homeboy Industries, located in East

Los Angeles, California, provides alternatives to current and former gang involved
DIFFERENCES IN EXECUTIVE FUNCTIONING 85

individuals. Services include counseling, education, tattoo removal, job training, and

placement.

Making the decision to leave a gang and its lifestyle is not a simple or immediate

decision. In fact, it is a long recovery process. Although individuals may not be active in

their gang, they may live in the same communities and interact daily with former gang

affiliates. For safety measures, it was decided to identify both current and former gang

members as those involved with Homeboy Industries. For purposes of this study, current

gang members will be operationalized as those individuals that have not been active in

their gang for less than one year. Former gang members will be operationalized as those

individuals that have not been active in their gang for more than two years.

Informed Consent

All participants will be fully advised of rights and responsibilities in an informed

consent prior to participation (See Appendix A). Included in the informed consent will

be an incentive of $20 upon completion of the testing measures.

Sampling Frame

All participants are estimated to be between the ages of 25 and 50 years.

Literature demonstrates that the frontal lobe does not finish developing until the 20s or

even 30s (Raine, 2002b). Most individuals remit antisocial behavior by early adulthood,

and those who persist delinquency into adulthood will perform poorer on cognitive tests

(Donnellan et al., 2000; Huesmann et al., 2009; Moffit, 1993). The population sample

will be limited to this age group in order to gather data within the framework of these two

phenomena as it relates to gang involvement.

DIFFERENCES IN EXECUTIVE FUNCTIONING 86

Confidentiality

To protect the confidentially of the participants, the collection of identifying

information shall be limited. Data will be collected using a demographic questionnaire

including date of birth, ethnic group/race, gender, SES, use of drugs, head injuries, and

current prescription of psychotropic medication(s). Specific identifying information such

as name, address, or social security number, will not be collected. The signature of the

participant will be required on the informed consent. The identifiable information will be

kept separate from the results to ensure confidentiality. The researcher will inform the

participants that the results of their measures will not be able to be linked back to them

because no identifying information will be listed on the measures. Participants will be

informed that the data will be kept locked for three years, and will then be appropriately

destroyed. The researcher and researcher's direct supervisor are the only individuals

allowed access to the data.

Geographic Location

Participants reside in the Southern California region, likely in lower SES areas.

Instrumentation / Validity and Reliability

Behavior Rating Inventory of Executive Function–Adult Version (BRIEF-A)

The BRIEF-A is a standardized measure that captures adults' views of their

executive functions, or self-regulation, in their everyday environment in two formats:

self-report and informant report. The test was designed to measure a wide variety of

developmental, systemic, neurological, and psychiatric disorders such as attention

disorders, learning disabilities, autism spectrum disorders, traumatic brain injury,

DIFFERENCES IN EXECUTIVE FUNCTIONING 87

multiple sclerosis, depression, mild cognitive impairment, dementias, and schizophrenia.

The BRIEF-A is composed of 75 items within nine non-overlapping, theoretically and

empirically derived clinical scales that measure various aspects of executive functioning

including the Inhibit, Self-Monitor, Plan/Organize, Shift, Initiate, Task Monitor,

Emotional Control, Working Memory, and Organization of Materials scales. These

clinical scales form two broader indexes: Behavioral Regulation Index (BRI) and

Metacognition Index (MI), which form the overall summary score, the Global Executive

Composite (GEC). The BRIEF-A also includes three validity scales (Negativity,

Inconsistency, and Infrequency). The BRIEF-A was standardized on a diverse

population that mirrored the U.S. population according to key demographics including

age, gender, ethnicity, education, and geographical regions. One thousand and fifty

participants for the self-report and 1,200 participants for the informant report were

recruited through Internet sampling and ranged in age from 18-90 years. The BRIEF-A

has been demonstrated to be reliable and valid. Internal consistency for the self-report

normative sample is strong (alpha range = .73-.90 for clinical scales; .93-.96 for BRI and

MI indexes, and the GEC), and equally strong for the Informant Report normative sample

(alpha range = .80-.93 for clinical scales; .95-.98 for BRI and MI indexes, and the GEC).

Internal consistency was also high for the mixed clinical/healthy sample (alpha range =

.80-.94 for clinical scales; .96-.98 for BRI and MI indexes, and the GEC), as well as the

Informant Report mixed clinical/healthy sample (alpha range = .85-.95 for clinical scales;

.96-.98 for BRI and MI indexes, and the GEC). Test-retest reliability ranged from .82-.93

over an average interval of 4.22 weeks for the Self- Report Form (n = 50) clinical scales,

and.93 for the BRI and MI and .94 for the GEC. Test-retest reliability for the Informant
DIFFERENCES IN EXECUTIVE FUNCTIONING 88

Report Form (n = 44) clinical scales ranged from .91-.94 over an average interval of 4.21

weeks, and was .96 for the BRI, MI, and GEC. Correlations between Self-Report ratings

and Informant Report ratings on interrater agreement ranged from .44-.68 for the clinical

scales and from .61-.63 for the BRI and MI indexes and the GEC. With regards to

convergent validity, the Self and Informant Report Form of the BRIEF-A scales, indexes,

and GEC, demonstrated significant correlations with the Frontal Systems Behavior

Scale™ (FrSBe), Dysexecutive Questionnaire (DEX), and Cognitive Failures

Questionnaire (CFQ). Although the BRIEF-A was not used in any of the literature cited

for this study, this data concludes that the BRIEF-A is a structurally valid and reliable

instrument.

Wisconsin Card Sorting Test (WCST)

The Wisconsin Card Sorting Test (WCST) is an objective measure that assesses

abstract reasoning and ability to shift cognitive strategies in response to environmental

changes. The WCST provides objective measures for both overall ability as well as

sources of difficulty (e.g., inefficient initial conceptualization, perseveration, failure to

maintain set and inefficient learning across several stages of the test). Clinically, the test

is widely used in patients with acquired brain injury, neurodegenerative disease, or

mental illness such as schizophrenia. It has been considered a measure of executive

function because of its reported sensitivity to frontal lobe dysfunction.

The WCST was standardized on individuals aged six and a half to 89 years.

Eight hundred and ninety-nine subjects were aggregated from six distinct samples: 453

normal children and adolescents in public schools in a large urban area of southeastern
DIFFERENCES IN EXECUTIVE FUNCTIONING 89

United States; 49 18 year olds and their friends in an urban area of the southwestern

United States; 150 individuals from Texas and Colorado, between 15 and 77 years; 50

subjects from Colorado between 58 and 84 years; 124 commercial airline pilots (119

from Colorado, five from Washington, D.C.), and 73 adults from independent living

retirement communities and general communities in the Detroit, Michigan area, 24 to 65

years.

The WCST has been demonstrated to be both reliable and valid. In studies with

adults, interscorer reliability had intraclass correlation coefficients (rICC) that ranged from

.75 to .97 for Perseverative Responses, Perseverative Errors, and Nonperseverative

Errors. Consistency of scorers, or intrascorer reliability, had intraclass correlation

coefficients (rICC) that ranged from .91 to .96. In studies with children, interscorer

reliability had intraclass correlation coefficients (rICC) that ranged from .895 to 1.000 for

Perseverative Responses, Perseverative Errors, and Nonperseverative Errors. Intrascorer

reliability had intraclass correlation coefficients (rICC) that ranged from .828 to 1.000.

The WCST has four stimulus cards and 128 response cards, or 64 response cards

in the computerized version. Participants are asked to match the response card presented

to them to one of the stimulus cards above them. They have the option to match

according to color (red, blue, green, or yellow), form (cross, circle, triangle, or star), or

number of figures (one, two, three, or four). The test will change the matching category

without notice and it is up to the respondent to follow and determine what is being asked

to match.
DIFFERENCES IN EXECUTIVE FUNCTIONING 90

The WCST allows the clinician to assess the following frontal lobe executive

functions: strategic planning, organized searching, utilizing environmental feedback to

shift cognitive sets, goal oriented behavior, and modulating impulsive reactions. The

WCST is one of the leading executive functioning tests to measure abstract ability and

flexibility, and was used in the following studies cited in the literature review: Bergeron

and Valliant, 2001; Dolan and Anderson, 2002; Giancola et al., 2006; Raine et al., 2005.

For purposes of this study, the computerized version of this test will be administered

(Heaton, 1993).

Demographic Questionnaire

The demographic questionnaire will be developed to capture participant

characteristics. Information collected will include date of birth, ethnic group/race, gender,

SES, drug use, head injuries, and current prescription of psychotropic medication(s) (see

Appendix C). The questionnaire will be utilized to describe the sample. Furthermore, the

frequencies, means, and standard deviations will be analyzed on the demographic

variables to describe the characteristics of the sample.

Data Collection

Upon arrival to the designated testing area, which will be in a quiet, climate

controlled setting and conducive to completing the measures, all participants will be

asked to complete an informed consent prior to data collection, which will include their

rights and responsibilities in the study. Participants will be informed that they are being

asked to take part in a study regarding certain planning, organizing, and goal oriented

processes. The researcher will provide the participants with an opportunity to ask
DIFFERENCES IN EXECUTIVE FUNCTIONING 91

questions. All questions will be answered to the best of the researcher's knowledge.

Once all questions are adequately addressed, the researcher will comprehensively review

the informed consent form required to enact the study, and place in a manila envelope

once signed. Each participant will then be given the demographic questionnaire to

ascertain participant characteristics. Once the demographic questionnaire is completed, it

will be kept in a separate manila envelope and the BRIEF-A will be administered. The

BRIEF-A protocol will be placed in a third manila envelope separate from the informed

consent and demographic questionnaires. The WCST will then be administered via

laptop, furnished by the researcher in an online version of this measure. Once all

participants complete the measures, they will be debriefed on the purpose of the study

and any last questions will be addressed. It is estimated that the administration period

will take approximately five weeks, administering the measures to an average of eight

participants per week.

Data Analysis

All data will be analyzed using PASW statistics software version 18.0. The

hypothesis will be analyzed using a Multivariate Analysis of Variance (MANOVA).

This study will examine two dependent variables, the BRIEF-A and WCST, and one

independent variable with two levels, current and former gang members. Dependent

variables depend on independent variables; a MANOVA will test whether the

independent variable simultaneously explains a significant amount of variance in the

dependent variable. A MANOVA has five assumptions: independent random sampling,

homogeneity of variance, multivariate normality, linearity and multicollinearity, and

DIFFERENCES IN EXECUTIVE FUNCTIONING 92

sensitivity to outliers. Wilks’ Lambda will compare the mean of the BRIEF-A and the

WCST, and a resulting F statistic will be examined for main effects. Significance level

will be set at p=.05. If significant main effects are found, post hoc tests will be ran to

examine the IV’s independently to determine which independent variable differs.

Summary

This research study will examine data collected via a self-report of The Brief

Rating Inventory of Executive Function (BRIEF-A), as well as an objective measure of

the Wisconsin Card Sorting Test (WCST), by current and former gang members. Using a

MANOVA, the researcher will examine any differences in executive functioning between

the two groups. A MANOVA will enable the researcher to examine multiple dependent

variables to gain a stronger measure of differences in executive functioning between

current and former gang members. Following sections will discuss findings and provide

information on future research.

DIFFERENCES IN EXECUTIVE FUNCTIONING 93

CHAPTER IV

Results

Participants in this study consisted of 40 males; 20 current and 20 prior gang

members affiliated with Homeboy Industries in East Los Angeles, California. For

purposes of this study, current gang members were defined as those that have not been

active in a gang for less than one year. Former gang members were defined as those

individuals that have not been active in a gang for more than two years. Ages ranged

from 24 to 53 years old (M=36). Furthermore, 70% of participants were Hispanic, 25%

African American, 2.5% Asian, and 2.5% Other (more detailed demographic information

is located in Appendix D).

The present study compared executive functioning between current and former

gang members. The goal was to examine whether any differences in executive

functioning might help explain a person’s decision or ability to exit the gang lifestyle.

The researcher hypothesized there would be significant differences in executive

functioning between current and former gang members. It was hypothesized former gang

members would exhibit greater executive functioning abilities. The self-reported Global

Executive Composite (GEC) on the BRIEF-A (M=54.775, SD=10.202) was compared to

the number of categories completed on the WCST (M=3.7, SD=2.312) (see Appendix A,

Table 1). This hypothesis was not supported by the data. The former gang members did

not demonstrate greater abilities in executive functioning than current gang members. A

one-way MANOVA revealed an insignificant multivariate main effect for the current

versus former gang members, Wilks’ Lambda = .999, F (2, 37) = .017, p > .05, partial eta
DIFFERENCES IN EXECUTIVE FUNCTIONING 94

squared = .001. Power to detect the effect was .052 (see Appendix A, Table 2). Thus,

the hypothesis was not confirmed. This suggests that former gang members do not have

significantly better executive functioning abilities than current gang members.

Although no significant main effects were found, a post-hoc analysis of between

subjects effects was performed to examine the BRIEF-A and WCST independently to

determine if one better measures executive functioning differences among current and

former gang members. The main effect of the between subjects variables of the BRIEF-

A and the WCST was not significant using a critical p of .05, the GEC on the BRIEF-A

with F (1, 1.225) = .011, p > .05, partial eta squared = .000. Power to detect the effect

was .051. The number of categories completed on the WCST had F (1, .100) = .018, p >

.05, partial eta squared = .000. The power to detect the effect was .052 (see Appendix A,

Table 3). This suggests neither the BRIEF-A, or the WCST, demonstrate significant

differences in executive functioning between current and former gang populations.

An exploratory analysis, performed with a one-sample t-test, was conducted on

each clinical scale of the BRIEF-A (Inhibit, Shift, Emotional Control, Self-Monitor,

Initiate, Working Memory, Plan/Organize, Task Monitor, and Organization of Materials),

as well as the Behavioral Regulation Index (BRI), Metacognition Index (MI), and Global

Executive Composite (GEC), to compare current and former gang members with the

population normed on the BRIEF-A. This was done to identify any one area of executive

functioning the gang populations may perform better or worse in than the general

population. There was a significant difference in eight out of the nine factors of the

BRIEF-A, as well as the BRI, MI, and GEC scores, that suggests the two populations
DIFFERENCES IN EXECUTIVE FUNCTIONING 95

performed poorer than the general population (see Appendix B, Table 4). There was one

scale, Organization of Materials, in which the two groups performed better than the

general population. On the BRIEF-A, higher t-scores indicate poorer performance.

There was a significant difference on the Inhibit scale (M=56.475, SD=10.70)

between current and former gang members and the normed population t(39) = 3.826, p =

.000. The Inhibit scale measures a respondent’s ability to inhibit or resist an impulse and

to manage behavior at the appropriate time. This difficulty may be seen in adults with

traumatic brain injuries. Individuals with poor impulse control may display inappropriate

physical responses to others and have a general failure to think before acting (Roth,

Isquith, & Gioia, 2005). The results of the exploratory analysis suggest current and

former gang members have more difficulty than the general population in this executive

function.

There was a significant difference on the Shift scale (M=56.63, SD=10.00)

between current and former gang members and the normed population t(39) = 4.189, p =

.000. The Shift scale measures one’s ability to move from one situation, activity, or

aspect of a problem to another with relative ease. Aspects of this function include

transitioning, flexibility in problem solving, and alternate attention. Ultimately, it is the

ability to shift both behaviorally and cognitively in response to changes in the

environment. Individuals with difficulty in this domain may respond to situations with

anger or anxiety. Deficits may range from efficiency of problem solving to perseverative

problem solving behaviors. They may be described as stubborn, rigid, or inflexible, as

well as demonstrating difficulty in generating problem solving ideas. Results of the

DIFFERENCES IN EXECUTIVE FUNCTIONING 96

exploratory analysis suggest current and former gang members have more difficulty

appropriately responding to changes in the environment than the general population.

There was a significant difference on the Emotional Control scale (M=53.275,

SD=9.05) between current and former gang members and the normed population t(39) =

2.29, p = .0280. The Emotional Control scale measures a respondent’s ability to

modulate emotional responses. Those with difficulty in this area may demonstrate

overreactions to seemingly minor situations. This area may be described as someone

with mood lability, or one who goes from zero to 100 in 60 seconds. Results of the

exploratory analysis suggest current and former gang members have more difficulty

modulating emotions than the general population.

There was a significant difference on the Self-Monitor scale (M=54.525,

SD=10.42) between current and former gang members and the normed population t(39) =

2.746, p = .009. The Self-Monitor scale measures one’s ability to follow their own

behavior and maintain awareness of the effect of his or her behavior on others. In

interactions, this may manifest in one not understanding why another individual may be

upset with them. Results of the exploratory analysis suggest current and former gang

members have more difficulty than the general population in their awareness of the effect

of their behavior on others.

There was a significant difference on the Behavioral Regulation Index (BRI)

(M=56.025, SD=10.56) between current and former gang members and the normed

population t(39) = 3.608, p - .001. The BRI represents an individual’s ability to

appropriately regulate his or her behavior and emotional responses. It is composed of the
DIFFERENCES IN EXECUTIVE FUNCTIONING 97

Inhibit, Shift, Emotional Control, and Self-Monitor scales. One’s ability to manage

emotions and recognize changes in the environment may be a precursor to effective

problem solving (Roth, Isquith, & Gioia, 2005). Results of the exploratory analysis

suggest current and former gang members have more difficulty than the general

population in their overall ability to regulate emotions and transition between changes in

the environment.

There was a significant difference on the Initiate scale (M=53.175, SD=8.86)

between current and former gang members and the normed population t(39) = 2.266, p =

.029. The Initiate scale measures one’s ability to begin activities and generate ideas or

problem solving strategies. It does not necessarily indicate disinterest, opposition, or

poor motivation; rather they may want to succeed at something but find it difficult to start

and may require prompting. This is often a significant difficulty for those with severe

frontal lobe brain injury (Roth, Isquith, & Gioia, 2005). Difficulty with initiation may

also be a secondary consequence to other executive functioning deficits, like poor

organization. Results of the exploratory analysis suggest current and former gang

members have more difficulty than the general population in initiating tasks.

There was a significant difference on the Working Memory scale (M=55.275,

SD=12.48) between current and former gang members and the normed population t(39) =

2.673, p = .011. The Working Memory scale measures one’s ability to hold information

for the purpose of completing tasks or generating a response to something. Individuals

with impaired working memory have difficulty remembering things even for a short time,

lose track of what they are doing, or forget altogether what they were instructed to do.
DIFFERENCES IN EXECUTIVE FUNCTIONING 98

An integral part of working memory is the ability to maintain attention and performance

over time (Roth, Isquith, & Gioia, 2005). Difficulties in this area involve maintaining an

activity for an age-appropriate amount of time. Results of the exploratory analysis

suggest current and former gang members have more difficulty than the general

population in holding information and attention for long periods.

There was a significant difference on the Plan/Organize scale (M=55.575,

SD=10.86) between current and former gang members and the normed population t(39) =

3.247, p = .002. The Plan/Organize scale involves two cognitive components to an

individual’s ability to manage current as well as future-oriented tasks. The plan

component involves the ability to anticipate future events, implement goals, and develop

steps to achieve said goals. The organize component involves the ability to have a level

of order to the actions or information to achieve goals. The way information or tasks are

organized can impact how the information or actions are learned and implemented.

Adults with organizational problems may approach tasks haphazardly or become easily

overwhelmed. They may have difficulty maintaining order in their environment.

Difficulty with planning and organizing plays an integral role in many cases of executive

functioning impairment (Roth, Isquith, & Gioia, 2005). Results of the exploratory

analysis suggest current and former gang members have more difficulty than the general

population in planning and organizing information and tasks.

There was a significant difference in the Task Monitor scale (M=55.875,

SD=10.747) between current and former gang members and the normed population t(39)

= 3.457, p = .001. The Task Monitor scale measures a person’s awareness of their own
DIFFERENCES IN EXECUTIVE FUNCTIONING 99

successes and failures in problem solving. People with difficulty in this area fail to

appreciate their errors when approaching problems or tasks, or may catch errors at the

last minute. Results of the exploratory analysis suggest current and former gang

members have more difficulty than the general population in their awareness of their

mistakes in problem solving.

There was a significant difference on the Organization of Materials scale

(M=45.50, SD=7.296) between current and former gang members and the normed

population t(39) = -3.901, p = .000; however, current and former gang members

performed significantly better than the population norm on this measure of executive

functioning. The Organization of Materials scale measures one’s ability to keep order in

everyday life with respect to orderliness of work and living spaces. This scale relates to

the ways in which a person organizes his/her world. Adults that have difficulty in this

area may demonstrate poor efficiency at home or work because they often misplace

things and do not have them readily available for use. Results of the exploratory analysis

suggest current and former gang members perform better than the general population in

organizing their world and its contents.

There was a significant difference on the Metacognition Index (MI) (M=53.30,

SD=9.608) between current and former gang members and the normed population t(39) =

2.172, p = .036. The MI represents the respondent’s ability to use planning and

organization to solve problems while keeping these efforts in active working memory. It

is composed of the Initiate, Working Memory, Plan/Organize, Task Monitor, and

Organization of Materials scales. The results of the exploratory analysis suggest current
DIFFERENCES IN EXECUTIVE FUNCTIONING 100

and former gang members have more difficulty than the general population in cognitively

managing attention and problem solving.

Finally, there was a significant difference on the GEC score (M=54.775,

SD=10.20) between current and former gang members and the normed population t(39) =

2.96, p = .005. The Global Executive Composite (GEC) is a summary score that

incorporates all of the clinical scales of the BRIEF-A. Overall, the results of the

exploratory analysis suggest current and former gang members have more difficulty than

the general population in most aspects of executive functioning.

Results of this study concluded that current gang members, identified as those

who have not been active in a gang for less than one year, and former gang members,

identified as individuals who have not been active in a gang for more than two years, do

not demonstrate significant differences in executive functioning. However, on the self-

report measure of the BRIEF-A, it was suggested that the combined current and former

gang population performed worse than the general population on various elements of

emotional control and cognitive flexibility, planning, and organizing.

DIFFERENCES IN EXECUTIVE FUNCTIONING 101

CHAPTER V

Discussion

Research has measured and suggested factors that contribute to or predict youth

joining gangs, as well as various differences between gang members and non-gang

members (Craig, Vitaro, Gagnon, & Tremblay, 2002; Hill, Howell, Hawkins, & Battin-

Pearson, 1999; Lahey, Gordon, Loeber, Stouthamer-Loeber, & Farrington, 1999;

Maxmen, Ward, & Kilgus, 2009). As Lahey et al. (1999) suggested, it is possible that

some of these differences may arise after a youth has entered a gang, and may not be a

contributing factor in their decision to join. In considering social mimicry, a question

remains whether gang members join gangs to attain social status and/or protection in their

community, or whether there is something more biogenic or etiological.

Factors that lead to gang involvement are important in being able to take

proactive measures to provide early interventions for youth; however, factors that lead to

gang desistance in adulthood are equally important. This research analyzed whether

there may be scientific factors that contribute to adults leaving a gang, specifically

executive functioning and one’s ability to reason, plan, and make decisions. The decision

to leave a gang, as does any life changing event, requires reasoning, planning, and

execution.

The frontal lobe, or executive functioning area of the brain, controls emotions,

urges, and knowledge of right and wrong; it is the impulse control center, and imposes

appropriate inhibitions. It is also the part of the brain that organizes information, enables

planning, and completion of goals. This research did not provide any significant results

indicating former gang members have better executive functioning skills than current
DIFFERENCES IN EXECUTIVE FUNCTIONING 102

gang members. As the data suggested, the two groups performed relatively similar. It is

important to explore other factors that may contribute to adults exiting a gang.

As Gibbs (2000) suggested, many gang members tire of the gang lifestyle, and

age out at approximately 30 years of age. However, the explanation cannot be that

simple. What does “age out” mean? Future research may explore possible scientific

explanations for how individuals involved in gangs ultimately reach the point of desisting

from the lifestyle. Young men who join gangs generally come from a lower

socioeconomic community, and are surrounded by adverse conditions, stress, and

oftentimes, chaos. This has psychosocial implications, but also biological implications.

They become attuned to surrounding stimuli. As van Goozen, Fairchild, and Harold

(2008) suggested, stressful events that occur early in life play an important role in

“programming” the stress systems. This may result in the adapting and down-regulating

of the stress-response system, and may include avoidance of chronic arousal and energy

expenditure that could result in serious pathophysiological consequences. As a result of

this lowered stress-response system, van Goozen, Fairchild, and Harold (2008) posited

that antisocial individuals are less likely to avoid stressful situations and are more fearless

and less concerned about possible negative consequences. When few prosocial

alternatives are available to mitigate this change in the response system, immunity to

stressful situations becomes the method of interacting with the world. This ingrained

style of engagement and interaction becomes more difficult to change, and executive

functioning may be impaired as a result.

As the exploratory analysis demonstrated, current and former gang members

performed more poorly than the general population in eight areas of executive
DIFFERENCES IN EXECUTIVE FUNCTIONING 103

functioning: inhibit impulses; shifting between environments and stimuli; emotional

control; monitor behavior; regulate behavior; ability to initiate ideas or activities; hold

information in their working memory; planning; and problem solving. This supports

research that suggests aggression and delinquent behavior are affected by impairments in

executive functions (Bergeron & Valliant, 2001; Dolan & Anderson, 2002; Giancola et

al., 2006; Raine et al., 2005; Seguin et al., 1995). Other literature found a connection

between self-control, aggression, and executive functioning (Beaver et al., 2009; Beaver

et al., 2010).

Literature demonstrates that the frontal lobe does not finish developing until the

twenties or even thirties (Raine, 2002b). Most individuals remit antisocial behavior by

early adulthood, and those who persist delinquency into adulthood will perform poorer on

cognitive tests (Donnellan et al., 2000; Huesmann et al., 2009; Moffit, 1993). ). As the

analysis further demonstrated, the current and former gang members continue to perform

more poorly than the general populations into their 30s, 40s, and 50s. The fact that the

two groups are seeking assistance in leaving the gang lifestyle supports the idea that

aggression and delinquency may be mitigated by improving executive function (Giancola

et al., 2006), or improving environmental influences (Craig et al., 2002; Hill et al., 1999;

Lahey et al., 1999). It further supports possibilities of early interventions from a holistic,

psychosocial and biogenic perspective.

Furthermore, as Moffit (1993) demonstrated, two neuropsychological deficits

empirically associated with antisocial behavior are verbal and executive functions.

Verbal deficits may include difficulties with receptive listening and reading, problem

solving, expressive speech and writing, and memory. Executive functioning deficits
DIFFERENCES IN EXECUTIVE FUNCTIONING 104

include problems with attention and impulsivity. Piquero (2001) also determined that

poor verbal ability has a significant relation to likelihood of engaging in delinquent

behavior. Goldberg (2009) suggested two evolutionary developments occurred

simultaneously: the emergence of language and the explosive rise of executive functions,

or more specifically, the frontal lobes. Goldberg (2009) posited their simultaneous

emergence serves a purpose. Language has a generative power to build new constructs;

however, this ability may depend upon the frontal lobes to make use of and manipulate

the language constructs into more complex operations. Young men who become

involved in gangs at an early age impair the development of executive functioning

capacities. As was questioned in Goldberg’s (2009) study, the question remains whether

poor verbal ability can be attributed to poor cognitive functioning, or the result of a poor

education that did not allow the subjects the opportunity to build verbal ability that is

afforded others.

There were two sources of continuity Moffitt (1993) suggested are reasons why

life-course persistent delinquents are impervious to change. One is failing to learn

prosocial alternatives to behavior. Moffitt suggested that once the deviant behavior

begins, a lack of recourse to more prosocial alternatives keeps them on the course toward

delinquency into adulthood. They miss out on opportunities at every stage of

development due to rejection from peers and adults. Second is becoming entangled in a

deviant lifestyle due to crime’s consequences. Poor self-control and impulsivity increase

the risk that antisocial youth will make irreversible decisions that close the doors of

opportunity. Once a male is immersed in the gang lifestyle, little research has examined

what factors may contribute to their decision to leave. Research abounds with
DIFFERENCES IN EXECUTIVE FUNCTIONING 105

taxonomies and theories that separate those involved in delinquent behavior, such as

gangs, and desist in their childhood, or continue a path of delinquent behavior for the

duration of their life course (Huesmann, Dubow, & Boxer, 2009; Moffitt, 1993; Raskin,

White, Bates, & Buyske, 2001).

As Moffit’s (1993) trajectory theory suggests, a person that engages in

delinquency from childhood through adulthood is unlikely to assume prosocial tendencies

after several decades of antisocial behavior. The participants in this research have

demonstrated they are making strides in desisting from a gang lifestyle and discovering a

more prosocial way of living. Future research might explore scientific or biogenic factors

that may contribute to the desistance of delinquent behavior, specifically gangs. It might

be important to examine cognitive processes that enable individuals to reason, plan,

organize, and execute a desire to desist from gangs, and regulate emotions and impulses

when this life has allowed them, and at times required them, to interact in maladaptive

ways, and how these processes might differentiate them from those unable to or unwilling

to leave.

Limitations

This research is not without faults or limitations. Findings by this researcher

should be examined in light of the following limitations, which may assist in directing

future research.

Internal Validity

Internal validity is the confidence placed on the cause and effect relationship in a

scientific study. There existed multiple confounding variables, or extraneous variables,

DIFFERENCES IN EXECUTIVE FUNCTIONING 106

that may have correlated (positively or negatively) with both the independent and

dependent variable. The sample size included two groups at Homeboy Industries: 20

current gang members, defined as those who have not been involved in a gang for less

than one year, and 20 former gang members, identified as those individuals not active in

a gang for more than two years.

The limitation lied in being able to safely gauge current gang membership;

therefore, individuals who have had the shortest time seeking assistance to desist from the

gang lifestyle, were identified as the current group. This may have influenced the results

of this study and may not be an accurate reflection of executive functioning in individuals

currently and actively involved in gangs. Their involvement in the organization to seek

help in desisting from gang involvement may put them more closely related to the former

gang member group. It would be advantageous to have a true current group in future

studies.

Another confounding factor is that the level of executive functioning prior to gang

involvement was not able to be obtained. It is not known if there were any impairment in

executive functioning prior to gang involvement, or if involvement in the deviant

behavior and lack of prosocial alternatives, contributed to executive functioning

impairment.

Another limitation was found within the demographic information provided by

participants. Several participants indicated they have had multiple head injuries resulting

from gunshot wounds, fights, and car accidents. Any effect of these head injuries and

loss of consciousness were not accounted for. In addition, 100% of the participants

indicated they abused substances throughout their life. It may have been helpful to
DIFFERENCES IN EXECUTIVE FUNCTIONING 107

inquire about the duration, extent, and pattern of drug use, specifically whether they used

substances as a social function or when they were sad or excited, as well as their method

of ingesting the substances. This may have impacted frontal lobe capacities and should

be taken into account for future research.

External Validity

External validity regards the general (causal) inferences in scientific studies. The

relatively small sample size (N = 40) obtained from a single non-profit organization and

geographical location with idiosyncratic features, may have limited the ability to

generalize to a population. Therefore, the researcher cannot conclude that the

relationships found in this study, apply to all gang members, or other geographical

locations. In essence, further sampling should be conducted in future studies, to produce

transferability and generalizability.

DIFFERENCES IN EXECUTIVE FUNCTIONING 108

References

Arfániarromo, A. (2001). Toward a psychosocial and sociocultural understanding of

achievement motivation among Latino gang members in U.S. schools. Journal of
Instructional Psychology, 28(3), 123-136.

American Psychiatric Association (2000) Diagnostic and Statistical Manual of Mental

Disorders: Fourth Edition Text Revision (DSM-IV-TR). Washington, DC:
American Psychiatric Association.

Anderson, E. (1999). Code of the Street: Decency, violence, and the moral life of the
inner city. New York: Norton.

Craig, W., Vitaro, F., Gagnon, C., & Tremblay, R. (2002). The road to gang membership:
Characteristics of male gang and nongang members from ages 10 to 14. Social
Development, 11(1), 53-68.

Beaver, K., DeLisi, M., Vaughn, M., & Wright, J. (2010). The intersection of genes and
neuropsychological deficits in the prediction of adolescent delinquency and low
self-control. International Journal of Offender Therapy and Comparative
Criminology, 54(1), 22-42.

Beaver, K., Shutt, J., Boutwell, B., Ratchford, M., Roberts, K., & Barnes, J. (2009).
Genetic and environmental influences on levels of self-control and delinquent
peer affiliation: Results from a longitudinal sample of adolescent twins. Criminal
Justice and Behavior, 36(1), 41-60.

Beaver, K., Wright, J., & Delisi, M. (2007). Self-control as an executive function:
Reformulating Gottfredson and Hirschi's parental socialization thesis. Criminal
Justice and Behavior, 34(10), 1345-1361.

Bergeron, T., & Valliant, P. (2001). Executive function and personality in adolescent and
adult offenders vs. non-offenders. Journal of Offender Rehabilitation, 33(3), 27-
45.

Beyers, J., Bates, J., Pettit, G., & Dodge, K. (2003). Neighborhood structure, parenting
processes, and the development of youths' externalizing behaviors: A multilevel
analysis. American Journal of Community Psychology, 31(1-2), 35-53.

Blair, R. (2007). Aggression, psychopathy and free will from a cognitive neuroscience
perspective. Behavioral Sciences & the Law, 25(2), 321-331.
DIFFERENCES IN EXECUTIVE FUNCTIONING 109

Caspi, A., Elder, G. H., & Bem, D. J. (1987). Moving against the world: Life-course
patterns of explosive children. Developmental Psychology, 23(2), 308-313.

Dolan, M., & Anderson, I. (2002). Executive and memory function and its relationship to
trait impulsivity and aggression in personality disordered offenders. Journal of
Forensic Psychiatry, 13(3), 503-526.

Donnellan, M., Ge, X., & Wenk, E. (2000). Cognitive abilities in adolescent-limited and
life-course-persistent criminal offenders. Journal of Abnormal Psychology,
109(3), 396-402.

Dupéré, V., Lacourse, É., Willms, J., Vitaro, F., & Tremblay, R. (2007). Affiliation to
youth gangs during adolescence: The interaction between childhood psychopathic
tendencies and neighborhood disadvantage. Journal of Abnormal Child
Psychology: An official publication of the International Society for Research in
Child and Adolescent Psychopathology, 35(6), 1035-1045.

Giancola, P., Roth, R., & Parrott, D. (2006). The mediating role of executive functioning
in the relation between difficult temperament and physical aggression. Journal of
Psychopathology and Behavioral Assessment, 28(4), 211-221.

Gibbs, J. (2000). Gangs as alternative transitional structures: Adaptations to racial and

social marginality in Los Angeles and London. Journal of Multicultural Social
Work, 8(1/2), 71-99.

Goldberg, E. (2009). The new executive brain: Frontal lobes in a complex world. New
York: Oxford University Press.

Gottfredson M. & Hirschi, T. (1990). A General Theory of Crime. Standford, CA:

Stanford University Press.

Heaton, R., Chelune, G., Talley, J., Kay, G., Curtiss, G. (1993). Wisconsin Card Sorting
Test Manual, Revised and Expanded. Lutz, FL: Psychological Assessment
Resources, Inc.

Hill, K., Howell, J., Hawkins, J., & Battin-Pearson, S. (1999). Childhood risk factors for
adolescent gang membership: Results from the Seattle Social Development
Project. Journal of Research in Crime and Delinquency, 36(3), 300-322.

Huesmann, L., Dubow, E., & Boxer, P. (2009). Continuity of aggression from childhood
to early adulthood as a predictor of life outcomes: Implications for the adolescent-
limited and life-course-persistent models. Aggressive Behavior, 35(2), 136-149.
DIFFERENCES IN EXECUTIVE FUNCTIONING 110

Kupersmidt, J., Coie, J., & Howell, J. (2004). Resilience in children exposed to negative
peer influences. Investing in children, youth, families, and communities:
Strengths-based research and policy (pp. 251-268). American Psychological
Association.

Lahey, B., Gordon, R., Loeber, R., Stouthamer-Loeber, M., & Farrington, D. (1999).
Boys who join gangs: A prospective study of predictors of first gang entry.
Journal of Abnormal Child Psychology: An official publication of the
International Society for Research in Child and Adolescent Psychopathology,
27(4), 261-276.

Maxmen, J., Ward, N., Kilgus, M. (2009). Essential Psychopathology and Its Treatment.
W.W. Norton Company & Inc.

McMahon, S., Felix, E., Halpert, J., & Petropoulos, L. (2009). Community violence
exposure and aggression among urban adolescents: Testing a cognitive mediator
model. Journal of Community Psychology, 37(7), 895-910.

Moffitt, T. (1993). Adolescence-Limited and life-course persistent antisocial behavior: A

developmental taxonomy. Psychological Review, 100(4), 674-701

National Gang Intelligence Center. (2009). National Gang Threat Assessment 2009
(NGIC Product No. 2009-M0335-001). Retrieved from http://www.fbi.gov/stats-
services/publications/national-gang-threat-assessment-2009-pdf

Piquero, A. (2001). Testing Moffitt's neuropsychological variation hypothesis for the

prediction of life-course persistent offending. Psychology, Crime & Law, 7(3),
193-215.

Pratt, T., Turner, M., & Piquero, A. (2004). Parental socialization and community
context: A longitudinal analysis of the structural sources of low self-control.
Journal of Research in Crime and Delinquency, 41(3), 219-243.

Raine, A. (2002a). Biosocial studies of antisocial and violent behavior in children and
adults: A review. Journal of Abnormal Child Psychology, 30(4), 311-326.

Raine, A. (2002b). Annotation: The role of prefrontal deficits, low autonomic arousal and
early health factors in the development of antisocial and aggressive behavior in
children. Journal of Child Psychology and Psychiatry, 43(4), 417-434.
DIFFERENCES IN EXECUTIVE FUNCTIONING 111

Raine, A., Moffitt, T., Caspi, A., Loeber, R., Stouthamer-Loeber, M., & Lynam, D.
(2005). Neurocognitive Impairments in Boys on the Life-Course Persistent
Antisocial Path. Journal of Abnormal Psychology, 114(1), 38-49.

Raskin White, H., Bates, M., Buyske, S. (2001). Adolescence-limited versus persistent
delinquency: Extending Moffitt’s hypothesis into adulthood. Journal of Abnormal
Psychology, 110(4), 600-609.

Ratchford, M., & Beaver, K. (2009). Neuropsychological deficits, low self-control, and
delinquent involvement: Toward a biosocial explanation of delinquency. Criminal
Justice and Behavior, 36(2), 147-162.

Roth, R., Isquith, P., Gioia, G. (2005). Behavior Rating Inventory of Executive Function
– Adult Version Professional Manual. Lutz, FL: Psychological Assessment
Resources, Inc.

Séguin, J., Pihl, R., Harden, P., Tremblay, R., & Boulerice, B. (1995). Cognitive and
neuropsychological characteristics of physically aggressive boys. Journal of
Abnormal Psychology, 104(4), 614-624.

Tuvblad, C., Grann, M., & Lichtenstein, P. (2006). Heritability for adolescent antisocial
behavior differs with socioeconomic status: gene–environment interaction.
Journal of Child Psychology & Psychiatry, 47(7), 734-743.

Valdez, A., Kaplan C., Codina E. (2000). Psychopathy among Mexican American gang
members: A comparative study. International Journal of Offender Therapy and
Comparative Criminology, 44(1), 46-58. Retrieved from E-Journals database.

van Goozen, S., Fairchild, G., & Harold, G. (2008). The role of neurobiological deficits
in childhood antisocial behavior. Current Directions in Psychological Science,
17(3), 224-228.

Vigil, J. (2003). Urban violence and street gangs. Annual Review of Anthropology, 32,
225-242.
DIFFERENCES IN EXECUTIVE FUNCTIONING 112

Appendix A

Table 1
MANOVA Descriptive Statistics

Current_Former Mean Std. Deviation

BRIEF_GEC 1.00 54.6000 12.57985
2.00 54.9500 7.43728
Total 54.7750 10.20178
WCST_# categories 1.00 3.6500 2.49789
2.00 3.7500 2.17340
Total 3.7000 2.31162

Table 2
Multivariate Tests

Effect Value F Hypothesis Error Sig. Partial Observed

df df Eta Power
Squared
Curr_Form Wilks’ .999 .017 2.000 37.000 .983 .001 .052
Lambda

Table 3
Tests of Between-Subject Effects

Source Dep Type Df Mean F Sig. Partial Observed

Variable
III Sum Square eta Power
of Squared
Squares
Curr_Form BRIEF_GEC 1.225 1 1.225 .011 .915 .000 .011
WCST_#cat .100 11 .100 .018 .893 .000 .018
DIFFERENCES IN EXECUTIVE FUNCTIONING 113

Appendix B

Table 4
One-Sample Statistics

Mean Std. Deviation Std. Error Mean

Inhibit 56.4750 10.70343 1.69236
Shift 56.6250 10.00176 1.58142
Emotional Control 53.2750 9.04685 1.43043
Self_Monitor 54.5250 10.42184 1.64784
BRI 56.0250 10.55993 1.66967
Initiate 53.1750 8.86042 1.40096
Working Memory 55.2750 12.47970 1.97321
Plan_Organize 55.5750 10.86015 1.71714
Task Monitor 55.8750 10.74694 1.69924
Organization of Mat. 45.5000 7.29594 1.15359
MI 53.3000 9.60822 1.51919
GEC 54.7750 10.20178 1.61304

Table 5
One-Sample Test

Test Value = 50
95% Confidence Interval
of the Difference
t df Sig. (2-tailed) Mean Lower Upper
Difference
Inhibit 3.826 39 .000 6.47500 3.0519 9.8981
Shift 4.189 39 .000 6.62500 3.4263 9.8237
Emotional Control 2.290 39 .028 3.27500 .3817 6.1683
Self_Monitor 2.746 39 .009 4.52500 1.1919 7.8581
BRI 3.608 39 .001 6.02500 2.6478 9.4022
Initiate 2.266 39 .029 3.17500 .3413 6.0087
Working Memory 2.673 39 .011 5.27500 1.2838 9.2662
Plan_Organize 3.247 39 .002 5.57500 2.1018 9.0482
Task Monitor 3.457 39 .001 5.87500 2.4380 9.3120
Organization of Mat. -3.901 39 .000 -4.50000 -6.8334 -2.1666
MI 2.172 39 .036 3.30000 .2271 6.3729
GEC 2.960 39 .005 4.77500 1.5123 8.0377
DIFFERENCES IN EXECUTIVE FUNCTIONING 114

Appendix C

Demographic Questionnaire

1. DOB ______ Age ______

2. Race/ethnicity (Circle the one that best describes you)

Caucasian African American
Hispanic Asian American
Mixed Native American
Other

3. Education: What is the highest grade that you have completed?

5th grade ____ 6th grade ____ 7th grade ____
8th grade ____ 9th grade ____ 10th grade _____
11th grade____
You have your High School Diploma/GED _____
Some Trade School ____
Some Community College _____

4. Have you ever been knocked unconscious? For example, have you been in a fight
or car accident and you lost consciousness for an extended period of time.
YES ____ NO ____
If yes, how many times? ______

5. Have you ever taken prescribed medication by a doctor for mental health
reasons?
YES ____ NO ____
If yes, what? ___________________________________________________________
When was the last time? ___________________________

6. Have you ever taken the following drugs: Please check all that apply.
___ Methamphetamine (Crank, Beannies, Black Beauty, Chrome, Cristina, Bling Bling)
___ Hallucinogens (LSD or PCP; Acid, Mellow Yellow, Cactus, Mexican Mushrooms)
___ Ecstacy (E, X, Adam, Roll)
___ Marijuana (Weed, Mary Jane)
___ Inhalants (Air blast, Huffing, Moon gas, Air Fresher)
___ Cocaine (Crack, Blow, C, Marching Powder, Nose Candy)
___ Heroin (Big H, China White, Mexican Brown, Smack, Witch Hazel, White Horse)
___ Alcohol
___ Prescription Drugs
___ Other (Please list)_____________________________________________________
When was the last time you used any of these substances?
_______________________________________________________________________
DIFFERENCES IN EXECUTIVE FUNCTIONING 115

Appendix D

Demographic Table

Variable N %

Age
24 4 10
25 1 2.5
26 2 5
27 3 7.5
28 1 2.5
29 2 5
30 1 2.5
31 1 2.5
32 0 0
33 0 0
34 1 2.5
35 1 2.5
36 4 10
37 4 10
38 4 10
39 1 2.5
40 0 0
41 0 0
42 1 2.5
43 1 2.5
44 0 0
45 0 0
46 0 0
47 1 2.5
48 1 2.5
49 3 7.5
50 1 2.5
51 0 0
52 0 0
53 1 2.5
54 0 0
55 0 0
56 0 0
57 1 2.5
DIFFERENCES IN EXECUTIVE FUNCTIONING 116

Variable N %

Ethnicity
African American 10 25
Hispanic 28 70
Asian 1 2.5
Other 1 2.5

Highest Grade Completed

5th 1 2.5
6th 0
7th 1 2.5
8th 0
9th 0
10th 4 10
11th 8 20
High School 14 35
Diploma/GED
Some College/Vocational 12 30

Knocked Unconscious
Yes 22 55
No 18 45

Substance Abuse
Yes 40 100
No 0 0
DIFFERENCES IN EXECUTIVE FUNCTIONING 117

Appendix E

Informed Consent

Alliant International University, Irvine Campus

2855 Michelle Drive, Ste. 300
Irvine, CA 92612
(949) 812-7446

INFORMED CONSENT AGREEMENT

Differences in Executive Functioning between Current and Former Gang Members

You are being asked to participate in a research study. However, before you give your consent to be a
volunteer, we want you to read the following and ask as many questions as necessary to be sure that you
understand what your participation will involve.

INVESTIGATOR:
Heather Farris
Noor Damavandi, Ph.D., Supervisor

PURPOSE OF THE RESEARCH: You are invited to participate in the research project entitled,
Differences in Executive Functioning between Current and Former Gang Members, which is being
conducted at Alliant International University and the California School of Forensic Studies, under the
direction of Sean Sterling, Ph.D. The purpose of this study is to examine any changes that take place in the
brain, such as the ability to reason or make plans that might offer an explanation for deciding to leave or
remain in a gang.

PROCEDURES TO BE FOLLOWED DURING THE RESEARCH: This research will take place in the
offices of Homeboy Industries. You will be asked to complete a demographic questionnaire and a
questionnaire in which you will evaluate your views on your own reasoning and problem-solving. You will
review a list of statements that describes people’s behavior and indicate whether you have had any
problems with these behaviors over the past month. Examples of statements in the self-report measure are
“I have problems getting started on tasks” and “I have emotional outbursts for little reason”. The last
measure you will be asked to complete is an online test that will evaluate your ability to plan and recognize
changes in the environment. You will be shown decks of cards and asked to match a card with one shown
on the screen. Your participation will take approximately 1 hour.

RISKS: There is minimal risk to participation in this study. Answering the questions in the self-report as
well as the measure you will be tested on might cause you to feel upset or anxious. If so, you may stop at
any time.

BENEFITS OF THE RESEARCH: An incentive of $20 will be provided for your participation. You must
complete this entire study in order to receive the incentive. Others may benefit by learning about the results
of this research.

ALTERNATIVES TO THIS RESEARCH: You have the alternative not to participate in this research.
DIFFERENCES IN EXECUTIVE FUNCTIONING 118

CONFIDENTIALITY: You have a right to privacy and all information identifying you will remain
confidential (private), unless otherwise required by law. Confidentiality will be maintained by keeping the
consent forms with signatures and demographic questionnaires separate from self-reports and test data,
which will not include names or any other identifiable information and will be presented to others only
when combined with other responses. The researcher will follow the American Psychological Association
Ethical Standards including those for Research with Human Subjects. The results of this study may be
published in scientific journals, or be presented at professional meetings as long as you are not identified
and cannot be identified from it. However, it is possible that under certain circumstances data could be
subpoenaed by court order.

QUESTIONS ABOUT THE RESEARCH: If you have questions regarding this research project or your
participation, you may call Heather Farris at 714-318-2975. You may also contact Dr. Noor Damavandi at
949-812-7478. Should you have any additional concerns, please contact the Institutional Review Board at
Alliant International University at (858) 635-4448 during normal working hours.

PARTICIPANT RIGHTS AND RESEARCH WITHDRAWAL: Your participation in this study is

voluntary. If you choose not to participate in this study, there will be no penalty or loss of benefits to which
you are otherwise entitled and your relationship with Alliant International University or Homeboy
Industries will not be affected. In addition, you may discontinue participation at any time without any
penalty or loss of benefits. You may also refuse to answer any questions you do not wish to answer.

We have tried to explain all the important details about the study to you. If you have any questions that are
not answered here, the investigator will be happy to give you more information.

SIGNATURE AND ACKNOWLEDGMENT:

My signature below indicates that I have read the above information and I have had a chance to ask
questions to help me understand what my participation will involve. I agree to participate in the study until
I decide otherwise. I acknowledge having received a copy of this agreement and a copy of the SUBJECT'S
BILL OF RIGHTS. I have been told that by signing this consent form I am not giving up any of my legal
rights.

Signature of Research Participant Date

Heather Farris 714-318-2975

Researcher’s Name (Print Clearly) Contact phone number

Researcher’s Signature Date

Noor Damavandi, Ph.D. 949-812-7478

Name of Supervisor or Chair (Print Clearly) Contact phone number