- .rx“.
5%“:

  
   
 
    

   

r
.‘F

.:“-2&"";':.

  

  
 
 
 
 
   

 

      
 

 
 
 
  
   

‘“ 1“”
sx‘w‘xwu .‘ .
' “7‘ 3:. “1:541;

“1‘34. 9».

         
  

   

$3.. ‘

' '§_ ‘4‘“- k
{(51% "k ”air-m "
«34...;
‘ {473 f

272““ um, |*(&“§,_

   

. - |. __;.l..i ,
:JJL £173. -flgy";
A.

 
 
 
  
 
  
    

 

 
  

W1
”A “W Jlr-A. \L' ~v‘ A
"in?“{fi'vfi “945.15?
my. my:

.. a. .

      
   

 

   

3:-

~44}.
1D,.“ ‘:
l .
3‘

     
    
 
     

m ‘3!“ w ~ \1
15-h .' - .4
'~\ as»; (12".. “wt-Mt?—
’aujd .u» <~'J

    

     

, :wxrmgw
‘

i‘3~\€-78
L- --,
in ’

 

 

 

 

 

 

 

 

     

"Ht-s33

MICHIGAN sure u

ll/I/l/lI/l/II/l/Ill/ll "”7?

////I/// Ill/l’z’lll/l/ll/l/fl

(”Jim

This is to certify that the

thesis entitled

An Exploratory Study of School Suspension,
Social Bonding and Juvenile Delinquency

presented by v

Angela Micaela Brown

has been accepted towards fulfillment
of the requirements for

Master of Science degree in Criminal Justice

 

% W
l / Major professor

Date #4 aéz/q7/
l/ U '

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

LIBRARY

Michigan State
University

 

PLACE IN RETURN BOX to remove We checkout from your record.
To AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

L.

MSU Is An Affirmative Amen/Equal Opportunity Institution
tm

 

 

 

 

anal

 

AN EXPLORATORY STUDY OF
SCHOOL SUSPENSION, SOCIAL BONDING AND
JUVENILE DELINQUENCY
By

Angela Micaela Brown

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

School of Criminal Justice

1991

65¢— 5?07 i

ABSTRACT

AN EXPLORATORY STUDY OF
SCHOOL SUSPENSION, SOCIAL BONDING AND
JUVENILE DELINQUENCY

By

Angela Micaela Brown

This thesis studies school suspension and the elements of social
bonding, as described by Travis Hirschi, and their relationships with
Juvenile Delinquency. The elements of focus were: school
ATTACHMENT; school COMMITMENT; school INVOLVEMENT; and
BELIEF in educational success. Social Bond Theory, along with current
research in the area of school suspension are reviewed. The sample was
taken from the 1977 National Youth Survey, which is a longitudinal study
of delinquency and drug use among American youth. The findings of this
research suggest that school suspensions are significantly related to
delinquency and that suspensions weaken the bonds, particularly the
BELIEF element of the bond and that weakened bonds may contribute to

delinquency. Further research in this area is, however, recommended.

ACKNOWLEDGEMENTS

Numerous delinquency theories exist. Some try to explain
relationships that exist among delinquents, while most all delinquency
theories want to know: why some youth do or some do not commit
deviant acts. Rather humorously, Travis Hirschi asks: “What is there
about criminologists that leads them to ask, why do they do it?” (Hirschi,
1969, 231). The answer to that question for me is quite simple. I’ve
longed to understand the delinquency that I was involved in during
adolescence. I still ask myself the question: “Why did I do it?”

As I later studied the various delinquency theories, I would often
search for answers to the question: “Why was I delinquent?” Thus I
must thank the entire Michigan State University Criminal Justice
Department faculty and staff for their encouraging my quest for answers.
I will be forever honored to have been so privileged to be a student of
such a renowned program and to have learned from the geniuses of the
MSU Criminal Justice Professors. I would like to conclude with special
dedication and thanks to committee members: Dr. Merry Morash, Dr.
Rosie Epynyoung, and Dr. Vincent Hoffman. Additionally, I must thank

the Grand Blanc. School System for suspending me so frequently. During

high school, I was suspended repeatedly and am not sure exactly how
many times I was suspended. It was the unforgettable suspension
experiences and their related effects that were the driving force behind
mylifelong quest to study suspension and my campaign to end the

frequent and unnecessary use of suspension and expulsion.

iii

TABLE OF CONTENTS

LIST OF TABLES ............................... x

Chapter: ‘

ONE. INTRODUCTION .' ......................... 1
The Problem .......................... 1
The Research Hypotheses .................. 6
Theoretical Framework—Introduction To Theory . . . . 6
Criticism of the Social Control Theory ......... 9
Overview ............................ 10

TWO. REVIEW OF THE LITERATURE ................. 11
School Suspensions: Are They Helping Children? . . . . 12
Methodology .......................... 13
Findings ............................ 15
Strengths and Weaknesses of the C.D.F. Survey . . . . 19
Suspended Students—Suspended Learning ........ 21
Methodology .......................... 22
Findings ............................ 23
Strengths and Weaknesses of Study ............ 26
Social Bonding and Delinquency ............ -. 27

iv

The Sample and The Data .................

Delinquency Defined .....................

School Attachment ......................
Like School ......................
Care What Teachers Think. .............

School Commitment (proofs of dedication to school)
Educational Aspirations and Expectations
Index of Achievement Orientation. ........
General School Aptitude ...............

School involvement .....................
Time Spent in and Concern for Homework.
Attendance at School Activities ..........
Participation in Interscholastic Athletics.

Belief ..............................
Respect for Police. ..................
Lack of Respect for the Law ..............

Acceptance of School Authority. ..........

The Causal Chain ..........................

Summary ................................

THREE. DESIGN OF THE STUDY ...................

Introduction ..........................

Sample Population and Data Collection Procedures

28

33

33

34

34

36

37

37

39

41

41

43

45

48

48

48

The Variables ......................... 50

The Dependent Variable ................ 50
Non-Delinquent. ................ 50
Occasional Delinquent. ............ 50
Serious Delinquent. .............. 51

The Independent Variables. ............. 52
School Suspensions. .............. 52

Non-Suspended Student ........ 53
Suspended Student ........... 53
School Attachment. .............. 53
Good Student ............... 54
Good Impression ............. 54
Nobody at School Cares. ....... 54
Belong at School. ........... 54
Lonely at School. ........... 55
School Commitment. ............. 55

Educational Aspirations and

expectations. ........... 56

General School Aptitude. ....... 56
Index of Achievement Orientation. . 56
School Involvement. ............. 57

Time Spent Doing School Work. . . 57

Time Spent on School Athletics. . . 58

vi

Time Spent on Other

School Activities ......... 58
Belief. ...................... 59
Conventional Norms and Attitudes. . 59
Acceptance of School Authority. . . 60
Testable Hypotheses ......................... 60
Introduction. .......................... 60
School Suspension. .................. 61
School Attachment. .................. 61
School COmmitment. ................. 62
School Involvement. ................. 62
Belief in Educational Success. ........... 63
Conclusion. .......................... 63
Data Collection Analysis ...................... 63
FOUR. THE FINDINGS .......................... 67
Introduction .......................... ' 67
General Descriptions of Sample Population ....... 68
General Description of the Suspended Student ..... 69
General Description of the Delinquent Youth ...... 70
General Description—Delinquency Typology ...... 72
The Dependent Variables .................. 73
Core Variables ........................ 73
Questionnaire I: Analysis of Suspension .......... 75

vii

Questionnaire 111: Analysis of Suspension ....... 77

Analysis of Social Bonds and Delinquency ....... 80
Questionnaire I Analysis ............... 80
Questionnaire II Analysis .............. 83
Questionnaire III Analysis ............. 86

Analysis of Suspension and Delinquency By Waves . . 90

Wave I ......................... 90

Wave II ......................... 91

Wave III. ........................ 91

Wave IV. ........................ 92

Wave V. ......................... 92

Regression Analysis Findings ............... 93

Review of the Findings ................... 94

FIVE. IMPLICATIONS AND SUMMARY _. . . . . . . . ...... 96
Summary and Discussién .................. 96

Conclusion ........................... 101
Implications for Future Research ............. 104
APPENDIX .................................. 107
BIBLIOGRAPHY .............. . ................ 135

viii

LIST OF TABLES

C.D.F. Survey - Reasons for Suspensions .............

Significant Relationships of Variables to
Delinquency Wave I ...........................................

Significant Relationships of Variables to
Delinquency Wave II ..........................................

Significant Relationships of Variables to
Delinquency Wave III .........................................

Significant Relationships of Variables to
Delinquency Wave IV ..........................................

Significant Relationships of Variables to
Delinquency Wave V ...........................................

Significant Relationships of .Variables to
Suspension Wave I ............................................

Significant Relationships of Variables to
Suspension Wave III ..........................................

Relationships of Suspension and Delinquency
by Waves ..........................................................

Two Step Regression Analysis ...........................

CHAPTER ONE
INTRODUCTION

The Problem
The role of the American family has been changing. The extended

family is disintegrating. Nuclear families often live hundreds of miles
away from their relatives. It is common to observe families in which
both the mother and father work. Infants and children are often left in
the care of baby-sitters and older children and frequently left on their
own. Due to economic and career priorities, families are spending less
time with their children thus making the ROLE OF THE SCHOOL
MONUMENTALLY IMPORTANT. Currently, the majority of adolescents
in the US are concentrated in public schools and because of its ever-
inereasing impact on the entire population, the educational system should
be of prime concern.

Evidence indicates that delinquency may be a consequence of a
multitude of factors, however a number of studies suggest that negative
school experiences can lead to youthful misbehavior and delinquency
(Stinchcombe, 1964; Polk and Schafer, 1972; Elliott and Voss 1974).
There exists an aspect of negative school experiences, namely school
suspensions and its relationship to delinquency that this researcher desires
to study. Charles B. Vergon (Associate Professor and Attorney) of the

University of Michigan School of Education, provides the most current

2

and complete picture of suspension in a recent study entitled
“Disciplinary Actions in Michigan Public Schools: Nature, Prevalence and
Impact 1978-1986” (1990). The problem Vergon found was that Michigan
ranked 6th in the nation with the highest suspension rates for public
school students. Students suspension rates in Michigan are 43% higher
than the national average.

Using data collected from school districts across the country by the
US Office for Civil Rights, Vergon analyzed the nature, prevalence, and
impact of various disciplinary measures for the state and for
systematically selected samples of Michigan school districts for various
years between 1978 and 1986. Included in the 1986 sample were 115
Michigan districts including urban, suburban and rural school systems.
Similar analysis of the nature, prevalence and impact of suspensions and
expulsions were carried out for the midwest and the US, allowing
comparisons between Michigan’s practices and broader regional and
national patterns. Nearly 97% of the Michigan districts studied, reported
using suspensions.

Suspended students, some of whom were suspended more than once,
numbered 102,000 in 1986 or 6.9% of Michigan K-12 population, but
suspensions were concentrated in 12 to 13% of the junior and senior high
school population. With a Michigan suspension rate of nearly 70 students
per thousand, as contrasted to a national rate of 50, Michigan students

were 43% more likely to experience suspension than were their

3

classmates nationwide. Only five states suspended a greater proportion
of their school-age population than did Michigan. These states were
Maryland, Florida, Louisiana, Delaware, and South Carolina.
Furthermore, substantial variations emerged among the districts studied
in Michigan. Although the suspension rate for the entire state was 70
students per thousand, it ranged from a low of 0 in four districts to a
high of 311 students per thousand in one community. This research also
examined the impact of disciplinary actions on various student
populations to ascertain if some were at greater risk of suspension than
others. In Michigan like the US and the Midwest, males accounted for
approximately seven out of every ten suspensions. This means that males
were approximately twice as likely to be suspended than their female
counterparts. Minority students as a group, and African-American and
Hispanic students in particular were suspended at higher rates than whites
in Michigan. The suspension rates for minorities was 141 as compared
to 56 students per thousand for non-minority students. By far the highest
incidence of suspension involved African-Americans who were suspended
at a rate of 167 students per thousand, followed by hiSpanics at a rate of
100 per thousand. African-Americans had the highest suspension rates
in the US with Michigan blacks being almost twice as likely to
experience suspension as their black counterparts, nationwide.

Over the eight-year period studied, the suspension trend in

Michigan climbed steadily. Additionally, Vergon found that the use and

4

duration of suspensions varies widely from district to district within the
state based on local determinations of appropriate minimums and
maximums for various types of misconduct. Thus, in some districts
suspensions may be limited to three days or less, while in other districts,
students may be removed from school or their regular classes for up to
the balance of the school year or even longer. In conclusion, Vergon
defines the problem as being: the frequent, increasing and unnecessary
use of suspension, particularly in Michigan, disproportionate suspension
rates for minorities, particularly for blacks. inconsistent use of suspension
and expulsion, thus leaving it up to individual school districts, what a
student is suspended for and how long.

Admittedly, Vergon did not examine the impact suspension had on
delinquency. An extensive body of literature has been developed on the
theme of “Negative School Experiences” or “School Failure” as a cause
of adolescent misconduct or delinquent behavior (e.g., Hargreaves, 1967;
Sehafer and Polk, 1967; Toby, 1957; Stinchcombe, 1964; Hirschi, 1969;
Polk and Sehafer, 1972; Kelley and Balch. 1971; Kelly, 1971, 1974;
Rhodes and Reiss, 1969; Elliott, 1966). Yet little study has been done
to examine the relationship between SCHOOL SUSPENSION AND
DELINQUENCY. Additionally, Vergon’s study, which provides the most
recent suspension data available, is limited through its use of Office of
Civil Rights data. O.C.R. data does not informnwhat students were

suspended for or how long. O.C.R. data is ‘a‘xeollection of basic

5

information. It tells how many students were suspended one or more
times but does not disclose the number of students that were suspended
repeatedly or more than once.

THE PROBLEM is twofold. Firstly, the use of suspension and
expulsion in American public schools is frequent and unnecessary.
Secondly, there is little research directly examining whether a
relationship exists between school suspensions and delinquency. Thus,
there exists a great NEED for this study, particularly since the majority
of our public schools still use suspension regularly. The PURPOSE of
this study is quite simple—to examine the frequency and necessity of
school suspensions in American public schools and to investigate if a
relationship exists between school suspensions and delinquency.
Furthermore, the purpose of this study is to develop a prediction model
on juvenile delinquency based on Travis Hirschi’s Social Bond Theory of
Juvenile Delinquency. This model will examine the youth’s attachment
to school as well as their involvement and commitment to school and

belief in educational goals.

The Research Hypotheses
The implications expected to be found in this study, are stated in
the broad form of hypotheses. These hypotheses are restated in testable

form in Chapter 3.

6

1. A negative relationship exists between a youth’s attachment,
commitment, involvement and belief in educational goals and school
suspension.

2. A negative relationship exists between a youth’s attachment,
commitment, involvement, and belief in educational goals and
delinquency.

3. A positive relationship exists between school suspensions and
juvenile delinquency.

4. Suspended students have a higher rate of delinquency, than non-
suspended students.

5. An increase in suspension one academic year, results in an increase

in delinquency the following year.

Theoretical Framework—Introduction To Theory

Numerous theories of delinquency causation exist, as was pointed
out in the fourth edition of “Juvenile Delinquency Concepts and Control,”
written by Robert Trojanowiez and Merry Morash, published in 1987. In
“Juvenile Delinquency Concepts and Control,” the popular assumptions
of numerous theories were reviewed. Travis Hirschi’s Social Bond
Theory was found to have the strongest supportingresearch. Thus, it
offers one of the best explanations of the causes of delinquency, at least

amongst the contemporary sociological theories of criminality.

7

Hirschi supports a Social Control Theory which assumes that an
individual has the free will to violate the law and that delinquent acts
result when an individual’s bond to conventional society is weak or
broken. Hirschi maintains that youth do not take part in illegal
activities, because of strong bonds to conventional society. Elements of
the bond consist of ATTACHMENT of the individual to others (i.e.,
parents, teachers), COMMITMENT to conventional lines of action (i.e.,
education or other legitimate goals), INVOLVEMENT in conventional.
activities and BELIEF in legitimate values, that of conventional society.
Basically, the more tightly bonded a person is to the conventional social
groups, the higher the level a person obtains of each element, the less
likely they will commit delinquent acts.

Hirschi provides a description of the elements of the bond to the
conventional society and attempts to show how each of these elements is
related to delinquent behavior and how they are related to each other.
Hirschi defines the elements of the bond as follows: ATTACHMENT is
defined by a child’s sensitivity to the opinions of significant others, such
as parents, teachers, and friends. ATTACHMENT to the parent is seen
as being the most important. If children do not care or think about
reactions of their parents, parental control over the child is seriously
reduced. Secondly, COMMITMENT is viewed as a rational component
in conformity which is measured by a child’s investment in conventional

society. Are children dedicating themselves to conventional activities

8

such as getting an education, building up a business or other legitimate
goals. Thirdly, INVOLVEMENT is measured by gauging the amount of
time spent in conventional activities, such as homework and athletics.

And the final element is BELIEF. To what extent does a child believe
in the norms of the conventional society. What attitude of respect does
the child maintain toward the rules of society. The more a child believes
in the morality of the law, the less likely the child will be delinquent.

A relation among the elements was seen by Hirschi: “In general,
the more closely a person is tied to conventional society in any of these
ways, the more closely he is likely to be tied in other ways. The person
who is attached to conventional people is for example, more likely to be
involved in conventional activities and to accept conventional notions of
desirable conduct” (Hirschi, 1969, 27).

Although attachment and commitment to parents was seen as most
significant, attachment and commitment to the school was also found to
be significant. Hirschi studied the elements of the social bond in relation
to education. A child’s ATTACHMENT to the school was measured by
the child’s sensitivity to the opinions of the teachers. COMMITMENT
was measured by the child‘s educational aspirations and expectations.
INVOLVEMENT was measured by the amount of time spent doing
homework and BELIEF was measured by the child’s acceptance of the
school’s authority. Hirschi presented a simple causal chain and examined

data relevant to it. The causal chain runs from academic incompetence,

9

to poor school performance, to disliking of school, to rejection of the

school’s authority, to the commission of delinquent acts.

Criticism of the Social Control Theory

Social Control Theory has some weaknesses and can explain only
some of the differences between youths in levels of delinquency. It does
not predict serious delinquency over minor delinquency. And it does not
explain why bonds are developed in some but not all youth. Social
Control Theory does not explain the more serious criminal behavior done
by adolescents or their subsequent adult criminality. Admittedly, it is
not this researcher’s interest to predict early adult criminality, rather it
is to investigate delinquency which occurs during adolescence and of
particular concern during ages when school suspensions are applied.

Social Control Theory is also criticized for overestimating the
significance of involvement in conventional activities and underestimating
the role played by ones peers. “Several studies have revealed that peer
group, and in some cases a delinquent peer group to which a youth is
strongly committed, plays an important part in the development of
delinquent behavior” (Trojanowiez and Morash, 1985, 59). Some of these
voids have been filled by revised and refined theories. One particularly
successful effort was made when Social Control Theory was integrated
with Strain and Learning theories, producing a new integrated theory

which was well supported by research (Elliott, Huizinga and Ageton,

10

1985). Hirschi’s theory is also criticized for the use of English and Math
grades as a measure of grade point average. Furthermore, the use of
grade point averages as an indication of intelligence is highly debated
amongst researchers, as is whether intelligence levels are predictive of
delinquency. In spite of the previously described weaknesses, Hirschi
developed and empirically tested an explanation of delinquency. His
findings provided considerable evidence to support the Social Control

explanation of delinquency.

91w

In Chapter 2, the Pertinent Literature is reviewed. In Chapter 3,
the Design of the Study is detailed. In Chapter 4, the Analysis of the
Results is provided and in Chapter 5, Summary and Conclusions are
drawn. The data identifying the most recent suspension trends
(Frequency) was already discussed in the problem section of this first
chapter. Beginning in Chapter 2, (The Literature Review), two studies

are reviewed for the purpose of characterizing the suspended student and

. identifying the necessity of suspension. Additionally, Hirschi’s Social

Bonding Theory is reviewed in detail.

CHAPTER 2

REVIEW OF THE LITERATURE

Numerous studies have evaluated the various programs that have
developed as alternatives to school suspensions. These studies however,
only look at the successfulness of the alternative programs and do not
specifically look at suspensions or their effects. For example, the
Portland Oregon Public Schools Research and Evaluation Department
instituted and studied a four-year pilot program—P.A.S.S. (Positive
Alternatives to School Suspension). Reports were issued in 1984 and
1986. These reports systematically analyzed whether the alternative
suspension programs were successful. As stated in the report, the main
goal of the P.A.S.S. program was to reduce out-of-school suspensions in
order to “obviate the negative consequences of school suspensions.” The
name of the program—Positive Alternatives to School Suspensions and
the study itself, implies that school suspensions are negative. Yet the
implication that school suspensions are negative is not supported by
research.

Most alternative suspension programs (like delinquency programs)
are based on some assumption or theory. In the case of alternative
suspension programs, it is assumed that school suspensions are negative.
In fact, some studies go so far as to say that school suspensions are

11

12

contributing to or furthering delinquency. Of the many alternative
suspension programs this researcher reviewed, the researchers neglected
to analyze the relationship between suspension and delinquency. Reason
being, they were already acting on the belief that school suspensions are
negative or cause delinquency. They therefore focused their attention on
evaluating the successes of the alternative program.

As previously stated, most of the existing research is on alternative
programs to school suspension. However, two studies were found that.
looked specifically at suspension. One large national study on school
suspension was conducted by the Children’ Defense Fund in 1975 and a

study on suspensions in the Philadelphia Public School was done in 1982.

School Suspensions: Are They Helping Children?
This report was published in 1975 by the Children’s Defense Fund

(C.D.F.) of the Washington Research Project. The study relies on two
primary sources of data: (1) the results of an extensive household survey
of children out of school which was published earlier by the C.D.F. in
“Children out of School in America” (1974), and (2) C.D.F. analysis of
information submitted by school districts to the Federal Office for Civil
Rights (O.C.R.). In the 1974 report, nearly two million children,
according to 1970 U.S. Census data, missed all or substantial portions of
their school year. C.D.F. concluded that they were out of school for the

most part, not by choice but because they had been excluded. One of the

13

most common mechanisms for putting children out of school is
exclusionary discipline in one form or another—whether it is called
suspension, expulsion, voluntary withdrawal, blocking, barring, temporary
dismissal or a cooling-off period. C.D.F. found that most of those
disciplinary exclusions were discretionary acts, hidden from public view
and that many suspensions were unnecessary, made no educational sense
and deserted the interests of the children involved. In many cases, short
term disciplinary exclusions added up to a significant loss in schooling

and caused youngsters to drop out of school permanently.

Methodology

The 1970 U.S. census indicated where most serious problems of
non-enrollment were and what impact of location, income, parental
education and minority status was on the likelihood of children being out
of school. However, census data does not tell who these children are,
why they are out of school, and what it means to them. The C.D.F.
decided to do a door-to-door survey in various parts of the country in
order to answer those questions. The report includes the results of the
C.D.F. survey of 8,500 households in nine states. Out of these nine
states, 30 areas were surveyed. Areas to monitor were selected by a
balance of factors, such as: regional variation, racial and ethnic mix,
urban and rural populations and different income levels. The nine states

selected were: (1) Alabama, (2) Colorado, (3) Georgia, (4) Iowa, (5)

14
Kentucky, (6) Maine, (7) Massachusetts, (8) Mississippi, and (9) S.

Carolina.

Monitors visited at least every fourth household. When there were
no answers or refusals, the monitor then visited the preceding household.
All household interviews were conducted in confidence and non
information on individuals visited was released to anyone outside of the
Children’s Defense Fund. C.D.F. calculated children out of school by
two different measures: (1) children who missed 3 or more consecutive
months of school, and (2) children who missed 45 days of school whether
or not it was consecutive. All monitors used an identical household
questionnaire prepared by C.D.F.

Suspension data submitted to the Office of Civil Rights (O.C.R.)
was also analyzed. Detailed analysis of O.C.R. suspension data covered
more than 2,800 school districts and 24,188,681 students. The districts
reporting to the OCR. are not a random sample of all districts in the
nation. Rather they are a census of all school districts containing
substantial minority enrollments. Furthermore, the set of districts
represents more than 50% of the total school enrollment in the nation and

almost 90% of the total minority school enrollment.

Findings
At the time of the study, C.D.F. found that public school

suspensions are rampant all over America and deprive hundreds of

15

thousands of school children annually of needed education. Use, grounds,
procedures and the lengths of suspensions vary widely between schools,
even in schools within the same district. The vast majority of school
suspensions in C.D.F. survey were for non-dangerous, nonviolent offenses
which did not have a seriously disruptive effect on the educational
process. Sixty-three point four percent of the suspensions were for non-
dangerous reasons, such as truancy and tardiness (24.5%), behavior
problems (13.6%), arguments (8.5%), and other; smoking, drugs, alcohol,
dress code, misc. punishment related offenses, i.e., failed to write 1,000
word essay or stay after school (16.8%). Approximately 1/3 of the
suspensions did involve fighting, however, all but a minuscule proportion
were fights between students and did not involve violence against school
faculty or staff. Less than 3% of the suspensions were for destruction
of property, use of alcohol or drugs or other criminal activity. In some
surveyed areas, the percentage of students suspended for truancy and
tardiness ranged from 30 to 50%. In short, according to C.D.F. survey, .
. most students were suspended for non-dangerous offenses. Sixty-three
point four percent of the suspensions were for infractions of school rules,

with almost 25% of which was for truancy and tardiness.

 

16

 

C.D.F. Survey - Reasons for Suspensions:

Fighting (physical conrtact) 36.6% 1

Truancy and tardiness 24.5%

Other: smoking, drugs, alcohol, g 63.4%

Misc. punishment related 16.8% Non-dangerous

Offenses

Behavior problems 13.6%

Arguments 8.5% ,

 

 

 

 

The largest number of suspended students were to be white,
however, proportionately more black children were suspended. At the
secondary level and overall, black students were suspended almost twice
as much as white students. The C.D.F.’s survey corroborates O.C.R.’s
data regarding the disproportionate suspension of minority students.
Blacks in the C.D.F. survey were suspended more than three times as
often as white students and black students more frequently received
multiple suspensions. Additionally, the O.C.R. data indicated that black
students were suspended longer than whites, however, the C.D.F. survey
did not find a racial pattern in the duration of suspensions.

C.D.F. found suspensions to be common in 'all types of school
districts, large city districts, large suburban districts and small urban and
suburban districts. Approximately 1 in every 13 secondary students were
suspended at least once during the 1972-73 school year. Suspension

figures under-state the problem of children out of school in America

17

because it does not include: (1) voluntary withdrawal or push outs, (2)
children put in to special education classes, (3) transfers, (4) drop outs
and (5) those in jail. C.D.F. found that suspended student is likely to
be: poor, from families receiving AFDC or other public assistance, black,
male, in secondary school and a child from single-parent (female~headed)
family.

As previously mentioned, the vast majority of suspensions in
C.D.F.’s survey were for non-dangerous, nonviolent offenses which did,
not have a seriously disruptive effect on the educational process. Thus,
the C.D.F. feels that suspensions are not necessary except in a small
minority of cases to maintain order. “The great majority of suspensions
do not serve any demonstrated valid interest of children or schools.
Instead they harm the child and jeopardize their education. Suspension
pushes children and their problems into the streets, thereby causing more
problems for them and the rest of us," (C.D.F. school suspension
interview).

C.D.F. further maintains that suspensions take away educational
time that may cause marginal, weak or poorly motivated students to drop
out permanently. Suspensions contribute to juvenile delinquency by
putting unsupervised children and those with problems into the streets to
do things they might not have done if they’d been in school.
Suspensions by themselves, do not solve behavior problems, they do not

help children. Suspension simply returns a child to class again with the

18

same problem. Not a single school official interviewed by C.D.F.
contended that suspensions help children.

Suspension seldom benefits children. Usually the child cannot
make up missed schoolwork or tests, resulting in the child getting further
behind. Additionally, C.D.F. maintains a labeling theory perspective, in
that suspension often labels a child as a troublemaker. This label causes
teachers, school officials and other students to foster expectations that
breed misbehavior. The labelling process does sometimes carry over from
one teacher to another. The teacher expects a certain kind of behavior,
namely rebellious behavior or negative behavior from a youngster, and
with that kind of expectation as a preset, the C.D.F. maintains that the
youngster naturally is reinforced into producing that kind of behavior.

According to the C.D.F., the use of suspensions in public schools
has reached mammoth proportions, thus making the issue of suspension
important because the majority of students in America attend public
schools. School officials have traditionally argued that the school acts
“in loco parentis”—as a parent substitute. Contradiction is shown when
school officials act so often to throw a child out of school. Few parents
would take such action. Even the courts have recognized the
“suspensions of just a few days could work substantial harm to a child,"
(Shankley v. Northeast Index. School District). As a result of the in~

depth survey and the harmful conclusions reached, the C.D.F. concludes

19

that reassessment of the underlying rationale and effectiveness of

suspension as an educational tool must be undertaken.

Strengths and Weaknesses of the C.D.F. Survey

The report provided by the C.D.F. titled “School Suspensions: are
they helping children,” is a good descriptive study based on a large
sample. Recall that the study is based on two sources of data: (1) an
extensive household, door-to-door survey (8,500 households), and (2)
analysis of information submitted by school districts to the Federal Office
of Civil Rights (National Census Data). Thus the study utilizes two
primary sources of data from which to draw conclusions about
suspension. The study provides a good characterization of the “suspended
student.” That is, the study characterizes the student who is suspended
most often. The O.C.R. data was not a random sample, rather it was a
census of all school districts (with substantial minority enrollments).
However, due to reporting problems, the census figures are a conservative
count of children out of school in America, thus O.C.R. data likely
understates the problem.

The C.D.F. survey of course provides the benefits common to self-
report studies. The questionnaire was uniform and interviews were
conducted in confidence with assured anonymity, hopefully resulting in
truthful answers by respondents. The C.D.F. study raises a lot of issues

about the effectiveness of school suspensions and brings to the surface

20

the extent of the problem of children out of school in America, at least
as it was in the mid-1970s. The C.D.F. survey illuminated the wide
spread use of suspensions and confirmed the patterns of racial
discrimination.

The C.D.F. drew a lot of conclusions for which there were no
empirical tests. For example, C.D.F. concludes that suspensions
contribute to delinquency by allowing unsupervised children into the
streets to commit delinquent acts. This seems logical, however, the
activities of children during suspension should be investigated not just
assumed. Another conclusion made by C.D.F. fell along the lines of the
Labelling Perspective wherein it is believed that the suspension/labelling
process by school officials, teachers and other students, fosters an
expected negative behavior. The student is reinforced into producing
negative behavior or delinquency as a result of the labelling process.
Again this conclusion is not empirically tested by C.D.F. or at least
backed up by some of the existing labelling theories, most of which have ,
since been unsubstantiated by recent theorists. The C.D.F. survey is
strictly a descriptive study which provides a good data base for further

investigation.

21
Suspended Students—Suspended Learning

Another report on suspensions was provided in a study by the
Philadelphia public schools entitled “Suspended Students—Suspended
Learning” (1982). This report was a collective effort by staff and
volunteers in a Philadelphia’s Parent Union, conducted over a four-year
period. The parents union analyzed the suspension data that school
districts are required to submit to the U.S. Office of Civil Rights
(O.C.R.). Each school in the system must report on a yearly basis the
number, race and sex of students who have been suspended at least once.
Through this analysis, the parent’s union found that: since 1977,
Philadelphia’s Public Schools have been suspending on the average almost
15% of its students and almost 25% of secondary students at least once.
Overall suspension rates have increased for both minority and non-
minority students. This report presents data on the scope and nature of
the suspension problem in Philadelphia Public Schools.

The report begins by noting two particular court decisions that have
determined what is considered to be proper and lawful use of
suspensions. In Goss vs. Lopez (95 S.Ct. 729; 1075) the right of
students to due process in all suspension cases was established. The
Supreme Court ruled that a suspension is, in effect, the denial of access
to a public education which is guaranteed by the state. In Jones vs.
Gillespie (Philadelphia, Pa. 1970) the process by which a student is

suspended is spelled out in greater detail. The critical issue in these

22

decisions is that although they establish the mechanism by which
suspensions may be levied, they do not in fact resolve the problem of
what constitutes a fair, reasonable and legitimate suspension, state law
is equally vague. Thus the ruling leaves school systems with the
discretion to determine: what are suspendable offenses, which students
will be suspended and for how long. And for the most part, the decision

to suspend a student is left up to individual principals.

Methodology

In examining 72 school handbooks, this study found a wide range
of behaviors were considered serous enough to warrant suspension. Some
of the so-called “misbehaviors” seem absurd, for example: illegal
employment, chewing gum or wearing a hat to class. Also what seems
to be common practice is the use of suspension for repeated acts of
misconduct as well as for the serious offense. Based on the belief that
suspensions were frequently used and abused, the examiner's first
intention was to document the extent to which suspensions were used as
a disciplinary measure by analyzing both the district-wide and the school-
by-sehool rates of suspension. The second intention was to ascertain if
suspension rates vary from school to school and identify those schools
that had a very high or very low suspension rates or dramatic changes

over the last two years. Thirdly, why were students being suspended

23

from school? Data were gathered by interviewing school personnel and
analyzing suspension data submitted to the Office of Civil Rights.

For the 1979-80 and 1980-81 school years, and for every school for
which data were available, the study linked the numbers of students
suspended to the appropriate enrollment figures, as well as to other
school characteristics such as the proportion of the student enrollment
from families on aid to families with dependent children, student and
teacher absentee rates, and the California Achievement Test Scores.
Overall suspension rates (total suspensions/total enrollment), the
suspension rates’for white students (number of whites suspended/white
enrollment) and the non-white suspension rates were calculated for the
entire district and for the individual schools, as well as to compare
variations in these rates across schools with similar or dissimilar
characteristics. The purpose of this report was to get a perspective on

the magnitude of the problem that exists city-wide.

Findings

Fifteen percent of all students, 16.8% of all minority students and
10.08% of all non-minority students were suspended at least once during
the school year. And the pattern of increase over a four year period
(1977-81) showed that the district’s suspension rate increased 3.2% with
the increase being greater for minority students than for non-minority, 3.7

compared to 1.5%. In 1980-81 the districts minority suspension rate was

24
the highest (18%) while the white rate was only 10.8%. Additionally,

these rates of suspension, for both minority and non-minority students,
appeared high when compared to the national rates. Minority students in
Philadelphia were suspended at a rate that is almost three times the
national rate for that group. As expected, the rates of suspension
differed markedly by school level. Almost 25% of Philadelphia
secondary students compared to approximately 6.3% of elementary
students were suspended, with suspension rates increasing for both
secondary and elementary students. There also was evidence of
increasing racial disproportion in the rates of suspension. All suspension
rates reached their highest point in the 1980-81 school year.

In some schools, suspensions were virtually non-existent, while in
others, suspension appears to be a very common occurrence. Of the 68
secondary schools in this study, 34 or exactly 1/2, suspended 25% or
more of their students at least once during the school year. Of these, 13
suspended 40% or more of their students at least once, and 6 schools
suspended 50% or more of their student population at least once. The
point is, that even within districts, there is wide variation in the use of
suspension.

Another concern of this study was to identify and describe school
characteristics associated with the different rates of suspension. Schools
with low suspension rates were characterized by a high degree of

community and parent involvement in the school. They were also

A

25

characterized by a strong personal belief of the principal in the
importance of instruction rather than control. Suspensions were measures
of last resort to be used only when all else had failed. Schools with
high suspension rates were characterized by a low level of parental or
community involvement. Standards were designed to emphasize control
in the school. Suspensions were used as the most common means of
implementing discipline. And schools whose suspension rates have
changed significantly were primarily associated with changes in
personnel.

The report concludes by highlighting some programs that were
developed to eliminate suspension. (Programs not emphasizing “control”
but emphasizing “positive discipline”). Again reference to suspension is
made in negative terms. The trend and belief by many educators is that
suspension is counterproductive and is considered “Negative disciplines.”
One North philadelphia high school principal explained that he began
realizing that “Suspensions just don’t work because they made everybody .
. feel worse. Students feel angry and rejected. They are rejected by the
teacher and principal, they are teased by fellow students and further
punished by angry parents. Rapport between student, home and
authorities in the school breaks down” (Principal Rogers, interviewed
1980-81 school year). It is believed by these authors that student
misconduct and delinquency can and often does reflect a sense of

frustration that students experience when their educational needs are not

26

met. The institution should adjust to meet the needs of the students,
rather than adjusting students (by suspending) to meet the needs of the
institution. Students must be in school to learn, suspension should be a

last resort.

Strengths and Weaknesses of Study

This report, like the one provided by the Children’s Defense Fund,
is basically a descriptive study. Also its analysis is limited to the
Philadelphia School District. However, some of its conclusions were
similar. Similarly: there exists a wide variety of offenses which are
suspendable, suspension rates vary as do the offenses for which a student
can be suspended, even within the same district, also minority students
were suspended at a rate higher than non-minority students. This study
analyzed O.C.R. data which is limited to the number, race and sex of
students suspended at least once during the school year. O.C.R. data
does not provide what the student was suspended for and it is unknown
how many were suspended more than once. The other main source of
data for the Philadelphia study was claimed to be drawn from interviews
with “key school personnel.” However, these “key personnel” were not
identified in the report. Were they teachers? principals? or janitors?
If the key personnel interviewed were school administrators, it is likely

they would not be too receptive to allowing the parent union’s

27

investigators (“outsiders”) to question their use of suspension as

disciplinary measure.

Social Bonding and Delinquency

The effects Social Bonding have on juvenile delinquency will now
be discussed through in-depth review of Hirschi’s Social Bonding Theory.
In a large body of data on delinquency collected in California, Travis
Hirschi contrasts the assumptions of three major perspectives on’
delinquency. Namely: (1) strain, (2) control and (3) Cultural deviance
theories. In 1969 Hirschi published his major findings in “Causes of
Delinquency.” Hirschi begins by outlining the assumptions of these
theories. Then by drawing from many sources, he criticizes strain and
cultural deviance theories and maintains a social control perspective
through which he advocates a Social Bonding Theory.

Traditionally, theorists study why some people break the law or
why some adolescents participate in delinquency. Hirschi focuses on why

some do not take part in illegal activities. Basically, Hirschi feels that

 

adolescents do not take part in illegal activities because of strong bonds
to conventional society. The elements of the bond as described by
Hirschi consist of ATTACHMENT of the individual to others,
COMMITMENT to conventional lines of action, INVOLVEMENT in
conventional activities and BELIEF in legitimate or conventional values.

Recall that control theory maintains that the higher the level a person

28

obtains of each of these elements, the less likely they are to commit
delinquent acts. The more tightly bonded an adolescent is to
conventional social groups, the less likely delinquent.
The Sample and The Data

Hirschi’s study sets out to explain why some adolescents do not
break the law. The sample was drawn from 17,500 students entering the
eleven public junior and senior high schools in Western Contra Coast
County, California in the Fall of 1964. The population was stratified by
race, sex, school and grade. The subjects were selected randomly for
inclusion in the sample. A stratified probability sample of 5,545 students
was produced: 1,479 black males, 2,126 non-black males, 1,076 black
females and 864 non-black females. Of the 5,545 students in the original
sample, complete data were eventually obtained on 4,077, or 73.5%. The
data came from 3 sources: (1) school records; (2) the questionnaire (self-

report survey) completed by the students; and (3) police records.

Delinguency Defined

“Delinquency is defined by acts, the detection of which is thought
to result in punishment of the person committing them by agents of the
larger society” (Hirschi, 1969, 47). Delinquency was measured by many
items on the questionnaire which dealt with delinquency or deviant
behavior. Specifically six items were included in a delinquency scale:

(1) petty larceny; (2) grand larceny; (3) auto theft; (4) vandalism

A

29

(malicious mischief); (5) assault and battery; and (6) all other offenses
that are commonly thought to result in punishment by agents of the larger
society if detected.

According to Hirschi, sex, race, social class, neighborhood,
mother’s employment, the broken home, size of family and so forth, are
the things of which most empirical studies, textbooks and theories of
delinquency are constructed (traditional variables). A strong relation
between race and delinquency can be interpreted by almost any theorist.
Hirschi feels that these variables should be treated as intervening
variables, rather than as direct causes of delinquency. Thus it is
necessary to control these variables in subsequent analysis. Hirschi uses
two traditional variables (social class and race) to describe the
distribution of delinquency in the population. In sum, there is no
important relation between social class as traditionally measured and
delinquency. Social class differences with respect to self-reported
delinquency were very small. The slight differences in self-reported
delinquency among race are not large, and do not make a relation
between race and self-reported delinquency logically necessary.
Furthermore, Hirschi feels that there is no reason to believe that the
causes of crime among blacks are different from those among whites.
Hirschi speculates that black-white official differences may be

exaggerated by differential police activity.

30

Hirschi further comments on some other traditional variables and
delinquency, namely: age, mother’s employment; size of family and
ordinal position (birth order) and broken homes. A positive relation
between AGE and delinquency exists, with middle adolescence being the
time period of maximum delinquent activity. It is also noted that
delinquent activity of dropouts declines after they leave school, further
supporting a relationship between schools and delinquency. Hirschi
found that the mother’s employment in relation to delinquency was not
very strong. However, there exist much consistent research that as
family size increases, the likelihood of delinquency increases. Children
from large families are more likely than children from small families to
have committed delinquent acts. Overcrowding in Hirschi’s sample was
found unrelated. However, family size is related to academic
achievement which is related to delinquency. Also, the middle child was
most likely to commit delinquent acts. Finally self-reports revealed that
differences in behavior between children in broken and unbroken homes, .
, is not great. Only a weak relationship favored intact homes. Yet “the
broken-home, causes delinquency” concept is firmly ingrained in society,
as is the labelling perspective. Both of these ideas are refuted by recent
research.

The family and the school are nearly always the conventional
socializing agencies. Although attachment and commitment to parents

was seen as being the most significant, attachment and commitment to the

31

schools was also found to be significant. The basic elements of the
social bond (attachment, commitment, involvement and belief) will be

reviewed here with particular focus on the schools.

School Attachment

Recall that school attachment is measured by a child’s sensitivity
to the opinions of the school officials. Specifically, does the child like
schools? Or care what teachers think about him? According to Hirschi:
“the boy who does not like school and who does not care what teachers
think of him is to this extent free to commit delinquent acts....if a person
feels no emotional attachment to a person or institution, the rules of that
person or institution tend to be denied legitimacy....the child who doesn’t
like school or does not care what teachers think, is likely to believe that

the school has no right to control him” (Hirschi, 1969, 127).

Like School. Hirschi found the relation between liking school and
delinquency to be very strong, represented by a correlation of .18. Over
79% of the boys committing two or more delinquent acts were found
among the 55% who were indifferent to or dislike school. 49% of the
boys who dislike school have committed two or more delinquent acts in
the previous year, compared to only 9% of those who like school. In
1985 Richard Lawrence found similar results. Tests for significant
differences between the means in school performance scores of the

delinquent and non-delinquent youths, resulted in statistical significant

32

differences. All of the items measuring school performance showed
significant differences between the two groups. Particularly the mean
school performance scores for delinquents liking school was 3.44
compared to 3.90 for nondelinquents, resulting in a t of 2.96 (P <
.01)(Richard Lawrence, 1985). Hirschi also found that ability and
performance (measured by English Average Marks) is related to attitudes
toward school, (Liking school), the correlation being .23. Similarly, Polk
and Schafer found that failing students are more likely to see the school
in negative terms. That is, the failing student is more likely to down-
rate the public school system and an increased likelihood of agreeing that
school is “dull and boring.” Only 33% of D/F students versus 60% of
A/B students rated the public school system as “very good.” And 46%
of D/F students agreed that “school is dull and boring” compared to 23%

of A/B students.

Care What Teachers Think. The question “Do you care what
teachers think of you?” is a central measure of attachment to school.
Hirschi found that “the less a boy cares about what teachers think of
him, the more likely he is to have committed delinquent action. This
relation too is very strong and is congruent with control theory” (Hirschi,
1969, 123). This was particularly evident with self-reported delinquent
acts wherein 36% of those without concern for teacher’s opinions had

committed two or more delinquent acts, whereas only 12% of those with

33

a lot of concern for teacher’s opinions, committed two or more delinquent

acts.

School Commitment roofs of dedication to school

School commitment was measured by the child’s educational
aspirations and expectations. Commitment to education was also
measured by several items on Hirschi’s questionnaire. These questions
formulated an “Achievement Orientation Index,” that is questions
designed to measure those that are committed to educational success.
Additionally, a child’s commitment to the school was measured by

general school aptitude.

Educational Aspirations and Expectations. Admittedly, it is
difficult to measure the intensity of aspirations. Hirschi measured
educational ASPIRATIONS by asking: “How much schooling would you
like to get eventually?” Hirschi concludes that the higher the student’s
educational aspirations, whether the student is black or white, the less
likely the student is to commit delinquent acts (by both self-report and
official measures). Similarly, Polk and Schafer concluded that the
educational aspirations of the uncommitted delinquent youth are lower
than those of the successful and committed non-delinquent youth (Schools
and Delinquency, 1972). Hirschi measured EXPECTATIONS by asking:
“How much schooling do you actually expect to get eventually?”

Aspiration levels in the sample were high. Unfortunately, educational

34

expectations in the sample were as high as aspirations. Thus producing
only minor discrepancies as the difference between aspiring to graduate
from college and the expectation of attaining only some college as
insufficient to produce delinquency. In other words, the discrepancies in
educational aspirations and expectations are not important antecedents of

delinquency in Hirschi’s sample.

Index of Achievement Orientation. Those committed to
educational success as evidenced by current efforts should be least likely,
according to control theory, to commit delinquent acts. Several items in
Hirschi’s questionnaire tap this kind of commitment. For example: “I try
hard in school” and “how important is getting good grades to you
personally?” Similar items were combined to produce an Index of
Achievement Orientation. Hirschi found a relation between achievement
orientation and self-reported delinquency. He concludes that the
ambitious (those that try hard in school, those that personally feel it is
important to get good grades, etc.) are much less likely than the

unambitious to have committed delinquent acts.

General School Aptitude. Hirschi measured General School
Aptitude by: the students academic competence; self-rating on school
ability and intelligence. The best measure of success in school is
undoubtedly grade~point average. However, overall grade-point averages

were not available when Hirschi’s analysis was conducted. However,

35

grades in required subjects such as English and Math were available for
most students in the sample. Hirschi measured academic competence by
analyzing DAT scores and English Average Marks.a The correlation
between Differential Aptitude Test Verbal Scores and average marks in
English is .39. Also the prediction that average marks are more closely
related to delinquency than test scores is confirmed. Hirschi concludes
that the better a student does in school, the less likely he is to have
committed delinquent acts and the less likely to have been picked up by
the police.

Richard Lawrence also found that the delinquent group scored
significantly lower on the intelligence scale (GPA). This finding is
supported by other research as well (Cattell and Cattell, 1975; Pierson
and Kelly, 1963, Polk and Schafer 1972). Lawrence found the mean GPA
for non-delinquents to be 3.50 compared to 2.565 for delinquents (t=6.02
p < .001). In Lawrence, several of the school performance variables
were significantly predictive of official delinquency. The discriminent.
function was statistically significant (Wilk’s Lambda = .51; X2 = 110.8;
df = 13; p = .00) and it accounted for 49% of the total variance
explained. The variable the most predictive of official delinquency status
was “lower intelligence” .64 and “lower grades” .37, ranking fourth in
order of their contribution.

Hirschi also suggests that the more competent a student thinks he

is the less likely the student is to commit delinquent acts. Thirteen

36

percent of those who consider themselves “among the best” in school
ability have committed two or more delinquent acts, while 35% of those
who consider themselves “below average” in school ability have
committed two or more delinquent acts. Hirschi concludes that the more
academically competent a boy is and/or the more competent he sees

himself to be, the less likely he is to be delinquent.

School involvement

The school does more than prepare students for the future. It acts
also as a holding operation; it attempts to engross and involve students
in activities. Historically, “idle hands” have been viewed as the “devil’s

workshop.’ In other words, unoccupied youths are likely to get involved
in mischief and of course the more time unoccupied, the greater the
likelihood of delinquency. Hirschi points out that as defined,
delinquency requires very little time, negating the “idle hands are the
devil’s workshop” theory. Furthermore, Hirschi states that the leisure
activities (such as TV, reading, hobbies, sports, etc.) of the boys in the
sample are collectively unrelated to the commission of delinquent acts.
However, all of the indicators of time spent in a concern for homework,
are reasonably strongly related to the commission of delinquent acts.

Also, attendance at school activities and participation in interscholastic

athletics are reviewed.

37

Time Spent in and Concern for Homework. School involvement
is traditionally measured by time spent in and concern for homework.
Hirschi provides several indicators of attention to homework on the
questionnaire. A relation between time devoted to homework and
delinquency is evident. Of those committing one or more delinquent acts,
64% dedicated less than a half hour per day doing homework. As
compared to 34% who spent 1 and 1/2 hours or more per day studying.
In other words, the less time studying, the greater the chances for.
delinquency.

In a study by Polk and Schafer, the question was asked “In the
average week, how many hours do you spend at schoolwork at home?”
(homework). Polk and Schafer concluded that the “uncommitted youth,
it would appear, is characterized by behavioral withdrawal from school.
He does not study, he receives poor grades, and does not participate in
activities” (Polk and Schafer, 1972, 85). Thus, delinquency may be a
function of the lack of commitment to school at least for some of the

uncommitted youths.

Attendance at School Activities. All of the items measuring
school performance in Lawrence’s study, showed significant differences
between delinquents and non-delinquents. Non-delinquents showed more
attendance at school activities than did delinquents. The mean score of

non-delinquent’s attendance at school activities was 1.56 compared to .91

 

38

for delinquents (t = 3.51 and p < .001). The zero order correlations
between school performance and self-reported delinquent behavior provide
support for a relation between two variables. Lack of participation in
school activities is associated with greater likelihood of involvement in
delinquent behavior. Also, several of the school performance variables
were significantly predictive of official delinquency adjudication. The
variables most predictive of official delinquency status were (in the order
of their contribution): lower intelligence, more school rule violations, and
less attendance at school activities. Discriminant function coefficient for
participation in school activities being .32 p < .001. These findings
provide support for Hirschi’s (1969) claim that attachment to school and
involvement in school activities provide a containment against
involvement in delinquent behavior.

In the study by Polk and Schafer (1972), the question was asked
“In the average week, how many evenings do you spend at school-related
activities?” Recall that Polk and Schafer conclude that the uncommitted
delinquent youth is characterized by behavioral withdrawal from school,
he does not study, he receives poor grades, and he does not participate
in school activities. And that delinquency among at least some youth
may be a function of the lack of commitment to school. Also, Polk and
Schafer’s interpretations lean toward the view that commitment is
temporally prior to delinquency, but admit that the data of this study do

not support this sequence over any alternative sequence.

39

Participation in Interscholastic Athletics. Interscholastic athletics
are supported by community members and school administrators partly out
of the belief that participation in sports is an effective deterrent to
delinquency. Polk and Schafer draw from numerous theories to predict
that participation in interscholastic athletics will have a deterring
influence on delinquency. For example, athletes should be delinquent
less often than non-athletes on the basis of the theoretical reasoning of
Differential Association. Wherein delinquency is contingent upon the
amount of exposure to deviant subgroups and inversely with exposure to
conforming influences. Coaches usually set strict standards of behavior
for on and off the field. Most athletes internalize these standards and
exert pressure on other athletes to conform as well. The athlete is more
likely to be exposed to strong conforming influences than the non-athlete.
Albert Cohen, Arthur Stinchcombe, Kenneth Polk and Walter Schafer
contend that delinquency is often the result of rebellion against the
school. For Polk and Schafer, rebellion sometimes rises out of failure.
For some youths, school is frustrating, an outcome of which is overt
rebellion in the form of illegal behavior. For successful athletes, school
is less likely to be a source of frustration than for the comparable non-
athlete. In short, participation in interscholastic athletics can be expected
to exert a deterring or negative influence on delinquent behavior. In
testable terms, athletes should be delinquent less often than comparable

non-athletes. In Schafer’s sample, as predicted. athletes are less often

A

4o

delinquent than non-athletes. Ninety-three percent of the athletes were
considered nondelinquent with the remaining 7% of athletes delinquent.
Comparatively, 83% of the non-athletes were considered non-delinquent

with the remaining 17% of non-athletes as being delinquent.

M

The final element of the social bond is belief. Beliefs are the
major variables in most sociological explanations of delinquent behavior.
To some, beliefs appear to be the key independent variable; delinquency
results when the norms have not been internalized. Within control theory
tradition, the relevant beliefs are more or less universally held within the
society, but are variably “neutralized” or explained other ways. For
control theorists; delinquency is not caused by beliefs that require
delinquency but is rather made possible by the absence of (effective)
beliefs that forbid delinquency. Since this study is concerned with
delinquency and the schools, belief and respect for the law will be

reviewed in addition to a child’s acceptance of school authority.

Respect for Polige. We have seen the effects of attachment to
parents and teachers. Hirschi found an equivalent relation between
“respect for the police” and delinquent behavior. Hirschi concludes that
lack of respect precedes delinquent acts and does not simply follow from
contact with the police. The relation between self-reported delinquency

and respect for police was evident in Hirschi’s sample. The question was

41

asked “I have a lot of respect for the Richmond Police?” Responses
ranged from strongly agree to strongly disagree. Forty-five percent of
those that did not respect the Richmond Police, reported having
committed two or more delinquent acts. Comparatively, 12% of those
having a lot of respect for the police, reported having committed two or
more delinquent acts, and 71% of those with a lot of respect for the

police reported no delinquency involvement.

Lack of Respect for the Law. Lack of respect for the police

presumably leads to lack of respect for the law. Hirschi found “attitude
toward the law” to be strongly related to self-reported delinquency.
There is variation in the extent to which boys believe they should obey
the law, and the less they believe they should obey it, the less likely
they are to do so. In Hirschi’s sample, the question was asked “It’s
alright to get around the law if you can get away with it?”. Forty one
percent of those that “strongly agreed” and 45% of those that “agreed”,
reported committing two or more delinquent acts. Comparatively, 15%
of those that “disagreed” and 9% of those that “strongly disagreed”

reported committing two or more delinquent acts.

Acceptance of School Authority, In regard to the school, beliefs
are measured by a child’s acceptance of school authority. Hirschi
measured feelings about the scope of the school’s legitimate authority by

asking the question “It is none of the school’s business if a student wants

42

to smoke outside of the classroom?”. Hirschi also measured self-reported
delinquency by age at which cigarette smoking began. Hirschi suggests
although smoking is a better predictor of delinquent activity than the
attitudinal measure, the attitudinal measure is independently related to
delinquency.

Acceptance of school authority was measured by “school rule
violations” and “truancy” in Richard Lawrence’s study (1985). The mean
score for school rule violations for non-delinquents was .73 as compared.
to 1.67 for delinquents (t = 5.40 and p < .001). Additionally, the mean
score for truancy for non-delinquents was .97 as compared to delinquents
1.44 (t = 2.39 and p < .05). School rule violations and truancy seem to
be predictive of involvement in self-reported delinquent behavior. To
determine the relative amount of delinquent behavior explained by school
performance and personality factors together, multiple regression analysis
was performed using the predictors having significant zero-order
regression coefficients. Number of school rule violations, attachment to
school and truancy were found to be the most important school
performance variables. The variables most predictive of official
delinquency status were (in the order of contribution): lower intelligence,
more school rule violations (-.436 P < .001), less attendance at school
activities...more truancy (.27 p < .01). Again, these findings provide

support for Hirschi’s (1969) claim that attachment to school and

43

involvement in school activities provide a containment against

involvement in delinquent behavior.

The Causal Chain

Hirschi concludes with a discussion of the effects of school
experience on delinquency. Hirschi presented a simple causal chain and
examined data relevant to it. The causal chain runs from: ACADEMIC
INCOMPETENCE, TO POOR SCHOOL PERFORMANCE, TO DISLIKING
OF SCHOOL, TO REJECTION OF SCHOOL’S AUTHORITY, TO THE
COMMISSION OF DELINQUENT ACTS. All statistical relations relevant
to this chain have been presented, and all are in fact consistent with it.
It was shown that the higher a boy’s score (the better a student does in
school), the less likely to have committed delinquent acts and the less
likely to have been picked up by police. Furthermore, the more
academically competent a boy is and/or the more competent he sees
himself to be, the less likely he is to be delinquent, regardless of his
position in the opportunity.

If the link between ability and performance and delinquency is the
bond to the school, then ability and performance must be related to
attitudes toward school. The correlation between Differential Aptitude
Test Verbal Scores and Liking School is .11. The correlation between
English Average Marks and Liking School is .23. The statistical data to

this point are consistent with the causal chain. Recall that 49% of the

44

boys who dislike schOol have reported committing two or more
delinquents acts in the previous year, compared to only 9% of those who
like school. In control theory, delinquency is not seen as compensation
for previous frustration and failure; it is not seen as an alternative route
to some remote goal. Boys who do badly in school reduce their interest
in school (they may of course actually come to hate it) and are thus free
to this extent to commit delinquent acts.

The boy who does not like school and who does not care what
teachers think of him is to this extent free to commit delinquent acts.
The less a boy cares what teachers think of him, the more likely he is to
have committed delinquent acts. Thirty six percent of those that didn’t
care very much what teachers thought, reported having committed two or
more delinquent acts compared to only 12% of those who care a lot about
what teachers thought. The items used to measure attachment to the
school, “Do you like school?” and “Do you care what teachers think of
you?” are, as would be expected, substantially related to an item designed
to tap feelings about the scope of the school’s legitimate authority; “it’s
none of the school’s business if a student wants to smoke outside of the
classroom”. This item, in turn, is very strongly related to delinquency.
Thirty eight percent of those that “strongly agreed” as well as those that
“agreed,” reported two or more delinquent acts compared to only 12% of

those that “disagreed” and 11% of those that “strongly disagreed.”

45

Beginning, then, with variation in academic competence, we have
traced a path through attachment to the school and support of the
school’s authority to delinquency. Let us now look at the simultaneous
effects of the variables. First, the set of school items, accounts for more
of the variance in delinquency than the father-mother items combined.
(The coefficient of multiple correlation is .41, whereas for the father-
mother items it is .36.) Second, as was expected, academic competence
has little direct effect on the commission of delinquent acts when the
effects of school performance and attitudes toward the school are
removed. Third, the effects of academic performance are not eliminated
when attachment to the school is taken into account. Fourth, when the
effects of academic competence, grades, and attitudes toward the school
are taken into account, the relation between self-perceived school ability
and self-reported delinquency disappears. Fifth and finally, the three
attitudinal measures, retain an effect on delinquency when the effects of
the competence and performance variables are removed. See Table l in _
the Appendix for self-reported delinquent acts and selected school items

(Hirschi, 1969, 130).

M

This literature review was designed to inform the reader of the
nature and use of suspensions as they are used in American public

Schools. It was discovered in three major studies of school suspension,

46

that suspensions are rampant all over America, and found to be common
in all types of school districts. Use, grounds, procedures, and lengths of
suspensions vary widely between schools, even in schools within the same
district. The vast majority of school suspensions were non-dangerous,
non—violent offenses; 63.4% of the reasons for suspension in one study
were for truancy and tardiness, behavior problems, arguments, smoking,
drugs, alcohol, dress code and miscellaneous punishment-related offenses
(non-dangerous reasons). In all the studies, minority students were
suspended at a higher rate than non-minority students. In addition to the
frequent use of suspensions, the problem with suspension is, they serve
no demonstrative valid interest of children or school and may in fact
harm the child by getting further behind. Suspensions may cause
marginal, weak or poorly motivated students to dropout permanently.
Additionally, it is believed that suspensions may contribute to juvenile
delinquency by putting unsupervised children into the streets.
Additionally, this literature review was designed to incorporate
theory. Thus an overall look at the elements of the social bond and their
relationship to juvenile delinquency was reviewed. Travis Hirschi found
the elements of the social bond in relation to education as being:
measured by the child’s sensitivity to the opinion of the teachers
(ATTACHMENT to the school); measured by the child’s educational
aspirations and expectations (COMMITMENT to the school); measured by

the amount of time spent in school activities, especially doing homework

47
(INVOLVEMENT in school); and measured by the child’s acceptance of

the school’s authority (BELIEF in educational goals). In short, it was
found that the higher the level a student obtains of each element, the less
likely they will commit delinquent acts. These concepts will be
incorporated into the “Design of study,” to be described next, in Chapter

3.

CHAPTER 3

DESIGN OF THE STUDY

Introduction

This chapter will present the methodology used in the study to test
whether a relationship exists between school suspension and delinquency.
And to test the relationship between juvenile delinquency and the social
bonding model. In this chapter, the sample population and data
collection procedures, the variables, the testable hypotheses, and the

subsequent data analysis techniques will be described.

Sample Population and Data Collection Procedures

The data for this study were made available (in part) by the Inter-
University Consortium for Political and Social Research, Ann Arbor,
Michigan. The data for the National Youth Survey (N.Y.S.) was
originally collected by Delbert Elliott beginning in 1976. The National
Youth Survey is a longitudinal study of delinquency and drug use among
American youth. The study design called for an initial survey in 1977
with a national probability sample of youths aged 11 to 17. The total
youth sample was initially interviewed between January and March of

1977 concerning their involvement in delinquent behavior and drug use

48

49

during the calendar year, 1976. Second, third, fourth and fifth surveys
were conducted during this same time period in successive years. By the
third survey (1979), the panel was 15 through 21 years of age; the
National Youth Survey employed a probability sample of households in
the continental United States in 1976 based upon multistage, cluster
sampling design. Seventy-six primary sampling units were selected, with
probability of selection being proportional to size. This sampling
procedure resulted in the listing of 67,166 household of which
approximately 8,000 were selected for inclusion in the sample. All
youths living in the selected households who were 11 through 17 years
on December 31, 1976 and were physically a mentally capable of being
interviewed were eligible respondents for the study. The selected
households generated an estimated 2,360 eligible youths. Of these, 635
(27%) did not participate in the study due to (1) parental refusal, (2)
youth refusal, or (3) an inability to make contact with the respondent.
The remaining 1,725 (73%) agreed to participate in the study, signed
informed consents, and completed interviewed in the initial (1977)
survey. The loss rate from any particular age, sex or racial group
appears to be proportional to that group’s representation in the
population. Further with respect 'to these characteristics, participating
youth appear to be representative of the total 11 through 17 year-old
youth population in the US as established by the US Census Bureau.

Finally, the respondent loss over the first three surveys was relatively

50
small. The loss rate for the 1978 survey was 4% (N = 70), and for the

1979 survey, the cumulative loss increased to 6% (N = 99). The
representiveness of the sample has not been affected in any serious way

by the loss over the first three surveys.

The Variables

The Dependent Variable. Juvenile Delinquency is measured
through self-reports of the number of times the student committed certain
acts within the last twelve months (Christmas a year ago to the Christmas
just past). Eighteen specific delinquent acts make up seven categories of
offenses: (1) drug offenses; (2) weapon offenses; (3) Breaking and
Entering a building or car; (4) vandalism; (5) theft, major and minor; (6)
assault, excluding sexual; and (7) robbery. The eighteen acts that make
up these categories are listed on the next page. Delinquency is further
subdivided into non-delinquent, occasional delinquent and serious

delinquent.

Non-Delinquent. A non-delinquent is defined as a surveyed youth,
between the ages of 11 and 19 , claiming to have not participated in any
of the seven categories of delinquency during the calendar year prior to

the survey. That is, Christmas a year ago to the Christmas just past.

Occasional Delinquent. An occasional delinquent is defined as a

Surveyed youth, between the ages of 11 and 19, claiming to have

51

participated in one or two acts of delinquency during the calendar year.

Serious Delinquent. A serious delinquent is defined as surveyed
youth, between the ages of 11 and 19, claiming to have participated in
three or more acts of delinquency during the calendar year prior to the
survey.

Categories of Delingpencv
1. Drug offenses

1.Marijuana sales
II. Weapon offenses

2.carrying hidden weapons
111. Breaking and entering

3.building or car
IV. Vandalism

4.school property

5.family property

6.other property

V. Theft

7.stole motor vehicle

8.stole more than $50

9.stole $5 to $50

10.Stole less than $5

VI. Assault and battery

52

11.hit teacher

12.hit parent

13.hit students
14.gang fights
15.aggravated assault

VII. Robbery

l6.strong-arm student
l7.strong-arm teacher

18.strong-arm others

The Independent Variables. The independent variables are school
suspension, and the elements of the social bond: school attachment,
school commitment, school involvement, and belief in school authority.

These variables will be discussed in further detail.

School Suspensions. Are measured through self-reports of the
number of times suspended within the past twelve months (Christmas a
year ago to the Christmas just past). Suspension is divided into non-
suspended and suspended students with the total number of suspensions
calendar year, per student (frequency) is accounted for. School

suspension is hypothesized to be positively related to delinquency.

53
Non-Suspended Student

A non-suspended student is defined as a junior or high school
student, ranging from 6th to 12th grade, between the ages of 11 and 18
years of age, claiming never to have been suspended from school during
the past twelve months prior to survey. The category response for

“never” been suspended last year is (0).

Suspended Student
A suspended student is defined as a junior or high school student,
ranging from 6th to 12th grade, between 11 and 18 years of age, claiming
to have been suspended from school at least once in the past twelve
months prior to survey. The category responses: (1), (2), (3), (4),
(5)...corresponds to the number of times suspended in the past twelve

months.

School Attachment. Is measured by five categories which depict
whether a child is sensitive to teacher’s opinions and whether a child
likes school. Specifically, how important is it to the child to have
teachers think of them as a “good student” and how important is it to
make a “good impression,” rather than tell the truth to teachers.
Additionally, whether or not a child likes school is measured by certain
feelings the child has about school, such as: feeling like NOBODY AT
SCHOOL CARES about me or feeling like I DON’T BELONG AT

SCHOOL, or feeling LONELY AT SCHOOL. Unfortunately, the direct

54

question “Do you like school?” was not asked in the National Youth
Survey. Sensitivity to teacher’s opinions and liking school are

hypothesized to be negatively related to delinquency.

Good Student. Sensitivity to teacher’s opinion is measured by
asking the student: “How important is it to you, to have teachers think
of you as a good student?” The category scales are: very important (5);

somewhat important (3); and not important at all (1).

Good Impression. Sensitivity to teacher’s opinion is further
measured by asking the student: “Making a good impression is more
important than telling the truth to teachers?”. The category scales are:
strongly agree (1); agree (2); neither agree or disagree (3); disagree (4);

and strongly disagree (5).

Nobody at School Cares. Whether a child likes school was
measured by certain feelings a child has about school. The question was
asked “I often feel like nobody at school cares about me?”. The response
scales are: strongly agree (1); agree (2); neither agree or disagree (3);

disagree (4); and strongly disagree (5).

Belong at School. Whether a child likes school was also measured
by a feeling of not belonging at school. The question was asked: “I

don’t’ feel as if I belong at school?”. The response scales are: strongly

55

agree (1); agree (2); neither agree or disagree (3); disagree (4); and

strongly disagree (5).

Lonely at School. Whether a child likes school is further
measured by a child’s feeling of loneliness at school. The question was
asked: “Even though there are lots of kids around, I often feel lonely at
school?”. The response scales are strongly agree (1); agree (2); neither
agree or disagree (3); disagree (4); and strongly disagree (5). A
combined “agreement” to two or more of the child’s feelings about school

(measures c, d, or e) will be construed to qualify as not liking school.

School Commitment. Is first measured by the child’s educational
ASPIRATIONS and EXPECTATIONS, which is measured by the
importance placed on going to college (aspirations) and the belief in
chances for finishing college (expectations). Secondly, GENERAL
SCHOOL APTITUDE is used to measure school commitment. General
School Aptitude is defined by the student’s academic competence, that is;
self-rating on school ability and intelligence. Intelligence is measured
by a student’s grade point average. Thirdly, school commitment, those
committed to educational success is further measured by ACHIEVEMENT
ORIENTATION. Several items from the questionnaire are used to tap
commitment, these items will be explained in the Index of Achievement
Orientation. It is hypothesized that school commitment is negatively

related to delinquency.

56

Educational Aspirations and expectations. Aspirations are
measured by asking: “How important is it to you to go to college?”. The
response scales are: very important (5); somewhat important (3); and not
important at all (1). Expectations were measured by asking: “What do
you think your chances are of completing a college degree?”. The

response scales are: good (3); fair (2); and poor (1).

General School Aptitude. Academic competence is first measured.
by the student’s self-rating on school ability and intelligence. Three
measures of self-rating on school ability are found by asking the student:
first, how well they think they are doing or whether they think they are
a good student. Secondly, how well they think they are at having a high
GPA and thirdly, how well they think they are doing in hard subjects.
The response scales are: very well (5); ok (3); and not well at all (1).
Intelligence is measured by asking the question: “What is your grade
point average?”. The response scales are: mostly A’s (5); mostly B’s (4);

mostly C’s (3); mostly D’s (2); and mostly F’s (1).

Index of Achievement Orientation. Several items are used to tap
commitment to educational success. “How important has you school work
been to you?”. “How important is it to you to do well even in hard
subjects?”. “How important is it to you to have a high grade point
average?”. And “how important is it to you to do your own school work

without help from anyone?” The response scales are: very important (5);

57

somewhat important (3); and not important at all (1). The previously
listed four items combine to produce an Index of Achievement
Orientation. The higher the student scores on the achievement index; the
more committed they are to educational success and theoretically, the
more committed a child is to school, the less likely the child is to be

delinquent.

School Involvement. Since Hirschi found that “leisure activities”
were collectively unrelated to delinquency, the focus here will be on time
spent in SCHOOL ACTIVITIES, particularly time spent in doing
SCHOOL WORK. Additionally, interest and participation in ATHLETICS
or other school activities, is also used to measure a students involvement

in school.

Time Spent Doing School Work. Time spent on school work was
measured by the amount of time spent studying during the weekday
afternoons and evenings and during the weekends. “On the average, how
many afternoons during the school week, from the end of school to
dinner, have you spent studying?”. And “on the average, how many
evenings during the school week, from dinner to bedtime, have you spent
studying?”. For both questions, responses range from 0 to 5.
Additionally, the question was asked: “On the weekends, how much time
have you generally spent studying?”. The response scales are: a great

deal (5); quite a bit (4); some (3); not to much (2); and very little (1).

58

Time Spent on School Athletics. First it is ascertained whether
a student is a member of any athletic teams at school. Available
responses are: yes (2) or no (1). Then involvement in athletics was
measured by the amount of time spent on athletics during the weekday
afternoons and evenings and on weekends. “On the average, how many
afternoons during the school week, from the end of school to dinner,
have you spent on team activities?”. And “on the average, how many
evenings during the school week, from dinner time to bed time, have you
spent on team activities?”. For both questions, responses range from 0
to 5. Additionally, the question is asked: “On the weekends, how much
time have you generally spent on team activities?”. The response scales
are: a great deal (5); quite a bit (4); some (3); not too much (2); and

very little (1).

Time Spent on Other School Activities. It is first ascertained if
the student has taken part in any activities at school, other than athletics,
such as; service clubs, recreational or hobby clubs, student government,
newspaper and or yearbook. The responses are: Yes (2) or No (1). Then
involvement in this activities is measured by the amount spent on these
activities during the weekday afternoons, evenings and on weekends. “On
the average, how many afternoons during the school week, from end of
school to dinner, have you spent on these activities?”. And “On the

average, how many evenings during the school week, from dinner time to

59

bed time, have you spent on these school activities?”. For both
questions, responses range from 0 to 5. Additionally, the question is
asked: “On weekends, how much time have you generally spent on these
school activities?”. The response scales are: a great deal (5); quite a bit

(4); some (3); not too much (2); and very little (1).

Bel—k; Belief is the final element of the social bond. Recall that
Hirschi maintains that conventional beliefs are more or less universally
held within the society. Through the use of the National Youth Survey
data, this belief is measured by asking the respondent their feelings about
cheating on school tests, drug use and delinquency (CONVENTIONAL
NORMS AND ATTITUDES). Additionally, conventionally beliefs are
measured by BELIEF AND RESPECT FOR THE LAW and ACCEPTANCE
OF SCHOOL AUTHORITY. It is hypothesized that absence of such
beliefs leaves the individual free to commit delinquent acts.
Conventional beliefs, particularly belief in educational goals, is

negatively related to delinquency.

Conventional Norms and Attitudes. To ascertain whether
conventional beliefs are universally held within the society as postulated
by Hirschi, data from the National Youth Survey was used to measure
attitudes toward deviance. Specifically eight questions are asked: “How
wrong is it for someone your age to cheat on school tests?” “Purposely

damage other people’s property?” “Use marijuana or hash?” “Steal

60

something worth less than $5?” “Hit or threaten to hit someone?” “Break
into a vehicle or building to steal something?” “Sell hard drugs such as
heroin, cocaine or LSD?” Or “Steal something worth more than 350?”.
The response scales are: very wrong (4); wrong (3); a little bit wrong
(2); and not wrong at all (1). The higher a student’s cumulative score
for the eight items, the higher or stronger a student’s conventional

beliefs.

Acceptance of School Authority. Determination of acceptance of
school authority is measured by school rule violations and truancy. From
the National Youth Survey Data, school rule violations are measured by
asking: “How many times in the last year have you: ‘cheated on school
tests?’ ‘Hit or threatened to hit a teacher or other adult at school?’ ‘Hit
or threatened to hit other students?’ or ‘purposely damaged or destroyed
property belonging to a school?”. Truancy is measured by asking: “How
many times in the last year have you skipped classes without an

excuse?”. Frequencies are calculated for each question.

Testpble Hypothgses

Introduction. The research hypothesis is' divided into five
elements. School suspension and the elements of the social bond: school
attachment, school commitment, school involvement and belief in

educational success. Each element has several predictors.

61

School Suspension. School Suspensions are indicated by the
number of times suspended in one year.

Null Hypothesis - There is no relationship between school

suspension and the elements of the bond.

Directional Hypothesis - School suspensions are negatively related

to a youth’s attachment, commitment, involvement and belief in

schooL

Null Hypothesis - There is no relationship between school.

suspensions and delinquency.

Directional Hypothesis - School suspensions are positively related

to delinquency.

School Attachment. Five indicators make up the school attachment
predictor of juvenile delinquency. The first two are: (1) concern that
teacher’s think respondent is a good student and (2) making a good
impression. Additionally, “Not liking school” is also an indicator of
school attachment. Not liking school is comprised of three predictors:
(3) feeling nobody at school cares, (4) feelings of not belonging at school
and (5) feeling lonely at school.

Null Hypothesis - There is no relationship between the first two

attachment predictors, “concern for teacher’s opinions” and “making

a good impression” and delinquency.

62

Directional Hypothesis - Concern for teacher’s opinions and making
a good impression are negatively related to delinquency.

Null Hypothesis - There is no relationship between liking school
and delinquency.

Directional Hypothesis - Liking school is negatively related to

delinquency.

School Commitment. School commitment is comprised of three
commitment predictors: (1) educational aspirations and expectations, (2)
general school aptitude and (3) achievement orientation.

Null Hypothesis - There is no relationship between school

commitment and delinquency.

Directional Hypothesis - School commitment is negatively related

to delinquency.

School Involvement. Three indicators make up the school
involvement predictor of juvenile delinquency. (1) Time spent in and
concern for homework, (2) athletics, and (3) other school activities.

Null Hypothesis - There is no relationship between school

involvement and delinquency.

Directional Hypothesis - School involvement is negatively related

to delinquency.

63

Belief in Educational Success. Belief is comprised of two
predictors: (1) belief in conventional norms and attitudes and (2)
acceptance of school authority.

Null Hypothesis - There is no relationship between conventional

beliefs and delinquency.

Directional Hypothesis - Conventional beliefs are negatively related

to delinquency.

Null Hypothesis - There is no relationship between acceptance of

school authority and delinquency.

Directional Hypothesis - Acceptance of school authority, which is

measured by the number of school rule violations, is positively

related to delinquency.

Conclusion. Each of these elements and predictors will be
independently tested and analyzed. The results of this testing will be

reported in Chapter 4.

Data Collection Analysig

The original collection of data on which this thesis is based is the
result of the National Youth Survey. “Survey research is probably the
best method available to the social scientist interested in collecting
original data for describing a population too large to observe directly”
(Earl Babbie, 1983, 209). The analysis of this data will be secondary

analysis. The main advantage of secondary analysis is of course, its

64

avoidance of the enormous expenditure of time and money which a survey
entails. The analysis will be accomplished through univariate, bivariate
and multivariate analysis.

Univariate Analysis will be used to examine only one variable at
a time. Univariate analysis gives summary statistics on the distribution
of the values of both the dependent and independent variables.
Univariate analysis describes the units of analysis of a study and, in this
case are a sample drawn from a larger population, thus allowing one to
make descriptive inferences about the larger population. These summary
statistics are accomplished by using the FREQUENCIES command on
discrete variables (nominal and ordinal levels of data measurement) and
the DESCRIPTIVE command on continuous variables (interval and ratio
levels of data measurement).

In contrast to univariate analysis, subgroup comparisons constitute
a kind of BIVARIATE ANALYSIS, in that two variables are involved.
While univariate analysis and subgroup comparisons focus on describing
the people (or other units of analysis) under study, bivariate analysis
focuses on the variables. Bivariate analysis will be used to analyze the
relationship between two variables, giving tests of significance and
measuring the association between delinquency and the independent
variables. The options to be used for measuring the association between
two variables are: Lambda, Gamma and Pearson’s Product-moment

Correlation. LAMBDA is an appropriate measure of association to be

65
used in the analysis of two NOMINAL variables. GAMMA is an

appropriate measure of association to be used in the analysis of two
ORDINAL variables and PEARSON’S r is an appropriate measure of
association to be used in the analysis of two INTERVAL or RATIO
variables.

MULTIVARIATE ANALYSIS is the third analysis technique to be
used. Instead of one independent variable, like in bivariate analysis, this
technique introduces more than one independent variable. Instead of
explaining" the dependent variable on the basis of a single independent
variable, an explanation is sought through the use of more than one
independent variable, thus allowing the researcher to check for the
spuriousness of the relationship. That is, are the dependent and
independent variables causally related? Or is this relationship the result
of the effects of some third variable.

One method of analysis to be used is MULTIPLE REGRESSION
ANALYSIS, where the dependent variable is affected simultaneously by
several independent variables. Multiple regression analysis’provides a
means of analyzing such situations. A second method of analysis is
provided by FACTOR ANALYSIS, which is a different approach to
multivariate analysis than regression analysis. Factor analysis is an
efficient method of discovering predominant patterns among a large
number of variables. Factor analysis will be used to create a typology

of delinquency. The data file contains several indicators of delinquency.

66

Each item should provide some indication of delinquency but none of
them alone will give a perfect indication. All of these items, moreover,
should be highly correlated empirically.

The STATISTICAL SIGNIFICANCE of a relationship that is
observed in a set of sample data is expressed in terms of probabilities.
The level of significance of this study is set at a .05 level. Which
means the probability of a relationship as strong as the observed one
being attributable to sampling error alone is no more than 5 in 100. In.
other words, the probability of sampling error is discounted when the

stated level of significance is less than five chances in a hundred.

CHAPTER 4

THE FINDINGS

Introduction

The findings presented in this chapter are an accumulation of
univariate, bivariate, and multivariate analysis. Beginning with a
GENERAL DESCRIPTION of: the sample population; the suspended
student; the delinquent youth; and a delinquency typology. Furthermore,
school suspension and the elements of the social bond and their
relationships with delinquency are reviewed. Recall that for this study,
the elements of the social bond are focused on the bonding process in the
educational atmosphere. Thus, school ATTACHMENT, school
COMMITMENT, school INVOLVEMENT, and BELIEF in educational
success are evaluated.

After the general descriptions, the effects of the school suspension
on the social bonds are examined in bivariate analyses. Then the
elements of the bond are tested for relationships with delinquency.
Together, school suspension and select bond elements are combined as
predictor variables and are regressed with delinquency. Regression

analysis is used to assess the relative importance of the various predictor

67

68

variables in their contribution to variation in the criterion variable.
Specifically to determine if the core variables have an effect on total
delinquency, Tables 2 through 10 in the appendix present the results of
these analyses within each questionnaire. The findings for each
questionnaire will be separated and presented in separate sections. Also

the results of hypotheses testing will be disclosed.

General Descriptions of Sample Population
Through the use of UNIVARIATE ANALYSIS, summary statistics

described the sample population by running frequencies. The sample in
this study initially consisted of 1725 junior high and high school aged
youth selected from across the United States. Their age at the first
interview (1977) ranged from 11 to 17. The ages of the sample are
basically evenly distributed, with each age group containing between 13
and 15% of the total sample population. The sample consisted of 53.2%
males (n = 918) and 46.8% females (n = 807). Seventy-eight point nine
percent (n = 1361) of the sample population was white; 15.1% (n = 260)
black; 4.4% (n = 76) hispanic; and 1.6% (n = 28) other. The “parent or
guardian interview” information was not available for all waves, however,
certain family characteristics were described in the first wave. The
principal wage earner in the family (at Wave 1) was the father 70.4%,
with 40.7% of the family incomes being $14,000 per year or less. In the

first wave, 82.1% reported'living in single family homes, 7.3% in

69

apartments, 4.5% in multi-family homes, and 3.3% in mobile homes.
Finally, 18% of the sample population reported being on some kind of

public assistance.

General Description of the Suspended Student

Each year, an average of 13% of the sampled students were
suspended at least once. Male students were consistently suspended twice
as often as female students (suspended once) and three times more often
than females who were suspended two or more times. Recall that the
majority of the students were white (78.9%), the remaining 21.1% were
minorities, 15.1% of which were black. Although the minority students
were only l/5th of the sample population, minorities were suspended at
a rate twice that of white students, and three or four times the rate of
white students who were suspended two or more times. For the first two
waves, it was discovered that the older the student, the larger the

percentage of students suspended, as was evident for students suspended

two or more times in Wave 1: (age 11 = .4% suspended), (age 12

1.6%), (age 13 = 5.2%), (age 14 = 5.8%), (age 15 = 6.8%), (age 16

9.3%), (age 17 = 8.2%). However, by the third wave, (ages 13-19), the
suspension rates did not consistently increase with age and showed no
significant relationship.

As might be expected, high school students were more likely to be

Suspended than junior high students; twice as often. Additionally, as

70

student grade point averages went down, the percentage of students
suspended went up. For example, of the students that got mostly A’s in
Wave 1, 3% were suspended two or more times; (mostly B’s = 2.7%
suspended), (mostly C’s = 9.4%), (mostly D’s = 12%), and (mostly F’s
= 14%). This pattern held constant for all waves and for both
categories of suspension; (suspended once or two or more times). Also,
suspended students on Public Assistance were suspended at a rate twice
that of students not on Public Assistance. Finally, suspended students
were more likely from homes with an annual income $14,000 per year or
less. Of the students that were suspended two or more times, almost 68%
came from homes with an annual income of $14,000 or less. Twenty-
eight point four percent came from middle income families ($14,000-
$26,000), and only 3.7% of the suspended students came from upper
income families ($26,000 or more). In short, the suspended student is
likely to be: male, minority, high school students from homes with
incomes of $14,000 or less, on Public Assistance, and likely to be

suffering failing grades.

General Description of the Delinquent Youth

The original eighteen specific acts of delinquency as detailed on
page 51 were further categorized into seven major forms of delinquency:
1. drug sales; 2. carrying hidden weapons; 3. breaking and entering into

building or car; 4. theft: major and minor; 5. vandalism; 6. assault

71

excluding sexual; and 7. robbery. On the average for all waves,
approximately 50% of all of the youth reported at least one act of
delinquency. Twice as many of those involved reported three or more
acts of delinquency. Males reported delinquency involvement twice that
of females for occasional and repeated offenses. In addition to males
being at least twice as likely to be involved in every category of
delinquency, males were five times more often than females to be
involved in weapons, breaking and entering, and robberies.

With regard to race, there were no significant differences in the
percentages of delinquency involvement as reported by the different races.
According to the self-reports; blacks, whites, and hispanics are basically
equally involved in delinquency. It is this issue that civil liberty unions
and various minority advocates scream about when prisons, jails, and
juvenile courts are disproportionately represented by minorities. Recall
that initially, the ages ranged from 11 to 17 with similar percentages of
each age category (13 to 15%). It would appear in looking at the rate
of serious delinquency (three or more acts), the percentage of self-
reported delinquency increases with age up to 16; then continues to
decrease with age. Although the significance level indicates a
relationship (P = .011). The significance and weak relationships were
constant throughout the waves.

There were no significant differences in the percentages of junior

high and high school students that reported delinquency involvement.

72

Almost equal percentage of either school level reported delinquency. For
example, in Wave 1, 64.7% of junior high students, and 61.9% of high
school students reported delinquency involvement. Of course, by the
fourth and fifth wave, most of the youth were high school aged.
Furthermore, as student grade point averages decreased, the percentage
of students reporting delinquency involvement increased. For example,
in Wave I (mostly A’s 48.1% delinquent), (mostly B’s 62.1%), (mostly
C’s 70.4%), (mostly D’s 74.4%), and (mostly F’s 78.6%). This pattern
held constant for all waves.

Finally, students reportedly involved in one or more acts of
delinquency were rather evenly distributed amongst the family income
levels which were divided into three categories: 1. $14,000 or less, 2.
$14,001 to 26,000, and 3. $26,001 or more. Similar percentages of each
income level reported delinquency involvement. In short, there were no
significant differences in delinquency rates amongst the social class and
races. Additionally, there was no difference in the rates of delinquency
reported by junior high and high school students. However, a weak
relationship between age and delinquency was shown. Also, the

delinquent youth was likely to be male with failing grades.

General Description—Delinquency Typology

Recall that eighteen specific acts make up the seven categories of

delinquency: 1. drug sales; 2. weapons; 3. breaking and entering; 4.

73

vandalism; 5. theft; 6. assault; and 7. robbery. Frequency distributions
showed that assaultive behavior, particularly fight among students was the
number one form of delinquency; followed by vandalism, then theft. This

pattern held constant for all waves.

The Dependent Variables

For the first part of the analysis, the elements of the social bond
are viewed as dependent variables for the purpose of examining the
effects of school suspension. Thereafter, the dependent variable is
delinquency wherein the effects of the bond elements on delinquency are
examined. A factor analysis was done on all the different types of
delinquency for each weave. Factor analyses revealed no significant
differences between the types of delinquency involvement by the subjects
in this study. This finding allowed the researcher to COMPUTE all these
delinquency types together to create the category “DELINQ.” DELINQ
measures the total amount of delinquency involvement reported by each
subject in the survey. The reported delinquency was further RECODED
into three different categories: “occas delinq” - one or two acts of
delinquency in the last twelve months; “serious delinq” - three or more

acts of delinquency in the last twelve months.

Core Variables
The variables that were common to all the questionnaires are school

suspension and the elements of the social bond: attachment, commitment,

74
involvement, and belief. Under school ATTACHMENT, the core variables

are concern for teachers’ opinion and liking school. “likesch” is a
variable obtained by computing the questions measuring students’ attitude
about school: feels lonely at school; no one at school cares; and don’t
belong at school. LIKSCH was recoded into: HATESCH, NEITHER, AND
LIKESCH.

Under school COMMITMENT, the core variables are educational
aspirations and expectations. Specifically, importance of going to college
and chances for completing college. Also, general school aptitude or
self-rating on ability resulted in the variable “selfrate” which was
developed by computing how well a student thought he or she was doing;
at being a good student; at having a high GPA; and how well are doing
in hard subjects. Furthermore, commitment was measured by an
achievement index, developed by computing the importance of:
schoolwork; GPA; and doing own school work without help.

Under school INVOLVEMENT, time spent in, and concern for: _
. homework, school athletics, and involvement in other school activities are
the core variables. Finally, under BELIEF in educational success, the
core variables are: conventional norms and attitudes; and acceptance of
school authority. Acceptance of school authority was measured by the
number of school rule violations, specifically: cheating on tests; skipping
class; damaging school property; striking a teacher; and fighting with

other students.

75
nggtionnaire 1: Analysis of Suspension and Social Bonds
The first element to be examined is school ATTACHMENT. This

category is made up of five variables. The first two measure a youth’s
sensitivity to teachers’ opinions. A relationship was found between
suspension and a youth’s concern that the teacher thinks the youth is a
good student and making a good impression with teachers. The next
three variables measure a youth’s attitude about school: 1. no one at
school cares; 2. feel don’t belong at school; and 3.’ feel lonely at school.
Since the original survey didn’t ask, these variables were used to measure
whether a student like school. A negative relationship between
suspension and the first two variables was found. However, the third
variable, “feel lonely at school,” was not significantly related.

The next element to be examined is school COMMITMENT. The
first two variables of commitment measure youths’ educational aspirations
and expectations. Negative relationships were found between suspension
and the importance of going to college and the chances for completing
college. This was particularly evident with suspended students wherein
25.7% of suspended students felt it was not important to go to college,
wherein 12.7% of non-suspended students reported it was not important
to go to college. Next, general school aptitude is made up of variables
measuring self-rating on ability and GPA. “SELFRATE” is a variable
created by factor analysis after three variables were found to measure

basically the same thing. The questions were asked: 1. How are you

76

doing at being a good student?; 2. How are you doing at having a high
GPA?; and 3. How are you doing in hard subjects? A negative
relationship was found between suspension and self-ratings on ability (P
= .00, X2 = 85.4, Gamma = -.56). The youth’s level of commitment to
school was also measured by asking: I. How important is school work?;
2. How important is it to do well in hard subjects? Suspension was
found to be negatively related to both measures: (P = .00, X2 = 34.0,
Gamma = -.40), (P =00, X2 = 20.8, Gamma = -.30). The suspended
student is likely to be less committed to school than the non-suspended
student.

School INVOLVEMENT was also looked at. This category was
made up of variables measuring time spent studying, time spent on school
athletics, and time spent on other school activities. The amount of time
spent on 1. Weekday afternoons; 2. Weekday evenings; and 3. Weekends
for each activity was accounted for. Time spent on athletics and other
school activities were not significant. However, time spent studying
during weekday evenings and on weekends were significant. A negative
relationship was found between suspension and involvement in studying.
That is, suspended students are likely to study less than non-suspended
students. This was apparent where 70% of suspended students studied
very little during weekday evenings compared to 50% of non-suspended

students.

77

The last element in the questionnaire to be looked at is BELIEF.
This category is made up of thirteen variables, eight of which measure
conventional norms and attitudes. Youth were asked whether they
thought eight specific acts were or were not wrong. Negative
relationships were found between suspension and all the measures of
conventional norms and attitudes (see BELIEFS, Table 1 in the
Appendix). The next variable to be looked at is Acceptance of school
authority, measured by the number of school rule violations. Suspension
was positively related to the number of times cheated on school tests, hit
a teacher or other students, damaged school property, and skipped class.
Suspended students reported more school rule violations than non-
suspended students. It would appear that the null hypothesis of no
relationship between suspension and the elements of the social bond is
rejected. Suspensions are negatively related to a youth’s attachment,
commitment, involvement, and belief in school. That is, the more

suspended, the weaker the bonds to school.

Questionnaire III: Analysis of Suspension and Social Bonds
The first two variables of the ATTACHMENT element measure the

youth’s concern for their teachers’ opinions. A relationship was found
between suspension and a youth’s concern that the teacher thinks the
youth is a good student, and making a good impression with teachers.

Next, certain attitudes about school were used to measure whether a child

78

liked school. Only one of the three variables was significant. A
negative relationship between suspension and belonging at school was
found (P = .00, X2 = 10.9, Gamma = -.29).

The next element to be examined is school COMMITMENT. The
first two variables measure a youth’s educational aspirations and
expectations. Negative relationships were found between suspension and
the importance of going to college and chances for completing college.
Also, suspensions were found to be negatively related to a youth’s self-
rating on ability. Suspended students were more likely to rate themselves
as not being a good student and not doing well on GPA than non-
suspended students. Additionally, suspensions were found to be
negatively related to Grade Point Average (P = .00, X2 = 63.5, Gamma
= -.48).

The next variable under commitment was created by factor analysis
- “ACHIVINDX.” This variable measures a youth’s level of commitment
by asking: I. How important is school work?; 2. How important is it to ,
. do well in hard subjects?; 3. How important is a high GPA? A negative
relationship was found between suspension and ACHVINDX. The
suspended student was more likely to find school work unimportant and
was found to be less committed to school than the non-suspended student.

Next, only the variables measuring time spent studying were found
to be significant INVOLVEMENT variables. Again, school athletics and

other school activities were insignificant. Time spent studying during

79

weekday afternoons and evenings were significant. A negative
relationship between suspension and studying was found. That is, the
suspended youth was likely to study less than non-suspended students.

The last element in the questionnaire to be looked at is BELIEF.
Conventional norms and attitudes and acceptance of school authority
make up the Belief element. Once again, negative relationships were
found between suspension and all the measures of conventional norms and
attitudes. Secondly, a relationship was found between suspension and
acceptance of school authority which was measured by the number of
certain school rule violations. Suspension was positively related to
school rule violations. This was particularly evident wherein 63.7% of
suspended students were involved infighting; whereas only 29% of non-
suspended students reported involvement in fighting. Also, 71.5% of
suspended students reported skipping classes compared to 38% of non-
suspended students.

Again, it would appear the school suspensions are negatively
related to a youth’s attachment, commitment, involvement, and belief in
school. That is, suspended students are less attached, less committed,
less involved in school, and maintain less conventional norms and
attitudes, and acceptance of school authority than non-suspended students.
These patterns held constant for all waves. These similar findings in

Waves IV and V will not be reviewed here, but are presented in the

80

Appendix. In the next part of the analysis, the relationship of the bond

elements and delinquency will be disclosed.

Analysis of Social Bonds and Delinquency

Questionnaire I Analysis

The first element to be examined under delinquency is school
ATTACHMENT. This category is made up of five variables. The first
two measure a youth’s sensitivity to their teachers’ opinions. A.
relationship was found between delinquency and a youth’s concern that
the teacher thinks the youth is a good student and making a good
impression with teachers. This was particularly evident with self-reported
delinquent acts wherein 69.0% of those without concern for teachers’
opinions had committed three or more delinquent acts, wherein only 36%
of those with a lot of concern for teachers’ opinions committed three or
more delinquent acts. The null hypothesis that there is no relationship
between the first two attachment predictors is rejected. The next
variables to be looked at are: 1. No on at school cares; 2. Feel don’t
belong at school; 3. Feel lonely at school. These variables were used to
measure whether a student liked school. A negative relationship between
delinquency and the first two variables was found. However, the third
variable, “Feeling lonely at school” was not significantly related (P =

.44, X2 = 11.9, Gamma = -.10).

81

The next element to be examined under delinquency is school
COMMITMENT. This category is made up of nine variables. The first
two variables to be looked at measure the youth’s educational aspirations
and expectations. Negative relationships were found between delinquency
and the importance of going to college and the chances for completing
college. Next to be looked at is general school aptitude which is made
up of variables measuring self-rating on ability and GPA. “SELFRATE”
is a variable created by factor analysis after three variables were found
to measure a youth’s self-rating on ability. Three questions were asked:
1. How are you at being a good student?; 2. How are you doing at
maintaining a high GPA?; and 3. How are you doing in hard subjects?
A negative relationship was found between delinquency and self-ratings
on ability (P = .00, X2 = 54.7, Gamma = -.23). This was particularly
evident with self-reported delinquent acts wherein 70% of those rating
themselves as not doing well committed three or more delinquent acts,
whereas only 31% of those that rated themselves as doing very well
committed three or more delinquent acts. Furthermore, Grade Point
Averages were found to be negatively related to delinquency (P = .00, X2
= 83.5, Gamma = -.27). That is, the better a student does in school, the
less likely the student is to be delinquent.

The next variable to be looked at is “ACHVINDX.” This variable
was created by factor analysis after four variables were found to measure

basically the same thing. The achievement index measured the youth’s

82

level of commitment to school by asking: I. How important is school
work?; 2. How important is it to do well in hard subjects?; 3. How
important is it to have a high GPA?; 4. How important is it to do own
school work without help? A negative relationship was found between
delinquency and achievement index. The importance of school was found
to be the most highly correlated of all the school commitment measures
(P = .00, X2 = 61.3, Gamma = -.30). The null hypothesis that there is
no relationship between school commitment and delinquency is rejected.
The subject’s commitment to school was found to be negatively related
to delinquency. That is, the more committed to school, the less likely
to be delinquent.

School INVOLVEMENT was also looked at. This category is made
up of nine variables. Time spent studying, time spent on school
athletics, and time spent on other school activities each contained three
variables. Specifically, each activity was measured by the amount of
time spent on the activity during 1. weekday afternoons; 2. weekday
evenings; and 3. weekends. None of the variables measuring athletics
and other school activities were significant. The null hypothesis that
there is no relationship between school activities-school athletics, and
delinquency is not rejected. HoweVer, a negative relationship was found
between delinquency and time spent studying. Thus, the more a student

studied, the less likely delinquent.

83

The last element of the social bond to be looked at in the
questionnaire is BELIEF. This category is made up of thirteen variables,
eight of which measure conventional norms and attitudes. Youth were
asked whether or not they thought eight specific acts were or were not
wrong. A negative relationship was found between delinquency and all
eight measures of conventional norms and attitudes. The next variable
to be looked at is acceptance of school authority, measured by the
number of school rule violations. The number of times cheated on school
tests, hit teacher or other students, damaged school property, and skipped
classes were all positively related to delinquency. That is, the more
school rule violations, the more likely delinquency. The null hypothesis
that there is no relationship between conventional beliefs and acceptance

of school authority and delinquency is rejected.

Questionnaire II Analysis

The first element to be examined under delinquency is school
ATTACHMENT. This category is made up of five variables. The first.
two variables to be looked at measure a youth’s sensitivity to their
teachers’ opinions. A relationship was found between delinquency and
a youth’s concern that the teacher thinks the youth is a good student and
is making a good impression with teachers. This was particularly evident
with self-reported delinquent acts wherein 65.9% of those without concern

for teachers’ opinions had committed three or more delinquent acts,

34

wherein only 24.4% of those with a lot of concern for teachers’ opinions
committed three or more delinquent acts. The null hypothesis that there
is no relationship between the first two attachment variables and
delinquency is rejected. The next variable to be looked at measures

9

whether a youth liked school. “No one at school cares,’ and “Feeling
lonely at school” were not significant. Another question was asked:
Does the student agree that he or she doesn’t belong at school? An
affirmative answer indicated that the student disliked school. A negative
relationship was found between delinquency and liking school. The
hypothesis that there is no relationship between liking school and
delinquency is rejected.

The next element to be examined under delinquency is school
COMMITMENT. This category is made up of eight variables. The first
two variables to be looked at measure the youth’s educational aspirations
and expectations. Negative relationships were found between delinquency
and the importance of going to college and the chances for completing
college. The next variable to be looked at is general schoOl aptitude
which is made up of variables measuring self-rating on ability and GPA.
“SELFRATE” is a variable created by factor analysis after three variables
were found to measure basically the same thing. The questions measured
the youth’s self-rating on ability. Three questions were asked: 1. How

are you doing at being a good student?; 2. How are you doing at having

a high GPA?; 3. How are you doing in hard subjects. A negative

85

relationship was found between delinquency and self-rating on ability (P
= .00, X2 = 15.4, Gamma = -.26). Furthermore, Grade Point Averages
were found to be negatively related to delinquency. Thus, the better a
student rated themselves on ability and the better a student does in
school, the less likely delinquent.

The next variable to be looked at is “ACHVINDX.” This variable
was created by factor analysis after three variables were found to
measure basically the same thing. The achievement index measured the.
youth’s level of commitment to school by asking: I. How important is
school work?; 2. How important is it to do well in hard subjects?; 3.
How important is it to do own school work without help? A negative
relationship was found between delinquency and achievement index. The
importance of school was found to be the most highly correlated of all
the school commitment measures (P = .00, X2 = 71.4, Gamma = -.38).
A negative relationship was found between delinquency and achievement
index. That is, the higher the score on the achievement index, the more
committed to school. The more committed to school, the less
delinquency. The null hypothesis that there is no relationship between
school commitment and delinquency is rejected.

School INVOLVEMENT was also looked at. This category is made
up of variables measuring time spent studying, time spent on school
athletics, and time spent on other school activities. Each activity was

measured by the amount of time spent on the activity during the weekday

86

afternoons, evenings, and on weekends. Again, none of the variables
measuring school athletics or other school activities were significant.
The null hypothesis that there is no relationship between school athletics
and other activities and delinquency is not rejected. However, a negative
relationship was found between delinquency and the amount of time spent
studying. The more a student studied, the less likely to be delinquent.

The last element of the social bond in the questionnaire is BELIEF.
This category is made up of ten variables; eight of which measure
conventional norms and attitudes. Youth were asked whether they
thought a certain conduct was right or wrong. A negative relationship
was found between delinquency and all eight measures of conventional
beliefs. The next variable to be looked at is the number of school rule
violations which were used to measure acceptance of school authority.
This category is made up of five variables. However, too much data was
missing for the frequencies: cheated on school tests; damage to school
property; and truancy. The two remaining school rule violations: 1.
frequency of hitting teacher; and 2. frequency of hitting others were both
positively relation to delinquency. The more school rule violations, the

more likely delinquent.

Questionnaire III Analysig

The first element to be examined under delinquency is school

ATTACHMENT. This category is made up of five variables. The first

87

two measure a youth’s sensitivity to their teachers’ opinions. A
relationship was found between delinquency and a youth’s concern that
the teacher thinks that the youth is a good student and is making a good
impression with teachers. This was particularly evident with self-reported
delinquent acts wherein 60.8% of those without concern for teachers’
opinions had committed three or more delinquent acts, whereas only
25.1% of those with a lot of concern for teachers opinions committed
three or more delinquent acts. The null hypothesis that there is no
relationship between concern for teachers opinions and delinquency is
rejected. The next variable to be looked at is LIKESCH: 1. No one at
school cares; 2. Feel don’t belong at school; 3. Feel lonely at school
were used to measure whether a student liked school. A negative
relationship between delinquency and the variables was found. The more
a youth liked school, the more attached to school, the less likely
delinquent. The null hypothesis that there is no relationship between
liking school and delinquency is rejected.

School COMMITMENT is the next element to be examined. This
category is made up of seven variables. The first two measure a youth’s
educational aspirations and expectations. Negative relationships were
found between delinquency and the importance of going to college, and
the chances for completing college. Next to be looked at is general
school aptitude which is made up of variables measuring self-rating on

ability and GPA. Self-rating was measured by asking: I. How are you

88

doing at being a good student; and 2. How are you doing at having a
high GPA. Negative relationships were found between delinquency and
the self-rating measures (P = .00, X2 = 56.5, Gamma = -.30) and (P =
.00, X2 = 74.4, Gamma = -.32), respectively. Furthermore, Grade Point
Averages were found to be negatively related to delinquency (P = .00, X2
= 103.1, Gamma = -.33). The better a student does in school, the less
likely delinquent.

The next variable to be looked at is “ACHVINDX.” This variable
was created by factor analysis after four variables were found to measure
basically the same thing. The achievement index measured a youth’s
level of commitment to school by asking: I. How important is school
work?; 2. How important is it to you to do well in hard subjects?; 3.
How important is it to have a high GPA?; and 4. How important is it to
do own school work without help? Again, the importance of school was
found to be the most highly correlated of all the school commitment
measures (P = .00, X2 = 102.8, Gamma = -.44). A negative relationship ,
. was found between delinquency and achievement index. The better a
student rated themselves, the better the GPA, the higher the score on the
achievement index, the more committed to school, the less likely
delinquent. The null hypothesis that there is no relationship between
commitment and delinquency is rejected.

School INVOLVEMENT was also looked at. Time spent studying,

time spent on school athletics, and time spent on other school activities

89

were measured by the amount of time spent on each activity during 1.
weekday afternoons; 2. weekday evenings; and 3. weekends. Again,
neither school athletics nor other school activities were significant. A
negative relationship was found between delinquency and the amount of
time spent studying, the less likely the youth is to be delinquent. The
null hypothesis that there is no relationship between school athletics or
other school activities and delinquency is not rejected.

The last element of the social bond to be looked at in the
questionnaire is BELIEF. This category is made up of thirteen variables,
eight of which measure conventional norms and attitudes. Youth were
asked whether they thought eight specific acts were right or wrong. A
negative relationship was found between delinquency and all eight
measures of conventional norms and attitudes. The next variable is
acceptance of school authority measured by the number of school rule
violations. School rule violations such as: cheating, truancy, fighting,
damaging school property, and hitting a teacher were found to be
positively related to delinquency. That is, the more School rule
violations, the more likely the youth is to be delinquent. The null
hypothesis that there is no relationship between conventional beliefs and
acceptance of school authority and delinquency is rejected.

For the first three waves, a youth’s attachment, commitment,
involvement, and belief in school are negatively related to delinquency.

That is, the stronger each of the elements of the bond, the less likely the

90

youth is to be delinquent. With some slight variations and, of course,
different significant levels, these findings were consistent with all the
questionnaires. In avoidance of being redundant, and perhaps even
boring, these similar findings in questionnaires IV and V will not be
reviewed here, but are presented in the Tables in the Appendix for the

reader’s review.

Analysis of Suspension and Delinquency By Waves

Wave 1. Since the probability is small (P = .00), the hypothesis
of no relationship between school suspension and delinquency is rejected.
School suspensions are positively related to delinquency. A chi square
of 95.0 shows the variables are adequately related with the strength of
the relationship indicated by the Gamma statistic of .59. School
suspensions were found to be positively related to delinquency. That is,
the suspended student is more likely to be delinquent than the non-
suspended student. This was particularly evident with self-reported
delinquent acts wherein 75% of the students suspended three or more
times and 70.3% of students suspended once or twice had committed three
or more acts of delinquency, wherein only 37.9% of the non-suspended
students reported committing three or more acts of delinquency.
Furthermore, suspensions in Wave I were found to be significantly related
to delinquency in Wave II (P = .00, X2 = 45.1, Gamma == .57).

However, the amount of delinquency by youth suspended in Wave I did

91

not significantly increase in Wave II. Also, the more often suspended,

the greater the percentage of delinquency involvement.

Wave II. Recall that an average of 13% of all students were
suspended any given year. In Wave II, more than half of the suspension
data was missing; therefore, with the sample being too small, analysis of

suspension and delinquency in Wave II was not done.

Wave III. Since the probability is small (P = .00), the hypothesis
of no relationship between suspension and delinquency is rejected.
School suspensions were found to be positively related to delinquency.
A chi square of 126.3 shows that the variables are adequately related
with the strength of the relationship indicated by the Gamma statistic of
.62. School suspensions are positively related to delinquency. Suspended
students are more likely to be delinquent than non-suspended students.
This was particularly evident with self-reported delinquent acts wherein
89.5% of the students suspended three or more times and 59.1% of
students who were suspended once or twice, had committed three or more
delinquent acts, whereas only 28.7% of non-suspended students committed
three of more delinquencies. Furthermore, suspensions in Wave III were
found to be significantly related to delinquency in Wave IV (P = .00, X2
= 97.0, Gamma = .54). However, the amount of delinquency by youth
suspended in Wave III did not significantly increase in Wave IV. Also,

the more often a student is suspended, the greater the percentage of

dcl
59.
sus

dc]

hl'j
rtj
dti

of

hy
re;
de
$11
811
PC
CV
511

Ill

1111

de

92

delinquency involvement. For example, 28.7% of non-suspended students,
59.1% of students suspended once or twice, and 89.5% of students
suspended three or more times reported three or more acts of

delinquency.

Wave IV. Again, the probability is small (P = .00). The
hypotheses of no relationship between suspension and delinquency is
rejected. School suspensions were found to be positively related to
delinquency (P = .00, X2 = 87.0, Gamma = .60). However, the amount

of delinquency in Wave IV did not significantly increase in Wave V.

Wave V. Again, the probability is small (P = .00). The
hypothesis of no relationship between suspension and delinquency is
rejected. School suspensions were found to be positively related to
delinquency (P = .00, X2 = 141.2, Gamma = .70). More suspended
students were found to be involved in delinquency than non-suspended
students. Also, the more often students were suspended, the greater the
percentage of students involved in delinquency. This was particularly
evident wherein 21.1% of non-suspended students, 58.5% of those
suspended once or twice, and 81.3% of students suspended three or more
times reported serious delinquency-involvement.

So far, the relationships of school suspension and the elements of
the bond, the bond elements and delinquency, and school suspension and

delinquency, have been reviewed. As a result of suspension and the bond

Cl:

35

93

elements being significantly related to delinquency, they-were combined
as predictor variables and regressed with delinquency. The next section

is the result of the regression analysis.

Regression Analysis Findings

A two-step regression analysis was conducted, see Table 10 in the
Appendix. After controls, the major predictor variables of the composite
measure of delinquency (Wave IV Delinquency) in the first step are:
school suspension and truancy. The major predictor variable being
suspensions. Prior delinquency (Wave I Delinquency) was not found to
be a significant predictor variable. The total amount of delinquency in
Wave IV was found to have an F of 5.1 (p S .01) and an R-Square of
.05.

At the second step of the regression analysis, a significantly greater
amount of variance in the composite measure of delinquency is explained.
After the interaction terms were introduced in step two, the total amount
of delinquency in Wave IV was found to have an F of 18.0 (p S .01) and '
i an R-Square of .29. At step two, the major predictor variables of the
composite measure of delinquency are: from the BELIEF element of the
bond, specifically school rule violations and conventional norms and
attitudes. Weak predictors include the COMMITMENT variables of self
rating on ability and importance of school work. School suspensions

were no longer a significant predictor of delinquency.

94

Review of the Findings

Through the use of bivariate analysis, negative relationships were
found between school suspension and the elements of the social bond.
Consistently in all waves, it was found that suspended students were less
attached, less committed and less involved in school and maintained less
conventional norms and attitudes than non-suspended students. Also,
suspended students maintained less acceptance of school authority
indicative by the number of school rule violations.

Secondly, through the use of bivariate analysis, relationships were
found between the elements of the social bond and delinquency.
Consistently in all waves, a negative relationship was found between a
youth’s attachment, commitment, involvement and belief in educational
success and delinquency. The higher the level a youth obtained of each
element, the less likely they were to commit delinquent acts. In
reference to the involvement element, only time spent studying was found
to be significant. School athletics and other school activities were not
significant in any waves. Thirdly, for Waves I, III, IV and V, school
suspension was individually tested for relationships with delinquency
through bivariate analysis. School suspensions were found to be
positively related to delinquency in each wave and for delinquency in the
following wave. However, the amount of delinquency by suspended youth

in one year did not significantly increase the next year. Suspended

95

students were more likely to be involved in delinquency than non-
suspended youth and the more suspended, the more the delinquency
involvement.

Finally, as a result of the bond elements and suspension being
significantly related to delinquency, they were selected as predictor
variables along with prior delinquency and were combined for regression
analysis with delinquency. In Wave IV, 5% of the variance in the
composite measure of delinquency was explained by the selected predictor .
variables in the first step of the regression. In the second step, 29% of
the variance in delinquency is explained. In the second step, the belief
and commitment elements of the bond explained away the affects of

suspension on delinquency.

CHAPTER 5

This chapter will summarize the thesis and will discuss the
implications of this study on the Social Bonding Theory, the implications

for policy and the implications for future research.

Summary and Discussion

School suspensions are still widely used in the United States,
particularly in public schools. University of Michigan Professor Charles
B. Vergon recently discovered that Michigan ranked sixth in the nation
for the highest suspension rates for public school students. Nearly 97%
of the Michigan districts studied, reported using suspension. Most of any
existing suspension data is collected by the U.S. Office of Civil Rights
(O.C.R.) and is limited to disclosing the number of student suspended.
The O.C.R. data does not indicate: the offense students were suspended
for and how long or how many students were suspended more than once.
Vergon also discovered that the suspension trend in Michigan has been
climbing steadily. In addition to the increasing trend in the use of
suspension, use and duration of suspension was found to vary widely

from district to district.

96

97

Vergon and other researchers such as the Children’s Defense Fund,
define the problem as being: the frequent, increasing and unnecessary use
of suspension, disproportionate suspension rates for minorities,
particularly for blacks and inconsistent use of suspension and expulsion.
Furthermore, the problem is complicated by the lack of empirical research
in examining the impact of suspension. This study is an attempt to
examine the effects of suspension on youth and particularly to look at the
relationships between suspension and juvenile delinquency.

Travis Hirschi’s Social Bond Theory was found to be one of the
best explanations of the causes of delinquency, at least amongst
contemporary sociological theories. Since Hirschi’s theory was found to
have such strong supporting research, Social Bond Theory was
incorporated into this study for the purpose of examining suspension and
juvenile delinquency. Hirschi studied the elements of the bond in
relation to education: ATTACHMENT, COMMITMENT, INVOLVEMENT
AND BELIEF. A child’s ATTACHMENT to school was measured by the
child’s sensitivity to teacher’s opinions and liking school.
COMMITMENT was measured by the youth’s educational aspirations and
expectations. Also General School Aptitude measured the student’s
academic competence, self rating' on school ability and intelligence.
INVOLVEMENT in school was measured by the amount of time spent on
school related activities and BELIEF in educational success was measured

first by conventional norms and attitudes, and acceptance of school

98

authority which is measured by the number of school rule violations.
According to Hirschi, the higher the level a youth obtains of each
element, the less likely they are to commit delinquent acts.

The concepts as provided by Travis Hirschi and the Social Bond
Theory were incorporated into the design of this study. The National
Youth Survey, a longitudinal study of delinquency in America was
originally collected by Delbert Elliott in 1977. The National Youth
Survey provided the data for this study. Following the concepts as
described by Hirschi, variables measuring the bond elements of
ATTACHMENT, COMMITMENT, INVOLVEMENT AND BELIEF, were
extracted from the National Youth Survey for further analysis. Through
the use of simple bivariate analysis, it appeared that school suspensions
weaken the social bonds, particularly the bonds to the school. The level
of attachment to these elements is lower with suspended students.
However, did suspensions weaken the bonds or did already weakened
bonds result in misbehavior which resulted in suspension? This issue of .
, time-order is problematic for many researchers and is not necessarily
resolved in this study. What can be said for certainty is that significant
relationships exists between suspension and the bond elements.

The elements of the bond were then individually examined as to
their relationship with delinquency. Through the use of simple bivariate
analysis, it appeared that a youth’s weak attachments, commitments and

involvement and beliefs resulted in delinquency. The level of attachment

99

to these elements is lower in delinquent youth than non-delinquent youth.
Again the issue of time order is problematic. Did weak attachments
result in delinquency? Or did delinquency result in weak attachments?
What can be said is that significant relationships exist between the bond
elements and delinquency. Furthermore, it is acknowledged that the
effects of poverty, unemployment, association with delinquent peers and
the like, are not evaluated. They are beyond the scope of this study and
are for future research endeavors.

School suspension was found to be positively related to delinquency

- .,_m.—,__-__

”RM-Jr

“um
um

in each wave and to delinquency in the followrng wave. Suspension was

v...-..._ .

 

.,__-—— -I ..
__h._..-w_.'o-oo "' —-4oce-‘.o

evaluated with delinquency in the following year in attempt to address
the time-order problem. Also, it is possible, although likely infrequent,

that a delinquent act could result in suspension. However, according to

research, the majority of suspensions were for non- dangerous reasons,

-.,... .n. .Mv
"- ‘M g”
“a... A.‘ nut.

such as: truancy, tardiness, behaviorproblems, arguments,smoking andm

'M—rv‘ --
W

other flirtemquent offenses, (Children’s Defense Fund Survey). Again
it is recognized that the relationships between suspension and delinquency
may be spurious. An attempt to deal with causality is dealt with in the
regression analysis.

It is suggested by the simple correlations that school suspensions
and the elements of the bond are strongly related to delinquency.

However, regression analysis revealed that school suspensions are only

related to delinquency through their effects on the bonds to the school.

100

A two-step regression analysis was conducted to determine whether
the correlations between suspension and delinquency could be attributed
to other variables, such as the control variables of prior delinquency
(Wave 1) or to elements of the bond. In the first step of the regression
analysis, when prior delinquency and bond elements were controlled,
school suspensions from Wave III were found to be associated with
delinquency in Wave IV (R2 = .05; F = 5.1; p S .01; Table 10). Also,
there was a low but significant Beta weight for truancy, indicating that.
truancy explained a very small amount of the variance in the composite
measure of delinquency. Before the interaction terms were considered,
school suspension (Wave 111) had the highest Beta weight, indicating that
it explained the greatest proportion of variance in the composite measure
of delinquency in Wave IV.

At the second step of the analysis, a significantly greater amount
of variance in the composite measure of delinquency is explained (R2 =
.29; F = 18.0; p S .01). The most highly predictive of delinquency was
no longer school suspensions, but from the COMMITMENT and
particularly the BELIEF elements of the bond. Under the Belief element,
acceptance of school authority, measured by the number of school rule
violations and conventional norms and attitudes had the highest Beta
weights, indicating that the Belief element of the bond was the most
highly predictive of the composite measure of delinquency. At the

second step it was discovered that school suspensions were not even

101

significant. While controlling for prior delinquency, the bond elements
explained away the affects of suspension on delinquency. Thus it is
possible that school suspensions weaken the attachments to the bonds, and
that weaken bonds, particularly Commitment and Belief, predict
delinquency.

The discovery of causal. relationships in scientific research is
always a difficult one. The first requirement in a causal relationship
between two variables is that the cause precede the effect in time. With
longitudinal data, this is easily accomplished. The second requirement
is that the two variables be empirically correlated with one another. In
this study, numerous correlations were shown. The third requirement for
a causal relationship is that the observed empirical correlation between
two variables, not be explained away as being due to the influence of
some third variable. As was evident in the regression analysis. The first
two requirements of causality are repeatedly conducted throughout this
study in bivariate analysis of suspension and the bond elements, the bond
elements and delinquency and suspension and delinquency. This provides
a good foundation for further research in that it is acknowledged that

various other predictor variables need to be considered.

Conclusion
First of all, this research is Secondary Analysis of Survey Data.

The major benefit of secondary analysis is of course the avoidance of

102

enormous expenditure of time and money. The key disadvantage involves
the recurrent question of validity. Do the questions originally asked
provide a valid measure of the variables this researcher wants to analyze.
This was particularly evident with one of the Attachment measures.
Hirschi found the relation between “liking school” and delinquency to be
very strong. This question of whether a youth liked school was not
specifically asked in the National Youth Survey. This researcher
computed a variable to measure whether a youth liked school. Three
questions were asked: (1) Feel lonely at school? (2) No one at school
cares? and (3) Don’t belong at school? It is recognized that these
variables may not actually measure whether a youth likes school.
However, the three questions are still a good measure of ATTACHMENT
to the school.

Another variable in question is with Hirschi’s measure of General
School Aptitude. Hirschi acknowledges that overall grade point averages
were not available when his analysis was conducted. Thus he used DAT
scores and English Average Marks. Hirschi’s use of english and math
grade as a measure of intelligence raises the issue of validity. Actual
grade point averages were available from the National Youth Survey,
however, the use of grade point averages as a measure of intelligence is
highly debated by researchers. Furthermore, consistent with criticisms of
the social bond theory, further analysis showed GPA as a measure of

intelligence was not a predictor of delinquency. All other measures

103

drawn from the National Youth Survey parallel Hirschi’s questionnaire
and theory.

Survey research is probably the most frequently used and is
probably the best method available to researchers interested in collecting
original data for describing a population too large to observe directly.
Recall that the National Youth Survey employed a probability sample of
households in the United States based upon Multistage Cluster Sampling
Design. Sampling error was reduced by the inclusion of a large sample,
initially 67,266 households of which 8,000 were selected for inclusion of
the sample. Furthermore, representativeness was obtained through the use
of probability selection, in that all members of the population had an
equal chance of being selected in the sample. Additionally, according to
Delbert Elliott, the sample of the National Youth Survey was
representative of the population from which it was selected as the
aggregate characteristics of the sample, closely approximated the
aggregate characteristics of the population.

The use of staff administered interviews in the survey provided for
a good return rate and ensured that the questionnaires were completed.
Of course, the respondents might be reluctant to report controversial or
deviant attitudes or behavior in a face to face interview, however
confidentiality was assured. Finally the question of reliability is
strengthened by the use of the social bond model, in that Hirschi’s

measures have proven their reliability in previous research. Additionally,

104

the National Youth Survey is a longitudinal study which provided a
consistent questionnaire for five consecutive years.

This research has examined the links between school suspension,
the social bond elements and delinquency. Initially there are significant
relationships between suspension and delinquency and the elements of the
bond and delinquency. However, when regressed with delinquency,
school suspension and many of the social bond elements lose their
significance. This was particularly true for the ATTACHMENT and
INVOLVEMENT elements of the bond. This study provides exploratory
evidence that suspension is related to delinquency by way of weakening
bonds which then may result in delinquency. However, Hirschi’s bond
theory as a prediction model is somewhat weak in that two of the four

main elements were not significant. Further research is recommended.

Implications for Future Research

This awareness of a statistical association between suspension and
delinquency has the potential to result in various policies and law.
Representative H. Lynn Jondahl is proposing legislation today to require
local districts that suspend or expel students, to provide alternative
instruction, ranging from makeup exams to special classes. This slow
trend of providing suspension alternatives is perhaps the result of the

recognition or suspicion that school suspension is related to delinquency.

105

The causes of delinquency are various and interrelated, as are the
delinquency theories. As a result of this study, it is recommended that
since conventional norms and attitudes were strong predictors of
delinquency, that theories with that emphasis, be used to further explore
the relationship of suspension and delinquency. Yet schools continue to
use suspension regularly. Its kind of like using an experimental drug
without knowing the possible side affects. Additionally, it is
recommended that other predictors of delinquency, such as delinquent.
peer association, be evaluated. Also, it would be ideal, time and money
permitting, if new data could be collected. As it is always questionable
whether results from studies done in the past can be applied to the
present. In conclusion, this researcher’s concern is with the frequent and
unnecessary use of suspension which is used in the majority of cases.
It is recognized that suspension is sometimes necessary in extreme cases
of violence, repeated disruption and with incorrigible youth. This
researcher is sympathetic to schools wherein they have no alternative for

such cases, however, these situations according to research are few.

APPENDIX

106

APPENDIX

TABLES 2-10

(pages following)

107

TABLE 2. Significant Relationships of Variables to Delinquency Have I

 

Not Occas. Serious
WAVE I Delinq Delinq Delinq
45 n X n X n df X2 4__gqggg
SCHOOL ATTACHMENT
Tchr think good stu (Sensitivity to tchr's opinion)
Not Imprt. 18.4 16 10.3 9 69.0 60
Somewhat 34.9 203 20.0 116 44.9 261
Very luprt. 39.1 401 24.1 247 36.7 376 12 81.4* “.19
Impress tchr foremost
Disagree 40.2 517 23.8 306 35.9 462
Neither 23.5 44 16.6 31 59.9 112
Agree 26.0 54 15.4 32 57.7 120 12 81.2* .34
None at sch care (Like school)
Agree/hatesch 28.1 39 21.6 30 46.9 69
Neither 23.9 33 20.3 28 55.8 77
Disagree/likesch 38.7 548 22.2 314 39.0 552 12 33.7* -.23
Don't belong at sch
Agree/hatesch 27.3 39 23.1 33 49.0 70
Neither 22.8 26 13.2 15 64.0 73
Disagree/likesch 38.6 555 22.5 324 38.6 555 12 45.7* -.27
SCHOOL COMMITMENT
Imprt go to college
so: imprt. 28.9 66 20.2 46 50.4 115
Somewhat 36.6 166 17.9 81 45.5 206
Very Imprt. 38.3 391 24.2 247 37.3 381 12 29.4* -.13
Chances complete
Poor 25.2 61 17.4 42 57.0 138
Pair 35.7 228 21.8 139 42.6 272
Good 40.3 326 23.5 190 35.8 290 12 44.7* -.18
Self rate as student (Gen. Sch. Apt./self rating on ability)
Not well 16.7 11 16.7 11 63.6 42
OR 34.9 328 22.0 208 42.9 403
Very well 44.0 264 24.0 144 32.0 192 8 85.2* '.23
Self rate GPA
Not well 18.0 37 15.1 31 65.9 135
Ok 37.3 350 21.8 205 40.8 383
Very well 43.1 207 26.3 126 30.4 146 12 94.2* -.27

108

 

TABLE 2 (cont'd). Not Occas. Serious
Delinq Delinq Delinq
EAVE I Z n 2 n, 2 347 df X2 gamma
SCHOOL COMMITMENT CONTINUED
Self rate hard subjs
Not well 24.9 42 14.8 25 59.8 101
Ok 37.2 428 22.3 257 40.3 463
Very well 41.3 136 25.2 83 33.4 110 10 42.9* -.19
Grade Point Average
Mostly P's 21.4 3 28.6 5 50.0 7
Mostly D's 25.7 19 14.9 11 59.5 44
Mostly C's 29.2 163 19.1 107 51.3 289
Mostly B's 37.5 288 24.2 185 37.9 292
Mostly A's 51.9 151 23.0 67 25.1 73 24 83.5* -.27
How imprt. sch. work (Achievement index)
Not Imprt. 21.6 24 13.5 15 64.9 72
Somewhat 27.9 87 20.2 63 51.3 160
Very Imprt. 40.1 509 23.2 294 36.6 464 12 61.3* -.30
Do well in hard subjs
Not Imprt. 31.8 14 13.6 6 52.3 23
Somewhat 28.9 179 21.0 130 49.8 309
Very Imprt. 41.5 427 22.9 236 35.5 365 12 81.5* -.24
How imprt high GPA
Not Imprt. 36.2 25 14.5 10 49.3 34
Somewhat 33.8 148 19.9 87 46.1 202
Very Imprt. 37.7 447 23.2 275 38.9 462 12 14.0* -.04
Do sch work w/o help
Not Imprt. 26.3 35 10.5 14 61.7 82
Somewhat 37.0 271 20.7 152 42.2 309
Very Imprt. 38.0 314 24.9 206 37.0 306 12 59.0* -.13
SCHOOL INVOLVEMENT
deay Afternoon study
Very Little 30.7 187 22.1 135 47.2 188
Some 34.6 234 21.2 143 44.2 299
Alot 48.8 198 23.2 94 28.1 114 4 47.1* -.20
Wkday Evenings study
Very Little 31.9 176 19.4 107 48.7 269
Some 37 .0 263 22 . 5 160 40 . 5 288
Alot 42.1 181 24.4 105 33.5 144 4 23.6* -.15
weekends study
Very little 32.1 336 21.5 225 46.1 482
Some 43.5 213 23.5 115 32.9 161
Alot 45.4 71 20.5 32 34.0 53 12 36.7* -.21

109

 

TABLE 2 (cont'd). Not Occas. Serious
Delinq Delinq Delinq
WAVE I X n X n X n. df X2
BELIEFS
Cheat on tests (Conventional norms & attitudes)
Not Wrong 23.7 86 13.8 50 62.0 225
Wrong 40.1 539 24.1 324 35.6 479 6 90.9*
Damage others prop
Not Wrong 2.2 1 17.8 8 77.8 35
Wrong 37.6 624 22.0 366 40.2 668 6 67.2*
Use Marijuana
Not Wrong 20.5 59 14.9 43 64.2 185
Wrong 39.9 565 23.4 331 36.6 518 6 82.3*
Steal < $5.00
Not Wrong 14.5 25 16.2 28 68.2 118
Wrong 39.2 600 22.6 346 38.1 584 6 80.9*
Assault
Not Wrong 12.1 28 21.6 50 65.5 152
Wrong 40.5 497 22.0 324 37.4 552 6 95.5*
Breaking & Entering
Not Wrong 7.4 2 18.5 5 74.1 20
Wrong 37.1 623 22.0 369 40.7 684 6 13.7*
Sell Hard Drugs
Not Wrong 9.4 3 12.5 4 78.1 25
Wrong 37.1 622 22.1 370 40.5 676 6 18.7*
Steal ) $50.00
Not Wrong 6.5 2 16.1 5 77.4 24
Wrong 37.2 623 22.0 369 40.6 680 6 18.4*
Lie to cops 4 frnds
Yes 20.9 45 11.2 24 67.0 144
Don't know 29.9 112 21.1 79 48.9 183
No 41.5 371 23.9 214 34.3 307 12 99.0*
Stop frnds go astray
No 16.9 12 14.1 10 66.2 47
Don't know 28.7 49 17.5 30 53.2 91
Yes 37.6 467 22.3 277 40.0 496 12 77.5*
Ok to break rules
Agree 14.9 40 19.4 52 65.7 176
Neither 23.1 46 22.1 44 54.8 109
Disagree 43.4 537 22.4 277 34.2 423 4 122.9*

.40

.72

.44

.53

.53

.60

.63

.65

.33

.45

110

 

TABLE 2 (cont'd). Not Occas. Serious
Delinq Delinq Delinq
WAVE I 3L n, lkjigv711 Z n df X2 4__g§gm§
BELIEFS CONTINUED (Acceptance of sch. authority)
Freq tests cheat (# school rule violations)
None 48.4 423 24.7 216 26.8 234
Once or twice 25.6 124 25.6 124 48.8 236
3 or more 22.2 77 9.8 34 67.1 233 12 218.4* .45
Freq hit teachers
None 39.6 625 23.1 365 37.3 589
Once 4.5 3 9.1 6 86.4 57
2 or more 1.6 1 1 6 1 96.8 60 4 145.1* .87
Freq hit students
None 70.2 625 17.4 155 12.4 110
once or twice 39.3 175 27.2 121 33.5 149
3 or more 16.9 63 18.0 67 65.1 242 4 429.2* .64
Freq damage sch prop
None 43.5 625 24.5 353 32.0 460
Once or twice 6.0 12 15.1 30 78.9 157
3 or more 2.9 2 O 0 97.1 68 4 267.0* .80
Freq skipped classes
None 43.4 508 24.3 285 32.3 378
Once or twice 29.5 75 21.7 55 48.4 123
3 or more 14.9 42 11.7 33 72.2 203 12 182.9* .46
Suspension
None 39.7 609 22.2 341 37.9 582
once or twice 9.4 12 19.5 25 70.3 90
3 or more 9.1 4 15.9 7 75.0 33 10 95.0* .59

111

TABLE 3. Significant Relationships of Variables to Delinquency Wave II

Not Occas. Serious
Delinq Delinq Delinq
WAVE I I 7. 1L 2 n X n df X2 gam

 

SCHOOL ATTACHMENT
Tchr think good stu
Not imprt.
Somewhat
Very Imprt.

Impress tchr foremost
Disagree
Neither
Agree

Don't belong at sch
Agree/hatesch
Neither
Disagree/likesch

SCHOOL COMMITMENT
Imprt go to college
Not Imprt.
Somewhat
Very imprt.

Chances complete
Poor
Fair
Good

Self rate as student
Not well
0k
Very Well

Self rate GPA
Not well
0k
Very well

Self rate hard subjs
Not well
Ok
Very well

Grade Point Average
Mostly F's
Mostly D's
Mostly C's
Mostly B's
Mostly A's

(Sensitivity to teacher's opinion)
14.6 6 19.5 8 65.9 27
43.3 117 19.3 52 37.4 101
52.5 200 21.8 83 24.4 93 14 41.0* -.27

52.2 285 19.8 108 27.8 152

20.5 15 27.4 20 47.9 35

30.6 22 20.8 15 47.2 34 14 68.3* .41
(Like school)

32.6 13 15.2 10 55.8 29

25.0 13 19.2 10 55.8 29

49.7 295 21.2 126 28.5 169 14 40.0* -.39

40.4 44 15.6 17 42.2 46
42.3 88 20.2 42 37.0 77
50.3 194 22.8 88 26.4 102 14 28.8* -.17

41.0 48 14.5 17 42.7 50
38.0 95 21.2 53 40.0 100
54.5 182 22.8 76 22.5 75 14 45.4* -.26

(Gen. Sch. Apt./Se1f rating on ability)

40.0 8 20.0 4 35.0 7

44.7 185 21.7 90 32.6 135

57.1 124 18.9 41 24.0 52 14 43.0* -.21

33.3 23 18.8 13 46.4 32
45.8 191 20.9 87 32.4 135
57.8 104 21.1 38 21.1 38 14 29.5* -.26

20.0 13 26.2 17 52.3 34
49.4 240 20.0 97 30.0 146
52.3 67 22.7 29 25.0 32 12 34.3* -.24

40.0 2 20.0 1 40.0 2
42.4 14 15.2 5 42.4 14
50.9 118 21.1 49 28.0 65
64.4 203 16.2 51 19.4 61
76.3 87 12.3 14 11.4 13 8 32.6* .30

112

 

TABLE 3 (cont'd). Not Occas Serious
Delinq Delinq Delinq
HAVE 11 j n Z n z I; df X2 M
SCHOOL COMMITMENT CONTINUED
How imprt. sch. work (Achievement index)
Not Imprt. 27.0 10 13.5 5 54.1 20
Somewhat 32.5 40 19.5 24 47.2 58
Very Imprt. 51.3 273 21.4 114 26.9 143 14 71.4* -.38
Do well in hard subjs
Not imprt. 23.1 3 0 0 69.2 9
Somewhat 37.9 92 21.4 52 39.9 97
Very imprt. 52.3 228 20.9 91 26.4 115 14 83.1* -.29
Do sch work w/o help
Not imprt. 30.6 11 5.6 2 61.1 22
Somewhat 42.1 127 20.2 61 37.1 112
Very imprt. 52.3 185 22.6 80 24.6 87 14 51.2* -.25
SCHOOL INVOLVEMENT
Wkday aftrns study
Very little 57.0 134 14.0 33 28.9 68
Some 62.2 166 18.4 49 19.5 52
Alot 64.0 121 17.5 33 18.5 35 4 9.2** -.11
Wkday eves study
Very little 54.4 124 15.4 35 30.3 69
Some 64.5 189 16.4 48 19.1 56
Alot 63.7 109 18.7 32 17.5 30 4 12.6* -.14
Weekends study
Very little 39.4 155 20.9 82 38.4 151
Some 53.3 121 20.7 47 26.0 59
Alot 65.3 47 19.4 14 15.3 11 14 29.6* -.30
BELIEFS
Cheat on tests (Conventional norms & attitudes)
Not wrong 26.3 42 18.8 30 55.0 88
Wrong 52.4 285 21.5 117 26.1 142 2 50.2* -.48
Damage others prop
Not Wrong 12.9 4 19.4 6 67.7 21
Wrong 48.0 323 21.0 141 31.1 209 2 20.1* -.62
Use marijuana
Not wrong 31.9 52 16.6 27 51.5 84
Wrong 50.8 275 22.2 120 27.0 146 2 34.6* -.39
Steal < $5.00
Not wrong 23.8 19 13.8 11 62.5 50
Wrong 49.4 308 21.8 136 28.8 180 2 36.7* .52

113

 

TABLE 3 (cont'd). Not Occas. Serious
Delinq Delinq Delinq
WAVE II 2 n Z n X n df X2 gamma
BELIEFS CONTINUED
Assault someone
Not wrong 17.4 12 26.1 18 56.5 39
wrong 49.6 315 20.3 129 30.1 191 2 28.2* -.51
B & E Bldg or car
Not wrong 8.3 2 12.5 3 79.2 19
Wrong 47.8 325 21.2 144 31.0 211 2 25.0* -.76
Sell hard drugs
Not wrong 11.1 2 22.2 4 66.7 12
Wrong 47.4 325 20.8 143 31.8 218 2 11.5* -.62
Steal > $50.00
Not wrong 0 0 18.8 3 81.3 12
Wrong 47.5 326 21.0 144 31.6 217 2 19.4* -.83
Lie to cops 4 frnds
Yes 27.8 35 13.5 17 58,7 74
Don't know 40.9 56 18.2 25 40.9 56
No 53.1 198 23.9 89 23.1 86 6 58.4* -.38
Stop frnds go astray
No 40.5 17 9.5 4 50.0 21
Don't know 41.3 31 14.7 11 44.0 33
yes 46.5 242 22.3 116 31.2 162 4 11.7* -.18
Ok break rules
Agree 27.4 26 21.1 20 51.6 49
Neither 27.2 22 17.3 14 55.6 45
Disagree 52.9 279 21.3 112 25.8 136 4 50.1* -.43
(Acceptance of school authority)
Freq hit teacher (# of school rule violations)
None 49.2 327 21.8 145 28.9 192
once 4.5 76 13.6 3 81.8 18
2 or more 0 O 0 0 100 83 4 66.7* .90
Freq hit students
None 79.6 327 13.1 54 7.3 30
once or twice 46.6 76 27.0 44 26.4 43
3 or more 18.5 24 17.7 23 63.8 83 4 224.9* .71

114

TABLE 4. Significant Relationships of Variables to Delinquency Wave III

 

Not Occas. Serious
Delinq Delinq Delinq
WAVE III 2 n 2 n 511, n df X2 ,1_s§mma
SCHOOL ATTACHMENT
Tchr think good stu (Sensitivity to teacher's opinion)
Not imprt. 24.7 24 14.4 14 60.8 59
Somewhat 44.3 294 17.9 119 37.8 251
Very imprt 58.6 443 16.3 123 25.1 190 4 70.3* -.31
Impress tchr foremost
Disagree 56.2 678 17.0 205 26.8 324
Neither 24.0 42 22.3 39 53.7 94
Agree 28.0 38 9.1 12 62.1 82 4 122.9* .17
None at sch care (Like school)
Agree/hatesch 34.8 24 21.7 15 43.5 30
Neither 40.0 46 20.9 24 39.1 45
Disagree/likesch 51.9 691 16.3 217 31.8 423 4 12.9* -.21
Don't belong at sch
Agree/hatesch 30.3 30 14.1 14 55.6 55
Neither 34.5 41 16.8 20 48.7 58
Disagree/likesch 53.3 691 17.0 221 29.7 385 4 45.1* -.37
Feel lonely at sch
Agree/hatesch 42.0 126 19.0 57 39.0 117
Neither 38.7 70 18.2 33 43.1 78
Disagree/likesch 54.7 566 16.0 166 29.3 303 4 27.6* -.20
SCHOOL COMMITMENT
Imprt go to college
Not imprt. 45.3 129 14.7 42 40.0 114
Somewhat 44.4 204 18.3 84 37.3 171
Very imprt. 56.0 486 15.8 137 28.2 245 4 24.4* -.17
Chances complete
Poor 40.1 113 15.6 44 44.3 125
Fair 45.7 244 19.1 102 35.2 188
Good 58.3 455 14.6 114 27.1 212 4 41.6* -.22
Self rate as student (Gen. Sch. Apt./Se1f rating on ability)
Not well 28.7 49 17.0 29 54.4 93
0k 51.7 538 17.5 182 30.8 321
Very well 61.6 165 14.9 40 23.5 63 4 56.5* -.30
Self rate GPA
Not well 24.3 37 17.8 27 57.9 88
Ok 51.6 475 16.8 155 31.6 6
Very Well 61.0 225 17.9 66 21.1 78 4 74.4* -.32

115

 

TABLE 4 (cont'd). Not Occas. Serious
Delinq Delinq Delinq
WAVE III 2 n 2 n4, 2 n df X2 gamma
SCHOOL COMMITMENT CONTINUED
Grade Point Average
Mostly P's 30.8 4 7.7 1 61.5 8
Mostly D's 17.5 11 12.7 8 69.8 44
Mostly C's 41.4 218 18.8 98 39.7 210
Mostly B's 53.5 350 17.6 114 28.9 190
Mostly A's 69.0 176 12.2 31 18.8 48 8 103.1* -.33
How imprt sch work (Achievement index)
Not imprt. 27.8 25 8.9 8 63.3 57
Somewhat 32.8 95 17.9 52 49.3 143
Very Imprt. 56.4 642 17.2 196 26.4 300 4 102.8* -.44
How imprt. high GPA
Not Imprt. 31.4 22 10.0 7 58.6 41
Somewhat 42.3 232 18.8 103 39.0 214
Very Imprt. 56.5 447 15.1 115 26.4 202 4 50.8* -.27
Do sch. work w/o help
Not Imprt. 34.4 33 14.6 14 51.0 49
Somewhat 42.9 282 19.3 127 37.8 249
Very Imprt. 58.5 447 15.1 115 26.4 202 4 50.6* -.27
Do well in hard subjs
th Imprt. 26.3 10 13.2 5 60.5 23
Somewhat 45.6 274 17.3 104 37.1 223
Very Imprt. 54.4 478 16.5 145 29.1 256 4 26.1* -.19
SCHOOL INVOLVEMENT
Wkday aftrns study
Very little 39.6 214 17.0 92 43.4 235
Some 54.1 315 15.6 91 30.2 176
Alot 59.1 233 18.0 71 22.8 90 4 52.5* -.25
Wkday eves study
Very little 39.6 177 19.2 86 41.2 184
Some 52.1 229 15.4 100 32.6 212
Alot 58.8 245 16.3 68 24.9 104 4 36.2* -.21
Weekends study
Very little 43.9 406 17.1 158 39.0 360
Some 58.3 250 16.1 69 25.6 110
Alot 64.2 106 18.2 30 4 47.5* -.28

17.6 29

116

 

TABLE 4 (cont'd). Not Occas. Serious
Delinq Delinq Delinq
WAVE III 45, n 2 n_ X 3;. df X2 g__g§ggg
BELIEFS
Cheat on tests (Conventional norms & attitudes)
Not wrong 30.3 121 16.5 66 53.1 212
Wrong 57.5 699 16.2 197 23.6 320 2 109.2* .47
Damage others prop
Not wrong 8.2 4 10.2 5 31.4 492
Wrong 52.1 817 16.5 258 81.6 40 2 55.6* .79
Use marijuana
Not wrong 35.6 196 15.3 84 49.1 270
Wrong 58.6 625 16.8 179 24.6 262 2 104.9* .42
Steal ( $5.00
Not wrong 25.0 55 8.6 19 66.4 146
Wrong 54.9 766 17.5 244 27.7 386 2 129.0* .58
Assault
Not wrong 22.5 27 11.7 14 65.8 79 .
Wrong 53.1 794 16.6 249 30.3 453 2 64.7* .57
Breaking & Entering
Not wrong 8.0 2 4.0 1 88.0 22
Wrong 51.5 818 16.5 262 32.0 509 2 34.9* .84
Sell hard drugs
Not wrong 12. 9 20.0 14 67.1 47
Wrong 52.5 812 16.1 249 31.4 485 2 47.3* .64
Steal > $50.00
Not wrong 2.9 1 17.1 6 80.0 28
Wrong 51.9 820 16.3 257 31.9 504 2 40.2* .80
Lie to cops 4 frnds
Yes 22.4 64 12.6 36 65.0 186
Don't know 46.6 199 17.3 74 36.1 154
No 61.6 553 17.1 153 21.3 101 4 198.3* .46
Stop frnds go astray
No 34.9 38 9.2 10 56.0 61
Dont know 45.5 95 15.8 33 38.8 81
Yes 52.9 683 17.0 220 30.1 389 4 34.3* .23
0k break rules
Agree 24.6 46 13.9 26 61.5 115
Neither 37.4 77 18.0 37 44.7 92
Disagree 57.1 697 16.2 198 26.6 325 4 112.6* .44

 

TABLE 4 (cont'd). Not Occas. Serious
Delinq Delinq Delinq
WAVE III 2 n X n Z n df X2 gamma

117

 

BELIEFS CONTINUED
Freq test cheated
None
Once or twice
3 or more

Freq hit teacher
None
Once
2 or more

Freq hit students
None
Once or twice
3 or more

Freq damage sch prop
None
Once or twice
3 or more

Freq skipped classes
None
Once or twice
3 or more

Suspension
None
Once or twice
3 or more

*p‘g .01.

(Acceptance of school authority)
(# of school rule violations)
64.5 606 14.0 131 21.5 202
36.0 127 24.1 85 39.9 141
27.2 88 14.5 47 58.3 189

54.4 821 16.8 253 28.8 435
7.9 5 7.9 5 84.1 53
6.8 3 4 5 2 88.6 39

76.2 821 12.2 131 11.8 127

39.4 115 26.4 77 34.2 100

15.9 39 16.7 41 67.5 166

57.4 821 17.5 251 25.1 359
15.6 21 79.3 107
2 0 1 92.0 46

63.5 596 15.4 144 21.1 198
39.3 95 19.4 47 41.3 100
29.8 130 16.5 72 53.7 234

55.2 786 16.1 229 28.7 409
22.1 34 18.8 29 59.1 91
2.6 1 7.9 3 89.5 34

4 208.7*
4 149.8*
4 447.5*
4 265.0*
4 178.0*
4 126.3*

.49

.84

.70

.85

.47

.62

118

TABLE 5. Significant Relationships of Variables to Delinquency Wave IV

Not Occas. Serious
Delinq Delinq Delinq
WAVE IV Z n X n X n df X2 gem

 

SCHOOL ATTACHMENT
Tchr think good stu (Sensitivity to tchr opinion)

th imprt. 43.2 41 11.6 11 44.2 42

Somewhat 51.8 324 17.1 107 30.9 193

Very imprt. 61.0 368 19.1 115 19.9 120 8 49.5* -.22
Impress tchr foremost

Disagree 59.2 652 17.8 196 22.8 251

Neither 38.8 54 16.5 23 44.6 62

Agree 31.3 25 16.3 13 52.5 42 8 61.3* .41
None at sch care (Like school)

Agree/hatesch 56.7 642 17.8 202 25.4 287

Neither 48.4 62 16.4 21 35.2 45

Disagree/likesch 46.0 29 15.9 10 36.5 23 8 29.3* .18

Dont belong at sch

Agree/hatesch 57.6 672 18.1 211 24.3 283

Neither 49.4 44 11.2 10 38.2 34

Disagree/likesch 25.4 17 17.9 12 56.7 38 8 58.1* .37

SCHOOL COMMITMENT

Imprt go to college

Not imprt. 49.7 155 18.6 58 31.4 98

Somewhat 55 .O 246 14 .8 66 30.2 135

Very imprt. 60.6 474 16.6 124 22.6 177 8 21.1* -.14
Chances complete

Poor 48.5 157 17.0 55 34.3 111

Fair 52.6 240 15.8 72 31.6 144

Good 62.7 469 16.6 124 20.6 154 8 36.3* -.20
Self rate as student (Gen. Sch. Apt./se1f rate ability)

Not well 15.6 5 15.6 5 68.8 22

Ok 54.3 430 17.2 136 28.5 226 ‘

Very well 63.5 256 20.1 81 16.1 65 6 57.3* -.27
Self rate GPA

Not well 29.2 42 18.8 27 52.1 75

Ok 56.8 460 18.9 153 24.3 197

Very well 69.9 216 13.6 42 15.9 49 8 90.3* -.37
Self rate hard subjs

Not well 25.0 29 20.7 24 54.3 63

Ok 57.2 541 18.4 174 24.3 230

Very well 67.3 152 12.8 29 19.9 45 6 69.0* -.34

119

TABLE 5 (cont'd). Not Occas. Serious
Delinq Delinq Delinq
WAVE IV 2 n X n X n df X2 ganlna

 

SCHOOL COMMITMENT CONTINUED
Grade point average

Mostly F's 18.8 3 12.5 2 68.8 11

Mostly D's 21.8 12 20.0 11 58.2 32

Mostly C's 45.4 202 19.6 87 35.1 156

Mostly B's 62.0 374 17.7 107 20.2 122

Mostly A's 70.3 142 12.4 25 17.3 35 8 98.6* -.35
How imprt sch work (Achievement index)

Not imprt. 27.1 23 16.5 14 55.3 47

Somewhat 42.4 106 21.2 53 36.4 91

Very imprt. 61.1 604 16.8 166 22.0 218 8 85.4* -.38
Do well in hard subjs

Not imprt. 28.6 10 17.1 6 51.4 18

Somewhat 53.0 260 17.5 86 29.5 145

Very imprt. 58.0 462 17.7 141 24.2 193 8 54.7* -.14
How imprt high GPA

Not imprt. 25.4 15 16.9 10 57.6 34

Somewhat 52.4 268 17.4 89 30.1 154

Very imprt. 59.8 450 17.8 134 22.2 167 8 43.5* -.22
Do sch work w/o help

Not imprt. 47.7 31 13.8 9 36.9 24

Somewhat 49.1 276 17.3 97 33.5 188

Very imprt. 61.2 426 18.2 127 20.5 143 8 52.2* -.23

SCHOOL INVOLVEMENT

Wkday aftrns study

Very little 45.0 197 17.8 78 37.2 163

Some 58.1 310 18.0 96 24.0 128

Alot 64.2 226 16.8 59 19.0 67 4 41.3* -.24
Wkday eves study

Very little 44.0 173 19.6 77 36.4 143

Some 57.7 319 15.9 88 26.4 146

Alot 63.7 240 18.0 68 18.3 69 4 39.5* -.23
Weekends study

Very little 48.0 366 18.8 143 32.9 251

Some 64.6 268 14.9 62 20.5 85

Alot 67.3 99 19.0 28 13.6 20 8 47.8* -.29

120

 

TABLE 5 (cont'd). Not Occas. Serious
Delinq Delinq Delinq
HAVE,IV Z n, Z n X n, df X2 4fi_g§gm§
BELIEFS
Cheat on tests (Conventional norms & beliefs)
Not wrong 37.3 114 18.6 57 44.1 135
Wrong 61.6 760 16.0 197 22.4 277 2 69.6* -.42
Damage others prop
Not wrong 10.5 4 15.8 6 73.7 28
Wrong 58.0 871 16.5 248 25.5 384 2 46.8* -.76
Use marijuana
Not wrong 44.1 240 17.5 95 38.4 209
Wrong 63.7 634 16.0 159 20.4 203 2 66.9* -.36
Steal < $5.00
Not wrong 19.7 34 15.6 27 64.7 112
Wrong 61.5 841 16.6 227 21.9 300 2 152.6* -.69
Assault
Not wrong 25.5 27 12.3 13 62.3 66
Wrong 59.1 847 16.8 241 24.1 346 2 74.5* -.60

Breaking & Entering

Not wrong 8.3 3 13.9 5 77.8 28

Wrong 58.0 871 16.5 248 25.5 384 2 51.2* -.80
Sell hard drugs

Not wrong 16.0 12 9.3 7 74.7 56

Wrong 58.9 863 16.8 247 24.3 356 2 93.2* -.75
Steal > $50.00

Not wrong 13.3 4 13.3 4 73.3 22

Wrong 57.6 871 16.5 250 25.8 390 2 35.2* -.74

Lie to cops 4 frnds

Yes 66.0 593 14.8 133 19.2 172

Don't know 54.5 193 20.3 72 25.1 89

No 30.3 87 17.1 49 52.6 151 4 145.5* -.40
Stop frnds go astray

Yes 38.1 40 14.3 15 47.6 50

Don't know 50.8 94 14.6 27 34.6 64

No 59.2 739 17.0 212 23.9 298 4 35.0* -.26
Ok to break rules

Agree 62.5 739 16.3 193 21.2 250

Neither 45.4 79 15.5 27 39.1 68

Disagree 30.4 56 18.5 34 51.1 94 4 68.3* .43

121

 

TABLE 5 (cont'd). Not Occas. Serious
Delinq Delinq Delinq
HAVE IV 5 n X n X n df X2 44_g§gg§
BELIEF CONTINUED (Acceptance of sch. authority)
Freq test cheats (# of sch. rule violations)
None 66.0 593 14.6 144 19.4 191
Once or twice 44.4 138 23.5 73 32.2 100
3 or more 35.2 86 15.2 37 49.6 121 4 125.8* .41
Freq hit teacher
None 60.0 875 16.8 245 23.2 339
Once 2.3 1 27.3 12 70.5 31
2 or more 0 O 2 6 1 97.4 37 4 164.9* .89
Freq hit students
None 76.3 875 11.0 126 12.7 146
Once or twice 45.1 102 27.9 63 27.0 61
3 or more 20.4 34 12.6 21 67.1 113 6 342.2* .66
Freq damage sch. prop
None 71.1 1011 13.5 190 14.9 210
Once or twice 16.7 15 21.1 19 62.2 56
3 or more 5.0 2 5.0 2 90.0 36 4 268.2* .84
Freq skipped classes
None 66.4 594 15.7 140 17.9 160
Once or twice 45.7 90 20.3 40 34.0 67
3 or more 42.4 191 16.4 74 41.1 185 4 102.1* .31
Suspension
None 61.5 884 16.0 219 22.6 310
Once or twice 21.7 30 23.9 33 54.3 75
3 or more 3.3 1 6.7 2 90.0 27 4 150.0* .67

*p.$ .01.

122

 

TABLE 6. Significant Relationships of Variables to Delinquency Wave V
Not Occas. Serious
Delinq Delinq Delinq
WAVE V Z n 2 n 3 n df X2 gamma
SCHOOL ATTACHMENT
Tchr think good stu (Sensitivity to tch's opinion)
Not imprt. 43.3 39 13.3 12 43.3 39
Somewhat 57.2 316 13.9 77 28.8 159
Very imprt. 69.0 358 11.6 60 19.5 101 4 33.1* -.26
Impress tchr foremost
Disagree 65.6 650 12.4 123 22.0 218
Neither 36.8 42 17.5 20 45.6 52
Agree 39.3 22 8.9 5 51.8 29 4 58.8* .47
Don't belong at sch (Like school)
Agree/hatesch 63.8 655 12.8 131 23.4 240
Neither 44.7 38 15.3 13 40.0 34
Disagree/likesch 42.6 23 9.3 5 48.1 26 4 28.3* -.36
SCHOOL COMMITMENT
Imprt go to college
Not imprt. 58.3 182 11.9 37 29.8 93
Somewhat 60.9 274 13.3 60 25.8 116
Very imprt. 66.4 484 11.4 83 22.2 162 4 8.8** -.11
Chances complete
Poor 54.0 167 12.9 40 33.0 102
Fair 62.4 271 12.4 54 25.1 109
Good 67.1 499 11.4 85 21.5 160 4 18.0* -.16
Self rate as student (Gen. sch. apt./se1f rating on ability)
Not well 27.3 6 9.1 2 63.6 14
0k 59.2 407 14.2 97 26.7 184
Very well 72.4 260 10.3 37 17.3 62 4 37.0* -.30
Self rate GPA
Not well 36.1 44 15.6 19 48.4 59
Ok 62.1 440 13.1 93 24.7 175
Very well 76.7 211 9.1 25 14.2 39 4 66.0* -.40
Self rate hard subjs
Not well 41.6 42 15.8 16 42.6 43
Ok 61.9 520 13.1 110 25.0 210
Very well 74.9 140 10.7 20 14.4 27 4 34.0* -.33
Grade Point Average -
Mostly F's 9.1 1 27.3 3 63.6 7
Mostly D's 32.0 17 18.0 18 50.0 25
Mostly C's 50.8 188 13.8 51 35.1 131
Mostly B's 67.0 368 12.6 68 20.4 112
Mostly A's 76.9 140 8.8 16 14.3 26 8 78.6* .36

123

 

TABLE 6 (cont'd). Not Occas. Serious
Delinq Delinq Delinq
WAVE V X n 2 n X n df X2___g§gm£
SCHOOL COMMITMENT CONTINUED
How imprt. sch. work (Achievement index)
Not imprt. 27.4 17 11.3 7 61.3 38
Somewhat 49.3 106 17.2 37 33.5 72
Very imprt. 66.8 592 11.7 104 21.4 190 4 66.8*
Do well in hard subjs
Not imprt. 33.3 11 9.1 3 57.6 19
Somewhat 55.3 244 16.1 71 28.6 126
Very imprt. 66.7 459 10.9 75 22.4 154 4 33.4*
How imprt high GPA
Not imprt. 30.9 17 21.8 12 47.3 26
Somewhat 58.8 280 12.4 59 28.8 137
Very imprt. 66.1 418 12.3 78 21.5 136 4 30.8*
Do sch. work w/o help
Not imprt. 48.6 34 15.7 11 35.7 25
Somewhat 54.0 256 15.0 71 31.0 147
Very imprt. 68.6 424 10.8 67 20.6 127 4 29.6*
SCHOOL INVOLVEMENT
Wkday aftrns study
Very little 49.7 178 13.7 49 36.6 131
Some 63.8 305 11.3 54 24.9 119
Alot 70.9 232 13.5 44 15.6 51 4 43.8*
Wkday eves study
Very little 48.4 153 13.0 41 38.6 122
Some 63.5 308 11.8 57 24.7 120
Alot 70.1 253 13.6 49 16.3 59 4 46.7*
Weekend study
Very little 57.2 332 12.7 80 34.6 218
Some 71.7 279 12.6 49 15.7 61
Alot 72.2 104 31.2 19 14.6 21 4 58.6*
BELIEFS
Cheat on tests (Conventional norms & attitudes)
Not wrong 48.1 142 12.9 38 39.0 115
Wrong 66.7 798 11.9 142 21.4 256 2 42.5*
Damage others prop
Not wrong 22.2 6 3.7 1 74.1 20
Wrong 63.8 934 12.2 179 24.0 351

2 35.6*

124

 

TABLE 6 (cont'd). Not Occas. Serious
Delinq Delinq Delinq
WAVE V Z n 3 n 2 n df X2
BELIEFS CONTINUED
Use Marijuana
Not wrong 49.3 291 13.7 81 36.9 218
Wrong 72.0 649 11.0 99 17.0 153 2 88.5*
Steal < $5.00
Not wrong 27.9 43 13.6 21 58.4 90
Wrong 67.1 897 11.9 159 21.0 281 2 111.6*
Assault
Not wrong 28.3 28 10.1 10 61.6 61
Wrong 65.5 912 12.2 170 22.3 310 2 78.1*
Breaking & entering
Not wrong 22.6 7 6.5 2 71.0 22
Wrong 63.9 933 12.2 178 23.9 349 2 36.0*
Sell hard drugs
Not wrong 27.8 27 14.4 14 57.7 56
Wrong 65.5 913 11.9 166 22.6 315 2 65.8*
Steal > $50.00
Not wrong 18.2 6 6.1 2 75.8 25
Wrong 64.1 934 12.2 178 23.7 346 2 46.8*
(Acceptance of school authority)
Freq cheated (# of school rule violations)
None 70.7 748 10.3 109 19.0 201
Once or twice 48.6 120 19.0 47 32.4 80
3 or more 38.7 72 12.9 24 48.4 90 4 109.8*
Freq hit teacher
None 65.9 948 12.7 183 21.4 308
Once 6.7 3 15.6 7 77.8 35
2 or more 0 O 0 100.0 7 4 106.2*
Freq damage sch prop
None 68.2 948 9.8 137 22.0 306
Once or twice 13.9 10 6.9 5 79.2 57
3 or more 7 1 2 10.7 3 78.6 22 6 216.0*
Freq skipped classes
None 72.4 631 10.3 90 17.2 150
Once 59.3 112 15.9 30 24.9 47
3 or more 45.6 196 14.0 60 40.5 174 4 101.6*

.43

.64

.63

.71

.61

.77

.43

.87

.84

.40

125

 

TABLE 6 (cont'd). Not Occas Serious
Delinq Delinq Delinq
WAVE V Z n Z In, Z n df X2 jgugg;
Freq suspended
None 67.3 910 11.7 158 21.1 285
Once or twice 24.5 26 17.0 18 58.5 62
3 or more 12.5 4 6.3 2 81.3 26 4 141.2* .70

*p $_.01.

126

 

TABLE 7. Significant Relationships of Variables to Suspension Wave I
Not 1 or more
Suspended Suspensions
WAVE I Z n Z n df X2 gamma
SCHOOL ATTACHMENT
Tchr think good stu (Sensitivity to tchr's opinion)
Not imprt. 4.3 66 12.3 21
Somewhat 33.7 515 40.4 69
Very imprt 62.0 949 47.4 81 2 26.5* -.30
Impress tchr foremost
Disagree 78.6 1194 57.3 98
Neither 10.7 163 15.8 27
Agree 10.7 163 26.9 46 2 45.1* .44
Nobody at sch cares (Like school)
Agree/hatesch 84.1 1287 79.1 136
Neither 7.7 118 12.8 22
Disagree/likesch 8.2 126 8.1 14 2 8.3** -.14
Dont belong at sch
Agree/hatesch 86.1 115 73.3 30
Neither 6.4 98 9.3 16
Disagree/likesch 7.5 1318 17.4 126 2 22.8* -.37
SCHOOL COMMITMENT
Imprt go to college
Not imprt. 12.0 185 25.7 45
Somewhat 26.8 411 26.3 46
Very imprt. 61.2 940 48.0 84 2 26.4* -.27
Chances complete
Poor 12.4 189 31.0 54
Pair 38.2 582 35.1 61
Good 49.4 752 33.9 59 2 46.0* -.34
Self rate as student (Gen. sch. apt./self rating on ability)
Not well 3.8 55 7.3 11
0k 57.5 841 69.3 104
Very well 38.8 567 23.3 35 2 16.1* -.35
Self rate GPA
Not well 10.6 156 31.7 51
0k 58.4 859 52.2 84
Very well 31.0 456 16.1 26 2 62.6* -.45
Self rate hard subjs
Not well 9.3 139 18.8 30
Ok 69.7 1043 70.0 112
Very well 21.0 314 11.3 18 2 19.5* -.34

 

127

TABLE 7 (cont'd). Not 1 or more
Suspended Suspensions
WAKE I Z n Z n df X2 gamma

 

SCHOOL COMMITMENT CONTINUED
Grade point average

Mostly F's .6 9 2 9 5

Mostly D's 3.8 59 9.2 16

Mostly C's 30.3 466 56.1 97

Mostly B's 46.6 717 28 9 50

Mostly A's 18.7 287 2 9 5 4 85.4* -.56
How imprt sch work

Not imprt. 5.9 90 12.8 22

Somewhat 17.1 261 30.2 52

Very imprt. 77.0 1178 57.0 98 2 34.0* -.40
Do well in hard subjs

Not imprt. 2.2 34 6.4 11

Somewhat 35.5 543 46.2 79

Very imprt. 62 3 054 47 4 81 2 20.8* -.30

SCHOOL INVOLVEMENT

Wkday evenings study

Very little 51.6 790 69.8 120

Some 21.9 335 14.5 25

Alot 26.5 405 15.7 27 2 28.0* -.30
Weekends study

Very little 60.7 928 72.7 125

Some 29.9 457 20.3 35

Alot 9.4 144 7.0 12 2 9.4* -.24

BELIEFS

Cheat on tests

Not wrong 19.4 299 37.1 65

Wrong 80.6 1241 62.9 110 1 29.5* -.42
Damage others prop

Not wrong 2.3 36 5.2 9

Wrong 97.7 1504 94.8 165 1 4.9** -.39
Use marijuana

Not wrong 15.1 232 34.5 60 .

Wrong 84.9 1307 65.5 114 1 41.6* -.49
Steal < $5.00

Not wrong 9.0 138 20.0 35

Wrong 91.0 1400 80.0 140 1 21.0* -.43

128

 

TABLE 7 (cont'd). Not 1 or more
Suspended Suspensions .
WAVE I 5 n Z n (if X2 M
BELIEFS CONTINUED
Assault
Not wrong 13.1 201 18.3 32
Wrong 86.9 1339 81.7 143 1 3.6** -.19
Breaking & Entering
Not wrong 1.4 21 3.4 6
Wrong 98.6 1519 96.6 169 l 4.3** -.43
Sell hard drugs
Not wrong 1.5 23 5.1 9
Wrong 98.5 1517 94.9 166 1 11.4* -.56
Steal > $50.00
Not wrong 1.4 21 5.7 10
Wrong 98.6 1519 94.3 165 1 16.7* -.62
Lie to cops 4 frnds
Yes 12.5 167 32.7 49
Don't know 25.2 337 25.3 38
No 62.4 835 42.0 63 2 47.1* -.40
Stop frnds go astray
No 4.3 58 8.7 13
Don't know 11.1 148 15.3 23
Yes 84.6 1133 76.0 114 2 8.6** -.26
Ok to break rules
Agree 74.4 223 56.6 47
Neither 11.1 171 16.6 29
Disagree 14.5 1143 26.9 99 2 26.0* -.35
(Acceptance of sch. authority)
Freq tests cheated (# of school rule violations)
None 53.1 817 35.6 62
1 or more 46.9 721 64.4 112 1 19.1* .34
Freq hit teacher
None 94.5 1455 74.3 130
1 or more 5.5 85 25.7 45 1 91.4* .71
Freq hit students
None 55.5 847 26.4 46
1 or more 45.0 692 73.6 128 1 51.2* .54

129

 

TABLE 7 (cont'd). Not 1 or more
Suspended Suspensions
MI Z n Z n df X2 M
BELIEFS CONTINUED
Freq damage sch prop
None 86.1 1326 66.9 117
1 or more 13.9 214 33.1 58 1 89.7* .48
Freq skipped classes
None 72.8 1121 31.4 55
1 or more 27.2 419 68.6 120 1 124.7* .70

 

130

TABLE 8. Significant Relationships of Variables to Suspension Wave III

 

Not 1 or more
Suspended Suspensions
WAVE III Z n Z n df X2 m
SCHOOL ATTACHMENT
Tchr think good stu
Not imprt. 4.9 66 17.3 32
Somewhat 43.4 582 45.9 85
Very imprt. 51.5 692 36.8 68 2 46.1* -.33
Impress tchr foremost
Disagree 80.9 1082 10.8 131
Neither 11.6 155 12.4 22
Agree 7 5 100 24.2 32 2 20.1* .28
Don't belong at sch (Like school)
Agree/hatesch 6.8 80 10.4 19
Neither 7.3 98 12.0 22
Disagree/likesch 86.7 1162 77.6 142 2 10.9* -.29
SCHOOL COMMITMENT
Importance of college
Not imprt. 16.8 240 24.0 46
Somewhat 28.4 406 30.7 59
Very imprt. 54.8 784 45.3 87 2 8.1* -.17
Chances complete
Poor 16.3 232 27.7 52
Fair 32.7 464 39.4 74
Good 51.0 723 33.0 62 2 25.1* -.31
Self rate as student (Gen. Sch. Apt./se1f rating on ability)
Not well 10.4 137 20.5 35
0k 70.2 925 70.2 120
Very well 19.4 255 9.4 16 2 21.5* -.36
Self rate GPA
Not well 9.6 124 17.8 29
Ok 63.3 815 67.5 110
Very well 27.0 348 14.7 24 2 18.0* -.33
Grade point average
Mostly F's .6 8 2.7 5
Mostly D's 3.4 46 9.2 17
Mostly C's 32.5 433 51.4 95
Mostly B's 44.8 598 32.4 60
Mostly A's 18.7 4.3 8 4 63.5* -.48

249

 

131

 

TABLE 8 (cont'd). Not 1 or more
Suspended Suspensions
WLAVE III Zr n Z t1 df X2 M
SCHOOL COMMITMENT CONTINUED
How imprt sch. work (Achievement index)
Not imprt. 5.0 67 12.4 23
Somewhat 18.0 241 27.0 50
Very imprt 77.0 1033 60.5 112 2 28.1*
Do well in hard subjs
Not imprt. 1.8 24 7.6 14
Somewhat 39.1 525 41.6 77
Very imprt. 59.1 792 50.8 94 2 23.9*
How imprt high GPA
Not imprt 3.6 48 11.9 22
Somewhat 35.3 473 42.7 79
Veryimprt. 61.1 819 45.4 84 2 33.6*
SCHOOL INVOLVEMENT
Wkday aftrns study
Very little 52.2 700 69.6 128
Some 20.8 279 12.5 23
Alot 27.0 362 17.9 33 2 27.4*
Wkday eves study
Very little 48.5 649 68.5 126
Some 22.4 300 16.3 30
Alot 29.1 390 15.2 28 2 41.9*
BELIEFS
Cheating
Not wrong 23.5 336 34.2 66
Wrong 76.5 1096 65.8 127 1 10.5*
Use marijuana
Not wrong 32.3 463 48.2 93
Wrong 67.7 970 51.8 100 1 19.0*
Steal < $5.00
Not wrong 12.4 177 23.8 46
Wrong 87.6 1256 76.2 147 1 18.9*
Assault
Not wrong . 96 12.4 24

6 7
Wrong 93.3 1337 87.6 169 l 8.1*

 

132

 

*p ( .01.

TABLE 8 (cont'd). Not 1 or more
Suspended Suspensions
WAVE III Z n Z n df X2 fi‘_g§gmg
BELIEFS CONTINUED
Breaking & Entering
Not wrong 1.1 16 4.7 9
Wrong 98.9 1415 95.3 184 14.01 -.62
Sell hard drugs
Not wrong 3.7 53 9.3 18
Wrong 96.3 1380 90.7 175 12.9* -.45
Steal > $50.00
Not wrong 1.5 21 7.3 21
Wrong 98.5 1412 92.7 179 27.0* -.68
Lie to cops 4 frnds
Yes 15.3 219 36.5 70
Don't know 27.1 387 22.9 44
No 57.5 821 40.6 78 52.0* -.35
Stop frnds go astray
No 6.1 87 12.0 23
Don't know 13.0 186 '13.0 25
Yes 80.9 1154 75.0 144 9.3* -.17
Ok to break rules
Agree 10.6 152 18.8 36
Neither 12.7 182 15.1 29
Disagree 76.7 1097 66.1 127 12.8* -.24
(Acceptance of sch. authority)
Freq tests cheat (# of school rule violations)
No cheating 60.2 863 43.5 84
1 or more 39.8 570 56.5 109 19.5* .32
Freq hit teacher
None 95.5 1368 77.7 150
l or more 4.5 65 22.3 43 86.3* .71
Freq hit students
None 71.0 1017 36.3 70 '
l or more 29.0 416 63.7 123 92.4* .62
Freq damage sch prop
None 90.7 1299 73.1 141
1 or more 9.3 133 26.9 52 52.5* .56
Freq skipped classes
None 62.0 889 28.5 55
1 or more 38.0 544 71.5 138 78.5* .60

133

TABLE 9. Relationships of Suspension and Delinquency by Waves

 

Not Occas. Serious
Delinq Delinq Delinq
Z n Z n Z n df X2 gamma
WAVE I
Suspension
None 39.7 609 22.2 341 37.9 582
Once or twice 9.4 12 19.5 25 70.3 90
3 or more 9.1 4 15.9 7 75.0 33 4 95.0* .59
WAVE III
Suspension
None 55.2 786 16.1 229 28.7 409
Once or twice 22.1 34 18.8 29 59.1 91
3 or more 2.6 1 7.9 3 89.5 34 4 126.3* .62
WAVE IV
Suspension
None 61.5 884 16.0 219 22.6 310
Once or twice 21.7 30 23.9 33 54.3 75
3 or more 3.3 1 6.7 2 90.0 27 4 150.0* .67
WAVE V
Suspension
None 67.3 910 11.7 158 21.1 285
Once or twice 24.5 26 17.0 18 58.5 62
3 or more 12.5 4 6.3 2 81.3 26 4 141.2* .70

 

*p 5 .01.

134

TABLE 10. Two Step Regression Analysis

 

Self Reported Delinquency Measures

 

STEP 1 STEP 2
Egedictor Variables r Beta T Beta T
SUSPENSIONS
Susp 1 .20* .01 .47 .01 .46
Susp 3 .14* .16 5.19* .02 .81
DELINQUENCY
Delinql .O7* .03 1.22 .01 .42
ATTACHMENT 1
Tchr opinion imprt -.04* .02 .99 .01 .51
Attitude about sch .O7* -.02 -.97 .02 -.94
COMMITMENT 1
Imprt of sch work -.05* .00 .32 .05 2.13**
Self rate on ability -.03* -.OO -.31 .02 1.01
INVOLVEMENT 1
# Eves studying -.O4** .01 .38 -.Ol -.52
BELIEFS 1
Attitude assault -.O8* -.05 -1.74*** -.05 l.83***
Attitude drug sales -.O6* .01 .38 .02 .81
Attitude damage property -.11* .00 .13 .03 1.36
Attitude steal < $5.00 -.07* -.02 -.86 -.02 -.82
Freq cheated on tests .05* .04 1.68*** .01 .69
Freq truant .O8* .06 1.93** .02 .92
ATTACHMENTS 4
Tchr opinion imprt -.10* .02 .95
Attitude about sch .12* .00 .19
COMMITMENTS 4
Self rate ability -.11* -.07 -2.85**
Imprt of scho work -.14* .00 .09
Imprt of college -.06* .03 .97
INVOLVEMENT 4
# Afternoons study -.12* -.03 -1.35
BELIEFS 4
Conv Norms & attitudes -.22* -.11 -3.30**
# Sch rule violations .22** .29 10.84**

*p < .01: **p g .05: ***p g .10.

R-Square=.05
F=5.1*

R-Square=.29
F=10.99*

BIBLIOGRAPHY

BIBLIOGRAPHY

Akers, R.L., Krohn, M.K., Lonza-Kaduce, L. and M. Radosevich. “Social
Learning and Deviant Behavior: A Specific Test of a General
Theory,” American Sociology Review. 44: 1979, 636-655.

Ball, R.A. “An Empirical Exploration of Neutralization Theory.”
Criminological Review. 4: 1976, 22-32.

Buehler, R.E., Patterson, G.R., and J.M. Furniss. “The Reinforcement of

Behavior in Institutional Settings.” Behavior Research and
Therapy. 4: 1966, 157.

Babbie, Earl. The Practice of Social Research. Belmont, Ca: Wadsworth
Publishing Co., 1983.

Children’s Defense Fund. School Suspension: Are They Helping Children.
Cambridge, MA: A washington Research Project Report, 1975.

Children’s Defense Fund. Children Out of School in America.
Cambridge, MA: A Washington Research Project Report, 1974.

Elliott, D.S., and H. Voss. Delinquency and Dropout. Lexington, MA:
D.C. Health Publications, 1974.

Elliott, D.S. “Delinquency, School Attendance and Dropout.” Social
Problems. 13:196, 307-314.

Elliott, D.S. Delinquency, Opportunity and Patterns of Orientations.
Ph.D. Dissertation, Seattle, WA: University of Washington Press,
1961.

Glueck, S. and E.T. Glueck. Unraveling Juvenile Delinquency.
Cambridge, MA: Harvard University Press, 1950.

Glueck, S., and E.T. Glueck. Physique and Delinquency. New York:
' Harper and Row, 1956.

135

Gold, M. Delinquent Behavior in an American City. Belmont, CA:
Wadsworth Publishing Co., 1970.

Gottfredson, G.D. Role Models, Bonding and Delinquency: An
Examination of Competing Perspectives. Baltimore, MD: Johns
HOpkins University, 1982.

Hargreaves, D.H. Social Relations in a Secondary School. London:
Routledge and Kegan Paul, 1967.

Hirschi, T. Causes of Delinquency. Berkley, CA: University of
California Press, 1969.

Jensen, G.F. “Parents, Peers and Delinquent Action: A Test of the
Differential Association Perspective.” American Journal of
Sociology. 78:1972, 562-575.

Johnson, R.E. Juvenile Delinquency and Its Origins. Cambridge, MA:
Cambridge University Press, 1979.

Johnstone, J.W. “The Family and Delinquency: A Reappraisal.” Youth
and Society. 14: 1981, 25-63.

Kachigan, S.K. Statistical Analysis: An Interdisciplinary Introduction to
Univariate and Multivariate Methods. New York: Radius Press,
1986.

Kelly, D.H. and R.W. Balch. “Social Origins and School Failure: A
Reexamination of Cohen’s Theory of Working Class Delinquency.”
Pacific Sociology Review. 14: 1971, 413-430.

Kornhauser, R.R. Social Sources of Delinquency. Chilcago, IL:
University of Chicago Press, 1978.

Lagrange, R.L., and HR. White. Age Differences in Delinquency: A Test
of Theory. Paper presented at the American Society of Criminology
meeting, Denver, Co., 1983.

Leeman, P. “Individual Values, Peer Values and Subcultural
Delinquency.” American Sociology Review. 33: 1968, 219-235.

Matza, David. Delinquency and Drift. New York, John Wiley, 1964.

136

Minor, W.W. “Techniques of Neutralization: A Reconceptualization and
Empirical Examination.” Journal of Research in Crime and
Delinquency. 18, 2: 1981, 295-318.

Polk, K. and W. Schafer. Schools and Delinquency. Englewood Cliffs,
NJ: Prentice-Hall, 1972.

Portland Oregon Public Schools Research and Evaluation Department.
Positive Alternatives to School Suspensions (P.A.S.S.). A Report,
1984.

Portland Oregon Public School Research and Evaluation Department.
Positive Alternatives to School Suspensions (P.A.S.S.). A Report, 1985-
86.

Reis, A.J. and A.L. Rhodes. “An Empirical Test of Differential
Association Theory.” Journal of Research in Crime and
Delinquency. l, 1964, 13-17.

Rhodes, A.L. and A.J. Reiss, Jr. “Apathy, Truancy and Delinquency as
Adaptions to School Failure.” Social Forces. 48: 1969, 12-22.

Schafer, W.E. and K. Polk. “Delinquency and the Schools.” Appendix
M to Task Force Report: Juvenile Delinquency and Youth Crime.
Washington, D.C.: US Government Printing Office, 1967.

9

Short, J.F. Jr. “Differential Association and Delinquency.’ Social

Problems. 4: 1957, 233-239.

Short, J.F. Jr. and EL. Strodtbeck. Group Process and Gang
Delinquency. Chicago, IL: University of Chicago Press, 1965.

Stinchcombe. Rebellion in a High School. Chicago, IL: Quandrangle,
1964.

Toby, J. “Social Disorganization and Stake in Conformity.” Journal of
Criminal Law, Criminology and Police Science. 48: 1957, 12-17.

Trojanowiez, R. and M. Morash. Juvenile Delinquency Concepts and
Control. Englewood Cliffs, NJ: Prentice Hall, 1987.

Vergon, Charles B. Disciplinary Actions in Michigan Public Schools:
Nature, Prevalence and Impact, 1978-86. A Report, 1990.

137

Voss, H.L. “Differential Association and Reported Delinquent Behavior:
A Replication.” Social Problems. 12: 1964, 78-85.

138