V. w
In.“
"4-

:. ..-

«.7
4‘ ‘4
I:
c

 
 

utw Rare}; 2

‘1‘ h :,K I .H .9 9

u23: ﬁ..§.¥z...w«n.z ..

.1... . ...zﬁmx. Can ..

. ‘ ...:
#4.. ..rainud.§m
n! “A I A

2%}.

..
I.
v ‘ .Nfu‘

 

‘
\xlr.m.mw3 .1

I .\..
\
1,

 

; kh- . .
...Wazl. 3...: an?
3 as: lax}: in...
q ..rlu..:u 3:

{3'9\

 

h. I). y
H. .1. ...
... Lmuar
.. $3...
... a

......“
.8 :15!!!
1.51..) }

 

 

 

‘W11
3’ :‘F.

‘ I. I:
..immmw

 

 

.).~

.‘s.
~ .. .K
.. 05V. 3

...» 1!...

Haul.

J. .Wnrt-a

  
  

t 3.1%...» 5..th ...:W. 1. l

5.....r4

 

 

 

It”... . I. ... .,..n. .V ‘. . ‘ :3). .. V . . I ‘
-harmuﬁq‘n.»“mafia .;.~‘.m.,..;‘“.._.3...u3£x._. .q.§...~_.o.a.....n_..p..¢ «(31.31... .....p......\...z.w Emmy.
II»
II I

3
QOQ1

This is to certify that the
dissertation entitled

Criminogenic Variation Among Gang and Non-Gang Offenders

presented by

Mengie Michaux Parker

has been accepted towards fulﬁllment
of the requirements for the

UBRARY
Michigan State
University

 

 

Ph.D. degree in Criminal Justice

 

 

Ma' F’rofessor’ ﬁlgnature

ﬂaw /5:. 5200 7
Date

MSU is an affinnative-action, equal—opportunity employer

- -.--—..-.-.--.-.-.-.-.-.---.-o-I-I—o-n-o-u-o—u-o-o-o-o-o-u-v-a-.-o-u— — -

A- ...-.-.-.-n-.-.-~-«-

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE ‘ DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/CIRC/Date0ue indd-p.1

 

 

CRIMINOGENIC VARIATION AMONG GANG AND NON-GANG
OFF ENDERS

By

Mengie Michaux Parker

A DISSERTATION
Submitted to
Michigan State University
In partial fulﬁllment of the requirements
For the degree of
DOCTOR OF PHILOSOPHY
School of Criminal Justice

2007

ABSTRACT

CRIMINOGENIC VARIATION AMONG GANG AND NON-GAN G
OFFENDERS

By
Mengie Michaux Parker

The purpose of this study was to determine if criminogenic variations could
be used to differentiate between gang members, non-gang member who displayed
deﬁant individualistic personality traits and non-gang member who did not
display deﬁant individualistic personality traits. The data were obtained ﬁom the
Indianapolis Lever-Pulling study conducted under National Institute of Justice
grant # 2003-U-CX-1038. This study conducted discriminate function analyses of
235 offenders who participated in the Indianapolis study.

Findings revealed that there was no statistical difference in the amount of
criminal justice system contact between gang members and non-gang members
who displayed deﬁant individualist traits. This lack of differential response also
extended to attitudinal variables in the study. However, the data showed that there
was no direct relationship between the number of criminal charges a respondent
incurred and the degree of deﬁant individualism displayed.

This study found that there was no signiﬁcant difference between the
perceptions of non-gang members who displayed deﬁant individualist personality
traits and non-gang members who did not display deﬁant individualist traits. The
study also suggests a positive correlation between the degree of deﬁant

individualism and the amount of post-intervention positive lifestyle change.

Dedicated

To

Kimberly Rosario Perez (ABD)
(1 976-2006)

iii

ACKNOWLEDGEMENTS

I would like to thank God for granting me the strength and endurance to
undertake this course of study. I would like to thank my aunt, the Honorable
Pattie S. Harrison and my family for offering me unwavering moral support
during my years in graduate school.

I would like to thank my dissertation committee members: Dr. Charles
Corley, Dr. Homer Hawkins, Dr. David Carter and Dr. Richard Thomas. Their
tireless hours of evaluation, consultation and correction, have propelled me to a
higher level of professionalism. I would also like to also thank Dr. Steve Chermak
for allowing me to use data from his primary study and Dr. J esenia Pizarro for her
professional consultation and support. I would like to especially thank Sky Gaeta
for her emotional support throughout this dissertation process.

I would like to thank my mentors from North Carolina Central University:
Dr. George Wilson, Dr. Harvey McMurray, Professor Joseph Sroka, Professor C.
Robert Fenlon, Mrs. Margaret Stanley and Professor Wendell Andrews, for laying
the foundation of everything that I have become. I am forever in your debt.
Finally, I would like to thank Kimberly Rosario Perez, for showing me that one

person can change the world.

iv

TABLE OF CONTENTS

LIST OF TABLES ............................................................. vii
CHAPTER-1
Introduction ................................................................... 1
Problem Statement ............................................................ 4
Study Hypotheses ............................................................. 8
Chapter Summary .......................................................... 9
CHAPTER-2
Logic of Conceptualization .................................................. 10
Gangs in the Age of Globalization .......................................... 12
Gangs as Organized Deﬁant Individualism ................................ 27
Desistance Under Deﬁant Individualism ................................... 31
Gangs as a Group Hazard .................................................... 34
Chapter Summary .......................................................... 41
CHAPTER-3
Methodology ................................................................... 43
Variables ........................................................................ 47
Statistical Tests ................................................................. 54
Chapter Summary .......................................................... 55
CHAPTER-4
Introduction to Analysis ...................................................... 57
Univariate Analysis ............................................................ 64
Demographic ................................................................ 64
Education and Income ...................................................... 69
Criminal Justice System Contact ......................................... 72
Correlation Analysis ........................................................... 76
Criminogenic Pattern ...................................................... 77
Zero-Inﬂuence Group ...................................................... 91
Deﬁant Individualist Group ............................................... 96
Gang Member Group ...................................................... 102
Discriminant Analysis ......................................................... 107
Chapter Summary .......................................................... 113
CHAPTER-5
Summary of Purpose ........................................................... 117
Summary of Literature ......................................................... 118
Summary of Methods ......................................................... 123
Summary of Findings ......................................................... 128
Discussion of Hypotheses ..................................................... 132
Study Limitations ............................................................... 136

Recommendations .............................................................. 138

Conclusions ..................................................................... 140
Attachments .................................................................... 142
Reference ....................................................................... 202

vi

LIST OF TABLES
Table 1- Demographic Analysis by Category ............................. 67
Table 2- Descriptive Means .................................................. 71

Table 3- Descriptive means Analysis
By Category ................................................................. 72

Table 4- Primary Criminogenic Analysis
By Category ................................................................. 73

Table 5- Secondary Criminogenic Analysis
By Category ................................................................. 76

Table 6- Bivariate Correlation
Primary Criminogenics .................................................... 80

Table 7- Bivariate Correlation
Secondary Criminogenics ................................................. 82

Table 8- Bivariate Correlation
Primary and Secondary Criminogenics ................................. 84

Table 9- Partial Correlation
Primary Criminogenics .................................................... 87

Table 10- Partial Correlation
Secondary Criminogenics ................................................. 89

Table 11- Partial Correlation
Primary and Secondary Criminogenics ................................. 91

Table 12- Bivariate Correlation (Zero-Inﬂuence Group)
Primary Criminogenics .................................................... 93

Table 13- Bivariate Correlation (Zero-Inﬂuence Group)
Secondary Criminogenics ................................................. 94

Table 14- Bivariate Correlation (Zero-Inﬂuence Group)
Primary and Secondary Criminogenics ................................. 96

Table 15- Bivariate Correlation (Deﬁant Individualist Group)
Primary Criminogenics .................................................... 98

vii

Table 16- Bivariate Correlation (Deﬁant Individualist Group)
Secondary Criminogenics ................................................. 99

Table 17- Bivariate Correlation (Deﬁant Individualist Group)
Primary and Secondary Criminogenics ................................. 102

Table 18- Bivariate Correlation (Gang Member Group)
Primary Criminogenics .................................................... 103

Table 19- Bivariate Correlation (Gang Member Group)
Secondary Criminogenics ................................................. 105

Table 20- Bivariate Correlation (Gang Member Group)
Primary and Secondary Criminogenics ................................. 107

Table 21- Discriminant Analysis (Zero-Inﬂuence and Deﬁant Individualists)
Secondary Criminogenics ................................................ 109

Table 22- Discriminant Analysis (Deﬁant Individualists and Gang Members)
Primary Criminogenics .................................................... 1 11

Table 23- Discriminant Analysis (Deﬁant Individualists and Gang Members)
Secondary Criminogenics .................................................. 112

viii

Shank—r!

“Introduction”

Gangs in the United States have become a pervasive criminological threat.
Their increase in the latter part of the twentieth century has been one of the more
consistent criminological trends identiﬁed by researchers. In the past, many
criminal justice and judicial practitioners have reported gangs in their respective
jurisdictions. Whether these jurisdictional administrators viewed gangs as a
serious problem or not, law enforcement ofﬁcials are becoming more forthright
about the growing presence of organized crime.

The greatest increase in gangs occurred between 1980 and 1995. In 1980,
there where approximately 286 cities with more than 2,000 gangs and close to
100,000 gang members (Jackson, 1999). In 1995, the gang representation had
increased to about 2,000 cities with more than 25,000 gangs and 650,000
members (Jackson, 1999). Results of the 2002, National Youth Gang Survey
suggest that this increase continued into the 21St century. In 2002 the number of
gangs in the US. increased to 21,500 with an active membership of 73 1 ,500 gang
members (01] DP, 2004). So consistent was the increase in gangs and crime that
the correlation has become almost axiomatic within social science. However,
there is a new trend that differs somewhat from the gang trends of the last two
decades.

The 2004 National Youth Gang Survey suggests that there was a decline in
the number of criminal justice agencies reporting gang problems (Egley Jr. and

Ritz, 2006). The 2004 survey results reveal slight increases in active gang

membership (760,000 members of 24,000 gangs in 2,900 jurisdictions) but these
increases were not statistically signiﬁcant (Ibid). Moreover of the 2,900
responding agencies only about 47% said that the gang problem in their
jurisdiction was getting worse (Egley Jr. and Ritz, 2006). The data ﬁorn the 2004
survey appears to suggest that the US. gang problem is beginning to dissipate.
Victim data supports a similar conclusion.

Based on the National Crime Victimization Survey (N CVS), gang violence
declined from 1.1 million violent victimizations in 1994 to only 341,000 in 2003
(Harrel, 2005). During that time, crime victims identiﬁed the alleged perpetrators
as gang members approximately 12% of the time (Ibid). Perpetrators were
identiﬁed as gang members in about 10% of robberies and 4% of the rapes
(Harrel, 2005). However there are three alternative explanations for these most
recent ﬁndings. Speciﬁcally, the 2004 National Youth Gang Survey ﬁndings
could have been inﬂuenced by criminal justice practitioners becoming more
acclimated to gangs and gang-related crime, (2) political intervention in the
agencies’ responses and or (3) changes within the gang culture itself.

It is important to note that the actual number of gangs and gang members
increased only slightly and the perception of gangs as a problem declined. The
ﬁrst alternative explanation for the decline in perception could be that criminal
justice ofﬁcials and practitioners are becoming accustomed to the dynamics of
gang-related crime. The discovery of gangs in a jurisdiction is no longer a serious
impediment to the daily operations of criminal justice agencies. Criminal justice

agencies can take full advantage of a host of gang investigator associations and

Internet based web sites to help address the impending gang phenomena. There is
a growmg wealth and availability of resources to alleviate the conceptualization
of gangs as a serious problem.

Another alternative explanation for the reduction in criminal justice
perceptions of gangs as highly problematic could be political inﬂuence. In many
locations, especially where jurisdiction leaders hold elected ofﬁce, it is
detrimental to an individual’s career to appear soft on crime. The appearance of
being unable to control crime in the jurisdiction may prompt some agency heads
to endorse departmental policies whereby gangs and gang-related issues are not
per se problematic. Subsequently, lower ranking individuals within the
bureaucracy may follow suit in recording and reporting criminal justice
information pursuant to perceptions of gang related activities.

The third alternate explanation for the reduction in the perception of the
gang phenomenon is that the fundamental nature of gangs could be changing.
Over the past two decades, there has been a great deal of sociological change in
the US. and the world. Criminologists admit that those same forces that impact
normative social structures also impact deviant structures. It is a theoretical error
to assume that gangs are passive systems that exist only to be acted upon. In
reality, gangs react not only to internal group dynamics but also to variations in
law enforcement strategies and the social contexts in which they exist. In short,
gangs evolve. It is the third alternative explanation that is explored by this study.

The boundaries between gang members and non-gang members are

blurring. As the behavior of gang members and individual offenders has become

more parsimonious, so too has their crimes. Due to technological advances such
as increased communication and the availability of personal computers, individual
offenders have the ability to plan and commit more sophisticated criminal
operation without the reliance on a stable criminal organization. The changing
mode of production in normative society has not gone unnoticed by deviant
members of society. Gangs can take advantage of the same networking models in
the commission of crime that legitimate business use to regulate international
trade. Formal membership is no longer the salient characteristic of this criminal
organization; yet traditional gang analysis relies heavily on membership as a
criterion for study inclusion. Due to this evolution in organized crime, the
traditional conceptualization of gangs must undergo a ﬁrndamental re-
conceptualization.

“Problem Statement”

Because of the change in gangs and gang-related dynamics, which is
causing gangs and non-gang members to exhibit parsimonious criminal behavior,
the differences between gang members and non- gang members must be re-
assessed. Without re-examining the differences between gang members and non-
gang members it may not be possible to craft effective interventions in addressing
organized criminality. Reliance on outdated paradigms threatens to stagnate
criminological progress in understanding the correlates of criminal careers,
thereby ensuring that no desistance is possible. This fundamental re-

conceptualization requires non-traditional methods of measurement and analysis

in order to more accurately examine the variation in criminal behavior across the
two groups.

Traditionally, gang studies have relied on the group hazard hypothesis as an
underlying assumption for the existence of gangs and gang-related crime. The
group hazard hypothesis assumes that deviance is essentially an aggregate concept
because groups induce, sustain and permit deviant behavior (Dentler and Erikson,
1959). This underlying assumption has led many researchers to focus on formal
membership as a criterion and category with which to correlate crime. However,
with the punctuated equilibrium of the gang culture, membership has become
more transient. F ormal membership is not as necessary and important as it might
have once been.

By networking with other criminals to conduct speciﬁc criminal operation,
individuals are relieved from the burden of maintaining large criminal
organizations. Once an operation is completed, all participants are free to go their
own way without the fear of betrayal or the burden of organizational maintenance.
In considering the evolution of organized crime, it is more accurate to couch the
analysis within the context of Organized Deﬁant Individualism.

Organized Deﬁant Individualism is a personality trait that is characterized
by simultaneously possessing mainstream social values and few resources with
which to achieve them (Sanchez-Jankowski, 2003). Deﬁant Individualism causes
people to undertake economic operations that can be either legal or illegal (Ibid).
Deﬁant Individualism also has an underlying dynamic that resists any attempt to

stop the Deﬁant Individualism (Ibid). This new gang conceptualization posits that

the gang is simply a grouping of people who have this personality trait. According
to Organized Deﬁant Individualism, formal membership does not hold the same
level of importance as in Group Hazard models. This new conceptualization
requires new methods of measuring gang-related crimes.

One method of measuring Deﬁant Individualism is on a continuum. Several
researchers (Y ablonsky, 1959; Hagadom, 1998; Morash, 1983; Thornberry,
Krohn, Lizotte and Weirschen, 1993) have recommended measuring gangs on a
continuum. The continuum allows researchers to examine variability more
accurately than by using dichotomous measures. On a continuum where gangs are
at one extreme and people with no Deﬁant Individualist characteristics are at the
other; it is possible to assess the fact that as an individual moves closer to the
gang end of the continuum he or she becomes more committed to the Deﬁant
Individualism personality. Formal membership in the gang is simply the ﬁnal step
in the process. A similar process toward general deviance was described by
Thrasher in his pivotal 1927 study of gangs and was reported in Morash (1983).
Organized Deﬁant Individualism does maintain provisions for group deviance but
it differs from the Group Hazard Hypothesis in group stability and function.

The Group Hazard Hypothesis suggests that gangs are more or less stable
structures in which the membership may be dynamic and constantly changing.
Conversely, the Organized Deﬁant Individualism theory supports the idea that the
dynamic nature of the gang and the membership are coterminous dynamics,
neither independent nor dependent on the other. The Group Hazard Hypothesis

clearly states that groups tend to induce, sustain and permit deviant behavior

(Dentler and Erikson, 1959). Under the Group Hazard ﬁarnework, the gang
functions as a workshop of deviance. It is within the conﬁnes of the workshop
that individuals discover, participate and receive support for various types of
criminal endeavors. The gang under the Organized Deﬁant Individualism
framework is much more akin to a tool; the gang is simply a way to accomplish a
goal, deviant or legitimate. Once a new conceptualization is formed, an
accompanying measurement model must be developed.

The purpose of the measurement model is to ensure that the analysis is
being conducted in a systematic and thorough manner. In examining a
phenomenon as diverse as gang crime, it is necessary to examine both gang crime
and gang attitudes toward crime. In order to achieve a more thorough assessment
of the Organized Deﬁant Individualism concept primary and secondary
criminogenics will be used.

Primary criminogenics are the actual rates of criminal behavior that an
individual commits. These measures can include several different offenses and
incidents. For example, the number of criminal convictions, the number of drug
arrests, the number of property arrests and the number of felony charges are all
primary criminogenics because they are either direct criminal behaviors or are the
result of direct criminal behavior. Conversely, secondary criminogenics are
attitudinal measures. Secondary criminogenics are not necessarily illegal since
they are attitude based. Secondary criminogenics such as attitude toward gun use
and attitude toward authority may affect how the individual views and interacts

with society, but are not direct criminal acts. The attitudes that support illegal

behavior are as important to understanding the new boundaries of gang behavior
as the behavior itself.

The study will use data from a sample of probationers which includes gang
members, individuals who exhibit Deﬁant Individualist traits and individuals who
display no criminal organizational inﬂuence. The ﬁmdamental nature of gangs has
changed to the degree that traditional conceptualizations and measurements are
insufﬁcient to draw a distinction between gang members and non-gang members.
Thus it is hypothesized that gang members and non-gang members who display
Deﬁant Individualist characteristics exhibit similar patterns of both criminality
and attitudes toward crime. Based on the literature review, the following
hypotheses were formulated.

“Study Hypotheses”

H I: There is no diﬁference in criminal justice contact between gang members
and Deﬁant Individualists that would denote a discriminant ﬁmction.

H 2: There is a direct relationship between the numbers of criminal charges
accumulated and the level of Deﬁant Individualism.

H3: There is no difference in the attitude toward gun use between deﬁant
individualists and subjects in the zero-inﬂuence group.

H4: There is an inverse relationship between Deﬁant Individualism and
positive trajectory shifts.

H 5: There is no diﬁference in the perception of gang criminality between gang

members and Defiant Individualists.

“Summary”

Despite the slight increase in gang membership reported by some law
enforcement agencies, crime victims are reporting relatively low incidents of
victimization by gang members. There is also no ofﬁcial data that show gang
crime to be decreasing. Ofﬁcial data, as well as gang research, often focuses on
membership as an inclusionary criterion. However, as gangs have changed they
have removed the necessity of membership.

Gangs are dynamic entities that are evolving in ways that make traditional
membership obsolete. This changing gang dynamic can affect research ﬁndings in
different ways. If ofﬁcial data and research are focusing on gang membership,
which is no longer as important to the gang culture as it once was, there might
appear to be only a slight increase when past years have shown large increases in
gang participation.

Conversely, if gangs are moving away from formal membership, victims
may still be preyed upon by individuals who are not actually gang members but
who commit gang-related crimes such as home invasions and drug sales. These
types of crimes require collusion but not necessarily formal membership. Victims
would be in a better position to know whether a perpetrator was in a gang or not.
Jurisdictions that fail to report gangs as a problem may do so because gangs have
become normalized to the jurisdiction or perhaps the law enforcement ofﬁcials
simply do not recognize the emerging gang structures. The term evolution is
somewhat misleading; it would be better to classify the change in gang structure

as punctuated equilibrium.

The punctuated equilibrium is an anthropological theory that can best be
described as the introduction or expansion of an existing organism beyond the
traditional applications. The interesting thing about punctuated equilibrium is that
the expansion of the traditional function of some organisms does not make all of
the organisms irrelevant. Just because some organisms evolve to more
sophisticated functions does not mean that the less sophisticated organisms cease
to exist. The changes occurring in the gang culture are characterized by the

reduction in reliance on formal membership.

10

MILL!
“The Logic of Conceptualization”

The systematic study of gangs, like most criminological concepts, varies
based upon the conceptualization of the phenomenon. Conceptualization
embodies more than a simple deﬁnition. Fear of crime, for example, does not
exist. Fear of crime has no chemical composition, no color, no size or shape and
no atomic mass. Fear of crime exists because individuals agree that it exists.
Therefore, conceptualizations of constructs must also contain underlying
assumptions about the phenomenon that color both the analysis and interpretation
of research ﬁndings. Gangs and gang-related crime are also a construct. The
relationship between conceptualization and construct can be either inductive or
deductive.

A deductive relationship between conceptualization and construct is most
commonly found in criminological studies. Deductive models test general theories
and apply them to speciﬁc groups or behaviors. Conceptualizations of
criminological constructs are oﬁen formed prior to any empirical tests designed to
verify an a priori theory. The deductive model is, of course, a valid paradigm and
heavily relied upon by social science researches. However, sometimes it is
necessary to use an inductive process to inform a conceptualization, such as when
phenomena undergo autonomic changes.

The inductive relationship between conceptualization and construct begins
with observation of the phenomenon and seeks to move from speciﬁc to general

statements about the phenomenon. The inductive process can be quite helpful in

11

re-examining phenomena after or during times of change. One element that makes
sociological research challenging is the fact that study subjects are not passive
back drops on which social processes are acted out; they can take an active role in
the production or reproduction of crime. This ability to be an active participant
allows individual the ability to create autonomic change or change that is
independent of any structural inﬂuence. This type of individually-inspired
autonomic change that has blurred the boundaries between gang member and non-
gang members. In order to understand how a re-conceptualization of gangs is
necessary, one must ﬁrst understand the societal changes that have produced the
need for the conceptual shift.

“Gangs in the Age of Globalization”

Globalization has had a profound effect on contemporary gangs. However,
prior to the examination of globalization’s effect on gangs, it is necessary to
clarify the concepts discussed. As a phenomenon, globalization is often
misunderstood and has many different conceptualizations. Therefore, for the
purpose of this analysis globalization is deﬁned as: The technologically produced
reduction in the spacio-temporal characteristics of civilization; whereby advances
in economics, communications and socio-cultural transfers are realized at the
individual level. It is necessary to ensure a thorough conceptual understanding of
globalization that goes beyond a simple deﬁnition. This deﬁnition allows
discussion of globalization in terms of both time and space.

One of the most notable outcomes of globalization has been the reduction

in time and space. The spacio-temporal reduction is only possible because of how

12

we as humans conceptualize distance. Distance, or space, is often viewed and
discussed in relation to time. When taking a ﬂight ﬁom North Carolina to
California, we often discuss the trip as a 10-hour ﬂight not as a 3,000-mile
journey. Therefore, by simply decreasing the amount of time it takes to move
from one place to another, we decrease the space. Geographical distance has not
been altered, only our perception of the geographical distance.
This spatial perception is not merely two-dimensional. In his second book

on globalization, Friedman (2005) uses the analogy that the world is ﬂat as a
method of demonstrating that not only is the perception of space and time
changing but also the perception of baniers and walls. Friedman (2005) argues
that not only is globalization shrinking the world but that it is leveling the playing
ﬁeld. Under globalization, this perceptual change in space and time is the direct
result of technological advances. Globalization can be seen as having two types of
advances: ﬁrst order advances (input) and second order advances (outcome).

Technological advances are ﬁrst-order advances and are somewhat of a
misnomer. The technological advances such as cellular phones, personal
computers, high-speed Internet capabilities and long-range super sonic air travel
are not actually advances. The technological revolution that has helped spurn
globalization is simply an extension of existing products. The extension allows
common products to be used in ways other than how they were originally
conceived. For example, advanced cellular phones not only can place and receive
calls anywhere in the world but they can also check e-mail, take high resolution

photographs, play music, surf the Internet and record video. These technological

13

advances, in turn, provide mechanisms with which to realize economic, socio-
cultural transfer and communication advances.

Economic, socio-cultural and communication gains are considered second-
order (output) advances and are similar in nature to the technological advances.
Second-order advances are also extensions or simpliﬁcations of previous
processes. International ﬁnance and trade had always existed. It is, of course,
much easier in a time where most business people can instantaneously access
ﬁnancial records via computer or hand held Blackberry device. There has always
been cultural transfer as people have migrated ﬁ'om one part of the world to
another. However, the regularity, speed and lower cost of migration add to the
relative ease of migration and subsequent cultural transfer. Globalization affects
many different aspects of people’s lives, including crime.

It seems only logical that a globalization effect would accrue to gangs.
Hobbs (2001) crystallizes the issue of globalization’s effect on local crime by
explaining that the historical backcloth, upon which organized crime is acted out,
is in the local working class communities. As local economies are changed by
globalization, through fragmentation or de-industrialization, organized crime will
also be transformed (Ibid). Globalization has two processes that most impact gang
crime at the local level: Democratization and the Network Enterprise.

The concept of Democratization (Friedman, 2000; Giddens, 2003) is an
extension of the basic globalization conceptualization. Globalization describes the
advances produced by technological irmovation, while democratization explains

to whom the advances accrue. While Friedman frames his discussion in more

14

general terms, Giddens (2003) speciﬁcally discusses the need for a democratizing
of democracy due to the members of society living in the same informational
environment as the people in power over them.

The shared informational environment can be seen in the media coverage
of war. In contemporary social conﬂicts, embedded reporters are able to show
real-time video of war. Due to advances in satellite communications, the same
level of information that government leaders have on battle tactics, causalities and
campaign success is also available to citizens. Questionable government conduct
is more difficult to keep secret due to the anonymity with which whistle blowers
can inform the public through a host of Internet applications. Citizens have direct
access to experts from around the world, which allows them to hear independent
assessments of governmental policies and are therefore not relegated to accepting
the ofﬁcial version of daily events.

Democratization is best conceptualized as: the process of obtaining access
to any advantage that was previously unattainable by the populace and once
reserved for the state. Under the auspices of globalization, democratization has
occurred in many different areas. When citizens have access to news with the
same speed as government ofﬁcials, democratization of information has occurred.
In nation-states where the citizens can purchase the same personal computers to
organize their work that government agencies use to conduct public planning,
democratization of technology has occurred. When citizens can hire a private
bodyguard force to provide 24-hour protection just like members of the Secret

Service, democratization of security has occurred. Perhaps the three most

15

inﬂuential elements in society that have impacted crime at the local level are the
democratization of technology and information as well as the democratization of
the mechanisms of war.

Personal computers, cellular phones, 24-hour banking, super sonic
international ﬂights and satellite communications have proliferated into the
mainstream populace at a rapid rate. As these technological innovations have
become common place, it would be only logical for the criminal populace to make
use of them as well. Criminals can avoid law enforcement wiretapping by using
unlocked Global Systems for Mobile communication (GSM) cell phones and
highly encrypted computer software. Criminals have an enhanced ability to ﬂee
prosecution with the increase of international travel. Criminals also have the
ability to select weapons that are on par with, or superior to, those of many law
enforcement agencies. The speed at which modern society adopts new technology
is different from past generations.

From the introduction of radio it took 40 years to obtain an audience of 50
million listeners in the United States; whereas the Internet only took 4 years to
obtain 50 million users after its introduction (Giddens, 2003). However,
technological proliferation is not the only issue in the age of globalization. The
nature and extended use of the Internet must be considered. This extended use of
the Internet has not gone unnoticed by government ofﬁcials. In an analogy of the
Internet’s effect on globalization, a United States Federal Communication
Commission advisor likened the Internet to Roman roads (Hardt and Negri,

2000). The information super highway provides ordinary citizens with the ability

16

to disseminate their ideas, beliefs, movements and cultures much the same way
that Roman roads spread Roman ideals.

Society has become so technologically interdependent that an entirely new
technology-based paradigm has emerged. The seminal work on the information
technology paradigm and network enterprise was written by Manuel Castells, who
is consistently sited by many scholars who study globalization and the new
technology paradigm (Audirac, 2003; Hobbs, 2001; Parker, 2002; Passas, 2000;
Tillman, 2002; Scholte, 2003; Hardt and Negri, 2000, Tomlinson, 2003; Dicken,
2003; Held, 2003 and Goldsmith, 2000).

Castells (2000) discusses the deﬁning characteristics of the new paradigm
as an information-technology paradigm. This new paradigm consists of many
different compatible technologies found in new gadgets that beneﬁt the user
because they are widely available and grow relatively less expensive over time
(Ibid). Castells (2000) also asserts that the new information-technology paradigm
is the foundation of the network enterprise. By using these new, more versatile
electronic tools, criminals can communicate and coordinate crimes more
effectively. The added features of ready availability and ever falling cost only
hasten and solidify the adoption of the new technology. There are ﬁve major
components of this information-technology paradigm: information as the raw
material, pervasive effect of the new technology, networking logistics, ﬂexibility
and convergence of speciﬁc technologies into a highly integrated system. These

components help develop an understanding of its applications to crime.

17

The ﬁrst characteristic of the information technology paradigm is that
information is the raw material (Castells, 2000). Within a society where a high
degree of reliance is placed on information, criminals are able to develop new
endeavors based on the manipulation of information. Unlike situations involving
drug deals or stolen property, information-related crimes are not limited by
logistics. There need not be a great deal of planning about where to hide stolen
identiﬁcations or bank accounts. Society is very familiar with the havoc that can
be wreaked by corporate espionage. When information is the basic raw material
of the new economy, it gives the average criminal the ability to pursue more
lucrative crime by becoming an information broker. One example of this
information transition occurred with a gang in the southeastern United States.

The organization in question, Blue Gang (A ﬁctitious name due to
conﬁdentiality concerns), is an international transplant to a medium sized city in
the southeastern United States. Shortlyaﬁer its migration, gang members begin to
sell a formula for making ‘black crack’. There are two primary formulas being
sold to local drug gangs. The ﬁrst formula (German) involves lacing the cocaine
mixture with pencil lead in order to turn the ﬁnished crack cocaine, black. The
second formula (Eastern European) uses iodine as an additive while cooking the
crack and will also turn the ﬁnished product black. The use of the speciﬁc
chemicals does not change the potency of the cocaine but it does prevent law
enforcement test kits from rendering a positive ﬁnding. If law enforcement test
kits carmot show a substance to be positive, the ofﬁcer cannot charge the suspect

with possession of the illegal substance. Additionally, the ofﬁcer must also be

18

able show that the substance had the same appearance of the drug in order to
charge the suspect with possession of a counterfeit substance.

Now, while members of the Blue Gang never sold drugs in this speciﬁc
jurisdiction, they sold the information on better methods of producing drugs to
other gangs for proﬁt. Thus, the Blue Gang capitalized on the informational
nature of crime to create a new type of market. The sophisticated nature of these
crimes is in stark contrast to the traditional conceptualizations of gangs as low-
level street criminals. The contemporary gang exercises an amazing amount of
agency and sophistication in designing criminal enterprises.

The second characteristic of the new information technology paradigm is
the pervasive effect of the new technology (Castells, 2000). Because our society
has fully embraced the advances in technology, criminals are better able to uses
these technologies in interesting ways. Criminal entities are better able to conduct
surveillance on potential targets as well as develop more intricate plans for
criminal operations. Computer hackers develop sophisticated programs designed
to commit cyber crime and evade law enforcement. In her 1998 study, Shelley
states that technology has changed the very nature of crime. One example is the
use of the Internet to recruit gang members. A North Carolina gang (name
withheld due to conﬁdentiality concerns) has had a recruitment website as far
back as 1998.

The third characteristic of the information-technology paradigm is the
networking logistic aspect (Castells, 2000). As we will discuss later, criminals are

networking with greater frequency than ever before. Cooperation among criminal

l9

entities is very natural due to the fact that they have the common problem of
evading law enforcement (McCusker, 2004). One example of the networking
capabilities of criminal entities in the technology paradigm occurred in October of
2000; a Sicilian Maﬁa group in conjunction with 20 other individuals created a
digital clone of the Bank of Sicily’s online component (McCusker, 2004). The
group planned to steal $400 million allotted to the bank by the European Union
and very possibly could have succeeded had the plot not been revealed by an
informant (Ibid). The real issue was not that the group tried and failed but that
they conceived of the plan in the ﬁrst place (McCusker, 2004).

Next the information-technology paradigm is characterized by ﬂexibility
(Castells, 2000). The ﬂexibility of contemporary crime is, to a great deal,
beholding to the networking logistic capabilities. Criminal entities now have the
ability to pull resources from any part of the planet as needed. An example of this
ﬂexibility network logistic dynamic occurred in the southeastern United States. In
1999 there was an ongoing conﬂict between two Hispanic gangs. A member of
the Yellow Gang (A ﬁctitious name for conﬁdentiality concerns) killed a member
of the Orange Gang (A ﬁctitious name for conﬁdentiality concerns) at an
apartment complex. Members of the Orange Gang shipped the dead body of its
member to El Salvador without ever notifying the authorities. The Orange Gang
then contracted an assassin from El Salvador to retaliate for the murder.

After killing two members of the Yellow gang, the assassin quickly
returned to El Salvador. It was only during the investigation of the second murder

that the full scope of the crime was uncovered. The most interesting aspect of this

20

case was the fact that the people in the apartment complex where the shooting
took place spread rumors of the murder but would not openly conﬁrm details of
the incident until long after the guilty parties had both ﬂed the country. Thus, this
local gang was demonstrating its ability to utilize global networking and resources
to carry out sophisticated criminal operation. In addition, we see that the
community was co—opted into keeping silent either through fear or overt loyalty to
the gang.

The ﬁnal characteristic of the new information-technology paradigm is
convergence of speciﬁc technologies into a highly integrated system (Castells,
2000). This new system, while designed for improved commerce and economic
growth, is quickly being embraced by the criminal rank and ﬁle. Career criminals
are becoming generalists with respect to the types of crime they commit.

The democratization of information is almost synonymous with the
technology but has enough variation to merit a separate discussion, albeit brief.
The multitude now has access to the same level of information as those in higher
levels of government in the traditional nation-state (Giddens, 2003). Criminals are
constantly devising new ways to use this access to new information. Identity theft,
cyber crimes and a whole host of other fraud-based crimes are on the increase in
the United States. The ready availability of the new information on social security
numbers, credit scores and even pay stubs enables criminals to commit ﬁnancial
crimes with a level of ease never before seen. Even banks are subject to

increasing ‘phishing’ attacks, which are designed to search for customer ﬁnancial

21

information. Perhaps the greatest problem associated with democratization is the
democratization of the mechanisms of war.

Weapons, just as anything else, have become comodiﬁed under
globalization. For the ﬁrst time in history, the individual has access to not only
traditional small arms but also weapons of mass destruction such as nuclear and
biological weapons. Weapons of choice include Man-Pad shoulder ﬁred missiles,
of which, 4,000 have gone missing from the former Iraqi arsenal (N aim, 2005).
The Iraqi loss is small in comparison to the total picture. According to the Small
Arms Survey, 100,000 Man-Pads are currently unaccounted for (Ibid).

The survey also revealed that at least 13 non-state groups own these
weapons (N aim, 2005). There are indications that these weapons are not mere
trophies or objects of discussion; these weapons are being used. In 2002 a Man-
Pad missile was ﬁred at an Israeli passenger plane as it departed Mombasa, Kenya
(N aim, 2005). The demand for this particular weapon is so high that companies in
Pakistan, North Korea, Egypt and Vietnam are now supplying additional groups
(Ibid).

Naim (2005) also states that Dr. A.Q. Khan, a nuclear arms dealer, is only
one of the ruthless, talented entrepreneurs who sell weapons internationally. The
demand for these more powerful weapons has even prompted some traditional
corporations to enter the arms trade. Elf, the French state owned oil company,
arranged the ﬁnancial backing for the Lissouba Regime to purchase $61.3 million
worth of light weapons from Iran, helicopters from Russia and the services of 40

Russian technicians (N aim, 2005). It is difﬁcult to separate the acquisition of the

22

weapons from the manner in which they will be used. Democratization has made
technology, information and the mechanisms of violence much more readily
available to citizens and criminals alike. Many gangs such as Latin kings, Mara
Salvatrucha and El Rukins have developed strategic alliances with international
non-state groups. These alliances create access to these uncontrolled military
weapons. Globalization has also impacted gangs through networked enterprises.
The use of network enterprises is increasing among criminals; however,

there is some general confusion about the concept. The confusion associated with
the network enterprise often stems from semantics. Criminology ﬁrst dealt with
network enterprises when combating gangs and organized crime. Organizations
that are involved in criminal network enterprises are sometimes referred to as
swarms, due to their loose organizational dynamics. The swarm organization is
formidable because it differs so much from traditional criminal structures.

Unlike Weberian structures, a swarm has no head and its members do not
necessarily need high levels of intelligence, it has no identiﬁable organizational
structure and it can form and disperse almost instantly (Hardt and Negri, 2004). It
is difﬁcult to use traditional law enforcement tactics when combating
organizations that use swarm structures. The formidable power of the swarm can
be seen even in nature. For example, although a single termite may not
necessarily be intelligent, a swarm of termites is an intelligent system (Hardt and
Negri, 2004).

The use of swarm structures becomes a force multiplier for an individual

criminal. The greatest threat to any criminal enterprise is the threat of betrayal,

23

such as the Sicilian Bank Fraud (McCusker, 2004). However, criminals who
properly network their crimes can avoid this threat because they effectively have
no permanent ‘members’ by which to be betrayed. The Internet also uses a type of
swarm intelligence and is almost impervious to attack from hostile extemalities
because of its composition of multiple singularities (Hardt and Negri, 2004). This
collaboration of singularities is what made the illegal downloading of music so
difficult for the record industry to curtail. In Hobbs’ discussion of the transition
from traditional British criminal ‘family ﬁrms’ to serious crime networks, he
provides an excellent assessment of the swarm.

In an environment where traditional neighborhoods have disintegrated
and family ﬁrms have lost their traditional notions of territory, a serious crime
network can operate as ﬂuid and ﬂexible marauders on changing terrain (Hobbs,
2001). Hobbs goes on to classify these coalitions of loosely structure collectives
as local social systems that could no longer rely on older forms of territorial
dominance seen in the 1950s and 19605. For the purpose of this response, the
swarm concept describes the actors and the network enterprise describes the
action.

The network enterprise is best deﬁned as: a type of enterprise in which the
system of means is composed of the intersections of segments of autonomous
systems of goals; thereby producing components that are paradoxically both
dependent and autonomous (Castells, 2000). Areas that result ﬁom ﬁagmented
working class neighborhoods and local labor markets are seen by some scholars

(Hobbs, 2001) as the ideal environment for both legal and illegal opportunities

24

due to the extra territoriality, support of ﬂexible networks and entrepreneurial
orientation. These areas are perfect for network enterprises. The network
enterprise is based on two principles: connectedness and consistency, and is
deceptively simple when applied to criminal enterprises.

Connectedness is the degree to which the network can facilitate noise-free
communication (Castells, 2000). By using swarm structure, a few intelligent
criminals can conduct very sophisticated criminal operations provided that they
can ensure their networks. With the technological proliferation previously
discussed, noise-free communication is more attainable. Criminals are able to
seamlessly integrate target acquisition, planning, criminal operation and dispersal
as effectively as any banker servicing a client’s account.

Consistency is the degree to which there is a shared interest between the
network’s goals and components (Castells, 2000). Consistency is the true
unknown element in examining these new criminal forms. The network enterprise
is revealing that the strangest of bed fellows can function well together. There are
already allegiances between Mexican and Colombian drug Cartels, between
Mexican and Chinese human trafﬁckers and between Colombian and Sicilian
drug trafﬁckers (McCusker, 2004). Perhaps the greatest example of consistency
within a networked criminal enterprise is the relationship between terrorist groups
and American street gangs. One of the better known terrorist-gang allegiance is
between Al Qaeda and the Mara Salvatrucha.

The Mara Salvatrucha (MS-l 3) is an international gang with deep

American roots. This gang came to the attention of Al Qaeda leaders after the

25

September 11th attacks and was quickly identiﬁed as an organization that could
facilitate the transport of weapons and humans into the US. illegally (W illiams,
2005). The networked enterprise so far has been very active. Between 2002 and
2004 thousands of ‘Special Interest Aliens’ have been smuggled into the US
from countries such as Saudi Arabia, Pakistan, Afghanistan, Egypt, Syria, Yemen
and Iraq (Williams, 2005). Not only are federal ofﬁcials aware of this network
enterprise, the authorities have renamed one particular crossing point ‘Arab Ally’
and another outside of Douglas Arizona; ‘Arab Road’ (Williams, 2005).

It is important to reiterate that the network enterprise is a legitimate
feature of our new information economy and is no more criminal than commerce
itself. It has, however, been re-invented to serve criminal purposes. These
legitimate innovations have produced an enhancement effect on gangs and gang-
related crime. There is a general consensus among criminologists (Stern, 2003;
Tillman, 2002; Williams, 2005; Clark, 2004; Finckenauer and Veroin, 2001, Baily
and Unnithan, 1994; Shelley, 1998; Somarajah, 2004; Passas, 2000) that
globalization has the potential to increase crime both globally and locally.

Local gangs can use the extended capabilities of their cellular phones and
laptop computers to more efﬁciently establish local narcotics rings. Conducting
counter surveillance on local law enforcement ofﬁcials is much easier with
cameras that can take 2 mega pixel photos of suspected under cover ofﬁcers and
e-mail them to fellow gang members. Highly motivated individuals can even
construct impromptu organizations for certain lucrative criminal operations; then

instantaneously disband them at the completion of the enterprise. Globalization

26

allows the criminal individual to truly function as a criminal mastermind and
conduct highly sophisticated crimes at the local level. The result of this increase
in networked enterprise and information-technology proliferation is gangs that
require different theoretical ﬁameworks with which to study them.

“Gangs as Organized Defiant Individualism”

Due to the changes in contemporary gang dynamics, this study proposes an
alternative conceptualization of gangs. Sanchez-Jankowski (2003) conceptualizes
gangs as organized deﬁant individualism. Deﬁant individualism is a personality
trait that is produced when an individual, usually lower income, simultaneously
possesses mainstream social values but has few resources with which to achieve
them (Sanchez-Jankowski, 2003). Deﬁant individualism causes people to
undertake economic ventures, legal or illegal and has an underlying dynamic that
resists any attempt to stop the deﬁant individualist (Ibid). Sanchez-Jankowski
(2003) goes on to say that deﬁant individualism can be viewed in both working
class and poor areas and that almost all gang members have it. He therefore
conceptualizes gangs as organized deﬁant individualism. Some of the more
salient features of gangs lend themselves to the organized deﬁant individualism
model.

According to deﬁant individualism, the manufacture, sale and delivery of
illegal drugs is the means by which deﬁant individualists realize mainstream
goals. Within the American society those values tend to be monetary, which
creates a nice ﬁt between means and ends. As we see with Levitt and Venkatesh

(2000), gang members in the study were driven by the prospects of future

27

earnings, not necessarily what they were making at the time of the study,
therefore, the gang represented a direct beneﬁt to members as Sanchez-Jankowski
(2003) claimed. However, this differs ﬁom the group hazard hypothesis because
the gang is not necessarily creating the illegality; it is simply the matrix in which
it is conducted.

Interestingly, rap music is another example of organized deﬁant
individualism. There are many cultural elements of the rap music industry that are
shared with the gang culture; however, the production, sale and delivery of Rap
music is legal. Most rap musicians construct and maintain large entourages, just
like gangs. There are well-publicized violent rivalries between various rap groups
and even rappers who live and represent one area of the country or another, just
like gangs. Rap musicians even kill one another over seemingly innocuous
rivalries, just like gang members. As described by deﬁant individualism, the
product in the rap industry is legal and the participants are going to any means to
prevent anyone from stopping the endeavor. Violence, unfortunately, is the most
common and reliable means for both gang members and rap musicians to use.

Gang members and rappers invariably use violence as a means of protecting
their endeavors, legal or illegal. Violence is simply expedient. It is important to
remember that a central element of deﬁant individualism is the lack of resources
to achieve social values. This study asserts that the lack of resources extends not
only to social networks or to economics but also to coping mechanisms. For
example, territorialism is often used as a variable through which to understand

gang-related conﬂicts. Violent clashes over territory are simply attempts to

28

protect market share in illegal markets or enterprises such as extortion rackets,
burglary, drugs or robbery. Rap musicians resort to violence in much the same
way.

When faced with a situation where one rap group insults another, the insult
is perceived as causing the receiving group to lose face, which equates to a loss of
credibility in the rap music industry. Credibility is a crucial element in the
marketing of any rap musician’s work. Violence is the rapper’s method of
protecting his or her enterprise. The reason there is an over representation of gang
members in the rap music industry is that gang members already understand the
organization of deﬁant individualism; rap music simply represents a new name to
the same game. Using the organized deﬁant individualist framework has several
advantages.

Organized deﬁant individualism (ODI) allows researchers to examine gang-
related crime on a continuum ﬁ'om having no elements of the ODI personality
trait to formal membership in a gang. The need for a continuum approach to gang
research has been called for by several researchers (Y ablonsky, 1959; Hagadom,
1998; Morash, 1983; Thomberry et al., 1993). Examining gang-related behavior
on a continuum can also provide an understanding of the temporal correlates of
gang crime.

For example, in the Thomberry et al. (1993) study, the authors examined
the crime rates of individuals before, during and after their membership in gangs.
Zhang et a1. (1999) found that gang membership had only a modest effect on

subsequent delinquency but that there was a strong positive correlation between

29

prior delinquency and gang membership. These ﬁndings were somewhat
supported by Gordon’s (2000) ﬁndings that individuals joined gangs over a period
of time and not in a spontaneous manner. The continuum approach also provides a
method of assessing organized deviance outside the frame of formal membership.

Despite the heavy reliance on formal membership, some researchers admit
that formal membership is not necessary for associative deviance (Lerman, 1968;
Winfree et al., 1994; Howell, 1994) and that there is sometimes more deviance
outside of the gang (Decker and Kempf-Leonard, 1991 ). There are two studies
(Morash, 1983 and Thomberry et al., 1993) that have similarities to this study.

In the 1983 study of gangs, Morash tested the peers’ delinquency as a
measure of the degree to which peer groups were like gangs. Here, the study uses
the innovation of the formal organization as a referent by which to assess other
associations. The study was based on the ideas of Fredrick Thrasher, who also
conceptualized gangs not as a group hazard but as only one element in the pursuit
of a free life (Morash, 1983). The study used several scales designed to assess
solidarity, activity orientation and gang-likeness. The peer’s gang-likeness was
statistically signiﬁcantly correlated with delinquency but was weak and accounted
for less than two percent of the variation in delinquency (Morash, 1983). The
current study differs from Morash (1983) in two ways: individual level measures
and expanded criminogenic correlates.

The current study examines the gang likeness on an individual scale rather
than the Morash (1983) structural measure of peer group likeness to gangs. Due to

the disaggregate trends seen in contemporary society, the assessment of deﬁant

30

individualism is more accurately measured on an individual level. The second
variation between the Morash (1983) study and the current one is the expanded
dependent variable. Morash (1983) examined delinquency as a correlate of gang-
likeness whereas this study uses both crime rates and attitudinal measures as
correlates to gang similarities.

Thombe et a1. (1993) examined the crime rates of individuals before,
dming and after their membership in gangs. The study also tested two model
assumptions of gang membership: the selection model and the social facilitation
model. The selection model theorized that gangs simply attracted people who
were already deviant and thereby produced higher crime rates. Conversely, the
social facilitation model asserted that the social structure of gangs promoted
increased deviance in people who were not necessarily deviant prior to joining.
Thomberry et a1. (1993) found that the selection model was not supported. The
relationship between these two diametrically opposed models is similar to the
models tested in the current study but there is one primary difference.

Thomberry et al. (1993) focuses on gang membership as a referent
category. Thomberry et a1. (1993) compares deviancy before, during and after
gang membership; this study contrasts deviancy of deﬁant individualism with that
of gang membership. Within the ﬁamework of deﬁant individualism,
criminological intervention becomes more difﬁcult.

“Desistance Under Defiant Individualism”
An additional dimension of organized deﬁant individualism is the problem

of desistence. Despite an increasing body of research into gangs and gang-related

31

deviance, there is a dearth of research into the process of desistance. Typically,
gang desistance is seen as a spontaneous occurrence in which individual gang
members undergo a miraculous transformation and awake as a member of
normative society. Not only are the cases of spontaneous life transformation
isolated and anecdotal, they are not replicable. Within the theoretical framework
of organized deﬁant individualism, desistence becomes more difficult.

Gang desistance is a unique phenomenon in that it actually entails not only
the desistance of behaviors but also the defection from a culture. The speciﬁc act
of an individual disassociating with a gang is typically insufﬁcient to promote the
type of lifestyle change necessary to insure the continued success of the individual
and guard against the possibility of recidivism. With respect to the organized
deﬁant individualism, the individual must defect from the culture that promotes
the existence of the gang. Under deﬁant individualism, gang membership is more
representative of a personal pathology than a socially facilitated one and will
therefore require a more systematic intervention. It is helpful to examine
desistance within the framework of the life course perspective.

Sampson and Laub’s life course perspective is arguably the most
important criminological theory to emerge in the last few decades. The traditional
view of crime posited that there was an inverse relationship between age and
crime. As an individual grew older he or she would commit less crime and
therefore ‘age out’ of crime (Siegel, 2004). The life course perspective, in general,

posits that the relationship between age and crime is not constant across the life

32

course due to several factors; two of which are homotypic continuity and
heterotypic continuity (Sampson and Laub, 1992).

Homotypic continuity refers to deviant behavior committed early in life that
prevents an individual from transitioning from a deviant life trajectory to a
normative life trajectory (Sampson and Laub, 1992). For example, a gang member
who is convicted of multiple felonies in his or her 205 will ﬁnd it much more
difﬁcult to desist from a deviant lifestyle in his or her 403. The trajectory shift or
transition is blocked by several factors. Rival gang members may seek revenge
against the desistor, thereby forcing the person to continue committing acts of
violence as a survival mechanism. Fellow gang members may also contribute to
the blocked transition by seeking favors of an illegal nature. Normative transitions
may also be blocked by members of normative society who refuse to employ the
individual because of his or her criminal history. Under these circumstances, the
age-crime curve is not consistent, in that the relationship then becomes positive;
as the individual increases in age so does the level of deviance.

Heterotypic continuity refers to behaviors that are learned early in the life
course, which are not necessarily mala in se but that lead to deviant or anti social
behaviors later in life (Sampson and Laub, 1992). For example, the gang culture
prizes aggression in gang members as a way of dealing with adversity. Through
overtly aggressive behavior, gang members gain status and rank within the gang.
However, in normative society this type of aggressive posturing is seen as hostile

and may block normative transitions.

33

Heterotypic and homotypic continuity may not be addressed by those with
the deﬁant individualist personality, because they may be less likely to view their
behavior as pathological. After all, to these individuals, they are only succeeding
the best way they can. Therefore, efforts to promote homotypic and heterotypic
discontinuity will probably be less effective on these individuals. Much of the
consensus as to the relationship between gangs and crime is the result of the
underlying assumption that gangs create a group hazard effect.

“Gangs as a Conceptual Group Hazard”

The group hazard effect is actually the combination of two different
concepts: the group hazard hypothesis and the group delinquency hypothesis.
Erickson’s (1973) group hazard hypothesis states that violating the law in groups
is more likely to ensure detection and ofﬁcial reaction than individual crime. The
group hazard hypothesis could be attributed simply to the fact that it is more
difﬁcult for groups to evade detection than for an individual to escape detection
(Erickson, 1973). Erickson’s group hazard is an extension of the commonly
accepted group delinquency hypothesis, which can be seen in earlier work.

Dentler and Erikson (1959) proposed three propositions that sought to
explain the aggregate dynamics of deviance. The ﬁrst proposition was that groups
tended to induce, sustain and permit deviant behavior (Dentler and Erikson,
1959). This ﬁrst proposition addressed the most salient issue of the gangs by
asserting that deviance is a central ﬁrnction of groups. With deviance playing such
a pivotal part in the group’s dynamics, it seems intuitive that the resulting decades

of gang research would rely heavily on the membership as a necessary criterion.

34

The second proposition states that deviant behavior functions in enduring
groups to help maintain equilibrium (Dentler and Erikson, 1959). The equilibrium
discussed refers to the gang’s ability to maintain its activities, such as robbery or
drug sales, at a certain level. The group uses deviance to ensure the organization
strengthens or removes weak members (Ibid). This equilibrium creates the ability
of the group to realize long-term growth and sustainability. In the early years of
the twenty-ﬁrst century, we see generational gang members and gangs that have
existed for decades.

The third proposition stated that groups will resist any trend toward
alienation of a member whose behavior deviates from the group standards
(Dentler and Erickson, 1959). The authors assert that in situations where groups
are faced with a member whose behavior violated the group’s standards, group
members will put pressure on that member in order to force the member to behave
in accordance with the group (Ibid). The rationality of the group is that there is
strength in numbers. Strength is diminished when members are alienated or
unnecessarily removed ﬁ’om the group. It is therefore important to maintain
membership at the highest levels possible. Group hazard and group delinquency
combine to produce the group hazard effect. Despite the earlier applications to
juvenile crime (Erickson, 1973) the group hazard effect forms the founding
assumptions for gang research. This group hazard conceptualization overlooks the
fact that gang-related crime is not relegated to juvenile actors and that even when
aggregate deviance is initiated during adolescence it may continue into adult

hood. The group hazard effect is supported by empirical ﬁndings.

35

As previously, stated there is a great deal of consensus among researchers
that gang members participate in crime and deviance at much higher rates than
non-gang members. The higher rates of crime cross several different domains
including violence, homicide and drug sales. Some researchers (Harper and
Robinson, 1999) have even observed higher rates of more general forms of
deviance such as sexual activity and substance abuse among juvenile gang
members. The relationship between gangs and homicide is, perhaps, the most well
documented correlate in gang research.

In a 1994 study, Hutson et a1. examined drive-by shootings of juveniles
under 18 years of age in the city of Los Angeles. From a sample of 677 incidents
recorded by police, the researchers found that 71% of the juveniles injured in
drive-by shootings in 1991 were gang members (Hutson et al., 1994). In a similar
study, Baily and Unnithan (1994) conducted an analysis of gang homicides in
California in order to determine if gang homicides were distinct from other
homicides. The larger California study found that gang homicide was
distinguishable from other forms of homicide and shared homogenous
characteristics (Baily and Unnithan, 1994). These studies are representative of
later homicide studies that produced similar ﬁndings in other locations including
Minneapolis (Kennedy and Braga, 1998), Boston (Braga et al., 1999) and St.
Louis (Decker and Curry, 2002). Despite the uniformity in ﬁndings, there is at
least one study whose ﬁndings differ ﬁom the consensus.

Brewer et al. (1998) found that gang homicides composed a relatively small

percent of the juvenile homicides in the city of Houston between 1990 and 1994.

36

There is, however, one problem with the data collection methodology which may
have produced this anomalous ﬁnding. The Houston study collected data from
newspaper articles and ofﬁcial Houston Police Department data in order to
triangulate the analysis (Brewer et al., 1998). The newspaper article label of
whether or not a homicide was gang-related was based on the ofﬁcial
investigation.

The Houston Police Department admittedly uses conservative criteria with
which to determine if a crime is gang-related or not (Brewer et al., 1998).
According to the Houston Police Department, a homicide is not considered gang-
related unless it is shown to be committed in ﬁrrtherance of the gang, or gang
motivated (Ibid). This has the effect of reducing the number of gang-related
homicides by narrowing the focus on the motive of the crime not the actors in the
crime. Other cities use more moderate classiﬁcation criteria and language. For
example, the concept of a gang-related crime versus a gang motivated crime. If
either the victim or the suspect is a gang member, that crime is considered gang-
related. This concept does not place a high degree of importance on the motive
for the crime, due to the fact that it is often difficult to discern the actual motives
for crime and individual gang members may commit crime under the color of
gang authority for personal gain. The ﬁndings supporting the gang-crime link are
just as robust when examining other forms of deviance.

Gang members have been linked to the sale of illegal drugs. Maxson (1995)
examined the drug sales of gangs in Pasadena, California and Pomona, California

and found that there was a substantial gang presence (26.7%) in the distribution of

37

cocaine in the two cities. The degree of non-cocaine sales by gang members was
much smaller (11.5%) and the total incidents of gang member drug sales was
much lower than the 90% predictions of local law enforcement ofﬁcers (Maxson,
1995). The ﬁndings were limited, due to many gang members escaping
identiﬁcation by either marginal or transitory gang involvement or simply by not
coming to the attention of ofﬁcers (Ibid). Illegal drugs also act as motivators for
gang members to maintain afﬁliation with the gang.

In a study of drug sales among gang members, Levitt and Venkatesh (2000)
found that despite earnings from drug sales being only slightly higher than the
legitimate labor market, the future prospects of drug earnings drove the desire for
gang membership. Drugs represent a realistic possibility for monetary success.
This establishes a stronger bond to the distribution of illegal drugs as a means to
advancements. Illegal drug sales is not the only advantage members derive ﬁom
the gang.

Hagan (1997) argued that delinquent subcultures tend to temporarily
insulate members ﬁom sources of distress. This insulatory effect helps to create
solidarity but it also creates additional problems for desistance. Hagan (1997)
describes a ‘sleeper effect’ whereby the gang member ﬁnds it difﬁcult to leave the
criminal subculture and pursue a normative life. The prior illegal activities
disqualify the gang member from participation in more traditional jobs. This is
similar to Sampson and Laub’s (1992) concept of homotypic continuity. Hagan

calls it the ‘sleeper effect’ because it does not present itself as a problem until

38

early mid-life. This interplay between internal gang behavior and external effect is
also seen in violent displays.

As far back as 1963, researchers Short and Strodbeck noticed that gang
members responded to internal challenges by engaging in more external crime.
This external acting out serves to reinforce the sleeper effect. This group bond has
manifested itself by many gang members referring to the gang as their family
(Ruble and Turner, 2000; Gordon, 2000). Other researchers (Schreck et a1, 2004)
have found that membership in deviant groups increases the amount of
victimization an individual may experience. The aforementioned ﬁndings do seem
to support the group hazard effect thoroughly. However, there are several
methodological issues that may explain the ﬁndings in support of the group
hazard effect.

The following discussion of methodological speciﬁcities is not meant to
imply that there are methodological errors in the cited studies. It is possible,
however, to detect a cumulative effect which may consistently produce ﬁndings
that support the group hazard effect. The ﬁrst methodological element of the
studies that supports the group hazard effect is the reliance on membership as a
salient criterion.

Many gang researchers (Maxson et al., 1998; Decker and Curry, 2000;
Winﬁee et al., 1994; Ebensen et al., 2001) focus their analyses on formal
membership as a correlate of crime. The problem is that formal membership is
embraced by individuals who have the strongest commitment to the gang culture.

This means that not only are the researchers missing the crime rates of individuals

39

who may simply be less committed to the gang culture but they are also ensuring
that the ‘gang’ data contains the most serious crimes committed with the most
frequency. When comparing these data to individuals who are non-members and
have less commitment to the deviant lifestyles, the non-member data will regress
toward the mean and produce an automatic statistically signiﬁcant variation
between the two groups. The problem of formal membership is compounded
when considering the second methodological speciﬁcity of ofﬁcial data.

Ofﬁcial data are notoriously unreliable. One problem that occurs with
ofﬁcial data is that ofﬁcers may inﬂate gang membership by misidentifying non-
gang members simply because they associate with other known members. As is
seen from the Maxson (1995) study, ofﬁcial data can also cause a deﬂation effect.
However, misidentiﬁcation tends to favor inﬂation of gang membership. Despite
criminal justice practitioners being in direct contact with gang members on a daily
basis, their encounters with gang members, outside of arrest procedures, is often
brief and frustrating. Gang members often adhere to a strict code of silence and
resist ofticers’ attempts to learn anything about the organizations. The result is the
ofﬁcer reverting to the defacto ‘safe’ assumption that an individual is a gang
member. The third methodological speciﬁcity is a culmination of the other two.

Due to the over-reliance on formal membership in gangs, some researchers
(Ebensen et a1. 2001; Bjerregaard, 2002; Thomberry et al., 1993) often measure
gangs as a dichotomous variable. Despite ﬁndings that support the idea that gang
members occupy various levels within gangs (Yablonsky, 1959 and Klein, 1971)

and that joining gangs is often a gradual process (Gordon, 2000), researchers still

40

conceptualize the gang as a dichotomous entity positing that an offender is either
a member or not a member. The dichotomous measure of gangs is not exhaustive
in the face of contemporary gang dynamics.

The group hazard effect and its derivative methodologies require
modiﬁcation when considering the evolving nature of gangs in the twenty-ﬁrst
century. Social context is often overlooked when examining criminological
phenomena. However, gangs and gang-related crime are evolving in new
directions as a result of the sociological changes. In order to develop a new
conceptualization of gangs and gang-related crime, it is necessary to consider the
societal shift affecting contemporary gangs.

“Summary”

The need for the re—conceptualization of gangs and gang-related crime is
based in the changing nature of society. To a great degree, Dentler and Erikson’s
(1959) traditional concept of group deviance has been subverted by the advances
of a global society. Where Dentler and Erikson (1959) posited the group as the
creator and sustainer of deviance, contemporary society reveals that the deﬁant
individualist personality can generate the same reliance on deviance as a method
of goal acquisition. The role of deviance maintenance explained by Dentler and
Erikson (1959) has now been replaced by culture. The culture associated with
deﬁant individualism serves to insulate its members from conversion to normalcy.
Likewise, the traditional risk of detection and apprehension (Erickson, 1973),
once associated with the group, has been alleviated by the reliance on formal

membership as a criterion for concern.

41

These substantive changes in the fundamental nature of gangs require
researchers and practitioners to revise the methods of studying gangs. It is no
longer sufﬁcient to simply focus on membership or other dichotomous
characteristics as the standard by which we are guided. Emerging research models
must examine a more holistic component of organized crime and organized

criminals.

42

Chapter III
“Methodology”

This study is a discriminate analysis of variations in criminogenics of both
gang and non-gang offenders. In designing the study research methods it was
important to draw concrete line of demarcation between the primary groups in
question. Many of the following analyses are composed of multiple tests. This
successive test and reﬁnement method is critical in order to accurately examine
the differences between the test groups. As previously stated, the
conceptualization of a phenomenon inﬂuences the measurement and interpretation
of that phenomenon.

Because the fundamental nature of gangs has changed it is necessary to re-
examine the differences between gang and non-gang actors. The traditional
reliance on membership as a criterion for study dose not serve researchers well
when examining the criminal careers of contemporary gangs. This study seeks to
examine the criminal behavior and attitudes of gang members, subjects who
exhibit deﬁant individualist traits and subjects who appear to have zero-inﬂuence
of either the gang culture or deﬁant individualism.

This study is a quantitative research project which uses secondary analysis
of pre-existing data to test the study’s hypotheses. This study addresses the
problem of the changing nature of gangs and gang-related crime. Gangs in the
modern social context have become less characteristic of the traditional ‘group
hazard’ conceptualization and much more representative of an organization of

deﬁant individualists.

43

Inherent in this shift is the shift in permanence of the group. Contemporary
gangs are more likely to exhibit shorter tenures of existence than traditional gangs
under the group hazard conceptualization. In addition, modern gangs are more
likely to undertake more sophisticated crimes by adopting conventional network
enterprise formats. Therefore, the individuals who comprise the membership of
the contemporary gang are not bound by formal membership and may repeatedly
go through stages of group afﬁliation and non-afﬁliation while simultaneously
carrying out criminal careers that are as violent as full time gang members. Based
on this problem, the following hypotheses were formulated.

Hypothesis one is that there is no difference in criminal justice contact
between gang members and deﬁant individualists. According to the theoretical
framework, contemporary gangs are transient aggregates of deﬁant individualists,
therefore, the criminal justice contact of the two groups should be similar.
Probationers in the deﬁant individualist category have the same motive and
opportunity to engage in crime as gang members and only lack consistent
membership.

The second study hypothesis is that there is a direct relationship between
the number of criminal charges accmnulated and deﬁant individualism. According
to Sanchez-Jankowski (2003), deﬁant individualism is a personality type. This
presents a greater problem for the ﬁeld of criminology because personalities are
often dynamic. It is possible to exhibit stronger personality traits over time. One
of the more problematic elements of the deﬁant individualist personality is the

disintermediation of the law. Deﬁant individualists pursue personal goals through

any means available, legal or illegal. Crime and the gang are only tools. Over
time, the deﬁant individualist may become more criminally oriented as his or her
legitimate opportunities are reduced.

The third hypothesis is that there is no difference in the attitude toward gun
use between deﬁant individualists and respondents who display no organized
criminal inﬂuence. Theoretically, the deﬁant individualist should not have a more
favorable attitude toward gun use than the probationer with no gang inﬂuences,
because deﬁant individualists are not conceptualized as being inherently deviant.
Crime to the deﬁant individualist is simply a tool of personal gain.

The fourth hypothesis is that there is an inverse relationship between deﬁant
individualism and positive trajectory shifts. A deﬁant individualist should remain
unaffected by criminological interventions such as lever-pulling, due to the nature
of the criminal behavior being the deﬁant individualist personality type. Altering
an individual’s personality requires more in-depth and personalized intervention
than is often possible in the criminal justice system.

The ﬁnal study hypothesis is that there is no difference in the perception of
gang criminality between gang members and deﬁant individualists. Due to the
close association of these two groups, the perceptions about gang criminality,
should be similar. The deﬁant individualist’s view of gang criminality
theoretically, should be more calibrated than that of the probationers who have no
gang inﬂuence.

This analysis is cross-sectional in nature and is not designed to study

longitudinal trends in gang and non-gang offending. Rather, this examination

45

seeks to explore the criminogenic variation among gang and non-gang offenders
post intervention.

The data were collected as part of grant # 2003-IJ-CX-1038 from the
National Institute of Justice. The purpose of the grant was to evaluate the
Indianapolis Lever Pulling Intervention. The data set includes both interview data
and respondents’ criminal histories. The study participants were comprised of
every felony probationer in the Indianapolis probation system. The probationers
had to meet several criteria prior to selection for the study. The probationers had
to be actively on probation for a felony offense and that offense had to be
speciﬁcally a drug offense, violent crime weapon offense or a property offense. A
study sample was drawn from consecutive sub samples of 1,000 probationers,
which were supplied each month. There were a total of six different pools of
probationers. Each of the sample pools was randomly assigned to one of three
groups: law enforcement meeting, community meeting or control group.

After the selection and randomization process, the study contained 540
probationers with 180 participants per group. Despite this preliminary study
count, the ﬁnal sample consisted of 235 participants. Ineffective notiﬁcation,
transportation problems and non-compliance with active probation requirements
were all reasons for the attenuation of the sample.

The dataset was comprised of 387 data points. Of the total data points, 195
were analyzed. Composite measures were produced from 66 of the data points.
Finally, 23 study variables, which best ﬁt the hypotheses, were selected. Nineteen

of the study variables were continuous in nature to allow for more sophisticated

46

analysis. Study variables that were rejected had large proportions of missing data.
For example, data on whether the respondent bought a gun under his or her own
name had 40% missing data, how often the respondents ﬁred guns had 99.1%
missing data and how many times the respondents carried a gun outside of the
home had 99.1% missing data.

“Variable Measures”

The dependent variables used in this analysis were divided into two classes:
primary and secondary criminogenics. The use of the criminogenic concept
supports the idea that there are multiple dimensions to criminal behavior. In
attempting to study criminal variation among groups, it is far too simplistic to
examine only criminal instances. Primary criminogenics were conceptualized as
ofﬁcial counts of a participant’s criminal activity. Secondary criminogenics are
composite measures that assess the participant’s attitudes toward various criminal
concepts. This study developed ﬁve measures of the participant’s criminal
activity. Each of these measures was operationalized by using the official counts
from the respondent’s criminal histories. The criminal activity variables are listed
and discussed below.

Number of Arrests. This variable is conceptualized as the actual count of
the instances in which the respondent was taken into police custody. This variable
was operationalized by using data from the respondents’ official criminal histories
found in the dataset.

Number of Violent Convictions. This variable is conceptualized as the

actual count of criminal convictions the respondent incurred due to violent

47

incidents such as assault and robbery. This variable was operationalized by using
data from the respondents’ official criminal histories found in the dataset.

Number of Property Convictions. This variable is conceptualized as the
actual count of criminal convictions the respondent incurred due to property-
related incidents such as vandalism and burglary. This variable was
operationalized by using data from the respondents’ official criminal histories
found in the dataset.

Number of Times on Probation. This variable is conceptualized as the
actual count of the instances in which the respondent was placed on some form of
probation as the result of a criminal charge. This variable was operationalized by
using data from the respondents’ ofﬁcial criminal histories found in the dataset.

Number of Times in the Department of Corrections. This variable is
conceptualized as the actual count of the instances the respondent was held in the
custody of the department of corrections. This variable was operationalized by
using data from the respondents’ ofﬁcial criminal histories found in the dataset.

Time in the Department of Corrections. This variable is conceptualized
as the actual count of the number of days the respondent spent in the custody of
the department of corrections. This variable was operationalized by using data
from the respondents’ ofﬁcial criminal histories found in the dataset.

Number of Charges. This variable was not considered a measure of the
respondent’s contact with the criminal justice system because the number of
charges a respondent accrues is not necessarily a function of ongoing contact with

the criminal justice system. However, the number of convictions and subsequent

48

times on probation could be considered cumulative functions of the number of
times arrested. Conversely, one arrest could result in multiple charges. This
variable was conceptualized as the number of charges accumulated by the
respondent. Operationalization was achieved by using data from the respondents’
ofﬁcial criminal histories. Secondary criminogenics were composite in nature.
Attitude Toward Gun Use. This variable is conceptualized as the degree to
which the respondent has a more or less favorable view of using a gun in conﬂict
situations. The scale originally contained 10 items but was reduced to 6 through
reliability analysis. The ﬁnal scale contained the following items: 1) Sometimes
situations get worse than they have to because someone pulls a gun, 2) I might ask
my ﬁiends to leave their guns at home when hanging out together, 3) If you need
a gun to ﬁt in with your ﬁiends, you’re hanging out with the wrong people, 4) If
you’re planning to go somewhere or do something you’d need a gun for, you’re
better off just staying home, 5) Carrying a gun is not worth the risk of getting in
trouble with the law and 6) It’s alright to have a gun to scare somebody or to
make sure they don’t give you trouble. The items were operationalized as follows:
l= Strongly Agree, 2= Agree, 3= Disagree, 4= Strongly disagree and 5= Don’t
know. Prior to reliability testing, question 6 was recoded in reverse order to
maintain the scale’s continuity. This scale produced the lowest reliability
coefﬁcient of any in the study (alpha= .578). A factor analysis revealed that all of
the items loaded with an Eigen value of at least .443. The range of the scale was
from 6, having a less accepting attitude toward gun use, to 24, having the most

accepting attitude toward gun use.

49

Perception of Gang Criminality. This variable is conceptualized as the
degree to which the respondent believes criminal behavior to be important to gang
members. This variable is a composite measure that contains the following items:
1) How important is murder to gangs, 2) How important is ﬁghting to gangs, 3)
How important is shooting to gangs, 4) How important is drug sales togangs, 5)
How important is drug use to gang and 6) How important is protecting turf to
gangs. The items were operationalized as follows: 1= not at all important, 2:
somewhat important, 3= moderately important, 4= important and 5= very
important. The scale produced a reliability coefﬁcient of .845. A factor analysis
revealed that all of the items loaded with Eigen values of at least .691. This scale
produced a range from 6, having a perception of low criminal importance to
gangs, to 30, having a perception of high criminal importance to gangs.

Positive Trajectory Shift. This variable is conceptualized as the degree to
which the respondent participated in post intervention, pro-social behavior. The
variable is a composite measure that originally contained 10 items. However, after
reliability analysis the scale contained the following items: 1) Since the meeting I
have gotten a job or job training, 2) Since the meeting I have gone back to school,
3) Since the meeting I have entered treatment, 4) Since the meeting I am going to
church, 5) Since the meeting I am attending counseling, 6) Since the meeting I
have contacted community leaders and 7) Since the meeting I have contacted
community organizations. This variable was operationalized as: 1= True and 0=
False. This scale produced a reliability coefﬁcient of .602, which was the second

lowest in the study. A factor analysis of the scale revealed that all of the items

50

loaded with an Eigen value of at least .473. The range of this scale was ﬁ'om 0,
making no positive trajectory shift, to 7, making the most positive trajectory shift.

Post Intervention Networking. This variable was conceptualized as the
degree to which the respondent contacted community-based supporters. The
variable was a composite measure that contained the following 6 items: 1) After
the meeting did you talk with family, 2) After the meeting did you talk with
spouse, girl/ boyfriend, 3) After the meeting did you talk with ﬁiends, 4) After the
meeting did you talk with co-workers, 5) After the meeting did you talk with
neighbors and 6) After the meeting did you talk with probation ofﬁcers. This scale
was operationalized as: 1= True and 0= False. The scale produced a reliability
coefﬁcient of .714. A factor analysis revealed that all the items loaded with an
Eigen value of at least .499. This scale had a range of 0, participating in no post-
intervention networking, to 6, participating in a high degree of post-intervention
networking.

Intervention Recall. This variable was conceptualized as the degree to I
which the respondent recalled elements of the intervention meeting. This is a
composite measure that contained the following items: 1) Remember that law
enforcement are cracking down on violent crime, 2) Remember that law
enforcement is cracking down on gun crime, 3) Remember that I can go to federal
prison for carrying a gun, 4) Remember that probation is watching behavior
closely, 5) Remember that law enforcement wants you to make good choices, 6)
Remember that community leaders have opportunities for you to get a job, 7)

Remember that community leaders are willing to help you in any way they can

51

and 8) Remember that I should stay out of trouble. These items were
operationalized as: l= True and 0= False. This scale produced a reliability
coefﬁcient of .906. A factor analysis revealed that all of the items loaded with an
Eigen value of at least .714. This scale produced a range from 0, having no recall
of the meeting, to 8, having total recall.

This study uses two independent variables as part of the discriminate
analysis. The predominant purpose of these variables is to provide a ﬁamework in
which to compare outcome variables.

Defiant Individualism Score (DIS). This variable is a composite measure
that originally contained 13 items but was reduced to 11 through reliability
analysis. The scale is composed of the following items: 1) Have you ever been a
member of a gang, 2) Have you ever been a member of a group, 3) Have you ever
thought of joining a gang, 4) Have you ever been recruited or pressured to join a
gang, 5) Have you ever hung out with gang members, 6) Have you ever drunk
alcohol or gotten high with gang members, 7) Have you ever vandalized
something with a gang member, 8) Have you ever stolen something with a gang
member, 9) Have you ever been attacked in a gang-related incident, 10) Have you
ever attacked someone in a gang-related incident and l 1) Do you have ﬁiends that
are gang members. The items on the scale were operationalized as: yes= 1 and
no =0. This scale produced a reliability coefﬁcient of .845. A factor analysis
revealed that all of the items loaded with an Ei gen value of at least .401. The
range of the scale was from 0, having no commitment to deﬁant individualism, to

11, having high commitment to deﬁant individualism.

52

Gang Membership. This variable is conceptualized as whether or not the
respondent was a member of a gang at the time of the interview. This variable was
operationalized as: 1= yes and 0= no.

Category. This variable is used primarily as a grouping variable through
which to compare ﬁndings. This variable is composed of three categories.
Category-1 (zero-inﬂuence group) consists of people who scored 0 on the DIS
index. Category-2 (deﬁant individualists) is composed of people who scored from
1-11 on the DIS index. Category-3 (gang members) is composed of people who
had missing data for the DIS index. In the original survey, a contingency question
asked respondents to verify whether or not they were current gang members. If
the respondent replied yes, he or she was instructed to skip the DIS index items.
Therefore, the respondents who have missing data for the DIS index comprise
category-3, the gang members group. These three groups allow for a comparison
of ﬁndings between gang members, non-members who display DIS behavior and
non-members who have no commitment to deﬁant individualism.

This study also examined the demographic characteristics of the sample.
The following variables were analyzed in order to summarize the sample
characteristics. The number of hours worked per week, Total income by legal
means, Total income by illegal means, Number of children, Highest grade
completed and the Respondent’s age were all operationalized by using count data
provided.

Gender was operationalized as: 0= female and ma1e= 1. Percent of the time

that the respondent was employed was operationalized as: 1: 100% of the time,

53

2= about 75% of the time, 3= about 50% of the time, 4= about 25% of the time
and 5= not employed. Respondent’s race was operationalized as: 1= White, 2=
Black, 3= Hispanic, 4= Asian, 5= Native American and 6= Other. Respondent’s
marital status was operationalized as: I: married, 2=living with partner, 3=
Widowed, 4= separated and 5= divorced.
“Statistical Tests”
The following tests will be used to test the study hypotheses.
H1: There is no dijference in criminal justice contact between gang members
and deﬁant individualists which would denote a discriminant ﬁmction.

This hypothesis will be tested across ﬁve variables: number of arrests (formal
criminal justice contact), number of convictions for violent offences (propensity),
number of convictions for property offenses (propensity), number of times on
probation (recidivism) and the length of days in the Department of Corrections
(severity). The measurement model will consist of a discriminant function
analysis between gang members and deﬁant individualists for each variable. The
model ﬁndings will determine whether or not the hypothesis is supported.

H 2: There is a direct relationship between the number of criminal charges

accumulated and the level of deﬁant individualism.

This hypothesis will be tested by computing a Pearson’s correlation of the two
variables. For the pru'pose of comparison, an additional point bi-serial correlation
between gang membership and number of criminal charges accumulated will also

be computed. The amount of attenuation due to dichotomization in the point bi-

54

serial correlation will be diagnosed and corrected using an extremeness of split
formula (Hunter and Schmidt, 2004: 36).
H 3: There is no diﬂerence in the attitude toward gun use between deﬁant
individualists and members of the zero-inﬂuenced group.
Hypothesis three will be tested using a discriminant function analysis of deﬁant
individualists and zero-inﬂuence groups for each dimension. The model ﬁndings
will determine whether or not the hypothesis is supported.
H4: There is an inverse relationship between deﬁant individualism and positive
trajectory shifts.
This hypothesis will be tested by computing a Pearson’s bi-variate correlation.
For the purpose of comparison, an additional point bi-serial correlation between
gang membership and positive trajectory shifts will also be computed as needed.
The amount of attenuation due to dichotomization in the point bi-serial correlation
will be diagnosed and corrected using an extremeness of split formula.
H 5: There is no diﬁerence in the perception of gang criminality between gang
members and deﬁant individualists.
Hypothesis ﬁve will be tested using a discriminant function analysis between
gang members and deﬁant individualists. The model ﬁndings will determine
whether or not the hypothesis is supported.
“Summary”
This study is signiﬁcant to criminology in three ways. This study creates a
new framework for measuring emerging organized crime patterns. By identifying

and differentiating organized crime that does not have a stable group base,

55

criminologists can assess the proportionality of new organized crime patterns. If
there are signiﬁcant criminological variations among respondents who exhibit the
deﬁant individualist personality, it may be possible to focus on deﬁant
individualism as an early predictor of an organized criminal lifestyle. This is a
vital ﬁrst step in crafting enforcement strategies that can effectively suppress
networked criminal enterprises.

A second contribution of this study is that it provides direction in research.
If there is signiﬁcant correlation between gang-related crime and the deﬁant
individualism personality type, as Sanchez-Jankowski (2003) asserts, research
should focus quantitatively on the early identiﬁcation of personality traits. This
study uses a proxy measure of deﬁant individualism, however, if the study ﬁnds a
signiﬁcant correlation between gang-related crime and deﬁant individualism,
additional research should develop speciﬁc indices that can assess deﬁant
individualistic traits at the earlier stages of personality development.

The ﬁnal contribution of this study is that it may help to formulate
intervention strategies. For example, if there is a signiﬁcant correlation between
the deﬁant individualist personality and age, researchers can develop age- graded
guidelines for intervention programs. As mentioned above, the early intervention
in the formation of the personality type may prove to be a more effective

intervention strategy than simple enforcement models.

56

Chapter IV

“Introduction to Analysis”

This study explores distinctions between individuals who display deﬁant
individualist personality traits and gang members with respect to criminogenic
attitude and behavior. The study addresses the problem that the ﬁrndamental
nature of contemporary gangs and gang-related crime has changed in ways that no
longer require reliance on formal membership to support ongoing criminal
lifestyles. The theoretical foundation for this study is Sanchez-Jankowski’s (2003)
concept of organized deﬁant individualism.

Sanchez-Jankowski (2003) posits that gangs are essentially an aggregation
of individuals who exhibit a personality type called deﬁant individualism. Thus,
the gang is an association of organized deﬁant individualists (Sanchez—Jankowski,
2003). Deﬁant individualists pursue socially valued goals through any means,
legal or illegal, and resist any attempt to stop their goal attainment. Due to a lack
of means with which to obtain these goals, deﬁant individualists often tum- to
crime. Sanchez-Jankowski (2003) goes on to assert that almost all gang members
have deﬁant individualist personalities. This creates gang members as a referent
category against which to compare other groups. However, not all persons who
possess deﬁant individualist traits may belong to gangs. Thus, actual gang
membership is not a prerequisite for this personality and it may well be the case
that persons with this personality are engaging in systematic, organized criminal
behavior without the designated criminal or gang label. From this theoretical

foundation, the study proceeds with ﬁve hypotheses.

57

The ﬁrst hypothesis articulates that there is no difference in criminal justice
contact between gang members and deﬁant individualists that would denote a
discriminant function. If Sanchez-Jankowski’s assertion, that almost all gang
members have deﬁant individualist personalities, is accurate then there should be
similar behavioral expressions found in both groups. Both groups should be
seeking mainstream goals but neither group should have adequate means to obtain
their goals. Therefore, there should be no discriminant criminal justice contact
between gang members and non-gang members who exhibit deﬁant individualist
traits.

The second hypothesis suggests that there is a direct relationship between
the number of criminal charges accumulated and the level of deﬁant
individualism The longer an individual maintains a criminal lifestyle; the
individual should logically accumulate more criminal charges.

The third hypothesis asserts that there is no difference in the attitude toward
gun use between deﬁant individualists and subjects in the zero-inﬂuence group.
Sanchez—Jankowski’s (2003) description of individuals with deﬁant individualist
personalities leads one to believe that they may embrace violence simply as a tool
through which to obtain or maintain goals. This suggests the possibility of deﬁant
individualists having a very similar view of gun use to that of individuals who
have no gang inﬂuence. It is possible the guns are viewed as being less important
than one might expect.

The fourth hypothesis asserts an inverse relationship between deﬁant

individualism and positive trajectory shifts. As an intervention strategy lever-

58

pulling is designed to conﬁ'ont habitual offenders about their criminal behavior,
which simultaneously presents them with opportunities to transition into more
normative behavioral patterns. Sanchez-Jankowski (2003) does not provide an
explanation by which the deﬁant individualist transitions out of the criminal
lifestyle and acquires prosocial values. Sanchez-Jankowski suggests that the
deﬁant individualist simply maintains the deviant lifestyle indeﬁnitely.
Presumably, this omission can be ﬁlled when considering Sampson and Laub’s
(1992) concept of homotypic continuity and Hagan’s (1997) sleeper effect.

Homotypic continuity is the process by which individuals ﬁnd it difﬁcult
to stop participating in deviant behavior due to prior deviant behavior. Hagan
(1997) describes the process as the ‘sleeper effect’ whereby the gang member
ﬁnds it difﬁcult to leave the criminal subculture and pursue a normative life. Prior
illegal activities have disqualiﬁed the gang member ﬁom participation in more
traditional jobs but the gang member does not recognize this as a problem until
early mid-life. Based on these interrelated concepts, the deﬁant individualists and
gang members should have similar difﬁculties transitioning into positive social
roles.

Hypothesis ﬁve holds that there is no difference in the perception of gang
criminality between gang members and deﬁant individualists. If Sanchez-
J ankowski’s (2003) assertion that almost all gang members have deﬁant
individualist personalities is accurate, then there should be no discriminant

difference between the perception of the importance of criminality between gang

59

members and deﬁant individualists due to the similar ideological foundation of
each group.

The data used to test these hypotheses is ﬁom an evaluation of the lever-
pulling program implemented in Indianapolis, Indiana. Lever-pulling is a
criminological intervention based on the ‘Cease Fire’ component of the Boston
Gun Project conducted in the 1990s (McGarrell, Chermak, Wilson and Corsaro,
2006). Lever-pulling is a focused deterrence strategy that is based on multiple
characteristics of and responses to offending (Ibid, 2006). A multi-agency work
group of criminal justice professionals identify and target habitual offenders.
These habitual offenders are required to attend notiﬁcation meetings where they
are advised that they will face signiﬁcant criminal justice sanctions if the
offenders do not stop engaging in certain criminal behaviors such as gun violence.
During the meetings, offenders are provided with networking opportunities

(McGarrell et al., 2006).
' The underlying purpose of the Indianapolis lever-pulling project was to
reduce homicides. The initial lever-pulling meetings began in 1998 and continued
through the summer of 1999 (McGarrell et al., 2006). Indianapolis conducted nine
lever-pulling meetings and eight follow up meetings. The treatment groups
consisted of 160 probationers and parolees (McGarrell et al., 2006). The total
sample was 23 5, which included the control group.

Despite a small sample size, the data were well suited to this type of
analysis. The dataset contains information about individuals who were on active

probation in Indianapolis. The dataset contains subjects from each of the

60

classiﬁcations (gang members, deﬁant individualists and zero-inﬂuenced
subjects) of offenders the study seeks to examine. The sample is constructed from
multiple selections of probationers participating in the lever-pulling program. The
sample was matched into three groups: a law enforcement treatment group, a
community treatment group and a control group.

The variables selected for the study are divided into two groups: primary
criminogenics and secondary criminogenics. Primary criminogenics is
conceptualized as variables that directly involve criminal behavior, such as the
number of violent arrests or the number of times on probation. Secondary
criminogenics are conceptualized as variables that may have an indirect
relationship to crime but are not illegal. Variables such as the individual’s attitude
toward gun use and intervention recall are secondary criminogenics. It is
important to examine both types of variables in order to determine if it was
possible to differentiate between gang members and non-gang members with
deﬁant individualist personalities.

If primary and secondary criminogenics discriminate between gang
members and deﬁant individualists, this suggests that perhaps primary
criminogenics would be adequate in differentiating between the two groups.
Additionally, this ﬁnding would suggest that formal membership imparts an extra
bonding factor that may not be readily identiﬁable but which allows the
researcher to differentiate between gang members and non-members, even if they

share the same personality traits. However, a non-signiﬁcant ﬁnding would

61

suggest that formal membership is less important in understanding sustained
deviant lifestyles than once supposed.

This study examines demographic variables as well as continuous and
composite measures pursuant to deviant behaviors. The demographic variables are
used to contextualize the sample. These include sex, race, age and marital status.
There are four employment-related variables in the study. Number of hours
worked per week, percent of the time that the respondent was employed, total
income by legal means and total income by illegal means are used to better
understand what proportion of the respondents were using crime as their primary
source of income. Similarly, the number of children and marital status are
examined in order to determine if the respondents had any pro-social bonds.
Highest grade completed is used to examine the education level of the sample
respondents.

This study examines ﬁve primary criminogenic variables. The number of
arrests is used to summarize the amount of law enforcement contact. However,
since being taken into custody does not automatically equate to criminal charges
being brought by an ofﬁcer due to probable cause constraints, the number of
charges is also examined. Criminal convictions represent an additional element to
criminal justice contact. Both the number of violent and property convictions are
examined as study variables.

It is also important to consider the outcomes of the criminal justice contacts

experienced by the study respondents. Number of times on probation and the

62

length of time in the department of corrections are both examined for this
purpose.

The study examines ﬁve secondary criminogenic variables also.

The attitude toward gun use is a composite measure that has a range from 6,
having a less accepting attitude toward gun use, to 24, having the most accepting
attitude toward gun use. Attitude toward gun use is a study variable that assesses
the anti-social ideology that may inﬂuence or correlate to the respondents’
primary criminogenic activity. The perception of gang criminality is a study
variable that tests the parsimony of criminal ideology between the various study '
groups. This variable is also a composite measure with a range from 6, having a
perception of low importance of crime to gangs, to 30, having a perception of
high importance of crime to gangs.

This study also uses three ﬁnal variables designed to assess the willingness
to desist criminal activity: positive traj ectory shift, post intervention networking
and post intervention recall. Positive trajectory shift is conceptualized as the
degree to which the respondent engages in non-deviant behaviors after
participating in the lever-pulling intervention. Positive trajectory shift has a range
ﬁom zero, making no positive trajectory shift, to 7, making the most positive
trajectory shift.

Post-intervention networking is conceptualized as the degree to which a
respondent engages in normative association after participating in the lever-
pulling intervention. Post-intervention networking has a range of 0, participating

in no post-intervention networking, to 6, participating in a high degree of post-

63

intervention networking. Post-intervention recall is conceptualized as the degree
or recall the respondent maintains after participating in the lever-pulling
intervention. Similarly, intervention recall has a range ﬁ'om 0, having no recall of
the meeting, to 8, having total recall.

“Univariate Analysis”

The univariate analysis consists of examining both the demographic
variables and study variables, in order to determine the level of normality in the
distribution. The univariate analysis also allows any speciﬁc trends that might
exist in the data to become readily apparent. Generalities and characteristics of the
entire sample are presented ﬁrst, followed by the traits of: the zero-inﬂuence
group, the deﬁant individualist group and the gang member group. The zero-
inﬂuence, deﬁant individualist and gang member groups are classiﬁcation groups.

The members of the zero-inﬂuence group are respondents who stated that
they were not gang members and who did not display any degree of deﬁant
individualist personality traits. The second of the classiﬁcation groups is the
deﬁant individualist group. The deﬁant individualists are respondents who are not
members of gangs but who display some deﬁant individualist traits similar to
gang members. The gang member group is comprised of respondents who
reported being in a gang at the time of the survey.

“Demographics”

The study sample contains a total of 235 subjects. There are 71 subjects in

the zero-inﬂuence group, 142 in the deﬁant individualist group and 21 subjects in

the gang member group. Overall, the sample is predominantly male (88.1 %,

64

n=207). This disproportionate male representation is displayed across the various
groups. The gang member group consists completely of male participants (100%,
n= 21). The largest proportion of female respondents is found in the zero-
inﬂuence group, which contained 18.1% (n=13). The overrepresentation of males
differs ﬁom the population parameter of Indianapolis. A full 51.4% (n= 393,114)
of the population is female (U .3. Census Bureau, 2005).

For a study of this type it is important to understand the demographics of
the individuals who make up the sample. A primary concern for American
criminologists has always been the racial or ethnic make up of offenders. An
analysis of the racial characteristics shows that the sample is predominantly Black
(69.8%, n=164). Whites comprise the second largest racial group (24.3%, n= 57)
with Hispanic, Native American and other groups representing negligible
proportions of the remaining sample. These patterns are visible in the all of the
individual groups.

Blacks comprise the largest proportion of the zero-inﬂuence group (61.1%,
n=44) as well as the deﬁant individualist group (74.5%, n= 105) and the gang
member group (71.4%, n= 15). White respondents retain their position as the
second most heavily represented group in the zero-inﬂuence classiﬁcation
(33.6%, n=24). Whites are also the second most represented group among the
deﬁant individualist group (20.4%, n=29) and in the gang member group (19%,
IF 4). The zero-inﬂuence and the deﬁant individualist groups contain no Native

American participants. These ﬁndings demonstrate that the two primary racial

65

groups in the sample are Black and White Americans. The disproportionate
representation of Black Americans is consistent with many criminological studies.

From examining the racial representation of the population of Indianapolis,
we see that the study sample is disproportionately African-American. Aﬁican-
American comprise 69.8% of the sample but only 25.5% (n= 195,044) of the
city’s population (U .S. Census Bureau, 2005). Conversely, Whites are only 24.3%
of the sample but are 66.3% (n= 507,520) of the population in Indianapolis (U .S.
Census Bureau, 2005). In Indianapolis, other racial groups comprise only 5.7%
(N= 44,568) of the population (U.S. Census Bureau, 2005).

This study also examines marital status as a way to assess the respondents’
bond to normative society. Marriage is often seen as an indicator of normative
stability. Individuals who maintain families are less likely to engage in systematic
deviance. By in large, the sample consists of people who have never been married
(45.2%, n=103). The second largest marital arrangement (20.6%, n= 47) is held
by those participants who were living with partners at the time of the survey. Only
13.6% (n=31) of the sample are married. These ﬁndings suggest that the majority
of the sample is living under a lesser commitment to a partner. ‘Live in’
relationships do not carry the same sociological responsibility as a formal
marriage; the most salient difference being the lack of legal recognition of the
‘live in’ relationship.

The zero-inﬂuence group also consists of individuals who largely have

never been married (47.1%, n= 33). There are, however, an equal number of

66

married (15.7%, n=1 1) and divorced respondents (15.7%, n=11). There is a

similar proportion of gang members who have never married (47.6%, n=10).

 

Table-1 Demographic Analysis by Category

 

 

 

 

 

 

 

 

 

 

Sex Male 207 88.1 59 81.9 127 89.4 21 100
Female 28 11.9 13 18.1 15 10.6 - -
Race White 57 24.3 24 33.3 29 20.4 4 19
Black 164 69.8 44 61.1 105 74.5 15 71.4
Hispanic 7 3.0 1 1.4 5 3.5 1 4.8
Native American 1 .4 - - - - 1 4 8
Other 5 2.1 3 4.2 2 1.4 - -
Marital Married 31 13.6 11 15.7 16 11.7 4 19.0
Liv. w/ Partner 47 20.6 12 17.1 29 21.2 6 28.6
Widowed 3 1.3 - - 3 2.2 - -
Separated 14 6.1 3 4.3 11 8.0 - -
Divorced 30 13.2 11 15.7 18 13.1 1 4.8
Never Married 103 45.2 33 47.1 60 43.8 10 47.6
Percent of Time 100% 116 49.4 39 54.2 68 47.9 9 42.9
Employed 75% 29 12.3 8 11.1 19 13.4 2 9.5
50% 31 13.2 10 13.9 19 13.4 2 9.5
25% 23 9.8 5 6.9 16 11.3 2 9.5
0% 36 15.3 10 13.9 20 14.1 6 28.6

 

 

Note: Table does not reﬂect missing values.
ZIG= Zero-inﬂuenced group
DIG= Deﬁant individualist group
GMG= Gang member group

In addition to the aforementioned demographics, there are other
characteristics of the sample that help to contextualize this study. The effects of
age can often create a variable effect in criminological studies. It is important to
determine if there are any age-related anomalies in the data that create an
intervening effect in the multivariate analyses. The mean age for study

participants is 31 years of age with a median age of 30 years (sd= 8.731). The

67

youngest person in the sample is 17 years old (.4%, n= 1) and the oldest person in
the sample is 58 years old (.9%, n=2).

The mean ages of the various group grow increasingly younger as one
moves ﬁom the zero inﬂuenced group (mean= 33.7) to the deﬁant individualist
group (mean= 30.9) and ﬁnally the gang member group (mean= 29.5). This
younger gang member contingent appears to support the concept that as
individuals age they ‘age out’ of certain types of deviance. The traditional
conceptualization of gangs is one of associations of criminal youth. The overall
sample, however, is not dissimilar to the population of Indianapolis with respect
to age. The median age of residents in Indianapolis was 34.8 years of age (U .8.
Census Bureau, 2005).

Education is also an important variable for contextualizing a sample within
normative social structures. Individuals with high levels of education typically
commit fewer criminal offenses and have access to more employment ﬂexibility
than those without higher levels of education. Most members of the sample do not
have college level education. Findings show that 84.2% (n= 102) of the study
participants have 12 years of education or less. Of these 84.2% with less than 12
years of education, 40.6% (n= 51) have only 11 years of education. This suggests
that approximately 40 % of the sample did not graduate from high school.
Education levels for all three groups are similar to the sample statistics and the
means of the other groups. The gang member group (mean= 11.3) has a slightly

lower mean level of education than the zero inﬂuenced group (mean= l 1.6).

68

Another measure of stability is the number of children. It is presumed, an
individual with children is participating in the normal functions of adulthood,
which suggests that the person is also not participating in deviant behaviors.
However, due to speciﬁc social patterns found in some segments of society, this
may not be a strong predictive measure. The presence of children does not
guarantee that the respondents are the primary care givers of their children.

The number of children each study participant has is also relatively small.
The mean number of children that each participant has is 1.8 (sd= 1.735). A full
27.4% (n=64) of the sample reports having no children. Gang members have a
slightly higher number of children (mean= 2.4) than either the zero-inﬂuence
group (mean= 2.0) or the deﬁant individualist group (mean= 1.7). 3

“Education and Income”

An individual’s employment is an important measure of lifestyle choice.
Statistics show that a little less than half (49.4%, n=116) of the sample have been
employed 100% of the time in the six months prior to the survey. Only 15.3%
(n=36) of the sample have not been employed at all in the six months prior to the
survey. Most of the participants from each group have also been employed 100%
of the time in the six months prior to the survey.

The zero-inﬂuence group has 54.2% (n= 39) of its members who had been
employed 100% of the time in the six months prior to the survey. The deﬁant
individualist group has a smaller proportion (47.9%, n= 68) of its members who

have been employed 100% of the time in the six months prior to the survey. The

69

gang member group has the smallest proportion of their respondents working
100% of the time in the six months prior to the survey (42.9%, n= 9)
(See Table-1).

The employed participants report wide variations in the number of hours
worked per week. At the time of the survey, only 1.2% (n= 2) of the respondents
report not working. Participants who work full time, 40 hours, numbered 32.9%
(n=54). Interestingly, 38.4% (n= 63) of the sample works more that 40 hours per
week. The number of hours worked suggested that the study participants, by in
large, are integrated into the community. The zero-inﬂuence group (mean= 43.1)
has a higher mean number of hours worked per week as compared to the deﬁant
individualist group (mean= 39.5). The gang member group has the lowest number
of mean hours worked per week (mean= 35.8).

An assessment of income source provides a way of determining whether or
not members of the sample have a substantial reliance on crime as a source of
income. Both legal and illegal monthly incomes are examined in order to make a
proportional comparison. With respect to income generated by legal endeavors,
14.7% (n=33) of the respondents report having no income despite the mean
income for the sample being $1,281 monthly (sd= 1,409). A ﬁrll 46.9% (n=119)
of the sample reports a legal income of less than $1,000 per month. The
distribution of income generated ﬁom illegal endeavors is negatively skewed due
to 92.3% (n= 217) of the respondents receiving no money ﬁ'om illegal means.

Over 95% of the sample (95.3%, n=223) earns less than $1,000 monthly from

70

illegal endeavors. These statistics suggested that the sample is not comprised of

career criminals who rely on crime as their primary source of income

 

 

 

 

 

 

 

 

 

(See Table-2).
Table-2 Descriptive Means Analysis
' Mean' Median SD‘ ‘ {N
is. ’ ' ' ' 31.6 "30.0 ' 2:751“ 232'”
Highest Grade Completed 11.5 12.0 1.667 234
Number of Children 1.8 2.0 1.735 234
Total Income- Legal 1,281 1,000 1,409 224
Total Income- Illegal 425 0 5,231 235
Number of Hours Worked 40.3 40.0 12.888 71
Weekly

 

 

 

 

 

 

 

Note: Table does not reflect missing values.

When examining the amount of income from legal sources, it is clear that
members of the zero-inﬂuence group earn more (mean= $1,491) than either the
deﬁant individualist group (mean= $1,249) or the gang member group (mean=
$861). The deﬁant individualist group, however, have the largest mean amount
earned ﬁ'om illegal endeavors (mean= $623). The zero inﬂuence group derive the
lowest mean amount of money from illegal endeavors (mean= $347). Due to the
ﬁndings by multiple researchers of higher crime rates among gang members, one
might have expected the gang member group to display larger incomes from

illegal endeavors but this is not the case (See Table-3).

71

 

Table-3

Descriptive Means Analysis by Category

 

 

 

 

 

 

 

 

 

 

 

 

 

., mm. ._. .
Age ZIG 33.7 32.0 9.547 71
DIG 30.9 29.0 8.565 142
GMG 29.5 29.0 5.937 21
Higher Grade ZIG 11.6 12.0 1.534 72
Completed DIG l 1.5 12.0 1.760 142
GMG 11.3 11.0 1.496 20
Number of Children ZIG 2.0 2.0 1.816 71
DIG 1.7 1.0 1.612 142
GMG 2.4 2.0 2.135 21
Total Income- Legal ZIG 1,491 1,000 1,564 64
DIG 1,249 1,000 1,362 139
GMG 861 350 1,137 21
Total Income-Illegal ZIG 347 0 294 72
DIG 623 0 6,714 142
GMG 428 0 1,121 21
Number of Hours ZIG 43.1 42.0 11.240 51
Worked Weekly DIG 39.5 40.0 12.950 101
GMG 35.8 40.0 17.198 12
Note: Table does not reﬂect missing values.
ZIG= Zero-inﬂuenced group
DIG= Deﬁant individualist group
GMG= Gang member group

“Criminal Justice System Contact”

Table-4 shows that throughout the examination of self-reported criminal
involvement, the gang member group is consistently higher than other groups. For
example, the mean number of convictions for violent crimes for the overall
sample is .54; however, the mean number of convictions for violent offenses
among gang members is .90. This is higher than either the zero-inﬂuence group
(mean= .36) or the deﬁant individualist group (mean= .57). A similar pattern is

observed in the number of convictions for property offenses, the number of times

72

 

on probation and the number of experiences further in the department of
corrections (See Table-4). This ﬁnding seems to support other studies that draw a
distinction between gang and non-gang offenders.

In analyzing the number of arrests, the mean numbers of arrests of the zero-
influence group (mean= 8.5) and the deﬁant individualist group (mean= 8.2) are
similar to the overall mean number of arrests (mean= 8.6). However, the gang
member group displays a larger number (mean= 11.3) of arrests. A similar pattern
is observed upon comparing offense charges. Again, the gang member group
displays a larger number of charges (mean= 20.1) than the zero-inﬂuence

(mean= 13.9) or individualist groups (mean= 13.8).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table—4 Primary Criminogenic Analysis by Category
Sample ZIG DIG GMG
Number of Arrests Mean 8.6 8.5 8.2 11.3
SD 7.127 6.219 6.778 11.155
Number of Charges Mean 14.4 13.9 13.8 20.1
SD 10.864 9.387 9.995 18.086
Number of Convictions- Mean .54 .36 .57 .90
Violent SD .948 .737 1.034 .889
Number of Convictions- Mean .77 .67 .78 1.0
Property SD 1.334 1.199 1.348 1.673
Times on Probation Mean 2.1 2.0 2.1 2.5
SD 1.420 1.346 1.409 1.720
Times in DOC Mean 1.1 .89 1.2 1.5
SD 1.095 .897 1.175 .978
Note: Table does not reflect missing values.
ZIG= Zero-inﬂuenced group
DIG= Defiant individualist group
GMG= Gang member group

 

The secondary criminogenics reveal similar variations in the responses of

the different group members. When analyzing the respondents’ attitude toward

73

gun use, Table-5 shows that generally the respondents have a less accepting
attitude toward gun use (mean= 8.5). The gun use scale ranges from 6, having the
least accepting attitude toward gun use, to 24, having the most accepting attitude
toward gun use. Both the zero-inﬂuence (mean= 8.4) and the deﬁant individualist
(mean= 8.4) groups have slightly lower mean scores on the attitude toward gun
use scale. However, the gang member group has a higher than average score
(mean= 9.0) (See Table-5).

Interestingly, gang members have a lower perception of the importance of
criminality to gang members than the other groups in the study (mean= 22.9).
This scale ranges from 6, having a perception of low criminal importance to
gangs, to 30 having a perception of high criminal importance to gangs. The
overall mean perception is 23.6. Zero-inﬂuence group members also report a
higher mean score (mean= 24.6) than deﬁant individualist members (mean=
23.2). Despite the slight variations in this variable, the ratings are toward the
higher end of the scale. The sample is fairly uniﬁed in its belief that crime is
important to gang members.

The amount of positive trajectory shift found in the sample is generally
small (mean= 1.7). The amount of positive trajectory shift is measured on a scale
of 0, being no positive trajectory shift, to 7, being the most positive trajectory
shift. Gang members show the smallest amount of positivetrajectory shift after
participating in the intervention program (mean= 1.3). Interestingly, the deﬁant
individualist group displays the largest amount of positive trajectory shift

(mean= 1.8).

74

The amount of post intervention networking is measured on a scale which
ranges from 0, engaging in no contact with community-based supporters, to 6,
engaging in the most contact with community-based supporters possible. The total
sample mean networking score is 2.0 which suggests that the sample overall did
not engage in a great deal of post intervention networking. The deﬁant
individualist and gang member groups display the same level of networking as the
total sample. The zero-inﬂuence group, however, is slightly less engaged in
networking than the other groups (mean= 1.8).

Post intervention recall of the sample is fairly moderate. The scale used to
measure the amount of information recalled from the intervention ranges from 0,
having no recall of the intervention topics, to 8, having total recall. The overall
sample has a mean recall score of 5.2. The zero-inﬂuence group displays the
highest recall score (mean= 5.5) while the deﬁant individualist group (mean= 5.0)
and the gang member group (mean= 5.3) have slightly lower scores (See Table-5).

The descriptive statistics of the demographic and study variables show that
the three groups are very similar to one another across the response categories
such as age, highest grade completed, number of children and the number of
hours worked weekly. This high degree of similarity is due to the purpose for
which the sample was originally collected. This sample was constructed in order
to test the Lever-Pulling intervention. The lever-pulling intervention utilized two
treatment groups and one control group. The study participants were randomly
placed into the law enforcement treatment group, the community treatment group

or the control group. For the purpose of insuring the most statistically valid

75

evaluation possible, the demographic characteristics of the individuals in the

 

 

 

 

 

 

 

 

 

 

 

 

 

groups were matched.

Table-5 Secondary Criminogenic Analysis by Category
Attitude Toward Mean 8.5 8.4 8.4 9.0
Gun Use SD 1.404 2.500 2.323 2.780
Perception of Gang Mean 23.6 24.6 23.2 22.9
Crime SD 5.803 5.444 5.878 6.301
Positive Trajectory Mean 1.7 1.5 1.8 1.3
Shift SD 1.586 1.254 1.771 1.154
Post Intervention Mean 2.0 1.8 2.0 2.0
Networking SD 1.734 1.570 1.826 1.891
Post Intervention Mean 5.2 5.5 5.0 5.3
Recall SD 2.940 2.940 2.972 2.86

Note: Table does not reﬂect missing values.
ZIG= Zero-influenced group
DIG= Deﬁant individualist group
GMG= Gang member group

“Correlation Analysis”

The next group of statistics is the bivariate correlations. Correlation Tables
6 through 8 establish the general relationship among the study variables. The
primary correlation matrices create a baseline of criminogenic behaviors.
Successive correlation matrices are computed in order to determine the stability of
the correlations across the sample groups and under varying conditions and assist
in furthering the goal of determining if the primary or secondary criminogenic
variables are suitable as tools with which to differentiate between gang members

and non-gang members with deﬁant individualists’ personality traits.

76

Table-6 shows the relationships among the primary criminogenic variables
in the study. Most of the correlations are intuitive in nature. For example, there is
a strong, direct correlation between the number of arrests and the number of
charges (r= .934, p=.000). This suggests that as the number of arrests increase so
does the number of charges that he or she incurs. The number of arrests also is
positively correlated with the number of property convictions (r= .519, p=.000)
and the number of times on probation (r= .692, p=000). The number of arrests is
positively correlated with the number of violent convictions (F. 204, p=.002) and
the number of times in the department of corrections (F .359, p=000). These
correlations, while statistically signiﬁcant, are moderately strong and suggest that
as the number of arrests increase so does the number of times the respondent is
held within the department of corrections (See Table-6).

Moreover, the number of charges is positively correlated with the number
of property convictions (r= .468, p= .000), the number of times on probation
(r= .657, p= .000) and the number of times a respondent had been held in the
department of corrections (r= 446, p=.000). All of theses relationships are direct,
indicating that as the number of charges increases the number of times on
probation, the number of property convictions and the number of times in the
department of corrections also increases accordingly.

“Criminogenic Patterns”
There are two other primary criminogenic variables that produce
statistically signiﬁcant but weaker correlations. Please note the direct correlations

between the number of charges and the number of violent convictions (r= .21 1,

77

=001) and being a gang member (r= .163, p= .012). These statistics suggest that
as the number of charges increase so does the number of violent convictions. In
addition, being a gang member is positively correlated with having higher
numbers of criminal charges.

Violent convictions are positively correlated with four other primary
criminogenic variables. Notice the positive, yet moderately weak, relationships
between the number of violent convictions and the number of property
convictions (r= .198, p= .002), the number of times on probation (r= .230, p=
.000) and the number of times in the department of corrections (r= .233, p= .000)
These relationships suggest that as the number of violent convictions increases so
does his or her number of property convictions, times on probation and times in
the department of corrections.

The fourth statistically signiﬁcant correlation involving the number of
violent convictions occurs with the deﬁant individualism score. The deﬁant
individualism score represents the degree to which the respondent displays deﬁant
individualistic traits. There is a positive, moderately weak relationship between
these two aforementioned variables (r= .158, p= .020). This suggests that
individuals who score higher on the deﬁant individualism scale also have higher
numbers of violent convictions. However, gang membership within itself is not
signiﬁcantly correlated with the number of violent convictions (r= .122, p= .062).
Initially, this seems to suggest deﬁant individualism may be more a predictor of
violence than actual gang membership. However, this conclusion cannot be made

until the multivariate analyses of these data are complete.

78

Moreover, the number of property convictions is signiﬁcantly correlated
with the number of times on probation (r= .465, p= .000) and the number of times
in the department of corrections (r= .490, p= .000). These ﬁndings suggest that as
the number of property convictions increases so does the number of times on
probation and the number of times the respondent is held in the department of
corrections. The number of times on probation and the number of times in the
department of corrections also show a direct, moderately strong correlation
(r= .352, p= .000). A ﬁnal primary criminogenic variable, times held within the
department of corrections, also produces a moderately weak, positive correlation
with the deﬁant individualism score (r= .141 , p= .039). This implies that
respondents who score higher on the deﬁant individualism scale also have been in

the department of corrections more often (See Table-6).

79

 

Table-6 Bivariate Correlation
(Primary Criminogenics)

 

ARR CHGS VCON PCON PROB XDOC DI GM
ARR 1.00 .934" .204" .519M .692" .359" .039 .120

 

CHGS 1.00 .211** .468“ .657" .446“ .036 .163*
VCON 1.00 .198* .230" .233" .158* .122
PCON 1.00 .465" .490" .090 .055
PROB 1.00 .352" .085 .091
XDOC 1.00 .141* .120
D1 1.00 a
GM 1.00
ARR= Number of arrests PROB: Number of times on probation
CHGS= Number of arrest charges XDOC= Number of times in DOC

VCON= Number of violent convictions DI= Deﬁant individualism Score
PCON= Number of property conviction GM= Gang membership

3 = Correlation could not be computed due to DI and GM being mutually exclusive variables.
**P< .01 *P 5 .05
Table-7 displays the results of the bivariate correlation among secondary

criminogenic variables. It is immediately apparent that the secondary
criminogenic variables (e. g. post intervention recall, post intervention networking,
attitude toward gun use and gang criminality) are not as consistently nor as
strongly correlated with one another as the primary criminogenic variables (i.e.
number of arrests, number of criminal charges, number of violent convictions,
times in department of corrections and times on probation). In fact, of the ﬁve
secondary criminogenic variables (attitude toward gun use, perception of gang

criminality, positive trajectory shift, post intervention networking and post

80

intervention recall) and two classiﬁcation variables (gang membership and deﬁant
individualism score) the correlation matrix produces only two statistically
signiﬁcant relationships.

These relationships include post intervention recall and post intervention
networking. There is a strong positive relationship between the amount of post
intervention recall and the amount of post intervention networking (r= .508,
p= 000). The statistic suggests that respondents who recalled the various
components of the lever-pulling presentations contacted more community-based
supporters. This also raises the question of program follow—up. It may be
necessary to include more follow-up in order to help participants increase their
program recall and thereby promote more community-based networking.

There is also a moderately strong positive relationship between the deﬁant
individualism score and the amount of positive trajectory shift (r= .200, p= .003).
This ﬁnding suggests that respondents who score higher on the deﬁant
individualism scale engage in more post intervention networking. An encouraging
explanation for this statistic is that individuals who have maintained deviant
lifestyles maybe more likely to transition out of them when presented with

interventions such as lever-pulling (See Table-7).

81

 

Table-7 Bivariate Correlation
(Secondary Criminogenics)

 

RCLL PINT POSS PGCR AGUN GM DI
RCLL 1.00 .508** .090 .092 .005 .003 .016

 

PINT 1.00 .118 .112 -.120 .013 .141
POSS 1.00 -.018 -.073 -.077 .200“
PGCR 1.00 -.087 -.034 -.045
AGUN 1.00 .058 -.076
GM 1.00 a
D1 1.00
RCLL: Degree of post intervention recall AGUN: Attitude toward gun use
PINT: Degree of post intervention networking GM= Gang membership
POSS= Amount ofpositive trajectory shift DI= Deﬁant individualism Score

PGCR= Perception of gang crime

a = Correlation could not be computed due to DI and GM being mutually exclusive variables.
**P< .01 *P 5 .05
A bivariate matrix of primary and secondary criminogenics is constructed to

examine whether relationships exist between the behavioral and attitudinal
measures. Table-8 displays these ﬁndings. There are only two statistically
signiﬁcant correlations. Two of the correlations displayed in Table-8 (post
intervention recall with post intervention networking and number of times in
DOC and number of property convictions) have been previously discussed
where it was noted that there is a positive correlation between the level of post
intervention recall and the amount of post intervention networking. This suggests
that respondents who recalled more components of the lever-pulling intervention

went on to take fuller advantage of those networking opportunities. The direct

82

relationship between number of times in the department of corrections and the
number of property convictions displayed in Table-8 ﬁrrther implies that
respondents who maintained criminal lifestyles and amassed increasing numbers
of property conviction were also held within the department of corrections more.
There is a weak inverse correlation between the number of convictions for
property offenses and the amount of post intervention recall (r= -.177, r= 036).
This statistic suggests that as the number of convictions for property offenses
increase the post intervention recall decreases. This suggests that individuals who
maintained criminal lifestyles, as evidenced by increased property convictions,
did not retain as much information about the lever-pulling intervention. Perhaps
this limited recall was due to a stronger commitment to a deviant lifestyle. In
addition to this ﬁnding, there is a weak direct relationship between the number of
times in the department of corrections and the amount of positive trajectory shift
(r= .130, p= .046). This suggests that respondents who have been in the
department of corrections more often undertake a more positive post intervention

trajectory shift (See Table-8).

83

 

Table-8 Bivariate Correlation
(Primary and Secondary Criminogenics)

 

RCLL PINT POSS PGCR PCON XDOC
RCLL 1.00 .508** .090 .092 -.177* -.124

 

PINT 1.00 .1 18 .112 -.057 .082

POSS 1.00 -.018 -.017 .130*

PGCR 1.00 .102 -.050

PCON l .00 .490"

XDOC 1.00
RCLL: Degree of post intervention recall PCON= Number of property conviction
PINT= Degree ofpost intervention networking XDOC= Number of times in DOC

POSS= Amount of positive trajectory shift
PGCR= Perception of gang crime

**P< .01 *P 5 .05

It is possible that the age of the respondents might play a part in their
criminal behavior. Therefore, in addition to the three correlation models reported
in Tables 6 through 8, a partial correlation matrix was computed controlling for
age. These ﬁndings are discussed below and displayed in Tables 9 through 11.

With four exceptions, the correlations between the primary criminogenic
variables displayed in Table-6 remain stable while controlling for the effects of
age on the sample. The direction, strength and signiﬁcance of the correlations are
unaffected by holding age constant. That is to say that age does not act as an
intervening force in the criminogenic dynamics of the sample. The greatest
change occurs is the correlation between the number of times on probation and

the number of violent convictions. In the base correlation matrix, this relationship

84

produces a signiﬁcant relationship (r= .23 0, p= .000), suggesting that respondents
who had more violent convictions also had been placed on probation more.
However, when controlling for age the correlation failed to attain statistical
signiﬁcance (r= .128, p= .199). Thus, age impacts the aforementioned relationship
denoting that younger offenders are perhaps too young to experience numerous
violence convictions and probation.

Another change in the model is seen in the correlation between the deﬁant
individualism score with the number of charges a respondent incurred. Although
both correlations fail to attain statistical signiﬁcance, the base model displays a
direct correlation (r= .03 6, p= .604) while the partial correlation controlling for
age produces an inverse relationship (r= -.025, p= .804). This suggests that when
the effects of age are constant, the respondent will have fewer criminal charges as
the deﬁant individualism score increases.

The correlation between the number of times in the department of
corrections and the deﬁant individualism score also changes when controlling the
effects of age. The base model shows a weak positive relationship between these
two variables (r= .141, p= .039). This suggests that respondents who scored
higher on the deﬁant individualism scale were held in the department of
corrections more often. However, when controlling for age, the relationship
failed to attain statistical signiﬁcance (r= .108, p= .281). The ﬁnal change in the
primary criminogenic correlation model occurs within the gang member variable.
No correlations are able to be computed with the gang member variable when

controlling for age (See Table—9). This occurs because age is negatively correlated

85

with gang membership. When controlling the effect of age, gang membership
cannot be computed because of the signiﬁcant role that age plays on the gang
member variable.

The overall impact of age on the primary criminogenic correlations is to
neutralize the life course dynamics. The speciﬁc life course dynamic in question
is the maintenance of a deviant lifestyle. As an individual continues in a criminal
trajectory, he or she is more likely to meet and associate with others who maintain
a similar lifestyle. Since younger respondents have not lived as long, they may not
have maintained criminal lifestyles long enough for these correlation patterns to
display themselves with the same level of statistical signiﬁcance seen in the larger
sample. This suggests that the baseline primary criminogenic correlations may
only be applicable in designing intervention programs for older offenders. There
is also a second effect of age on the statistics. The effect of age also diminishes
some statistical correlations in other statistical models.

Tables 9 and 10 show that when controlling for age, bivariate correlation of
gang membership can not be computed with either primary or secondary
criminogenics. Age is inversely correlated with gang membership. This inverse
correlation is supported in the literature as well as the model statistics. The earlier
demographic analysis (See Table-3) shows that participants in the gang member
group are younger than the participants in the zero-inﬂuence group or the deﬁant

individualist group.

86

 

 

Table-9 Partial Correlation- Controlling for Age
(Primary Criminogenics)

 

ARR CHGS VCON PCON PROB XDOC DI GM
ARR 1.00 .916" 233* .557" .688" .259" .008

 

CHGS 1.00 .286“ .547“ .640" .392“ -.025
VCON 1.00 .230" .128 .366” .226“
PCON 1.00 .558" .529" .031 .
PROB 1.00 .256" .055
XDOC 1.00 .108 .
DI 1.00

GM 1.00
ARR= Number of arrests PROB= Number of times on probation
CHGS= Number of arrest charges XDOC= Number of times in DOC

VCON= Number of violent convictions DI= Deﬁant Individualism Score
PCON= Number of property conviction GM= Gang membership

**P< .01 *P< .05
Table-10 displays the secondary criminogenic partial correlations

controlling for age. This model contains only two variations ﬁ'om the base model
displayed in Table-7. The ﬁrst variation occurs in the correlation between the
perception of gang criminality and the positive trajectory shift. Despite neither of
the correlations attaining signiﬁcance, the base model displays an inverse
correlation (r= -.018, p= .798). This suggests that as the respondent’s perception
of criminality as being important to gangs increases, their amount of positive
trajectory shift decreases. That is to say, respondents who displayed less positive
trajectory shifts after participating in the lever-pulling intervention thought that

crime was more important to gang members. However, in the partial correlation

87

model the correlation direction switches (r= .128, p= .201). This suggests that as
the respondent’s perception of criminality as being important to gangs increases,
their amount of positive trajectory shift also increases. That is to say that when
controlling the effect of age, respondents who participated in more positive
trajectory shifts thought that crime was more important to gangs.

The effect of age on the secondary correlation is similar to the effect on the
primary criminogenic correlations. Age neutralizes the maintenance dynamic of
life course. Younger respondents may have not had the opportunity to maintain or
desist deviant lifestyles according to the same patterns as other members of the
sample. Controlling the effect of age on the correlation between perceived gang
criminality and positive trajectory shift causes the direction of the correlation to
change. The partial correlation suggests that respondents who see criminality as
important to gangs also have less positive trajectory shifts.

The only other variation in the partial correlation of secondary
criminogenics is the inability to compute correlations with the gang member
variable. This occurs because age is negatively correlated with gang membership.
When controlling the effect of age, gang membership cannot be computed
because of the signiﬁcant role that age plays on the gang member variable (See

Table-10).

88

 

Table-10 Partial Correlation - Controlling for Age
(Secondary Criminogenics)

 

RCLL PINT POSS PGCR AGUN GM DI

RCLL 1.00 .558“ .065 .158 .710 . .034
PINT 1.00 .161 .154 -.042 . .098
POSS 1.00 .128 -.039 . 209*
PGCR 1.00 -.049 . -.041
AGUN 1.00 . -.073
GM 1.00 a
D1 1.00

 

RCLL: Degree of post intervention recall AGUN= Attitude toward gun use
PINT= Degree of post intervention networking GM= Gang membership

POSS= Amount of positive trajectory shift DI= Deﬁant Individualism Score
PGCR= Perception of gang crime

a = Correlation could not be computed due to D1 and GM being mutually exclusive variables.
**P< .01 *P 5 .05
Perhaps the greatest amount of variation between the base correlation
matrices and the partial correlations occurs when examining the partial correlation
between primary and secondary criminogenics. Table-11 displays the ﬁve
variations that occur when controlling for age. Immediately notable is the addition
of two extra variables, attitude toward gun use and number of times on probation.
These additional variables are included in this table because there is a statistically
signiﬁcant relationship that did not exist in any previous model.
There are two correlations that experience a directional change when

controlling for age. In the base model, positive trajectory shift produces an inverse

correlation with both perception of gang criminality (r= -.018, p= .798) and

89

 

number of property convictions (F -.017, p= .800). However, in the partial
correlation matrix, the relationship between positive trajectory shift and
perception of gang criminality (F .128, p= .201), as well as the relationship
between positive trajectory shift and number of property convictions (F .073,

p= .463), are both positive. This of course does not change the failure to attain
signiﬁcance. This directional shift, when holding the effects of age constant,
suggests that when age is factored out, offenders may engage in more positive
trajectory shifts if they perceive crime as being important to gang members and if
they have more property convictions. This varied correlation pattern would
necessitate different intervention strategies.

In addition to these directional changes of the previous variables, the
correlation between positive trajectory shift and the number of times in the
department of corrections (F .130, p= .046) loses statistical signiﬁcance in the
partial correlation (F .179, p= .072). This suggests that when age is held constant
the desire of career criminals to change their life trajectory may not be as readily
apparent.

Similarly, the positive correlation between the number of property
convictions and the number of times in the department of corrections (F .490,
p= .000) is no longer signiﬁcant when controlling for age (F .108, p= .281). The
partial correlation matrix also produces a moderate correlation between the
number of times on probation and attitude toward gun use (F .208, p= .036). This
statistic suggests that as the number of times on probation increases the

respondent’s attitude toward gun use becomes more favorable (See Table-1 1).

90

Thus, as age and probationary experiences increase attitudes toward gun use,

appears to become more neutralized.

 

Table-11 Partial Correlation- Controlling for Age
(Primary and Secondary Criminogenics)

 

RCLL PINT POSS PGCR PCON XDOC AGUN PROB

 

RCLL 1.00 .558" .065 .158 -.289* -.158 .037 -.142
PINT 1.00 .161 .154 -.083 .068 -.042 .043
POSS 1.00 .128 .073 .179 -.039 .010
PGCR 1.00 .037 -.095 -.049 -.026
PCON 1.00 .108 .072 .558"
XDOC 1.00 .108 .256"I
AGUN 1.00 .208“
PROB 1.00
RCLL= Degree of post intervention recall PCON= Number of property conviction

PINT= Degree of post intervention networking XDOC= Number of times in DOC
POSS= Amount of positive trajectory shift AGUN= Attitude toward gun use
PGCR= Perception of gang crime PROB= Number of times on probation

**P< .01 *P g .05
“Zero-Inﬂuence Group”

The purpose of computing sub-sample correlations is to identify the primary
and secondary criminogenic correlations among speciﬁc groups without the
affects of other groups confounding the results. The following analyses are
replications of the base correlation models stratiﬁed by the zero-inﬂuence, deﬁant

individualist and gang member groups in the study.

91

The ﬁrst group to be examined is the zero-inﬂuence group. When
examining the primary criminogenic correlations, the number of arrests is
strongly correlated with the number of charges incurred by the respondents
(F .907, p= .000). This suggests that as the number of arrests increases so does
the number of charges. This ﬁnding is intuitive in its nature. The number of
arrests was also signiﬁcantly correlated with the number of property convictions
(F .365, p= .002) and the number of times on probation (F .644, F .000). Both
of these correlations are direct, which suggests that as the number of arrests
increase so does the number of property convictions and the number of times on
probation (See Table-12).

Zero-inﬂuenced respondents display a strong direct correlation between the
number of charges they incurred and the number of times respondents were on
probation (F .612, F .000). As the number of charges increases so does the
number of experiences with probation. Among the zero-inﬂuence group, the
number of charges is also correlated with the number of property convictions
(F .326, p= .005) and the number of times in the department of corrections
(F .259, p= .028). These subsequent correlations are moderately strong and in the
same direction as the other primary criminogenic correlations.

The number of times on probation is correlated with both the number of
violent convictions (F .244, p= .039) and the number of property convictions (F
.460, p= .000). The direct nature of these correlations suggests that as the number
of property and violent convictions increases so does the number of times on

probation. The number of times in the department of corrections is moderately

92

correlated with the number of property convictions (F .397, p= .001) and the
number of times on probation (F .333, p= .004). As property convictions and the
number of times on probation increases so does the number of times in the

department of corrections.

 

Table-12 Bivariate Correlation - Zero Inﬂuence Group
(Primary Criminogenics)

 

ARR CHGS VCON PCON PROB XDOC
ARR 1.00 .907" .209 .365“ .644" .227

 

CHGS 1.00 .170 .326“ .612** .259*
VCON 1.00 .154 244* .189
PCON 1.00 .460** .397**
PROB 1.00 .333**
xnoc 1.00
ARR= Number of arrests PCON= Number of property conviction
CHGS= Number of arrest charges PROB= Number of times on probation

VCON= Number of violent convictions XDOC= Number of times in DOC

**P< .01 *P g .05
An examination of the secondary criminogenics ﬁnds that among the zero-
inﬂuence group, there is only one statistically signiﬁcant correlation. Speciﬁcally,
there is a positive, moderate correlation between post intervention recall and the
amount of post intervention networking (F .392, p= .004). This statistic suggests
that among respondents of the zero-inﬂuence group, post intervention networking
increases when respondents retain more information from the program (See

Table-13). This correlation suggests that it may be possible to increase the

93

amount of post intervention networking, simply by improving the participants’
recall of program components. Members of the zero-inﬂuence group may beneﬁt
ﬁ'om the incorporation of mnemonic devices or rhetoric in the lever-pulling
curriculum.

Revising the lever-pulling curriculum to include more phrases or slogans
that serve as mnemonic devices may increase the recall of the participants.
Additionally, the incorporation of some type of follow-up in the program may
also serve as an important link between the respondent’s recall of the program
components and the respondent’s networking with community-based supporters.
Through a follow-up stage, the program facilitators could remind the participants
of various components in the program and help the participants initiate the post

intervention contact with community-based supporters.

 

Table-13 Bivariate Correlation - Zero Inﬂuence Group
(Secondary Criminogenics)

 

RCLL PINT POSS PGCR AGUN
RCLL 1.00 .392" -.092 -.079 .082

 

PINT 1.00 -.049 -.O3l -.155
POSS 1.00 -.174 -.162
PGCR 1.00 -.267
AGUN 1.00
RCLL=Degree of post intervention recall PGCR= Perception of crime importance to gang

PINT= Degree of post intervention networking AGUN= Favorable attitude toward gun use
POSS= Amount of positive trajectory shift

**P< .01 *P 5.05

94

Table-14 displays the results of the primary and secondary criminogenic
correlation for the zero-inﬂuence group. This speciﬁc matrix produces no
statistically signiﬁcant correlations between primary and secondary criminogenic
variables. The two correlations displayed in Table-14 are previously discussed.
The positive correlation between post intervention recall and the amount of post
intervention networking (r= .3 92, p= .004) is also displayed in Table-13 where it
suggests that among respondents of the zero-inﬂuence group, post intervention
networking increases when respondents retain more information from the
program. ‘

The second correlation that is displayed in Table-14 occurs between the
number of times the respondent is held within the department of corrections and
the number of property convictions the respondent incurs. The number of times in
the department of corrections is positively correlated with the number of property

convictions (r= .3 97, p= .001); suggesting that as the respondents were convicted
I of property crimes more often, they were also held within the department of
corrections more often.

Table-14 shows that the primary and secondary variables are not
signiﬁcantly correlated with one another among the zero-inﬂuence group. The
result is that members of the zero-inﬂuence group had no interdependence of
primary and secondary criminogenic variables. The criminal behaviors exhibited
by members of this group were not correlated to any observable non-criminal

attitudes.

95

 

Table-14 Bivariate Correlation - Zero Inﬂuence Group
(Primary and Secondary Criminogenics)

 

RCLL PINT POSS PGCR PCON XDOC
RCLL 1.00 .392" -.092 -.079 -.245 -.176

 

PINT 1.00 -.049 -.O31 -.085 .063
POSS 1.00 -.174 -.137 .131
PGCR 1.00 -.006 -.O86
PCON 1.00 .397**
XDOC 1.00
RCLL=Degree of post intervention recall PGCR= Perception of crime importance to gang

PINT= Degree of post intervention networking PCON= Number of property convictions
POSS= Amount of positive trajectory shiﬁ XDOC= Times in department of corrections

**P< .01 *P _<_ .05
“Defiant Individualists”

The deﬁant individualist group was comprised of respondents who reported
not having an active membership in a gang at the time of the survey but who
exhibited a range of deﬁant individualistic traits. The bivariate correlations of
primary criminogenic variables produced statistically signiﬁcant relationships
among the deﬁant individualist group. There are strong direct correlations
between the number of arrests and the number of charges (r= .931, p= .000),
number of arrests and the number of property convictions (r= .524, p= .000) as
well as the number of arrests and the number of times on probation (r= .716, p=
.000). These statistics suggest that within the deﬁant individualist group, as the
number of arrests increase so does the number of property convictions, charges

and number of times on probation.

96

Respondents’ arrests also produces moderate positive correlations with the
number of violent convictions (r= .209, p= .012) and the number of times in the
department of corrections (r= .366, p= .000). Both subsequent correlations are
positive, which suggest that as the deﬁant individualist’s numbers of arrests
increases so does their numbers of violent convictions and numbers of times in
the department of corrections.

The number of violent convictions is moderately correlated with the number
of charges (r= .234, p= .005). The direction of the relationship suggests that as the
number of charges increases for deﬁant individualists so does the number of
violent convictions. The number of charges produces three additional statistically
signiﬁcant relationships among the deﬁant individualist sub-sample. The number
of charges produces direct correlations with the number of property convictions
(r= .463, p= .000), the number of times on probation (r= .658, p= .000) and the
number of times in the department of corrections (r= .488, p= .000)

(See Table-15).

Within the deﬁant individualist group, the number of violent convictions is
signiﬁcantly correlated with the number of times in the department of corrections
(r= .226, p= .007). This suggests that as the number of violent convictions
increased so did the number of times in the department of corrections. In addition
to this moderate relationship, the number of violent convictions is also correlated
to the number of times on probation (r= .209, p= .012) and the number of property

convictions (r= .182, p= .031).

97

The number of property convictions is correlated to both the number of
times on probation (r= .439, p= .000) and the number of times in the department
of corrections (r= .510, p= .000). Both correlations are direct, which suggests that
as the number of property convictions increased so did the number of times on
probation and the number of times in the department of corrections. As an
extension of this, there is a positive correlation between the number of times on
probation and the number of times in the department of corrections (r= .327,
p= .000). Taken as a whole, these positive correlations suggest that as members of
the deﬁant individualist group maintained criminal lifestyles, they increasingly
came into contact with various forms of the criminal justice system such as

probation and incarceration in the department of corrections.

 

Table-15 Bivariate Correlation - Deﬁant Individualist Group
(Primary C riminogenics)

 

ARR CHGS VCON PCON PROB XDOC
ARR 1.00 .931" .209* .524" .716** .366"

 

CHGS 1.00 .234" .463** .658** .488“
VCON 1.00 .182* .209* .226**
PCON 1.00 .439" .510**
PROB 1.00 .327**
XDOC 1.00
ARR= Number of arrests PCON= Number of property conviction
CHGS= Number of arrest charges PROB= Number of times on probation

VCON= Number of violent convictions XDOC= Number of times in DOC

**P< .01 *P 5 .05

98

Table-16 displays the correlations between secondary criminogenic
variables among deﬁant individualists. It is readily apparent that this matrix
produced only one statistically signiﬁcant relationship. There is a strong positive
correlation between the amount of post intervention recall, which denotes the
number of components the respondent remembers ﬁ'om the lever-pulling
intervention and the amount of post intervention networking (r= .576, p= .000),
which speaks to the degree to which the respondent made contact with
community-based supporters present at the lever-pulling program. This statistic
suggests that deﬁant individualists who recall more elements of the intervention
program also engage in more post intervention networking. This ﬁnding is similar

to the zero-inﬂuence group (See Table-16).

 

Table-l6 Bivariate Correlation- Defiant Individualist Group
(Secondary Criminogenics)

 

RCLL PINT POSS 'lPGCR AGUN
RCLL 1.00 .576“ .178 .214 -.001

 

PINT 1.00 .186 .179 -.001

POSS 1.00 .034 -.035

PGCR 1.00 .023

AGUN 1.00
RCLL= Degree of post intervention recall PGCR= Perception of gang crime
PINT= Degree of Post intervention networking AGUN= Attitude toward gun use

POSS= Amount of positive trajectory shiﬁ

**P< .01 *P 5 .05

99

The matrix of primary and secondary criminogenic variables only produces
two statistically signiﬁcant correlations for the deﬁant individualist group. There
is an inverse signiﬁcant correlation between the number of property convictions
and the amount of post intervention recall (r= -.232, p= .042). This suggests that
people with high numbers of property convictions have lower post intervention
recall, which reﬂects their ability to remember various components of the lever-
pulling intervention. This may be due to respondents with high property
convictions having stronger commitments to a criminal lifestyle. The high degree
of commitment to criminal lifestyles would be antithetical to the pro-social
messages and opportunities being offered at the intervention meetings.

Table-17 displays a correlation of primary and secondary criminogenic
variables. Due to the table containing both primary and secondary criminogenic
variables, some correlations are replicated from Table-15 and 16. For example,
the correlation between the degree of post intervention recall and the amount of
post intervention networking (r= .576, p= .000), which suggests that as the
respondent remembers more components of the lever-pulling intervention, the
respondent also contacts more community-based supporters.

Additionally the positive correlations between the number of criminal
charges and the number of property convictions (r= .463, p= .000) were
replicated. The positive correlation between the number of criminal charges and
the munber of times in the department of corrections (r= .488, p= .000) is also
displayed in both Table-17 and Table-15. In addition to these replicated

correlations, Table-17 also displays three new correlations.

100

A positive correlation exists between the number of property offenses and
the perception of gang criminality (r= .194, p= .023). This suggests that
respondents who have higher property convictions also believe that crime is more
important to gangs. It is also possible that these respondents have more violent
beliefs about gang behavior. There is a negative correlation between the number
of property convictions and the amount of post intervention recall (r= -.232, p=
.042). This statistic suggests that respondents who remembered more components
of the lever-pulling intervention also had fewer property convictions. This
relationship suggests that deﬁant individualists with fewer property crime
convictions may have less commitment to deviant lifestyles and were therefore
more responsive to the opportunities presented in the lever-pulling program.

The ﬁnal correlation displayed in Table-17 was the positive correlation
between the number of charges and the amount of positive trajectory shift (r=
.174, p= .039). This statistic suggests that deﬁant individualists who had more
criminal charges overall engaged in normative processes more often after
participating in the lever-pulling intervention (See Table-17). This ﬁnding may
speak to the desire of the deﬁant individualists to desist their criminal life

trajectories.

101

 

Table-17 Bivariate Correlation - Defiant Individualist Group
(Primary and Secondary Criminogenics)

 

RCLL PINT POSS PGCR PCON XDOC CHGS
RCLL 1.00 .576** .178 .214 -.232* -.118 -.025

 

PINT 1.00 .186 .179 -.013 .096 .039
POSS 1.00 .034 .026 .127 .174*
PGCR 1.00 . 194* .019 -.081
PCON 1.00 .510** .463"
XDOC 1.00 .488“
CHGS 1.00
RCLL= Degree ofpost intervention recall PGCR= Perception of gang crime

PINT= Degree of post intervention networking PCON= Number of property conviction
POSS= Amount of positive trajectory shift XDOC= Number of times in DOC

**P< .01 *P 5 .05
“Gang Member Group”

Table-18 displays the primary criminogenic correlations within the gang
member group. As with the other study groups, there are signiﬁcant correlations
between the number of arrests and the number of property convictions (r= .970,
p= .000), the number of times on probation (F .729, p= .000) and the number of
times in the department of corrections (r= .605, p= .004). These ﬁndings suggest
that as the number of arrests increases so does the number of property
convictions, the number of times on probation and the number of times in the
department of corrections.

The number of property convictions is also correlated with the number of

charges incurred (r= .687, p= .001). As the number of property convictions

102

increases so does the number of overall charges. The number of charges is, in
turn, signiﬁcantly correlated with the number of times on probation (r= .714,
p= .000) and the number of times in the department of corrections (r= .638,
p= .002).

There are two strong correlations between the number of property
convictions and the number of times on probation (r= .573, p= .007) and the
number of times in the department of corrections (r= .580, p= .006). Both of these
relationships are signiﬁcant. As the number of property convictions increases so
does the number of times on probation and the number of times in the department
of corrections. There is a correlation between the number of times on probation

and the number of times in the department of corrections (r= .539, p= .012).

 

Table-18 Bivariate Correlation - Gang Member Group
(Primary Criminogenics)

 

ARR CHGS VCON PCON PROB XDOC
ARR 1.00 .970" .134 .729" .696“ .605"

 

CHGS 1.00 .113 .687" .741" .638"
VCON 1.00 .336 .266 .123
PCON 1.00 .573** .580"
PROB 1.00 .5 39*
XDOC 1.00
ARR= Number of arrests PCON= Number of property conviction
CHGS= Number of arrest charges PROB= Number of times on probation

VCON= Number of violent convictions XDOC= Number of times in DOC

**P< .01 *P 5 .05

103

Secondary criminogenic correlations are displayed in Table-19. This matrix
displays only two statistically signiﬁcant correlations. There is a strong positive
correlation between the amount of post intervention networking, reﬂecting the
degree to which the respondent took advantage of community-based contacts
presented at the lever-pulling intervention and the amount of post intervention
recall, which reﬂects the number of components the respondents remember from
the lever-pulling program (r= .597, p= .038). As with other groups in the study, as
the amount of post intervention recall increases, so does the amount of post
intervention networking. This correlation suggests that it may be possible to
increase the amount of post intervention networking simply by improving the
participants’ recall of program components. Members of the zero-inﬂuence group
may beneﬁt ﬁ'om the incorporation of mnemonic devices or rhetoric in the lever-
pulling curriculum.

Revising the lever-pulling curriculum to include more phrases, slogans or
even visual components may increase the recall of the participants: Additionally,
the incorporation of some type of follow-up in the program may also serve as an
important link between the respondent’s recall of the program components and
the respondent’s networking with the community-based supporters. Through a
follow-up stage, the program facilitators could remind the participants of various
components in the program and help the participants initiate the post intervention

contact with community-based supporters.

104

An inverse correlation between the amount of post intervention networking
and attitude toward gun use (r= -.610, p= .046) is also shown in Table-18. This
suggests that respondents who have more favorable attitudes toward gun use also
have less post intervention networking. This statistic is by far the strongest
secondary correlation coefﬁcient among the gang member group. This suggests
that gang members have a stronger commitment to violence, as evidenced by their
more acceptable attitudes toward gun use and therefore engage in less post
intervention networking with community-based supporters (See Table-19). Gang
members with more favorable attitudes toward gun use are also more likely to
have a stronger commitment to the gang culture and less likely to engage in

positive transitions.

 

Table-l9 Bivariate Correlation - Gang Member Group
(Secondary Criminogenics)

 

RCLL PINT POSS PGCR AGUN
RCLL 1.00 .579* .182 -.165 -.196

 

PINT 1.00 .097 .187 -.610*

POSS 1.00 .011 -.167

PGCR 1.00 -.363

AGUN 1.00
RCLL= Post intervention recall PGCR= Perception of gang crime
PINT= Post intervention networking AGUN= Attitude toward gun use

POSS= Positive trajectory shift

**P< .01 *P _<_ .05

105

A bivariate correlation matrix of primary and secondary criminogenic
variables produces three statistically signiﬁcant relationships. Note that all three
of the variables are related to the attitude toward gun use. First, there is a strong
positive correlation between the attitude toward gun use and the number of
property convictions (r= .705, p= .002). Moreover, as the number of property
convictions increases, the attitude toward gun use becomes more favorable. The
only negative gun-related correlation in Table-20 is the relationship between the
attitude toward gun use and the amount of post intervention networking (r= -.610,
p= .046). Therefore, gang members who have more favorable views of gun use
engage in less contact with community-based supporters present at the lever-
pulling intervention. This ﬁnding suggests that gang members are somewhat
resistant to desisting deviant attitudes that may impact their criminal behavior.

The attitude toward gun use is also strongly correlated with the number of
arrests (r= .563, p= .023) and the number of charges (r= .549, p= .028). These
statistics suggest that as the number of charges and arrests increase, the attitude
toward gun use becomes more favorable. Gang members in the study appear to
have a much stronger commitment to gun use than either the deﬁant individualist
or zero-inﬂuence groups, as evidenced by more variables being signiﬁcantly
correlated with attitudes toward gun use. This favorable attitude toward gun use
may explain the traditionally higher rates of violence within the gang sub-culture

(See Table-20).

106

 

Table-20 Bivariate Correlation - Gang Member Group
(Primary and Secondary Criminogenics)

 

RCLL PINT POSS PGCR PCON ARR CHGS AGUN

RCLL 1.00 .579* .182 -.165 .222 .055 .062 -.l96

 

PINT 1.00 .097 .187 -. 194 -.l 16 -.007 -.610*
POSS 1.00 .011 -.502 .1 l 1 .207 -. l 67
PGCR 1.00 -.136 —. 170 -.223 -.363
PCON 1.00 .729" .687“ .705”
AR 1.00 .970" 563*
CHGS 1.00 .549“
AGUN 1.00
RCLL= Degree of post intervention recall PGCR= Perception of gang crime

PINT= Degree of post intervention networking PCON= Number of property conviction
POSS= Amount of positive trajectory shift XDOC= Number of times in DOC

**P< .01 *P 5 .05
“Discriminant Function Analyses”

The following analyses are designed to determine whether or not primary
and secondary criminogenic variables discriminate between various study groups.
A discriminant function analysis is a statistic used to test whether or not a group
of variables signiﬁcantly discriminate between two or more groups, thereby
producing a latent function. The discriminant function analysis allows the
researcher to determine if the overall model discriminates between the various
dependent groups as well as which individual component of the model contributes

most to the differentiation.

107

A discriminant analysis model is displayed in Table-21. This model
displays a discriminant analysis of secondary crirrrinogenic variables across the
zero-inﬂuence and deﬁant individualist groups. The Wilks’ Lambda statistic
(A: .941, p= .311) shows that the secondary criminogenic variables do not
discriminate between the zero-inﬂuence and the deﬁant individualist groups. The
function of the groups at the centroid (ZI = .337, DI= -.181), show that these two
groups were fairly close together. This closeness also denotes no discriminant
ﬁrnction (See Table-21).

From the standardized discriminant function coefﬁcients, it is clear that the
perception of gang crime [f(x)= .802] and the amount of positive trajectory shift
[f(x)= -.688] contribute the most to the secondary criminogenic differences
between these two groups. The structure matrix shows that perception of gang
crime (r= .678) and positive trajectory shift (r= -.560) also have the strongest
correlation to a latent function in this model despite the model not attaining
statistical signiﬁcance.

Interestingly, the weakest predictor of a discriminant function is the attitude
toward gun use [f(x)= -.030]. Attitude toward gun use also displays the weakest
correlation to the discriminant function (r= -.045). The canonical correlation for
Table-21 also shows that there is a weak relationship between the model groups
and secondary criminogenics (RC: .242). This weak canonical correlation shows
that secondary criminogenics are not necessarily the best way in which to

differentiate deﬁant individualists from respondents in the zero-inﬂuence group.

108

 

Table-21 Secondary Criminogenic Discriminant Analysis
(Zero-Inﬂuence and Deﬁant Individualists)

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

Mean SD Dis. f(x) Struct.
Attitude Toward Gun Z1 8.3 2.520 -.030 -.045
Use DI 8.4 2.349
Perception of Gang ZI 25.5 4.286 .802 .678
Crime DI 23.6 5.69
Post Intervention Z1 1.9 1.558 -.256 -.098
Networking DI 2.0 1.799
Positive Trajectory 21 1.4 1.180 -.688 -.560
Shift DI 1.8 1.851
Post Intervention Z1 5.3 3.190 .256 .172
Recall DI 5.0 2.963
Wilks’ A: .941 Sig.= .311 Rc= .242

Centroid Functions (ZI = .337, DI= -.181)

Table-22 displays the results of a discriminate function model of the
primary criminogenic variables. The variables are tested between deﬁant
individualist and gang member groups. The signiﬁcance level of the Wilks’
Lambda (p= .283), displayed in Table-22, shows that the model of primary
criminogenics does not discriminate between deﬁant individualists and gang
members; hence these data show little difference in the level of criminal justice
contact as evidenced by the number of arrests, number of violent convictions,
number of property convictions, number of times on probation and the length of
time held within the department of corrections.

The Wilks’ Lambda ranges from zero to one and functions as an F test of
signiﬁcance. If the model attains statistical signiﬁcance, then each individual
variable is assessed in order to determine which variable differs signiﬁcantly by
group. A Wilks’ Lambda of zero (0) is interpreted as the group means differ and

the groups are therefore different from one another.

109

However, a Wilks’ Lambda of one (1) suggests that the group means do not
differ and the two groups are more similar to one another. The model of primary
criminogenics, displayed in Table-21, produces a Wilks’ Lambda of .961. The
function of the groups at the centroid (D1 = -.077, GM= .518) denotes the distance
between the two groups. The closer the centroid functions are, the less discrete the
two groups. When the centroid ﬁrnctions are close together, this suggests that the
model variables do not discriminate between the two groups (See Table-22).

Table-22 also shows the results of the standardized discriminant ﬁmction
coefﬁcients for variables in the model. These coefﬁcients are partial in that they
do not show overlapping effects of the other variables in the model. The
standardized discriminant coefﬁcients denote the amount of discrimination that
each variable lends to the discriminant function. The two variables on which the
deﬁant individualist and gang member groups differ the most are numbers of
arrests [f(x)= .715] and number of violent convictions [f(x)= .515]. Therefore,
these two variables contribute the most to the ability to differentiate between
deﬁant individualists and gang members; although there is no statistically
signiﬁcant difference.

In discriminant analysis, the discriminant function is a latent variable that is
created as a linear function of the independent variables. The structure coefﬁcient
denotes the uncontrolled association between the independent variable and the
latent function. Table-22 shows that, again, the variable with the greatest
correlation to a latent function is the number of arrests (r= .689). Structure

coefﬁcients are interpreted like standard correlation coefﬁcients; therefore, it is

110

important not to confuse them with the model’s canonical correlation (R). The
canonical correlation expresses the relationship between the dependant variable
groups and the discriminant function.

An Rc= 0 would be interpreted as no relationship between the groups and
the discriminant function. Conversely, an Rc= 1 would be interpreted as a perfect
association between the dependent groups and the latent discriminant function.
Table-22 shows that there is a weak relationship between the model groups and
primary criminogenics (Rc= .197). This weak canonical correlation suggests that
perhaps primary criminogenics are not the best way in which to differentiate gang

members from deﬁant individualists.

 

Table-22 Primary Criminogenic Discriminant Analysis
(Deﬁant Individualists and Gang Members)

 

 

Mean ‘ sn ‘ Dis. f(x) struct';
. , . . . r _. -.Matrix7
DI 8.2 6.778 .715 .689
GM 11.3 11.155
Number of Conviction DI .5 1.034 .515 .553

.7..,.,

 

N timber of Arrests

 

 

 

 

 

 

 

 

 

 

 

Violent GM .9 .889

Number of Conviction D1 .7 1.348 -.290 .264
Property GM 1.0 1 .673

Number of Times on DI 2.1 1.409 -.064 .491
Probation GM 2.5 l .720

Length of Time in DOC DI 3757.5 3419.081 -.557 -.594
(Days) GM 2576.3 2764.157

Wilks’ A= .961 Sig.= .283 Rc= .197

Centroid Functions (DI = -.077, GM= .518)
Table-23 shows a discriminant analysis of secondary criminogenic variables
across the deﬁant individualist and gang member groups. The Wilks’ Lambda

shows that secondary criminogenic variables also do not discriminate between

111

deﬁant individualists and gang members (A= .956, p= .763). The group functions
at the centroid (D1 = -.076, GM= .461) are also relatively close and supports the
ﬁnding of no discriminant function (See Table-23).

From the standardized discriminant function coefﬁcients, we see that
attitude toward gun use [f(x)= .861] and positive trajectory shift [f(x)= -.476] are
the greatest contributors to differentiation between deﬁant individualists and gang
members. The structure matrix also conﬁrms that attitude toward gun use
(r= .851) and positive trajectory shift (r= -.461) have stronger correlations to a
latent ﬁrnction than the other variables in the model. The canonical correlation,
however, suggests that secondary criminogenics are not the best means with

which to differentiate between deﬁant individualists and gang members.

 

Table-23 Secondary Criminogenic Discriminant Analysis
(Deﬁant Individualists and Gang Members)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. . _Mean‘ SD,“ turner—Aswan?
2,; 24.2.3.5: . p. i .;- - "1 1’ Matrix
Attitude Toward Gun DI 8.4 2.349 .861 .851
Use GM 9.5 3.142
Perception of Gang D1 23.6 5.693 -.010 -.158
Crime GM 23.1 4.874
Post Intervention DI 2.0 1.799 .375 .124
Networking GM 2.1 2.040
Positive Trajectory DI 1.8 1.851 -.476 -.461
Shift GM 1.4 1.213
Post Intervention D] 5.0 2.963 -.115 .001
Recall GM 5.0 3.015
Wilks’ A= .956 Sig.= .763 R,= .186

Centroid Functions (DI = -.076, GM= .461)

112

“Summary of Findings”

This study contains several relevant ﬁndings with which to better
understand the criminogenic differences between gang members, deﬁant
individualists and offenders with no gang inﬂuence. The demographic analysis
shows that the overall sample contains more blacks than any other racial minority.
This representation of black respondents was disproportionate to the
representation of blacks in the population of Indianapolis, Indiana.

Males are also over represented in the sample. The general population of
Indianapolis is almost evenly distributed between males and females but the
sample was heavily male. Additionally, the ﬁndings show that a large percentage
of the sample is not married. This marriage ﬁnding is interesting when
considering the age distribution of the sample. The average age of the sample
respondent is 31 with a median age of 30.

Respondents in the sample also have relatively moderate education levels.
The majority (84.2%) of the sample has 12 years of education or less. While the
overall education level is not very high, the employment statistics are high. The
sample shows only a small number (1 .2%) of respondents as being unemployed at
the time of the survey.

An examination of the primary criminogenic variables shows that gang
members had consistently higher average offenses than other groups. These
higher average offenses can be seen in the number of arrests, the number of

charges, the number of violent convictions, the number property convictions, the

113

number of times on probation and the number of times in the department of
corrections.

An analysis of the secondary criminogenics reveals sirrrilar patterns to those
of the primary criminogenics. The attitude toward gun use shows that the sample
was generally less accepting of gun use. However, gang members are more
accepting of gun use than other groups in the sample. Interestingly, gang members
perceive crime as being less important to gang members than either of the other
two groups in the sample.

The amount of positive trajectory shift found in the sample is relatively low.
However, members of the deﬁant individualist group display the most positive
trajectory shift. Similarly, the sample shows low levels of post intervention
networking. The zero-inﬂuence group displays the least post intervention
networking among the study groups. Despite displaying the least post intervention
networking, the zero-inﬂuenced group had the greatest post intervention recall of
the three groups.

Among the bivariate correlation, the primary criminogenic variables
(number of arrests, the number of charges, the number of violent convictions, the
number property convictions, the number of times on probation and the number of
times in the department of corrections) produced 18 statistically signiﬁcant
relationships. These primary criminogenic relationships are intuitive in nature
such as the statistically signiﬁcant relationship between the number of arrests and

the number of charges incurred. The primary correlation matrix functions as a

114

baseline model against which to determine any ﬂuctuation across the study
groups.

Unlike the primary criminogenic correlation matrix, the secondary
criminogenic correlation matrix produced only two statistically signiﬁcant
relationships. A bivariate correlation matrix of both primary and secondary
criminogenic variables also shows only two statistically signiﬁcant relationships.

In addition to univariate, bivariate and multi-variate statistics, this study
tests ﬁve hypotheses that focus on the variations of both primary and secondary
criminogenic factors, which might provide a means by which deﬁant
individualists could be differentiated ﬁ'om gang members and offenders with no
gang-related inﬂuence. The ﬁrst study hypothesis is that there is no difference in
the criminal justice system contact between gang members and deﬁant
individualists, which would constitute a discriminate ﬁrnction. Based on the
preceding analysis this hypothesis is supported (See Table-22).

The second hypothesis is that there is a direct relationship between the
number of criminal charges and deﬁant individualism. Based on the preceding
analysis this hypothesis is not supported (See Table-6). The third hypothesis is
that there is no difference in the attitude toward gun use between the deﬁant
individualist and the zero-inﬂuence group, which could constitute a discriminant
function. Based on the preceding analysis this hypothesis is supported
(See Table-21).

The fourth hypothesis is that there is an inverse relationship between deﬁant

individualism and the amount of positive trajectory shift. Based on the preceding

115

analysis, this hypothesis is not supported (See Table-7). The ﬁfth hypothesis is
that there is no difference in the perception of gang criminality between gang
members and deﬁant individualists, which could constitute a discriminant
ﬁmction. Based on the preceding analysis this hypothesis is supported

(See Table-23). Subsequent explanations as to why the results occurred in this

manner are discussed in the concluding chapter.

116

Mﬁr—V
“Summary of Purpose”

This study addresses the changing nature of gangs in the United States.
Despite the slight increase in active gang membership (Eagley and Ritz, 2006),
victims reported fewer crimes being perpetrated by gang members (Harrel, 2005).
The reduction of reported crimes committed by gang members is dramatic in
proportionality. The number of violent victimizations decreased from 1.1 million
in 1994 to only 341,000 in 2003 (Harrel, 2005). From 1994-2003, crime victims
identiﬁed the alleged perpetrators as gang members approximately 12% of the
time (Ibid). Perpetrators were identiﬁed as gang members in about 10% of
robberies and 4% of the rapes (Harrel, 2005).This study posits that this seeming
inconsistency can be explained not only by criminal justice practitioners
becoming more acclimated to gangs, thereby over identifying perpetrators as gang
members, or political intervention in the agencies’ responses which re-deﬁne who
is a gang member but, most importantly, due to the changing nature of gangs.

This study posits that contemporary gangs and gang-related crimes are less
driven by formal membership. Thus, inconsistencies in reported gang crime are
more attributable to a blurring of boundaries between gang members and non-
gang members. This fundamental change in gang structure and purpose requires
examination in order to determine if traditional distinctions between the crimes of

gang members and non-gang members are still valid.

117

“Summary of Literature Review”

This study asserts that the contemporary gang transformation diminishes the
importance of formal membership and thus requires a re-conceptualization of the
gang phenomenon. This evolution of the gang is inspired by a current sociological
trend associated with globalization called the networked enterprise. Castells
(2000) explains that the networked enterprise creates a system where intersecting
segments are both dependent and autonomous at the same time. Under a system of
networked enterprises, gangs are becoming organizations with ﬂuctuating
memberships and fewer permanent associations. The lack of stable association
should not be misconstrued as weak associations.

Not only have some scholars (Hobbs, 2001) begun to study this gang
transformation but others (Hardt and Negri, 2004; Castells, 2000) have explained
how the general social transformation accrues to the criminal element in society.
Still, other researchers (McCusker, 2004; Williams, 2005) have demonstrated
increased networking of some well-known gangs. This structural and functional
transformation requires a theoretical foundation. As a theoretical foundation,
Sanchez-Jankowski (2003) provides an explanation for the re-conceptualization of
gangs. Sanchez-Jankowski (2003) posits that gangs have become agglomerations
of individuals who exhibit the deﬁant individualist personality trait.

Individuals who display this personality type seek to attain socially
prescribed goals by any means available. This goal-oriented pursuit is somewhat
impeded by a lack of resources with which to legally attain the desired outcomes;

therefore, the deﬁant individualist resorts to illegal means (Sanchez-Jankowski,

118

2003). In addition to the disintermediation of laws and social norms, the deﬁant
individualist is willing to go to any lengths to prevent the disruption of his or her
goal pursuit. The inherent resistant qualities of the new paradigm makes
intervention efforts that much more difﬁcult. Under this new paradigm, gangs
have become groups of organized deﬁant individualists who come together to
undertake criminal enterprises, which are structured as networked enterprises.
Hence, the need for formal membership no longer exists, since the association is
transient by nature.

Organized deﬁant individualism is a radical departure ﬁ'om the traditional
conceptualization of gangs. Traditionally, gangs have been viewed as socially
problematic due to the group hazard effect. There are two different concepts that
combine to form the group hazard effect: Dentler and Erikson’s (1959) group
delinquency hypothesis and Erickson’s (1973) group hazard hypothesis. The
group hazard hypothesis posits that group deviance is perceived as a greater threat
to society and therefore draws more attention ﬁ'om ofﬁcial social control agents
(Erickson, 1973-b). The group deviance hypothesis posits that groups tend to
induce, sustain and permit deviance (Dentler and Erikson, 1959). Together these
two concepts form a group hazard effect which views gangs as workshops of
deviance. Conversely, the organized deﬁant individualism hypothesis views the
gang as a tool, rather than the workshop. Gangs, under the deﬁant individualist
conceptualization, also create a problem for individuals seeking desistance from a

criminal lifestyle.

119

Gang desistance entails not only the desistance of behaviors but also the
defection from a culture. The speciﬁc act of an individual disassociating with a
gang is typically insufﬁcient to promote the type of long-term lifestyle change
necessary to insure the continued success of the individual and guard against the
possibility of recidivism. With respect to the organized deﬁant individualism, the
individual must alter his or her personality traits that promote the existence of the
gang. Under the deﬁant individualism conceptualization, gang membership is
more representative of a personal pathology than a socially facilitated pathology.
Promoting desistence may be much more difﬁcult for individuals who display a
deﬁant individualist personality type. Understanding contemporary gangs as
organized deﬁant individualism is a stark contrast to the traditional
conceptualization.

Traditionally, gangs are conceptualized as a group hazard. The group
hazard effect is a combination of two similar but different concepts: the groups
hazard hypothesis and the group delinquency hypothesis. Erickson’s (1973) group
hazard hypothesis states that violating the law in groups is more likely to ensure
detection and ofﬁcial reaction than individual crime. The group hazard hypothesis
could be attributed simply to the fact that it is more difﬁcult for groups to evade
detection than for an individual to escape detection (Erickson, 1973). The group
delinquency hypothesis is the second component of the traditional gang
conceptualization.

Dentler and Erikson (1959) proposed three propositions that sought to

explain the aggregate dynamics of deviance: groups induce and sustain deviance,

120

deviance maintains group equilibrium and groups resist alienation of a member
whose behavior deviates from the group standards. It is clear by the ﬁrst
proposition that deviance is viewed traditionally as a group pathology.
Additionally, the group is viewed as serving a maintenance role in deviance as
evidenced by the second and third propositions. Taken together, the group hazard
hypothesis and the group delinquency hypothesis construct a conceptualization of
gangs that places a great deal of focus on the group dynamics. With the alternate
conceptualization in place, this study proceeds to determine whether criminal
behaviors discriminate between gang members, individuals who display deﬁant
individualist personality traits and individuals with no gang inﬂuence.

Based on the literature, this study proposes ﬁve hypotheses. Hypothesis one
suggests that there is no difference in criminal justice contact between gang
members and deﬁant individualists. Under the traditional group hazard
conceptualization, deviance is maintained and promoted through membership in
the gang. A large number of studies have shown the statistically signiﬁcant '
difference in criminal behaviors between gang members and non- gang members.
However, if the organized deﬁant individualism conceptualization is valid, there
should not be any criminological differences between gang members and non-
gang members who display deﬁant individualism.

The second study hypothesis posits that there is a direct relationship
between the number of criminal charges accumulated and deﬁant individualism.
Over time, the deﬁant individualist may become more criminally oriented as his

or her legitimate opportunities are reduced even further by early criminal

121

offenses. The deﬁant individualist’s reliance on criminal pursuits should logically
increase as the commitment to the personality type increases.

The third hypothesis holds that there is no difference in the attitude toward
gun use between deﬁant individualists and respondents who display no organized
criminal inﬂuence. Since it is posited that deﬁant individualists are not
pathological in nature, but rather rational, it should logically follow that their
attitude toward gun use is much more utilitarian in nature. Deﬁant individualists
should not have a predisposition toward using guns more than individuals without
the deﬁant individualist personality.

The fourth hypothesis suggests that there is an inverse relationship between
deﬁant individualism and positive trajectory shifts. A deﬁant individualist should
remain unaffected by criminological interventions, such as lever-pulling, due to
the nature of the criminal behavior being the deﬁant individualist personality type.
Lever-pulling is a focused deterrence strategy that is based on multiple
characteristics of and responses to offending (McGarrell et al., 2006). During the
lever-pulling program, a multi-agency work group of criminal justice
professionals identify and target habitual offenders who are required to attend
notiﬁcation meetings (Ibid, 2006).

At these meetings, habitual offenders are advised that they will face
signiﬁcant criminal justice sanctions if the offenders do not stop engaging in
certain criminal behaviors such as gun violence. During the meeting, offenders
are provided with networking opportunities (McGarrell et al., 2006). Lever-

pulling attempts to promote criminal desistance in this way. Altering an

122

individual’s personality requires more in-depth and personalized intervention than
is often possible in the criminal justice system.

The ﬁnal study hypothesis is a null hypothesis offering no difference in the
perception of gang criminality between gang members and deﬁant individualists.
Due to the close association of these two groups, the perceptions about gang
criminality should be similar.

“Summary of Methods”

This study uses secondary data to test these hypotheses. The data are part of
a research study funded by grant # 2003-U-CX-1038 from the National Institute
of Justice. The purpose of the grant is to evaluate the Indianapolis lever pulling
intervention. The data set includes both interview data and respondents’ criminal
histories. The study participants were comprised of every felony probationer in
the Indianapolis probation system. The probationers had to meet several criteria
prior to selection for the study. The probationers had to be actively on probation
for a felony offense and that offense had to be speciﬁcally a drug offense, violent
crime weapon offense or a property offense. A study sample was drawn from
consecutive sub samples of 1,000 probationers which were supplied each month.
There were a total of six different pools of probationers. Each of the sample pools
was randomly assigned to one of the three groups: law enforcement meeting,
community meeting or control group.

After the selection and randomization process, the study contained 540
probationers with 180 participants per group. Despite this preliminary study

count, the ﬁnal sample consisted of 235 participants. Ineffective notiﬁcation,

123

transportation problems and non-compliance with active probation requirements
were all reasons for the attenuation of the sample.

The dataset is comprised of 387 data points. Of the total data points, 195
were analyzed. Study variables that were rejected had large proportions of
missing data. For example, data on whether the respondent bought a gun under his
or her own name had 40% missing data, how often the respondents ﬁred guns had
99.1% missing data and how many times the respondents carried a gun outside of
the home had 99.1% missing data. Twenty-three study variables which best ﬁt the
hypotheses were selected. Nineteen of the study variables were continuous in
nature to allow for more sophisticated analysis.

The dependent variables used in this analysis were divided into two classes:
primary and secondary criminogenics. Primary criminogenics were
conceptualized as ofﬁcial counts of a participant’s criminal activity such as
number of arrests, number of violent convictions, number of property convictions,
number of times on probation, number of times in the department on corrections,
number of days in the department of corrections and the number of criminal
charges. Operationalization of the primary criminogenics was achieved by using
data from the respondents’ official criminal histories. In addition to the primary
criminogenics, the study also examined secondary criminogenics.

Secondary criminogenics are composite measures that assess the
participant’s non-criminal attitudes or behaviors toward various concepts. The
study uses ﬁve secondary criminogenic variables. Attitude toward gun use is

conceptualized as the degree to which the respondent has a more of less favorable

124

view of using a gun in conﬂict situations. This composite measure contains 6
items. The items are operationalized on a ﬁve point Likert scale which produces a
range from 6, having a less accepting attitude toward gun use, to 24, having the
most accepting attitude toward gun use. This scale produces the lowest reliability
coefﬁcient of any in the study (alpha= .578). A factor analysis reveals that all of
the items load with an Eigen value of at least .443.

Perception of gang criminality is the next secondary criminogenic variable.
This variable is conceptualized as the degree to which the respondent believes
criminal behavior is important to gang members. This variable is a composite
measure that contains six items which were operationalized on a ﬁve-point Likert
scale. This scale produced a range from 6, having a perception of low criminal
importance to gangs, to 30, having a high perception of criminal importance to
gangs. The scale produced a reliability coefficient of .845. A factor analysis
revealed that all of the items loaded with Eigen values of at least .691.

Positive trajectory shift is conceptualized as the degree to which the
respondent participated in post intervention, pro-social social behavior. The
variable is a composite measure that originally contains seven items. This variable
was operationalized on a Guttrnan scale which produced a range ﬁ'om 0, making
no positive trajectory shift, to 7, making the most positive trajectory shift. This
scale produced a reliability coefﬁcient of .602, which was the second lowest in
the study. A factor analysis of the scale revealed that all of the items loaded with

an Eigen value of at least .473.

125

The fourth secondary criminogenic variable is post intervention networking.
This variable was conceptualized as the degree to which the respondent contacted
community-based supporters. The variable was a composite measure that
contained 6 items operationalized using a Guttrnan scale, which is composed of
dichotomous items. The scale produced a reliability coefficient of .714. A factor
analysis revealed that all the items loaded with an Eigen value of at least .499.
This scale had a range of O, participating in no post-intervention networking, to 6,
participating in a high degree of post-intervention networking.

The ﬁnal secondary criminogenic variable is intervention recall. This
variable was conceptualized as the degree to which the respondent remembered
elements of the intervention meeting. These items were operationalized using a
Guttrnan scale containing eight dichotomous items. This scale produced a range
ﬁ'om 0, having no recall of the meetings, to 8, having total recall. This scale
produced a reliability coefﬁcient of .906. A factor analysis revealed that all of the
items loaded with an Eigen value of at least .714.

This study also uses two classiﬁcation variables: deﬁant individualism
score and gang membership, which serve as the dependent variables in the
analysis. Deﬁant individualism score (DIS) is a composite variable that is
conceptualized as the degree to which an individual displays the deﬁant
individualist personality. The scale is composed of the following items: 1) Have
you ever been a member of a gang, 2) Have you ever been a member of a group,
3) Have you ever thought of joining a gang, 4) Have you ever been recruited or

pressured to join a gang, 5) Have you ever hung out with gang members, 6) Have

126

you ever drunk alcohol or gotten high with gang members, 7) Have you ever
vandalized something with a gang member, 8) Have you ever stolen something
with a gang member, 9) Have you ever been attacked in a gang-related incident,
10) Have you ever attacked someone in a gang-related incident and 11) Do you
have friends that are gang members. The items on the scale were operationalized
as: yes= 1 and no =0. This scale produced a reliability coefﬁcient of .845. A
factor analysis revealed that all of the items loaded with an Eigen value of at least
.401 . The range of the scale was from 0, having no commitment to deﬁant
individualism, to 11, having high commitment to deﬁant individualism.

The gang membership variable is conceptualized as whether or not the
respondent was a member of a gang at the time of the interview. This variable was
operationalized as: 1= yes and 0= no and is mutually exclusive with the deﬁant
individualism score. The study proceeds to examine the data at univariate and
multivariate methods including Pearson’s Correlation and Discriminant Function
Analysis. .

A discriminant ﬁmction analysis is a statistic used to test whether or not a
group of variables signiﬁcantly discriminate between two or more groups, thereby
producing a latent ﬁrnction. The discriminant ﬁmction analysis allows the
researcher to determine if the overall model discriminates between the various
dependent groups as well as which individual component of the model contributes
most to the differentiation. One of the most important statistics in discriminant

function analysis is the Wilks’ Lambda.

127

The Wilks’ Lambda ranges ﬁom zero to one and functions as an F test of
signiﬁcance. If the model attains statistical signiﬁcance then each individual
variable is assessed in order to determine which variable differs signiﬁcantly by
group. A Wilks’ Lambda of zero (0) is interpreted as the group means differ and
the groups are therefore different from one another. However, a Wilks’ Lambda
of one (1) suggests that the group means do not differ and the two groups are
more similar to one another. Using measures of central tendency, bivariate
correlation and discriminant function analysis, this study proceeds with the
analysis and hypothesis testing.

“Summary of Findings”

The demographic analysis shows that the overall sample contains more
blacks than any other racial minority. This representation of black respondents is
disproportionate to the representation of blacks in the population of Indianapolis,
Indiana. Males are also over represented in the sample. The general population of
Indianapolis is almost evenly distributed between males and females but the
sample was heavily male. Despite the average age of the sample respondent being
31 years of age (median age=30 years) a large percentage of the sample is not
married.

Respondents in the sample also have relatively moderate education levels.
The majority (84.2%) of the sample has 12 years of education or less. Despite the
relatively low education level, the employment statistics are high. The sample
shows only a small number (1.2%) of respondents as being unemployed at the

time of the survey. Additionally, approximately half (n= 49.4%) of the sample

128

was employed ﬁrll-time at the time of the survey. The univariate examination also
extends to the study variables.

A univariate examination of the primary criminogenic variables shows that
gang members had consistently higher average offenses than other groups. These
higher average offenses can be seen in the number of arrests, the number of
charges, the number of violent convictions, the number property convictions, the
number of times on probation and the number of times in the department of
corrections. An analysis of the secondary criminogenics reveals similar patterns to
those of the primary criminogenics. The attitude toward gun use shows that the
sample is generally less accepting of gun use. However, gang members are more
accepting of gun use than other groups in the sample. Interestingly, gang members
perceive crime as being less important to gang members than either of the other
two groups in the sample.

The amount of positive trajectory shift found in the sample is relatively low.
Generally, respondents in the sample did not experience a great deal of lifestyle
alteration after participating in the lever-pulling program. However, members of
the deﬁant individualist group display the most positive trajectory shift. Similarly,
the sample shows low levels of post intervention networking. The respondents
generally did not contact community-based supporters very much. The zero-
inﬂuence group displays the least post intervention networking among the study
groups. Despite displaying the least post intervention networking, the zero-
inﬂuenced group had the greatest post intervention recall of the three groups. The

study also includes bivariate correlations designed to determine how intuitive

129

criminal relationships vary across the zero-inﬂuence, deﬁant individualist and
gang members groups.

Among the bivariate con'elations, the primary criminogenic variables (the
number of arrests, the number of charges, the number of violent convictions, the
number property convictions, the number of times on probation and the number of
times in the department of corrections) produced 18 statistically signiﬁcant
relationships. These primary criminogenic relationships are intuitive in nature,
such as the statistically signiﬁcant relationship between the number of arrests and
the number of charges incurred. As the numbers of arrests increase, so do the
number of criminal charges. This correlation displays maintenance of criminal
lifestyles.

Unlike the primary criminogenic correlation matrix, the secondary
crinrinogenic correlation matrix is not intuitive. Because the secondary correlation
matrices correlated non-criminal attitudes and behaviors, the relationships are not
as predictable. The secondary criminogenic matrices show only two statistically
signiﬁcant relationships. There is a positive correlation between the amount of
post intervention recall and the amount of post intervention networking. This
ﬁnding suggests that respondents who remembered more components of the
lever-pulling program contacted more community-based supporters. The zero-
inﬂuence group, the deﬁant individualist group and the gang member group all
had the same positive correlation between the amount of post intervention recall

and the amount of post intervention networking. Perhaps the most noticeable

130

correlations can be seen in the primary and secondary matrix of the gang member
group.

The primary and secondary matrix of the gang member group produces four
signiﬁcant gun-related correlations. Attitude toward gun use, which measures
whether or not the respondent is more or less favorable of using a gun in conﬂicts,
was not signiﬁcantly correlated with any other variable in any other groups (zero-
inﬂuence group or deﬁant individualist group) other than among the gang
member group. First, there is a strong positive correlation between the attitude
toward gun use and the number of property convictions (r= .705, p= .002). As the
number of property convictions increases, the attitude toward gun use becomes
more favorable.

The attitude toward gun use is also strongly correlated with the number of
arrests (r= .563, p= .023) and the number of charges (r= .549, p= .028). These
statistics suggest that as the number of charges and arrests increase, the attitude
toward gun use becomes more favorable. Gang members in the study appear to
have a much stronger commitment to gun use than either the deﬁant individualist
or zero-inﬂuence groups, as evidenced by more variables being signiﬁcantly
correlated with attitudes toward gun use. This favorable attitude toward gun use
may explain the traditionally higher rates of violence within the gang sub-culture.
The only negative gun-related correlation is the relationship between the attitude
toward gun use and the amount of post intervention networking (r= -.610,
p= .046). Therefore, gang members who have more favorable views of gun use

engage in less contact with community-based supporters. This ﬁnding suggests

131

that gang members are somewhat resistant to desisting deviant attitudes that may
impact their criminal behavior. In addition to bivariate correlations, this study also
uses discriminant function analysis to examine whether or not there are
discemable differences between the zero-inﬂuence group, the deﬁant individualist
group and the gang member group and test the study hypotheses.

“Discussion of Study Hypotheses”

The ﬁrst study hypothesis states that there is no difference in the criminal
justice system contact between gang members and deﬁant individualists, which
would constitute a discriminate function. Based on the analysis in chapter four,
this hypothesis is supported. Table-22 displays the discriminant function analysis
of primary criminogenics between gang members and deﬁant individualists. The
model is not statistically signiﬁcant (A= .961, p= .283).

The primary criminogenic model examines ﬁve variables that are direct
elements of contact with the criminal justice system: number of arrests, number of
violent convictions, number of property convictions, number of times on
probation and length of time in the department of corrections. It is necessary to
include measures from every aspect of the criminal justice system in order to
produce an adequate picture of the overall contact with the criminal justice
system. The number of arrests assesses the law enforcement or primary contact
with the criminal justice system. Number of violent convictions and number of
property convictions assesses the variation in the judicial contact between the two

groups. The model also includes the number of times on probation and the length

132

of time in the department of corrections as an assessment of the correctional
aspects of the criminal justice system.

If any one of the variables in the primary criminogenic model is found to be
signiﬁcant, the Wilks’ Lambda for the entire model will attain statistical
signiﬁcance. The failure of the primary criminogenic model to discriminate
between gang members and deﬁant individualists suggests that these two groups
have similar criminal justice contact. As previously discussed in chapter two, it is
possible for deﬁant individualists to be misidentiﬁed as gang members, due to
their relatively close associations and criminal complacency in gang-related
crime.

The second hypothesis states that that there is a direct relationship between
the number of criminal charges a respondent incurred and the degree of deﬁant
individualism. Based on the analysis in chapter four, this hypothesis is not
supported. Table-6 shows that the relationship between deﬁant individualism and
the number of criminal charges incurred is not statistically signiﬁcant (r=. 03 9,
p=569). Even when controlling for the effects of age, the relationship fails to
attain signiﬁcance (r= -.025, p= .809).

According to Life Course Theory, these two variables should be directly
correlated. Life Course Theory posits that as individuals engage in criminal life
styles they ﬁnd it more difﬁcult to desist criminal activities, due to antisocial
decisions and behaviors made earlier in life. This linear function is called
Homotypic Continuity (Sampson and Laub, 1992). There is a plausible

explanation for this unexpected ﬁnding. Game Theory suggests that as an

133

individual repeatedly engages in any given activity he or she will discover which
tactics best provide the desired outcome. As the deﬁant individualist maintains a
criminal lifestyle, he or she may develop better technique for committing crimes
or avoiding detection. The deﬁant individualist may also develop larger more
advanced criminal networks with which to better avoid detection.

The third study hypothesis states that there is no difference in the attitude
toward gun use between the deﬁant individualist and the zero-inﬂuence group,
which could constitute a discriminant ﬁrnction. The analysis in chapter four
suggests that this hypothesis is supported. Table-21 shows that there is no
secondary criminogenic discriminant function between the zero inﬂuence and the
deﬁant individualist group (A= .941, p= .242).

This model contains not only a measure of the respondents’ attitudes toward
gun use but also their perception of gang crime, the respondents post intervention
networking, positive trajectory shift and post intervention recall. All of these
variables are facets of secondary behaviors that might impact the primary criminal
behavior. Respondents post intervention networking, positive trajectory shift and
post intervention recall speciﬁcally examine whether or not the respondents
transitioned toward prosocial activities. The model’s failure to attain statistical
signiﬁcance suggests that the both respondents in the zero-inﬂuence group and
the deﬁant individualists shared similar attitudes toward gun use and positive
transitions.

The fourth study hypothesis states that there is an inverse relationship

between deﬁant individualism and the amount of positive trajectory shift.

134

Therefore, the study supposes that respondents with higher deﬁant individualist
scores will engage in less pro-social behaviors after participating in the lever-
pulling intervention. Based on the preceding analysis, this hypothesis is not
supported but the relationship does attain statistical signiﬁcance. The relationship
between the deﬁant individualism and positive trajectory shift is positively
correlated (r= .200, p= .003). This positive correlation is also stable when
controlling for the effects of age (r= .209, p= .035).

It is possible that individuals who have greater deﬁant individualistic traits
require the prosocial opportunities provided by the lever pulling intervention and
take greater advantage of these services. This statistic is compatible with the
concept of homotypic continuity. The longer the individual maintains the
antisocial lifestyle, the greater the desire may be to make a prosocial transition.
This intervention program perhaps may be better suited to people with higher
levels of deﬁant individualism and not as a blanket program for all offenders.

The ﬁfth hypothesis states that there is no difference in the perception of
gang criminality between gang members and deﬁant individualists, which could
constitute a discriminant function. Based on the analysis in chapter four this
hypothesis is supported. T able-23 shows that the model of secondary
crirrrinogenics does not represent a discriminant function between the deﬁant
individualist and gang member groups (A= .941, p= .242).

The model in Table-23 is similar to the model in Table-22, in that it not
only contains measures of the respondents’ attitudes toward gun use but also their

perception of gang crime, the respondents post intervention networking as

135

evidenced by contact with community-based supporters and positive trajectory
shift, which entails engaging in more prosocial behaviors and post intervention
recall. Respondents’ post intervention networking, positive trajectory shift and
post intervention recall speciﬁcally examined whether or not the respondents
transitioned toward prosocial activities. The model’s failure to attain statistical
signiﬁcance suggests that the respondents in the deﬁant individualist group and
the gang member group shared similar secondary criminogenic attitudes and
behaviors.

“Study Limitations”

Despite generating several supported hypotheses, this study has three
limitations that should be addressed in subsequent replications. There are
limitations due to sample size, index reliability and the measure of deﬁant
individualism. While these limitations are not serious enough to render the study
invalid, the study could produce a more useful replication by strengthening these
areas.

The sample size in this study is relatively small (N= 23 5) compared to many
sample sizes in the criminal justice ﬁeld. This smaller than normal sample size
also produces unequal numbers of respondents in the three study groups. It is
important to remember that these data are not speciﬁcally collected for this study.
In order to produce ﬁndings that could be generalized to the population at large,
the sample size would need to be in the range of 1,000 to 1,500 cases. This would

provide an appropriate statistical power for extrapolating the ﬁndings to larger

136

groups. The sample size is not overly problematic, due to the exploratory nature
of this study.

There is also a limitation created by two of the secondary criminogenic
indices. The scale measuring the respondents’ attitudes toward gun use has the
lowest reliability coefficient of all the variables in the study (n= .578). It is easy to
understand why respondents may have succumbed to the Hawthorne Effect in
completing these questions. The Hawthome Effect occurs when a research
participant gives the response he or she thinks the researcher expects based on the
knowledge of the research project. Since these study respondents are probationers,
it is easy to see how they might have given answers they perceived as the ‘right’
answer or failed to complete the gun use-related questions altogether.

The second scale that contributes to the limitation was the positive
trajectory shift scale, which also produced a relatively low reliability coefﬁcient
(n= .602). It is possible that the respondents were suffering from the Hawthorne
Effect when completing these questions also. If the respondents who were on
probation had not taken full advantage of the opportunities to network in positive
ways, it is possible that they may have not answered the questions in this section
or answered in sporadic illogical patterns.

The measure of deﬁant individualism by a proxy scale also contributes to
the limitations of the study. The scale which measures deﬁant individualism is a
continuum that assesses closeness to gang members on a number of issues. The
logic of the scale is based on the Morash (1983) study in which groups were

assessed as more or less ‘gang like’. The criminological variations of these groups

137

are then analyzed in order to determine similarities to gang behavior. The study’s
deﬁant individualism scale essentially measures closeness to gangs.

Based on the theoretical assertion by Sanchez-Jankowski (2003), that
almost all gang members have deﬁant individualist personality traits, the deﬁant
individualism index in this study measures closeness to gang members as a proxy
for increased deﬁant individualism. In subsequent studies, researchers would need
to develop an individual index that measures deﬁant individualism as an isolated
trait not as a proxy. Perhaps a scale developed around the diagnostic model for
Oppositional Deﬁant Disorder would be more beneﬁcial. Oppositional Deﬁant
Disorder is a mental disorder seen in juveniles and is characterized by rebellion to
authority ﬁgures in the pursuit of personally valued goals.

“Recommendations”

Based on the preceding study ﬁndings and limitations, the following
recommendations are presented. First, a replication study is recommended. The
support for three of the ﬁve study hypotheses (and a fourth statistically signiﬁcant
hypothesis) justiﬁes additional inquiry. This study has demonstrated that neither
primary nor secondary criminogenics perform a discriminant ﬁrnction for gang
members versus deﬁant individualists. This inability to distinguish between gang
members and non-member deﬁant individualists suggests that more study is
needed to determine what other variables might be used to differentiate between
these two classiﬁcations of offenders. If it is not possible to differentiate the two

groups, subsequent research must measure the groups as similar in sociological

138

threat. Additionally, if a discourse of differentiation cannot be identiﬁed, the
reliance on formal membership as an inclusionary criterion should be abandoned.
Additional research is also needed to better understand the characteristics of
deﬁant individualism as it pertains to criminology. There are many questions that
this new criminal conceptualization raises. For example, how is the personality
developed among criminals, is it more prevalent in any given population, is it
debilitating or does it allow its subjects a degree of functionality? Most
importantly, can the personality trait be reversed or de-criminalized? All of these
questions need to be subjected to scientiﬁc rigor. It may be possible to glean
information from existing studies in other disciplines that address similar
concepts. This study provides a foundation from which to proceed in identifying
the salient characteristics of deﬁant individualism as it pertains to criminology.
Another recommendation is that subsequent studies of deﬁant individualism
develop and use a deﬁant individualism index. Developing an independent deﬁant
individualism scale provides for much stronger reliability and validity of
subsequent studies. Perhaps, the deﬁant individualism index could be created by a
more thorough examination of the oppositional deﬁant disorder literature.
Oppositional deﬁant disorder is a psychological disorder, diagnosed in children,
which has many of the same characteristics as deﬁant individualism. Although
oppositional deﬁant disorder is most commonly diagnosed in children, the
diagnostic tools may help to inform a deﬁant individualism index, which could be

used for adults.

139

A ﬁnal recommendation is for a larger sample size in subsequent
replications. Future studies should begin with an a priori power analysis which is
designed to reveal the sample size needed to be able to generalize ﬁnding to larger
populations.

“Conclusion”

This study shows that the contact with the criminal justice system does not
discriminate between gang members and deﬁant individualists. The gang member
versus non-gang member differentiations produced under the group hazard
models can be partially diminished by more accurately measuring all groups
involved in the emerging structure of the criminal culture. Deﬁant individualists
have the ability to maintain criminal lifestyles just as gang members do without
the need or burden of stable gang membership. Despite the criminal activities of
people with deﬁant individualism, there is little that is known about the
personality trait.

The study ﬁndings demonstrate that. the actions of deﬁant individualists are
not always readily predicted. This direct correlation between deﬁant
individualism and positive trajectory shift is somewhat antithetical to an expected
outcome. However, the ﬁnding suggests that there is hope for this emerging
threat. This ﬁnding suggests that perhaps we should not scrap the lever pulling
intervention strategies just yet. Perhaps, through this exploratory analysis we have
discovered that the intervention is useful when properly targeted.

Traditional conceptualizations about gangs and gang-related crime have not

kept pace with the criminal world. This study has shown that gang membership

140

does not necessarily delineate the greater threat and should not be considered the
criterion that deﬁnes criminal careers. As society in general continues to change
and take full advantage of social shifts, so do gangs. It is imperative that
criminologists not only adapt to these changes but that we understand and
anticipate them. It is the ability to anticipate crime that is the ﬁrst responsibility of

the criminologist.

141

Attachment

“Code Book”

List of variables on the working ﬁle

Name

GROUP Group
Measurement Level: Scale
Value Label

1 Law Enforcement
2 Community
3 Control

BADD
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

GENDER
Measurement Level: Scale
Value Label

.00 Female
1.00 Male

ATTEND
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

STAADD Status when getting address

Measurement Level: Scale
Value Label

1.00 Meeting Date
2.00 Discharged
3.00 Revoked

4.00 Absconded

142

Position

5.00 Transfer
6.00 Incarcerated
7.00 TRV

MCJA Interviewed at MCJA
Measurement Level: Scale »
Value Label

.00 No
1.00 Yes

CHEARD Heard of the LP meetings (control)
Measurement Level: Scale
Value Label

1.00 Yes
2.00 No

CWHOM Heard of LP meetings from whom?
Measurement Level: Scale
Value Label

1.00 Friends

2.00 Family

3.00 Probation

4.00 Law Enforcement
5.00 Community Leaders
6.00 Television/Radio

REMEMl Remember law enforcement cracking down
on violent crime
Measurement Level: Scale
Value Label

0 False
1 True

REMEM2 Remember law enforcement cracking
down on gun crime
Measurement Level: Scale
Value Label

0 False
1 True

143

REMEM3 Remember can go to federal prison is carry gun
Measurement Level: Scale
Value Label

0 False
1 True

REMEM4 Remember probation is watching behavior closely
Measurement Level: Scale
Value Label

0 False
1 True

REMEMS Remember law enforcement wants to make choices
Measurement Level: Scale
Value Label

0 False
1 True

REMEM6 Remember community leaders have opportunities
for you to get
Measurement Level: Scale
Value Label

0 False
1 True

REMEM7 Remember community leaders are willing to help
you in any way
Measurement Level: Scale

Value Label
0 False
1 True

REMEM8 Remember should stay out of trouble
Measurement Level: Scale
Value Label

0 False
1 True

144

ll

12

13

14

15

16

 

AFTl

AFI’Z

AFT3

AFT4

AFTS

AFT6

After the meeting did you talk with family
Measurement Level: Scale
Value Label

0 False
1 True

After the meeting did you talk with spouse,
girl/boy friend

Measurement Level: Scale

Value Label

0 False
1 True

After the meeting did you talk with friends
Measurement Level: Scale
Value Label

0 False
1 True

After the meeting did you talk with coworkers
Measurement Level: Scale
Value Label

0 False
1 True

After the meeting did you talk with neighbors
Measurement Level: Scale
Value Label

0 False
1 True

After the meeting did you talk with probation officers
Measurement Level: Scale
Value Label

0 False
1 True

145

17

18

19

20

21

22

SINCE]

SINCEZ

SINCE3

SINCE4

SINCES

SIN CE6

Gotten a job or job training
Measurement Level: Scale
Value Label

0 False
1 True

Gone back to school
Measurement Level: Scale
Value Label

0 False
1 True

Entered treatment
Measurement Level: Scale
Value Label

0 False
1 True
Going to church

Measurement Level: Scale
Value Label

0 False
1 True

Attending counseling
Measurement Level: Scale
Value Label

0 False
1 True

Missed meetings with probation
Measurement Level: Scale
Value Label

0 False
1 True

146

23

24

25

26

27

28

SIN CE7 Contacted law enforcement
Measurement Level: Scale
Value Label

0 False
1 True

SIN CE8 Contacted community leaders
Measurement Level: Scale
Value Label

0 False
1 True

SIN CE9 Contacted community organization
Measurement Level: Scale
Value Label

0 False
1 True

SINCEIO Asked probation officer for help
Measurement Level: Scale
Value Label

0 False
1 True

CHOICES Better choices because I attend the meetings
Measurement Level: Scale
Value Label

1 Strongly Agree
2 Agree

3 Disagree
4 Strongly Disagree
5 M Don't Know

147

29

30

31

32

33

PROMC Promise to crack down on gun crime

Measurement Level: Scale
Value Label

1 Strongly Agree

2 Agree

3 Disagree

4 Strongly Disagree
5 M Don't Know

PROMG Promise to Send to federal prison

Measurement Level: Scale
Value Label

1 Strongly Agree

2 Agree

3 Disagree

4 Strongly Disagree
5 M Don't Know

PROMISE Law enforcements agencies have followed

CLOPP

through on their pro
Measurement Level: Scale
Value Label

Strongly Agree
Agree

Disagree
Strongly Disagree
5 M Don't Know

smut—

Community leaders were willing to help me ﬁnd
opportunities
Measurement Level: Scale
Value Label

1 Strongly Agree

2 Agree

3 Disagree

4 Strongly Disagree
5 M Don't Know

148

34

35

36

CLPROM Community leaders followed through on their

THINK

HELP

promises
Measurement Level: Scale
Value Label

1 Strongly Agree

2 Agree

3 Disagree

4 Strongly Disagree
5 M Don't Know

Often think about the meetings
Measurement Level: Scale
Value Label

1 Very Frequently

2 Frequently

3 Somewhat Frequently
4 Not at all

Helpful were the meetings
Measurement Level: Scale
Value Label

Very Helpful
Helpful
Somewhat Helpful
Not at all Helpful

AWNI—

CONTINUE Meetings should continue

Measurement Level: Scale
Value Label

Strongly Agree
Agree

Disagree
Strongly Disagree
5 M Don't Know

Ale-t

DISCOUR Discourage people from breaking the law

Measurement Level: Scale
Value Label

1 Strongly Agree

2 Agree
3 Disagree

149

39

40

41

42

SCARE

TIME

JOB

4 Strongly Disagree
5 M Don't Know

Just scare tactics
Measurement Level: Scale
Value Label

1 Strongly Agree

2 Agree

3 Disagree

4 Strongly Disagree
5 M Don't Know

Do not have the time or money to follow through
Measurement Level: Scale
Value Label

1 Strongly Agree

2 Agree

3 Disagree

4 Strongly Disagree
5 M Don't Know

Can't provide me with a job where I can make money

Measurement Level: Scale
Value Label

Strongly Agree
Agree

Disagree

Strongly Disagree
5 M Don't Know

AWN-s

WATCH Watching probationers more now than before

Measurement Level: Scale
Value Label

1 Strongly Agree
2 Agree

3 Disagree
4 Strongly Disagree
5 M Don't Know

150

43

44

45

46

GETOUT Difﬁcult for arresttees to get out of the system
Measurement Level: Scale
Value Label

1 Strongly Agree
2 Agree

3 Disagree
4 Strongly Disagree
5 M Don't Know

BREAK Less likely to break the law because of the message
Measurement Level: Scale
Value Label

Strongly Agree
Agree

Disagree

Strongly Disagree
5 M Don't Know

rAMNl-t

HEARDO Heard other probationers talk about meetings
Measurement Level: Scale
Value Label

1 Strongly Agree
2 Agree

3 Disagree
4 Strongly Disagree
5 M Don't Know

CAREFUL You can get away with a lot of crimes
Measurement Level: Scale
Value Label

1 Strongly Agree
2 Agree

3 Disagree
4 Strongly Disagree
5 M Don't Know

PRYING Law enforcement does not have any business prying
Measurement Level: Scale
Value Label

1 Strongly Agree
2 Agree

151

3 Disagree
4 Strongly Disagree
5 M Don't Know

SERIOUS Law enforcement is serious about responding

to crime
Measurement Level: Scale
Value Label
1 Strongly Agree
2 Agree
3 Disagree

4 Strongly Disagree
5 M Don't Know

SUCCESS Successful you have been in doing crimes

Measurement Level: Scale
Value Label

1 Very Successful

2 Somewhat Successful

3 Somewhat Unsuccessful
4 Very Unsuccessful

MEETO Often meet probation officer at his ofﬁce

Measurement Level: Scale
Value Label

0 Not at all

Less than every month

Every month or almost every month
Every week of almost every week
Several times a week

Everyday or almost every day

UI-laszl-t

MEETH Probation ofﬁcer contact you at home

Measurement Level: Scale
Value Label

Not at all

Less than every month

Every month or almost every month
Every week of almost every week
Several times a week

Everyday or almost every day

MANNHG

152

52

53

54

55

MEETW Probatin officer contact you at work
Measurement Level: Scale
Value Label

TELE

UlbbJNl-sc

Not at all

Less than every month

Every month or almost every month
Every week of almost every week
Several times a week

Everyday or almost every day

Probation officer contact you by telephone
Measurement Level: Scale
Value Label

UIAOJNl—G

Not at all

Less than every month

Every month or almost every month
Every week of almost every week
Several times a week

Everyday or almost every day

CTREAT Participated in court ordered treatment
Measurement Level: Scale
Value Label

SWEEP

UI-kle-Ie

Not at all

Less than every month

Every month or almost every month
Every week of almost every week
Several times a week

Everyday or almost every day

How many times contact during a probation sweep

Measurement Level: Scale
Value Label

MANN-dc

Not at all

Less than every month

Every month or almost every month
Every week of almost every week
Several times a week

Everyday or almost every day

153

56

57

58

59

PCON How many times local police ofﬁcers contacted you
Measurement Level: Scale

FCON How many times federal law enforcement contacted
you
Measurement Level: Scale

PRCON How many times prosecutors contacted you
Measurement Level: Scale

CRCON How many times community representatives
contacted you
Measurement Level: Scale

CLCON How many times clergy contacted you
Measurement Level: Scale

POCON How many times probation contacted you
Measurement Level: Scale

CONFRONT Confronting someone on the street with a gun
Measurement Level: Scale
Value Label

Much more
Somewhat more
About the same
Somewhat less
Much less

UI-BMNI-i

ARRESTG Someone's risk of being arrestted
Measurement Level: Scale
Value Label

Much more
Somewhat more
About the same
Somewhat less
Much less

UI-BMNl-I

GUNPEN Legal penalities for illegally carring a gun
Measurement Level: Scale
Value Label

1 Much more
2 Somewhat more

154

60

61

62

63

65

66

67

68

3
4
5

About the same
Somewhat less
Much less

USEGUN Likely is it that you will use a gun
Measurement Level: Scale
Value Label

cit-amul—

LUSEGUN

Not at all likely
Somewhat unlikely
About the same
Somewhat likely
Very likely

Likelihood that you will use a gun

Measurement Level: Scale
Value Label

MANNH

RARREST

Much more
Somewhat more
About the same
Somewhat less
Much less

Risk of being arrested

Measurement Level: Scale
Value Label

MAle-s

Much more
Somewhat more
About the same
Somewhat less
Much less

RCONVICT Risk of being convicted
Measurement Level: Scale
Value Label

UIAbJNl—I

Much more
Somewhat more
About the same
Somewhat less
Much less

155

69

70

71

72

 

RPRISON Risk of going to prison
Measurement Level: Scale
Value Label

(II-Bub):—

CASE2

Much more
Somewhat more
About the same
Somewhat less
Much less

Measurement Level: Scale

GAR Chances of arrest
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

No Chance

Low Chance

Some Chance
Good Chance

High Chance
Completely Certain

GCON Chances of conviction
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

No Chance

Low Chance

Some Chance

Good Chance

High Chance
Completely Certain

GPRI Chances of going to prison
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

No Chance

Low Chance

Some Chance

Good Chance

High Chance
Completely Certain

156

73

74

75

76

77

 

AGUN Arrest for gun
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

N o Chance

Low Chance

Some Chance

Good Chance

High Chance
Completely Certain

ABURG Arrest for burglary
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

No Chance

Low Chance

Some Chance

Good Chance

High Chance
Completely Certain

ATHEFT Arrest for theft
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

N o Chance

Low Chance

Some Chance

Good Chance

High Chance
Completely Certain

ASDRUGS Arrest for selling drugs
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

APDRUGS Arrest for purchasing drugs

No Chance

Low Chance

Some Chance

Good Chance

High Chance
Completely Certain

Measurement Level: Scale

157

78

79

80

81

82

ACAR

Value Label

1.00 No Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

Arrest for stealing a car
Measurement Level: Scale
Value Label

1.00 N o Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

ABCHECK Arrest for writing a bad check

AROB

Measurement Level: Scale
Value Label

1.00 N o Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

Arrest for robbery
Measurement Level: Scale
Value Label

1.00 No Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

158

83

85

AASSAULT Arrest for assaulting someone

ARAPE

Measurement Level: Scale
Value Label

1.00 N o Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

Arrest for raping someone
Measurement Level: Scale
Value Label

1.00 No Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

AMURDER Arrest for murdering someone

PGUN

Measurement Level: Scale
Value Label

1.00 No Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

Prison for gun
Measurement Level: Scale
Value Label

1.00 No Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

159

86

87

88

89

PBURG Prison for burglary
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

N o Chance

Low Chance

Some Chance

Good Chance

High Chance
Completely Certain

PTHEFI‘ Prison for theft
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

No Chance

Low Chance

Some Chance

Good Chance

High Chance
Completely Certain

PSDRUGS Prison for selling drugs
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

PPDRUGS Prison for purchasing drugs

No Chance

Low Chance

Some Chance

Good Chance

High Chance
Completely Certain

Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

No Chance

Low Chance

Some Chance

Good Chance

High Chance
Completely Certain

160

90

91

92

93

PSCAR Prison for stealing a car
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

No Chance

Low Chance

Some Chance

Good Chance

High Chance
Completely Certain

PBCHECK Prison for writing a bad check
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

No Chance

Low Chance

Some Chance
Good Chance

High Chance
Completely Certain

PROB Prison for robbery
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

No Chance

Low Chance

Some Chance
Good Chance

High Chance
Completely Certain

PASSAULT Prison for assaulting someone
Measurement Level: Scale

Value Label

1.00 N o Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

161

94

95

96

97

PRAPE

Prison for raping someone
Measurement Level: Scale
Value Label

1.00 No Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

PMURDER Prison for murdering someone

SGUN

SBURG

Measurement Level: Scale
Value Label

1.00 No Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

Serious thing for gun
Measurement Level: Scale
Value Label

1.00 Warning
2.00 Arrest

3.00 Fine

4.00 Probation
5.00 Short Prison
6.00 Long Prison

Serious thing for burglary
Measurement Level: Scale
Value Label

1.00 Warning
2.00 Arrest

3.00 Fine

4.00 Probation
5.00 Short Prison
6.00 Long Prison

162

98

99

100

101

STHEFT Serious thing for theft

Measurement Level: Scale
Value Label

1.00 Warning
2.00 Arrest

3.00 Fine

4.00 Probation
5.00 Short Prison
6.00 Long Prison

SSDRUGS Serious thing for selling drugs

Measurement Level: Scale
Value Label

1.00 Warning
2.00 Arrest

3.00 Fine

4.00 Probation
5.00 Short Prison
6.00 Long Prison

SPDRUGS Serious thing for purchasing drugs

SSCAR

Measurement Level: Scale
Value Label

1.00 Warning
2.00 Arrest

3.00 Fine

4.00 Probation
5.00 Short Prison
6.00 Long Prison

Serious thing for stealing a car
Measurement Level: Scale
Value Label

1.00 Warning
2.00 Arrest

3.00 Fine

4.00 Probation
5.00 Short Prison
6.00 Long Prison

163

102

103

104

105

SBCHECK Serious thing for writing a bad check

SROB

Measurement Level: Scale
Value Label

1.00 Warning
2.00 Arrest

3.00 Fine

4.00 Probation
5.00 Short Prison
6.00 Long Prison

Serious thing for robbery
Measurement Level: Scale
Value Label

1.00 Warning
2.00 Arrest

3.00 Fine

4.00 Probation
5.00 Short Prison
6.00 Long Prison

SASSAULT Serious thing for assaulting someone

SRAPE

Measurement Level: Scale
Value Label

1.00 Warning
2.00 Arrest -
3.00 Fine

4.00 Probation
5.00 Short Prison
6.00 Long Prison

Serious thing for raping someone
Measurement Level: Scale
Value Label

1.00 Warning
2.00 Arrest

3.00 Fine

4.00 Probation
5.00 Short Prison
6.00 Long Prison

164

106

107

108

109

SMURDER Serious thing for murdering someone
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

Warning
Arrest

Fine
Probation
Short Prison
Long Prison

FGUN Federal court for gun
Measurement Level: Scale

Value Label

1.00 N o Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

FBURG Federal court for burglary
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00

No Chance
Low Chance
Some Chance
Good Chance
High Chance

6.00 Completely Certain

FSDRUGS Federal court for selling drugs
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

No Chance

Low Chance

Some Chance

Good Chance

High Chance
Completely Certain

165

110

111

112

113

 

 

FPDRUGS Federal court for purchasing drugs

FROB

FRAPE

Measurement Level: Scale
Value Label

1.00 No Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

Federal court for robbery
Measurement Level: Scale
Value Label

1.00 No Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

Federal court for rape
Measurement Level: Scale
Value Label

1.00 No Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

FMURDER Federal court for murder

Measurement Level: Scale
Value Label

1.00 No Chance

2.00 Low Chance

3.00 Some Chance

4.00 Good Chance

5.00 High Chance

6.00 Completely Certain

166

114

115

116

117

FELONGUN Felon legally carry a gun
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

FPENGUN Federal penalties for carrying a gun
Measurement Level: Scale
Value Label

.00 N o
1.00 Yes

HARSHER System has the harsher penalties
Measurement Level: Scale
Value Label

.00 State
1.00 Federal

CONPEN Consider the penalties for carrying a gun
Measurement Level: Scale
Value Label

1.00 Strongly Agree
2.00 Agree

3.00 Disagree

4.00 Strongly Disagree
5.00 M Don't Know

LEARN P Main way you learn about punishmens
Measurement Level: Scale
Value Label

1.00 Personal experience

2.00 Friends

3.00 People who have in the system

4.00 Television, Radio

5.00 'Other advertising, like bus signs, billboards
6.00 Family

7.00 Other

167

118

119

120

121

122

STOPYOU Most important thing that stop you from
using a gun
Measurement Level: Scale
Value Label

1.00 Chances of being arrested

2.00 Chances of going to state prison
3.00 Chances of going to federal prison
4.00 Concerns about your family

5.00 Concerns about your own safety
6.00 How you would be treated in prison

F PENPOS Federal penalty for a felon in possession
Measurement Level: Nominal

FPENUSE Federal penalty for using a gun in a crime
Measurement Level: Nominal

SPENPOS State penalty for a felon in possession
Measurement Level: Nominal

SPENUSE State penalty for a felon in possession
Measurement Level: Nominal

CASE3 Case Number 3
Measurement Level: Scale

TROUBLE Trouble to get gun
Measurement Level: Scale
Value Label

1.00 Almost impossible
2.00 Alot of trouble, but you could do it
3.00 Little of no trouble

EASYGUN Easy for felons to get gun
Measurement Level: Scale
Value Label

1.00 Strongly Agree
2.00 Agree

3.00 Disagree

4.00 Strongly Disagree
5.00 M Don't Know

168

123

124

129

130

131

132

133

134

GETWORSE Situations get worse pulls gun 135
Measurement Level: Scale
Value Label

1.00 Strongly Agree
2.00 Agree

3.00 Disagree

4.00 Strongly Disagree
5.00 M Don't Know

REDUCECO Reduce gun violence in my community 136
Measurement Level: Scale
Value Label

1.00 Strongly Agree
2.00 Agree

3.00 Disagree

4.00 Strongly Disagree
5.00 M Don't Know

STAYOUT Nothing you can do to stay out of a gun ﬁght 137
Measurement Level: Scale
Value Label

1.00 Strongly Agree
2.00 Agree

3.00 Disagree

4.00 Strongly Disagree
5.00 M Don't Know

OKTOSHT Ok to shoot somebody if they're about to hurt 138
or kill you
Measurement Level: Scale
Value Label

1.00 Strongly Agree
2.00 Agree

3.00 Disagree

4.00 Strongly Disagree
5.00 M Don't Know

CARGUN Need to carry a gun in my neighborhood 139
Measurement Level: Scale

Value Label

1.00 Strongly Agree

169

 

2.00 Agree

3.00 Disagree

4.00 Strongly Disagree
5.00 M Don't Know

FRIENDSG Ask my friends to leave their guns at home

when we hang out
Measurement Level: Scale
Value Label

1.00 Strongly Agree
2.00 Agree

3.00 Disagree

4.00 Strongly Disagree
5.00 M Don't Know

HANGING Hanging out with the wrong people

EDUC

Measurement Level: Scale
Value Label

1.00 Strongly Agree
2.00 Agree

3.00 Disagree

4.00 Strongly Disagree
5.00 M Don't Know

Improve my education
Measurement Level: Scale
Value Label

1.00 Strongly Agree
2.00 Agree

3.00 Disagree

4.00 Strongly Disagree
5.00 M Don't Know

NEEDGUN Need a gun, stay at home

Measurement Level: Scale
Value Label

1.00 Strongly Agree

2.00 Agree
3.00 Disagree
4.00 Strongly Disagree

5.00 M Don't Know

170

140

141

142

143

WORTHRIS Carrying a gun is not worth the risk 144
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00

Strongly Agree
Agree

Disagree

Strongly Disagree

5.00 M Don't Know

SCARESOM It alright to have a gun to scare somebody 145
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00

Strongly Agree
Agree

Disagree

Strongly Disagree

5.00 M Don't Know

EASYGGUN Easiest way for a convicted felon to get a gun 146
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

Have someone else purchase it for him

Buy froma gun store using a fake identiﬁcation
Buy a gun from somebody who sells guns illegally
Buy one from somebody he knows

Steal one

Borrow one from a friend

WHYOWN Why do people own guns in your neighborhood 147
Measurement Level: Scale
Value Label

1.00 For protection
2.00 For respect
3.00 For a job
4.00 For committing crimes
5.00 For sport/hunting
GINHOME Guns in your home 148

Measurement Level: Scale
Value Label

.00 No

171

1.00 Yes

GOUTHOME Guns outside the home
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

HMTGOUT How many times did you have guns outside

the home

Measurement Level: Scale

HOGOVT How often did you carry a gun outside the

home in the months

Measurement Level: Scale
Value Label

1.00
2.00
3.00

Everyday or almost everyday
Several times a week
Every week or almost every week

4.00 Less than every week

TIMESGl How many times per week 1
Measurement Level: Scale

TIMESG2 How many times per week 2
Measurement Level: Scale

MONTHG3 How many per month 1
Measurement Level: Scale

MONTHG4 How many per month 2
Measurement Level: Scale

FIREGUN How often did you ﬁre gun
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00

Never

One to ﬁve times
Once/twice month
Once/twice week
Almost everyday

172

149

150

151

152

153

154

155

156

THREATEN How often threatened with gun
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00

Never

One to ﬁve times
Once/twice month
Once/twice week
Almost everyday

SHOTAT How often were you shot at with gun
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00

Never

One to ﬁve times
Once/twice month
Once/twice week
Almost everyday

INJURED How many times injured with a gun
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00

Never

Once

Twice

3-5 Times

More than 5 times

WHENGET When did you get you last gun
Measurement Level: Nominal

KINDGUN What kind of gun was it
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00

Pistol

Derringer
Revolver

Riﬂe

Shotgun

Sawed off shot gun
Machine gun

Other type of gun

173

157

158

159

160

161

KINDOTH Other type of gun
Measurement Level: Nominal

LASTGUN Where did you get your last gun
Measurement Level: Scale

Value Label

1.00 Gun dealer

2.00 Retail or sporting good store
3.00 Pawn shop

4.00 Street dealer

5.00 Friend/girlfriend/family
6.00 Gang member

7.00 Stole it

8.00 Other

LASTOTH Last ,gun other
Measurement Level: Nominal

PREASON Primary reason you got the gun
Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

For protection

For respect

For work

As a gift

For committing crimes
Other reason

REASOTH Other reason
Measurement Level: Nominal

DIRECT Buy the gun directly
Measurement Level: Scale
Value Label

1.00
2.00

Directly
Someone bought

RELATION Person in relation to you
Measurement Level: Scale
Value Label

1.00
2.00

Girlfriend/boyfriend
Friend

174

162

163

164

165

166

167

168

3.00 Family
4.00 Gang member
5.00 Stranger

OWNNAME Did you buy it under you own name
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

PCHECK Did the seller do a police check on you
Measurement Level: Scale
Value Label

.00 N o
1.00 Yes

SELLFAM Sell to a family member
Measurement Level: Scale
Value Label

.00 False
1.00 True

SELLFRI Sell to a friend
Measurement Level: Scale
Value Label

.00 False
1.00 True

SELLGUND Sell to a gun dealer
Measurement Level: Scale
Value Label

.00 False
1.00 True

PAWN Pawn the gun
Measurement Level: Scale
Value Label

.00 False
1.00 True

175

169

170

171

1172

173

174

TRADEFM Trade it to a family member
Measurement Level: Scale
Value Label

.00 False
1.00 True

TRADEFD Trade it to a friend
Measurement Level: Scale
Value Label

.00 False
1.00 True

TRADEGUN Trade with a gun dealer
Measurement Level: Scale
Value Label

.00 False
1.00 True

GIVEFM Give it to a family member
Measurement Level: Scale
Value Label

.00 False
1.00 True

GIVEFD Give it to a friend
Measurement Level: Scale
Value Label

.00 False
1.00 True

THROW Throw it away
Measurement Level: Scale
Value Label

.00 False
1.00 True

PDRUGS Purchase drugs illegally

Measurement Level: Scale
Value Label

176

175

176

177

178

179

180

181

.00 No
1.00 Yes

HMTPDRUG How many times purchase drugs
Measurement Level: Scale

HOPDRUG How often did you do it in a month when
purchasing

Measurement Level: Scale
Value Label

1.00 Everyday or almost everyday
2.00 Several times a week

3.00 Every week or almost every week
4.00 Less than every week

CALCPDRG How many drug using the followups
Measurement Level: Scale

PDTDAYl How many times per day
Measurement Level: Scale

PDTWEEKI How many times per week 1
Measurement Level: Scale

PDTWEEK2 How many times per week 2
Measurement Level: Scale

PDMONTH3 How many per month 3
Measurement Level: Scale

PDMONTH4 How many per month 4
Measurement Level: Scale

SDRUGS Sell drugs illegally
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

HMTSDRUG How many times did you sell drugs
Measurement Level: Scale

177

182

183

184

185

186

187

188

189

190

191

HOSDRUG How often did you do it in a month when selling
Measurement Level: Scale
Value Label

1.00 Everyday or almost everyday
2.00 Several times a week

3.00 Every week or almost every week
4.00 Less than every week

CALCSDRG How many times sell drugs using the followups
Measurement Level: Scale

SDTDAYl How many times per day 1
Measurement Level: Scale

SDTWEEKl How many times per week 1
Measurement Level: Scale

SDTWEEK2 How many times per week 2
Measurement Level: Scale

SDMONTH3 How many per month 3
Measurement Level: Scale

SDMONTH4 How many per month 4
Measurement Level: Scale

CONSUMEA Consume alcohol
Measurement Level: Scale
Value Label

1.00 Three or more times a day
2.00 1-2 times a day

3.00 3-4 times a day

4.00 1-2 times a week

5.00 1-2 times a month

6.00 Once a year

7.00 Never

8.00 M Don't know

USEDRUGA Use any type of illegal drug
Measurement Level: Scale
Value Label

1.00 Three or more times a day
2.00 1-2 times a day

178

192

193

194

195

196

197

198

199

200

3.00 3-4 times a week
4.00 1-2 times a week
5.00 1-2 times a month
6.00 Once

7.00 Never

8.00 M Don't know

LMARI Marijuana in your life 201
Measurement Level: Scale
Value Label

1.00 Never

2.00 Once

3.00 Once in a while
4.00 Few times a month
5.00 Few times a week

LCOKE Coke in your life 202
Measurement Level: Scale
Value Label

1.00 Never

2.00 Once

3.00 Once in a while
4.00 Few times a month
5.00 Few times a week

LHEROIN Heroin in your life 203
Measurement Level: Scale
Value Label

1.00 Never

2.00 Once

3.00 Once in a while
4.00 Few times a month
5.00 Few times a week

LMETH Meth in your life 204
Measurement Level: Scale
Value Label

1.00 Never

2.00 Once

3.00 Once in a while
4.00 Few times a month
5.00 Few times a week

179

LACID Acid in your life
Measurement Level: Scale
Value Label

1.00 Never

2.00 Once

3.00 Once in a while
4.00 Few times a month
5.00 Few times a week

CASE4
Measurement Level: Scale

BURG Commit any burglaries
Measurement Level: Scale
Value Label

.00 N o
1.00 Yes

HMTBURG How many times commit burglaries
Measurement Level: Scale

HOBURG How often commit burglaries in months when
committing
Measurement Level: Scale
Value Label

1.00 Everyday or almost everyday
2.00 Several times a week

3.00 Every week or almost every week
4.00 Less than every week

BTIMESDI Times per day] burglaries
Measurement Level: Scale

BTIMESWI Times per weekl burglaries
Measurement Level: Scale

BTIMESW2 Times per week2 burglaries
Measurement Level: Scale

BMONTH3 Times per month3 burglaries
Measurement Level: Scale

180

205

206

207

208

209

210

211

212

213

BMONTH4 Times per month4 burglaries
Measurement Level: Scale

THEFT Commit any thefts
Measurement Level: Scale
Value Label

.00 N o
1.00 Yes

HMTTHEFT How many times commit theft
Measurement Level: Scale

HOTHEFT How often commit theft in months when
committing
Measurement Level: Scale
Value Label

1.00 Everyday or almost everyday
2.00 Several times a week

3.00 Every week or almost every week
4.00 Less than every week

TTIMESDl Times per day] theft
Measurement Level: Scale

TTIMESWl Times per weekl theft
Measurement Level: Scale

TTIMESW2 Times per week2 theft
Measurement Level: Scale

TMONTH3 Times per month3 theft
Measurement Level: Scale

TMONTH4 Times per month4 theft
Measurement Level: Scale

CAR Steal any cars
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

181

214

215

216

217

218

219

220

221

222

223

 

HMTCAR How many times steal cars
Measurement Level: Scale

HOCAR How often steal cars in months when committing
Measurement Level: Scale
Value Label

1.00 Everyday or almost everyday
2.00 Several times a week

3.00 Every week or almost every week
4.00 Less than every week

CTIMESDl Times per day] steal cars
Measurement Level: Scale

CTIMESWI Times per weekl steal cars
Measurement Level: Scale

CTIMESW2 Times per week2 steal cars
Measurement Level: Scale

CMONTH3 Times per month3 steal cars
Measurement Level: Scale

CMONTH4 Times per month4 steal cars
Measurement Level: Scale

BADCHECK Pass any bad checks
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

HMTBCHEC How many times pass bad checks
Measurement Level: Scale

HOBCHECK How often pass bad checks in months when
committing
Measurement Level: Scale
Value Label

1.00 Everyday or almost everyday
2.00 Several times a week

3.00 Every week or almost every week
4.00 Less than every week

182

224

225

226

227

228

229

230

231

232

233

BCTIMEDI Times per day] had check
Measurement Level: Scale

BCTIMEWl Times per weekl bad check
Measurement Level: Scale

BCTIMEW2 Times per week2 bad check
Measurement Level: Scale

BCMONTH3 Times per month3 bad check
Measurement Level: Scale

BCMONTH4 Times per month4 bad check
Measurement Level: Scale

BROB Commit any business robberies
Measurement Level: Scale
Value Label

.00 N o
1.00 Yes

HMTBROB How many times commit business robberies
Measurement Level: Scale

HOBROB How often commit business robberies when
committing
Measurement Level: Scale
Value Label

1.00 Everyday or almost everyday
2.00 Several times a week

3.00 Every week or almost every week
4.00 Less than every week

BRTIMEDI Times per day] business robberies
Measurement Level: Scale

BRTIMEWI Times per weekl business robberies
Measurement Level: Scale

BRTIMEW2 Times per week2 business robberies
Measurement Level: Scale

BRMONTH3 Times per month3 business robberies
Measurement Level: Scale

183

234

235

236

237

238

239

240

241

242

243

244

245

BRMONTH4 Times per month4 business robberies
Measurement Level: Scale

PROBBERY Commit any personal robberies
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

HMTPROB How many times commit personal robberies
Measurement Level: Scale

HOPROB How often commit personal robberies when
committing
Measurement Level: Scale
Value Label

1.00 Everyday or almost everyday
2.00 Several times a week

3.00 Every week or almost every week
4.00 Less than every week

PRTIMEDl Times per dayl personal robberies
Measurement Level: Scale

PRTIMEWI Times per weekl personal robberies
Measurement Level: Scale

PRTIMEW2 Times per week2 personal robberies
Measurement Level: Scale

PRMONTH3 Times per month3 personal robberies
Measurement Level: Scale

PRMONTH4 Times per month4 personal robberies
Measurement Level: Scale

FIGHTS Involved in any ﬁghts
Measurement Level: Scale
Value Label

1.00 Never

2.00 One to Five

3.00 Once/Twice Month
4.00 Once/Twice Week

184

246

247

248

249

250

251

252

253

254

255

5.00 Almost Everyday

CVICTIM Been a victim of crime
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

HMTCVIC How many times been a victim of crime
Measurement Level: Scale
Value Label

1.00 Once

2.00 2-3 Times

3.00 4-6 Times

4.00 More than 6 Times

VICTIMOF Most serious crime that you were a victim of
Measurement Level: Scale
Value Label

1.00 I had something stolen from me
2.00 My house was broken into

3.00 My car was stolen

4.00 I was robbed

5.00 I was assaulted

6.00 I was raped

7.00 Other

FFAMVIC Friends/Family been victims of violent crime
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

VICHOM Friends/Family been victim of homicide
Measurement Level: Scale

Value Label

.00 No
1.00 Yes

GAN GS How many gangs
Measurement Level: Scale

185

256

257

258

259

260

261

 

GMEMBERS How many gang members
Measurement Level: Scale

DAN GER] Most dangerous gang
Measurement Level: Nominal

DANGER2 Most dangerous gang
Measurement Level: Nominal

GMCGUNS How often do gang members carry guns
Measurement Level: Scale
Value Label

1.00 Never

2.00 Sometimes

3.00 Most of the time
4.00 Always

GAN GIMP How extensive is the impact of gangs on
neighborhood
Measurement Level: Scale
Value Label

1.00 N 0 Effect
10.00 Impacts every aspect

GMURDER How important is murder to gangs
Measurement Level: Scale
Value Label

1.00 Not at all important
5.00 Very Important

GFIGHT How important is ﬁghting to gangs
Measurement Level: Scale
Value Label

1.00 Not at all important
5.00 Very Important

GSHOOT How important is shooting to gangs
Measurement Level: Scale
Value Label

1.00 Not at all important
5.00 Very Important

186

262

263

264

265

266

267

268

269

 

GDRUGS How important is drug sales to gangs 270
Measurement Level: Scale
Value Label

1.00 Not at all important
5.00 Very Important

GDRUGU How important is drug use to gangs 271
Measurement Level: Scale
Value Label

1.00 Not at all important
5.00 Very Important

GTURF How important is protecting turf to gangs 272
Measurement Level: Scale
Value Label

1.00 Not at all important
5.00 Very Important

REGAN G Resist the pressures to get involved in gang activity 273
Measurement Level: Scale
Value Label

1.00 Very hard

2.00 Difﬁcult

3.00 Pressures are moderate, most youth resist them
4.00 No pressure

GNEIGH Gangs in your neighborhood 274
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

HMGNEIGH How many gangs are in your neighborhood 275
Measurement Level: Scale
Value Label

1.00 One or Two
2.00 Three to Five
3.00 Six to Ten

4.00 More than Ten

187

GMEMBER Are you a member of a gang
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

EVERBG Ever been a member of a gang
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

MGROUP Member of a group
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

EVERMG Ever been a member of a group
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

THOUGHT Thought of joining a gang
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

RECRUIT Recruited or pressured to join a gang
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

HUNGOUT Hung out with gang members

Measurement Level: Scale
Value Label

188

276

277

278

279

280

281

282

 

.00 N o
1.00 Yes

HIGHGM Drunk alcohol or gotten high with gang members
Measurement Level: Scale
Value Label

.00 N o
1.00 Yes

VANDAL Vandalized something with gang members
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

STOLEGM Stolen something with gang members
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

ATTACKGM Been attacked in a gang related incident
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

ATTGRI Attacked someone in a gang related incident
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

FGAN GM Friends who are gang members
Measurement Level: Scale

Value Label

.00 No
1.00 Yes

189

283

284

285

286

287

288

 

INTERACT Interact with somebody who is a member of
street gang
Measurement Level: Scale
Value Label

1.00 Never

2.00 Once/Twice Month
3.00 Once/Twice Week
4.00 Everyday

CASES
Measurement Level: Scale

FARREST Old were you when you were ﬁrst arrested
Measurement Level: Scale

HMTARRES How many times arrested in your lifetime
Measurement Level: Scale
Value Label

1.00 Once

2.00 2-5 Times

3.00 6-10 Times

4.00 11-20 Times

5.00 21-50 Times

6.00 More than 50 Times

HMTARR6 Times arrested in the last six months
Measurement Level: Scale
Value Label

.00 Never
1.00 Once
2.00 2-5 Times

3.00 6-10 Times

4.00 11-20 Times

5.00 21-50 Times

6.00 More than 50 Times

FCONVICT Old were you when you were ﬁrst convicted of
crime
Measurement Level: Scale

190

289

290

291

292

293

294

 

HMTCONV How many times have you been convicted of a
crime
Measurement Level: Scale
Value Label

1.00 Once

2.00 2-5 Times

3.00 6-10 Times

4.00 11-20 Times

5.00 More than 20 Times

VCLIFE Violent crimes have you committed in your lifetime
Measurement Level: Scale
Value Label

1.00 None

2.00 One

3.00 2-5

4.00 11-20

5.00 21-50

6.00 21-50

7.00 More than 51 Times

NVCLIFE Nonviolent crimes have you committed in your
lifetime
Measurement Level: Scale
Value Label

1.00 None

2.00 One

3.00 2-5

4.00 11-20

5.00 21-50

6.00 21-50

7.00 More than 51 Times

PDDRUGSL Purchased drugs in your lifetime -
Measurement Level: Scale
Value Label

.00 Never
1.00 One
2.00 2-5
3.00 6-10
4.00 11-20
5.00 21-50

191

295

296

297

298

 

6.00 More than 50 times

SDRUGSLI Sold drugs in your lifetime
Measurement Level: Scale
Value Label

.00 Never

1.00 One

2.00 2-5

3.00 6-10

4.00 11-20

5.00 21-50

6.00 More than 50 times

MREASON Main reason ﬁrst got involved in crime
Measurement Level: Scale
Value Label

1.00 Excitement

2.00 Friends

3.00 Money

4.00 Lost my Temper
5.00 Reputation

6.00 Other

MREASOTH Main reason other
Measurement Level: Nominal

LOCKEDUP How long have you been locked up in your
lifetime
Measurement Level: Scale
Value Label

1.00 0-6 months

2.00 6 months to 1 year
3.00 1 to 2 years

4.00 2-4 years

5.00 4-6 years

6.00 More than 6 years

FELONS How many felonies convicted of
Measurement Level: Scale

Value Label

1.00 One
2.00 2-3

192

299

300

301

302

303

 

3.00 4-6

4.00 7-10
5.00 1 1-15
6.00 16—25

7.00 More than 25

DATREAT Ever been in alcohol or drug treatment
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

FPRISON Family members served time in prison
Measurement Level: Scale

FRPRISON Friends have served time in prison
Measurement Level: Scale

FRFELONY Friends have felony convictions
Measurement Level: Scale

BIRTH What year were you born
Measurement Level: Scale

AGE
Measurement Level: Scale

SCHOOL High grade of school you completed
Measurement Level: Scale

MARRY Have you ever been married
Measurement Level: Scale
Value Label

.00 No
1.00 Yes

CMARRY Are you currently
Measurement Level: Scale
Value Label

1.00 Married

2.00 Living with a partner
3.00 Widowed

4.00 Separated

193

304

305

306

307

308

309

310

311

312

5.00

Divorced

6.00 Never married

CHILD How many children do you have
Measurement Level: Scale

TINCOME Total income by legal means
Measurement Level: Scale

WASJOB What was your job
Measurement Level: Nominal

TILLINC Total income by illegal means
Measurement Level: Scale

ILLEGAL Main source of illegal income
Measurement Level: Scale

Value Label

1.00 Selling drugs

2.00 Prostitution/Pimping
3.00 Robbery/Burglary
4.00 Selling Stolen Goods
5.00 Theft

6.00 Other

RACE Race or ethnic background
- Measurement Level: Scale
Value Label

1.00
2.00
3.00
4.00
5.00
6.00

White

Black or African American
Hispanic

Asian

Native American, or

Other

ROTHER Other Race
Measurement Level: Scale

EMPLOY Are you currently employed
Measurement Level: Scale
Value Label

.00 No

1.00

Yes

194

313

314

315

316

317

318

319

320

 

HWEEK How many hours do you work in a typical week 321
Measurement Level: Scale

TEMPLOY What percent of the time were you employed 322
Measurement Level: Scale
Value Label

1.00 100 percent

2.00 About 3/4 of the time
3.00 About 1/2 of the time
4.00 About 1/4 of the time
5.00 I was not employed

GAMBLE Gamble 323
Measurement Level: Scale
Value Label

1.00 Never

2.00 Once/Twice Month
3.00 Once Week

4.00 Everyday

CLUBS Clubs/Bars 324
Measurement Level: Scale
Value Label

1.00 Never

2.00 Once/Twice Month
3.00 Once Week

4.00 Everyday

SPORTS Organized sports 325
Measurement Level: Scale
Value Label

1.00 Never

2.00 Once/Twice Month
3.00 Once Week

4.00 Everyday

MOVIES Go to movies 326
Measurement Level: Scale

Value Label

1.00 Never
2.00 Once/Twice Month

195

 

3.00 Once Week
4.00 Everyday

CHURCH Go to church
Measurement Level: Scale
Value Label

1.00 Never

2.00 Once/Twice Month
3.00 Once Week

4.00 Everyday

CEVENT Community event
Measurement Level: Scale
Value Label

1.00 Never

2.00 Once/Twice Month
3.00 Once Week

4.00 Everyday

HANGOUT Hang out with friends
Measurement Level: Scale
Value Label

1.00 Never

2.00 Once/'1‘ wice Month
3.00 Once Week

4.00 Everyday

FILTER_$ uniquecode ~= code (FILTER)
Measurement Level: Scale
Value Label

0 Not Selected
1 Selected

GUNSALL self-report any gun activity
Measurement Level: Scale

DRUGSALL Use or Sale of Drugs
Measurement Level: Scale

NONVIOAL
Measurement Level: Scale

196

327

328

329

330

331

332

333

 

VIOALL Violent Crime combo
Measurement Level: Scale

LAWGROUP recoded group variable
Measurement Level: Scale

COMGROUP recoded community group variable
Measurement Level: Scale

CTLGROUP recoded control group variable
Measurement Level: Scale

ANY CRIME Admit any criminal activity
Measurement Level: Scale

TOTCRIME admit crimes added together
Measurement Level: Scale

FCGUNT chances of going fed extreme
Measurement Level: Scale

N EWGUN
Measurement Level: Scale

N EWGUN2
Measurement Level: Scale

RERACE recoded race
Measurement Level: Scale

REMAR current marriage situation
Measurement Level: Scale

COMPLY compliance (alcohol and address)
Measurement Level: Scale

RETREAT recoding of participation in court ordered
treatment
Measurement Level: Scale

REASS involved in any ﬁghts
Measurement Level: Scale

RETHREAT Threatened with gun dichotomous
Measurement Level: Scale

197

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

 

RESHOT Shot at with gun Recode
Measurement Level: Scale

REINJURE Injured with a Gun (Recode)
Measurement Level: Scale

RECARR recode of arrest in last six months
Measurement Level: Scale

REDUSE recode of any drug use
Measurement Level: Scale

RESTATAD meeting etc vs revoked inc abs
Measurement Level: Scale

REPOLCON dichotomous police contact
Measurement Level: Scale

REFEDCON times federal law e contact dichotomous
Measurement Level: Scale

REPROSCO dichotomous prosecutor contact
Measurement Level: Scale

RECOMCON dichotomous community contact
Measurement Level: Scale

RECLCON dichotomous clergy contact
Measurement Level: Scale

REPOCON dichotomous probation contact
Measurement Level: Scale

REMARRY remarry dichotomous
Measurement Level: Scale

REARRLF Dichotomous arrests in your life
Measurement Level: Scale
Value Label

.00 1, 2-5 Arrests
1.00 More than 6 arrests

REARR6M dichotomous arrested last six months
Measurement Level: Scale

198

349

350

351

352

353

354

355

356

357

358

359

360

361

362

 

Value Label

.00 Never
1.00 Arrested at Least Once

RECONV dichotomous reconviction variable
Measurement Level: Scale
Value Label

.00 Once
1.00 More than Once

RESTATUS dichotomous status at address
Measurement Level: Scale

COMLEORG contacted community leader or community
organization
Measurement Level: Scale
Value Label

.00 no
1.00 yes

RECTTRET dichotomous court ordered treatment
Measurement Level: Scale

N UMARRES Number of Arrests
Measurement Level: Scale

N UMARR_A Number of Arrest Charges
Measurement Level: Scale

VIOLENTC Number of Arrest Charges for Violent Offenses
Measurement Level: Scale

PROPERTY Number of Arrest Charges for Property
Offenses
Measurement Level: Scale

DRUGCHRG Number of Arrest Charges for Drug
Offenses
Measurement Level: Scale

199

363

364

365

366

367

368

369

370

371

 

ALCOHOLC Number of Arrest Charges for Alcohol
Offenses .
Measurement Level: Scale

WEAPONSC Number of Arrest Charges for Weapon
Offenses
Measurement Level: Scale

RESISTCH Number of Arrest Charges for Resisting Arrest
Measurement Level: Scale

OTHERCHR Number of Arrest Charges for Other Offenses
Measurement Level: Scale

NUMMIS Number of Misdemeanor Charges
Measurement Level: Scale

NUMFEL Number of Felongy Charges
Measurement Level: Scale

VIOLEN_A Number of Violent Convictions
Measurement Level: Scale

PROPCON Number of Property Convictions
Measurement Level: Scale

DRUGCON Number of Drug Convictions
- Measurement Level: Scale

ALCOHO_A Number of Alcohol Convictions
' Measurement Level: Scale

WEAPONCO Number of Weapons Convictions
Measurement Level: Scale

RESISCON Number of Resisting Convictions
Measurement Level: Scale

OTHERCON Number of Other Convictions
Measurement Level: Scale

FELCON Number of Felony Convictions
Measurement Level: Scale

MISCON Number of Misdemeanor Convictions
Measurement Level: Scale

200

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

 

TIMESPRO Number of Times on Probation 387
Measurement Level: Scale

MJCTIMES Number of Times in Marion County Jail 388
Measurement Level: Scale

DOCTIMES Number of Times in DOC 389
Measurement Level: Scale

DOCLEN GT Length of Time in DOC (Days) 390
Measurement Level: Scale

SCARNEW 391
Measurement Level: Scale

DISCALE 392
Measurement Level: Scale

GCAT 393
Measurement Level: Scale

AT2GUNU 394
Measurement Level: Scale

PREGCRIM 395
Measurement Level: Scale

POSSHIFI‘ 396
Measurement Level: Scale

PSTINNET 397
Measurement Level: Scale

PIRECALL 398
Measurement Level: Scale

GMEM 399
Measurement Level: Scale

201

 

Reference

Audirac, I. (2003). “Information-age Landscape outside the Deve10ped World:
Bangalore, India and Guadalajara, Mexico.” Journal of the American
Planning Association. Vol. 69, i1. Winter.

Bailey, G. W. & Unnithan, N. P. (1994). Gang homicides in California: A
discriminant analysis. Journal of Criminal Justice, 22(3), 267-275.

Bjerregaard, B. (2002). Self-deﬁnitions of gang membership and involvement in
delinquent activities. Youth & Society, 34(1), 31-54.

Braga, A. A., Piehl, A. M., & Kennedy, D. M. (1999). Youth homicide in Boston.
Homicide Studies, 3(4), 277-299.

Brewer, V. E., Damphousse, K. R., & Adkinson, C. D. (1998). The role of
juveniles in urban homicide: The case of Houston, 1990-1994. Homicide
Studies, 2(3), 321-339.

Castells, M. (2000). The Rise of the Network Society. Second ed. Blackwell
Publishing. Maulden, MA.

Decker, S. & Kempf-Leonard, K. (1991). Constructing gangs: The social
deﬁnition of youth activities. Criminal Justice Policy Review, 5(4), 271-
291.

Dentler, R. A. & Erikson, K. T. (1959). The functions of deviance in groups.
Social Problems, 7(2), 98-107.

Decker, S. H. & Curry, G. D. (2000). Responding to gangs: Comparing gang
member, police, and task force perspectives. Journal of Criminal Justice,
28, 129-137.

Decker, S. H. & Curry, G. D. (2002). Gangs, gang homicides, and gang loyalty:
Organized crimes or disorganized criminals. Journal of Criminal Justice,
30, 343-3 52.

Dicken, P. (2003). “A New Geo-economy”. In The Global Trgnsformation
Reader. Edited by Held, D and A. McGrew. Polity Press. Cambridge.

Egley, A. Jr. and Ritz, C. E. (2006). Highlights of the 2004 National Youth Gang

Survey. Ofﬁce of Juvenile Justice and Delinquency Prevention Fact Sheet.
Washington, DC: Office of Juvenile Justice and Delinquency Prevention.

202

 

Esbensen, F. A., Winﬁee, L. T., He, N., & Taylor, T. J. (2001). Youth gangs and
deﬁnitional issues: When is a gang 3 gang, and why does it matter? Crime
& Delinquency, 47(1), 105-130.

Erickson, M. L. (1973). Group violations and ofﬁcial delinquency: The group
hazard hypothesis. Criminology, 11(2), 127-160.
Friedman, T. (2000). The Lexus and the Olive Tree. Anchor Books. New York

Giddens, A. (2003-a). Runaway World. Rutledge Press. New York

Gordon, R. M. (2000). Criminal business organizations, street gangs, and ‘wanna-
be’ groups: A Vancouver perspective. Canadian Journal of Criminology,
42(1), 39-61.

Hagan, J. (1997). Deﬁance and despair: Subcultural and structural linkages
between delinquency and despair in the life course. Social Forces, 76(1),
1 19-134.

Hagedom, J. M. (1998). Gang violence in the postindustrial era. In M. Tonry &
M. Moore (Eds.), Youth Violence (pp. 365-419). Chicago: University of
Chicago Press.

Hardt, M and A. Negri. (2000). Empire. Harvard University Press. Cambridge,
Massachusetts.

Hardt, M and A. Negri. (2004). Multitude: War and Democracy in the age of
Empire. Penguin Press. New York.

Harrell, E. (2005). Violence by gang memberszl993-2003. National Youth Gang
Center Bulletin. Washington, DC: Bureau of Justice Statistics.

Harper, G. W. & Robinson, W. L. (1999). Pathways to risk among inner-city
Aﬁican-American adolescent females: The inﬂuence of gang membership.
American Journal of Community Psychology, 27(3), 383-405.

Held, D. (2003). “Cosmopolitanism: Taming Globalization.” In The Global
Transformation Reader. Edited by Held, D and A. McGrew. Polity
Press. Cambridge.

Hobbs, D. (2001). “The Firm: Organizational Logic and Criminal Culture on a
Shifting Terrain.” British Journal of Criminology. Vol. 41, i4. Autumn.

Howell, J. C. (1994). Recent gang research: Program and policy implications.
Crime & Delinquency, 40(4), 495-515.

203

 

Hutson, H. R., Anglin, D., & Pratts, M. J. (1994). Adolescents and children
injured or killed in drive-by shootings in Los Angeles. The New England
Journal of Medicine, 330(5), 324-327.

Kennedy, D. M. & Braga, A. A. (1998). Homicide in Minneapolis. Homicide
Studies, 2(3), 263-290.

Klein, M. W. (1971). Street gangs and street worm. Englewood Cliffs, NJ:
Prentice Hall.

Lerman, P. (1968). Individual values, peer values, and subcultural delinquency.
American Sociological Review, 33(2), 219-235.

Levitt, S. D. & Venkatesh, S. A. (2000). An economic analysis of a drug-selling
gang’s ﬁnances. The Quarterly Journal of Economics, 755-789.

Maxson, C. L. (1995). Street gangs and drug sales in two suburban cities.
Research in Brief. Washington, DC: US. Department of Justice, Ofﬁce of
Justice Programs, National Institute of Justice. NCJ 155185.

Maxson, C. L., Whitlock, M. L., & Klein, M. W. (1998). Vulnerability to street
gang membership: Implications for practice. Social Service Review, 70-91.

McCusker, R. (2004). “China, Globalization and Crime: A Potential Victim of Its
own Prospective Success.” Journal of Financial Crimes. Vol. 12, i1. Aug.

Morash, M. (1983). Gangs, groups, and delinquency. The British Journal of
Criminology, 23(4), 309-335. -

Naim, M. (2005). Illicit: How Smugglers. trafﬁckers and Copycats are Hﬁackrjrg
the Global Economy. Doubleday Publishing. New York.

 

Ofﬁce of Juvenile Justice and Delinquency Prevention.(2004). Annual Report

2003-2004. Washington, DC: Ofﬁce of Juvenile Justice and Delinquency
Prevention, NCJ 206630.

Parker, R. (2002). “The Global HIV/ AIDS Pandemic, structural Inequalities and
Politics of International Health.” American Journal of Public Health. Vol.
92, i3. Mar.

Passas, N. (2000). “Global Anomie, Dysnomie and Economic Crime: Hidden

Consequences of Neoliberalism and Globalization in Russia and Around
the World.” Social Justice. Vol. 27, i2. Summer.

204

 

Ruble, N. M. & Turner, W. L. (2000). A systematic analysis of the dynamics and
organization of urban street gangs. American Journal of Family Therapy,
28(2), 117-133.

Sanchez-Jankowski, M. (2003). Gangs and social change. Theoretical
Criminology, 7(2), 191-216.

Scholte, J. (2003). “What is ‘Global’ About Globalization.” In The Global
Transformation Reader. Edited by Held, D and A. McGrew. Polity
Press. Cambridge.

Schreck, C. J ., Fisher, B. S., & Miller, J. M. (2004). The social context of violent
victimization: A study of the delinquent peer effect. Justice Quarterly,
21(1), 23-48.

Shelley, L. (1998). “Crime and Corruption in the Digital Age.” Journal of
International Aﬂairs. Vol: 51, i2. Spring.

Somarajah, M. (2004). “Transnational Crimes: The Third Limb of the Criminal
Law.” Singapore Journal of Legal Studies. Dec. Pg. 390.

Short, J. F. & Strodtbeck, F. L. (1963). The response of gang leaders to status
treats: An observation on group process and delinquent behavior. The
American Journal of Sociology, 68(5), 571-579.

Stern, J. (2003). Terror in the Name of God: Why Religious Militants Kill.
Harper-Collins. New York.

Tomlinson, J. (2003). “Globalization and Cultural Identity” In The Global
Igansformation Reader. Edited by Held, D and A. McGrew. Polity
Press. Cambridge.

Thomberry, T. P., Krohn, M. D., Lizotte, A. J ., & Chard-Wierschern, D. (1993).
The role of juvenile gangs in facilitating delinquent behavior. Journal of
Research in Crime and Delinquency, 30(1), 55-87.

Tillman, R. (2002). Global Pirates: Fraud in the Offshore Insurance Industry.
Northeastern University Press. Boston.

Williams, P. (2005). The Al aeda Connection: International Terrorism.
Organized Crime and the Coming Apocalypse. Prometheus Books.
Amherst. New York.

Winfree, L. T., Mays, G. L., & Vigil-Backstrom, T. (1994). Youth gangs and

incarcerated delinquents: Exploring the ties between gang membership,
delinquency, and social learning theory. Justice Quarterly, 11(2), 229-256.

205

Yablonsky, L. (1959). The delinquent gang as a near-group. Social Problems,
7(2), 108-117.

206

 

 

     

HIG 51

1111111111111!

3 129

ERSITY ES

tilt]lllllltlll

45 9489