University of Southern California Dissertations and Theses

Delinquency As A Function Of Intrafamily Relationships

(USC Thesis Other)

Delinquency As A Function Of Intrafamily Relationships

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content This dissertation has been
microfilmed exactly as received 6 7 -1 3 ,7 6 4
VENEZIA, P eter Salvatore, 1931-
DELINQUENCY AS A FUNCTION OF INTRAFAMILY
RELATIONSHIPS.
U niversity of-Southem C alifornia, Ph.D ., 1967
Psychology, clin ical
U niversity M icrofilm s, Inc., A n n Arbor, M ichigan
Copyright by
Peter Salvatore Venezia
DELINQUENCY AS A FUNCTION OF
INTRAFAMILY RELATIONSHIPS
by
Peter Salvatore Venezia
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Psychology)
June 1967
UNIVERSITY O F S O U T H E R N CALIFORNIA
T H E G R A D U A T E S C H O O L
U N IV E R S IT Y PA R K
L O S A N G E L E S , C A L IF O R N IA 9 0 0 0 7
This dissertation, written by
.......................
under the direction of /r.lJSL.. Dissertation C o m
mittee, and approved by all its members, has
been presented to and accepted by the Graduate
School, in partial fulfillment of requirements
for the degree of
D O C T O R O F P H I L O S O P H Y
.....
Dean
. S
ACKNOWLEDGMENTS
It is with deep appreciation that I wish to ack
nowledge the guidance and inspiration provided by
Dr. Milton Metfessel. I am grateful to Dr. Constance
Lovell for her assistance during the terminal phases of
the research. I am indebted, also, to my wife, Sherri,
whose moral support and technical assistance helped to
surmount numerous obstacles.
TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS ........................................ ii
LIST OF TABLES.......................................... v
I. PROBLEM........................................ 1
Physical and Constitutional Causation
Biochemistry and the Offender
Psychological Factors as Causative
Socio-Cultural Theories
The Family's Influence upon Delinquency
Authority Relationships and Delinquency
II. HYPOTHESES TESTED .............................. 28
Delinquents and Nondelinquents Compared
FIT Scores and Possible Equivalents
III. PROCEDURE 30
The Experimental Groups
The Control Groups
The Family Information Test
Approach
IV. RESULTS AND DISCUSSION....................... 37
Normalizing of FIT Scores
FIT Reliability
Comparison of Delinquents and
Nondelinquents
Analysis of Data from Group I
Predictions to Group II; FIT Validity
Further Findings
Future Research
V. SUMMARY AND CONCLUSIONS 67
REFERENCES ................ 70
APPENDIX A ............................................. 77
APPENDIX B ............................................. 87
APPENDIX C ............................................. 91
APPENDIX D ................................. ..... 93
APPENDIX E ............................................. 94
APPENDIX F ............................................. 95
APPENDIX G ............................................. 96
APPENDIX H ............................................. 97
iv
LIST OF TABLES
Table Page
1. Summary Chart of Experimental and
Control Groups and Their Subgroups .... 40
2. Summary Chart of G I FIT Scores Corre
lated with All of the Variables
Studied....................................... 46
3. Summary Chart of the Variables Studied
Correlated with Group I Rehabilitation
Outcome....................................... 51
4. Summary Chart of Variables Studied Cor
related with Group II FIT Scores and
Outcome....................................... 55
5. Results of Predictions to Outcome in
Group I and Group II with Use of FIT
Criterion Measure from Group I ............ 58
6. Results of Predictions to Outcome in
Group II and Group I with Use of FIT
Criterion Measure from Group II ............ 60
7. Summary Chart of Predictions from G I
to G II, Independent of FIT Scores,
Based upon Objective Factors Related
to Delinquency.............................. 62
v
I. PROBLEM
This study was based on an interpretation of delin
quency as a function of intra-family relations that was
developed by the writer in the course of eleven years' work
with delinquents. Observations during this period sug
gested that youngsters who show the greatest degree of
commitment to delinquent values and the least response to
treatment are characterized by what may be termed premature
autonomy. They appear to regard themselves as adults
rather than children and to be attitudinally distant from
their families. It was also observed that, when asked for
information about various members of their families, they
did poorly in answering such questions. From this apparent
relationship grew the idea that a measure of family knowl
edge might be useful as an indicator of commitment to
delinquency and of potential for treatment. To test this
idea, such an instrument was developed and administered to
both nondelinquent and delinquent boys.
The term "delinquent" has been defined in varying
ways. In this research it refers to a minor below the age
of eighteen who has been adjudicated by the juvenile court
as (a) repeatedly committing acts which if committed by an
adult would be punishable as crimes or (b) displaying a
persistent pattern of incorrigible behavior. Incorrigibi
lity involves repeated instances of willful defiance of
parental, surrogated, or legally constituted authority.
Such individuals, admittedly, are not the only ones who
commit delinquent acts; but they are the focus of preven
tive and rehabilitative efforts.
Numerous approaches have also been made in attempt
ing to formulate the factors influencing the development of
delinquency (Cavan, 1964) . The brief overview that follows
illustrates the major ones that have been used: physical
and constitutional, biochemical, hereditary, psychological,
and socio-cultural.
Physical and Constitutional Causation
The role of biology as a factor in producing delin
quent behavior has been emphasized by Sheldon (1949). He
studied 200 delinquent youths in a private treatment insti
tution in Boston. As in his previous investigations, he
somatotyped these cases from photographs and studied their
personalities. He also explored their social and criminal
histories. Sheldon concluded that about half of the sample
became delinquent because of insufficiencies, mental and
medical, and because of psychopathic personality. Among
the remainder, he found what he believed to be significant
relationships between their delinquency and their somato-
types. However, Sheldon's study may be criticized on four
counts: (a) His sample was highly selected and small.
(b) He utilized no control group. (c) His findings hold
for only part of the sample, and thus do not lend them
selves to any systematic formulation as to delinquency
causation. (d) Other research has not supported his find
ings. For example, in their study of 2,000 delinquents,
Healy and Bronner (1936) concluded that physical attributes
had been a major factor in only four per cent of the cases.
Biochemistry and the Offender
Endocrinology is an infant science and much remains
to be discovered concerning the influence of the ductless
glands, and the hormones they secrete, upon behavior. From
evidence already at hand, however, it is clear that the
endocrines profoundly affect the constitution and the reac
tions of the individual. A study conducted at Sing Sing
to discover any relationship that might exist between
criminal behavior and biochemical composition has been re
ported by Podolsky (1955). The findings indicated an ap
parent correlation between excesses and deficiencies of
various hormones and specific types of criminal behavior.
Here, too, the study is weakened by the absence of con
trols. Also, it is equally logical to suspect causation in
the opposite direction: that long-standing reaction
4
patterns peculiar to the individual give rise to differen
tial effects on glandular output. In addition, conditions
such as nutrition, physical surroundings, and psychological
stress were not taken into account as factors contributing
to the results. Yet all these factors are known to affect
the endocrine system (Selye, 1956).
The Role of Heredity
Delinquency is a cultural phenomenon, determined by
socio-legal definitions and highly variable in time and
place. No definitive evidence exists to indicate that a
criminological predisposition, as such, is inheritable,
though considerable research has been conducted to investi
gate such inheritance. Montagu (1941) summarizes several
studies of identical twins and fraternal twins, in which
one or both members were criminal. Such studies have gen
erally concluded that the greater frequency of concordance
in identical twins than in two-egg twins implied an inheri
tance of criminality. However, the greater similarity in
behavioral history of the identical twins may be accounted
for, in large measure, by their closer companionship and
identification. The probability is that both encounter in
fluences leading to crime if one does. Although such
studies do not prove hereditary influence, neither does
this typical rejoinder of the environmentalists disprove
it. The question is unresolved by the research data.
Psychological Factors As Causative
The psychological approach of studying the individ
ual's personality, attitudes, and emotional conflicts in an
attempt to unravel the etiology of delinquency has turned
up some highly relevant data, but at the same time has run
into obstacles. Schuessler and Cressey (1950) surveyed 113
studies that had been made during the previous 25 years in
testing the personality characteristics of offenders. They
found that 42 per cent of the studies showed personality
differentiations in favor of the non-offenders while the
remainder were indeterminate in their conclusions.
Schuessler and Cressey criticized the tests employed, the
nonrepresentative character of the samples, the inadequate
control groups, the a priori hypotheses, and the interpre
tations derived from the data.
In the same year that the Schuessler and Cressey
survey was published, the Gluecks (1950) published an exten
sive psychodiagnostic study that may be used as an illustra*
tion of this approach. Delinquents and nondelinquents were
compared on the basis of test protocols and psychiatric
interviews, conducted independently. The two methods
agreed in showing significant differences between the two
groups with respect to numerous traits. For example, it
was concluded that delinquents were significantly more
socially assertive, defiant, ambivalent to authority, and
resentfuj of others, to cite only a partial list of the
6
differences found.
However, though the groups differed significantly,
there was considerable overlap between individuals from the
two groups. Also, only a minority of either delinquents
or nondelinquents clearly displayed the qualities noted.
Thus, the findings lack predictive utility, a characteris
tic shared with other psychodiagnostic studies of delin
quents. Additional objections to this method are that
delinquents, by virtue of their attitudes toward adults
and their life experiences, do not make amenable testing
subjects, and that psychodiagnosticians do not ordinarily
possess objective or experiential norms necessary for
effective testing of the delinquent (Jacks, 1962).
Psychological theories, as well as research ef
forts, run aground when dealing with the causes of delin
quency, and, typically, the foundering occurs against the
all-too-familiar rocks of prediction and primacy of causa
tion. The overwhelming number of individuals suffering
from the emotional handicaps postulated as causative never
become delinquent; and for each delinquent who does fit
the theory, the ideopathic approach must be employed to
explain how he became "disturbed." Thus, though these
formulations appear to explain a given delinquent, they
could not have predicted his delinquency. They add little
to the understanding of the etiology of delinquency other
than the concept that a youngster's experiences during
7
formative childhood years play a significant role in wheth
er or not he will be a delinquent.
Socio-Cultural Theories
These formulations all deal with gang-type delin
quency occurring primarily in lower-class, urban communi
ties. Thus, they are of limited usefulness when dealing
with the broad spectrum of delinquency.
One viewpoint sees the delinquent subculture as
arising from the gap between the aspirations of lower-class
boys and the means available to them to realize these
aspirations. According to this view, lower-class social
ization does not equip boys to perform adequately in
middle-class-dominated institutions such as school. Conse
quently, they suffer status deprivation and low self
esteem. The presence of large numbers of such boys in
effective interaction in urban areas leads to the genera
tion of group-held values which serve simultaneously to
recoup the loss of self-esteem and to insulate the boys
from further "status punishment." Thus, a delinquent sub
culture develops which values precisely what middle-class
institutions devalue; e.g., "hanging around" instead of
industriousness and aggressiveness instead of self-control.
This general viewpoint is associated with the work of Cohen
(1958), Cloward (1959), and Ohlin (1959).
A second major position sees the values and beliefs
8
of the street-corner group as simply the adolescent version
of "lower-class culture" (Shaw, 1942). This cultural sys
tem is described as valuing masculine toughness, sharpness,
and other traits that loom large in gang life. The empha
sis in this version is not upon the gang's culture being
derived from a reaction to the demands of middle-class
culture. Instead, it espouses the view that lower-class
culture as a more-or-less systematized body of beliefs,
values, "focal concerns," and even household form, has
existed in its own right for generations. It is character
ized by the female-based family as a very frequent house
hold form, and also by marked development of the one-sex
peer-group as an associational form at all ages.
A third approach stresses the element of age and
differential access to material and prestige goals (Block
and Neiderhoffer, 1959). According to this viewpoint, the
adolescent gang arises as an attempt to create and maintain
a set of status criteria to deal with the inability of
adolescents to gain access to the rights and privileges of
adults. It serves to fill the gap in our society between
childhood and adulthood. Group names, uniforms, tattoos,
and other symbols fill the same social and psychological
function as the transition rights of many primitive
societies.
A final example of socio-cultural theories empha
sizes the fact that effective control over the young is
9
made difficult or impossible in many urban areas because of
such factors as culture conflict between immigrants and
their children, rapid population movement, and the presence
of adult criminal groups (Tefferteller, 1959). Under these
conditions, delinquency becomes a group tradition trans
mitted from older to younger boys.
Several criticisms can be levelled at these socio
cultural approaches to delinquency etiology:
(a) Delinquency rates are increasing twice as fast
in rural as compared to urban areas. The increase of
arrests of persons under 18 years of age is 16 per cent in
cities with a population of 25,000 or under, while it is
only 8 per cent in cities of 25,000 or more (Perlman, 1959).
This is a fact for which these theories cannot account.
(b) The majority of youngsters in lower-class
neighborhoods are non-gang members and are law abiding.
This suggests that factors other than social-class position
and gang membership play a determining role, and research
has given support to this idea. Reckless, Dinitz, and Kay
(1959) showed that, although the socioeconomic status of
delinquent and nondelinquent boys in high delinquency areas
did not differ, home conditions manifested significant dif
ferences. Those of the nondelinquents presented a picture
of greater stability and cohesion and of better parent-
child relationships. Scarpitti (1965) found that negative
perceptions of middle-class values and feelings of blocked
10
opportunities did not result in delinquent behavior of
lower-class boys if other aspects of the personality per
mitted the boys to see themselves as capable. Spiller
(1965) studied two urban, lower-class gangs and found
little evidence to support the hypothesis that much delin
quency can be accounted for by the rejection of, or in
ability to attain middle-class goals. Briar and Piliavan
(1965) concluded from their research that the "delinquent
subculture thesis" and the general class of theories of
which it is a part are unable to account satisfactorily for
these crucial aspects of delinquency.
(c) Lower-class urban minority groups living in
high delinquency areas can and do have the lowest rate of
delinquency and crime (Robison, 1961). This has classical
ly been the case of the Chinese and Japanese in the United
States, despite outright discrimination against them by the
white majority. Compared to the indigenous population,
these Oriental groups emphasize: strong family ties, con
sistent discipline of children, maximum supervision of
their activities, and a demand that children respect and
obey parents and other representatives of authority (Woods,
1956). In addition, Oriental parents tend to provide their
children with attitudinal and behavioral models conducive
to law-abiding behavior in the latter (Hsu, 1953). Thus,
it appears that family traditions and child rearing prac
tices are much more closely related to delinquency than
11
are socioeconomic class, area delinquency rates, or ethnic
group membership, per se.
(d) There is considerable evidence that the lower-
class is not the exclusive generator of delinquency it is
purported to be. Several years ago in New York's West
chester County, one of the wealthiest neighborhoods in the
world, it was found that 151 of its youngsters, between
the ages of thirteen and nineteen, were involved in drink
ing, promiscuous sexual behavior, and use of narcotics.
These children repeatedly indulged in serious delinquent
behavior in spite of the fact that the community had orga
nized far more social and recreational facilities to keep
teenagers out of mischief than most communities in America.
In other words, serious delinquency had entered families
which had given their children every opportunity that money
could buy.
That delinquency runs through all strata of society
was given statistical support in a study made by Nye and
Olson (1958). Of 2,350 high school boys and girls examined
in half a dozen Midwestern communities, the researchers
found no significant differences in delinquent behavior in
the different socioeconomic strata. Their analysis indi
cated that the problem of delinquent behavior may be more
evenly distributed in the various socioeconomic strata than
the official records lead one to believe. There is re
search evidence to show that a majority of youngsters,
12
regardless of social class, break the law during adoles
cence (Cohen, 1955).
One reason why delinquency appears to be primarily
a lower-class phenomenon is that the concentration of wel
fare services is located in such areas. Social-welfare
workers work with minority, lower-class, urban groups, and
tend to make the judgment that juvenile delinquents come
from this group.
Two other factors help to create this impression.
The first is that by the nature of their aggressive acts,
lower-class delinquents call attention to themselves. The
forms of delinquency in the upper classes tend to be more
subtle, and the delinquent youngsters tend to be overtly
more conforming to societal norms and amenable to adult
direction. Secondly, in lower-class areas, a juvenile is
much more quickly picked up for an offense and much more
frequently referred to the court than are youngsters in
more privileged areas. "If two persons on different eco
nomic levels are equally guilty of the same offense, the
one on the lower level is more likely to be arrested, con
victed and sent to an institution," stated the late E. H.
Sutherland (1949, p. 29),
The Family's Influence upon Delinquency
The role of the family in connection with delin
quency has been considered from various aspects.
13
Considerable study has been made of the relationship be
tween broken homes and delinquent behavior. One compre
hensive study involved over 40,000 delinquents who appeared
in municipal courts during a five-year period (Monohan,
1957). The data indicated that not only were broken homes
related to delinquency but that type of broken home was an
important consideration. The investigator found that the
percentages of all cases in the recidivist class among
White boys increased from 32, where both parents were mar
ried and living together; to 38, where the father was dead
and the boy was living with his mother; to 42, where both
parents were dead; to 46, where the parents were living
apart; to 49, where the parents were divorced; to 55, where
the boy was living with his unmarried mother. The same
pattern prevailed for both sex and color group.
The Gluecks (1950), also, found that type of break
is important. In their samples, delinquency was correlated
most frequently with abandonment by unmarried parents,
desertion, temporary separation, or prolonged absence due
to incarceration of a parent. Among the nondelinquent
controls, the first break in the family had occurred most
frequently as a result of divorce or separation, death of
a parent, and prolonged absence due to illness.
In light of the evidence of a relationship between
delinquency and type of broken home, it has been postulated
that the association between the two can be accounted for
14
largely or entirely on the basis of defective family rela
tionships in broken homes. A number of authorities have
expressed this view* but Monohan (1957) expressed it the
most succinctly, "All in all, the stability and continuity
of family life stands out as the most important factor in
the development of the child. It would seem, therefore,
that the place of the home in the genesis of normal or de
linquent patterns of behavior should receive greater prac
tical recognition. The relationship is so strong that, if
ways could be found to do it, a strengthening and preserv
ing of family life among the groups which need it most
could probably accomplish more in the amelioration and
prevention of delinquency and other problems than any other
single program yet devised."
Merton (1957), too, placed primary emphasis upon
ongoing family relationships in the genesis of delinquency.
He indicated that the quality of a child's associations
within the family group greatly affects his conduct in
other groups, as his experience expands to include them.
The importance of the kind of relationships found
in the home was also emphasized by Nye (1958). He found a
small but significant difference in delinquent behavior
between youths from broken and unbroken homes but encoun
tered less delinquent behavior in broken than in unhappy
homes. In other words, the happiness factor was more
closely related to delinquency than was formal family
status.
In the Gluecks1 study (1950) a close relationship
was observed between delinquency and the affection and dis
cipline received in the home. Physical punishment was the
method of discipline favored by more than half of the
mothers and by two-thirds of the fathers of their sample of
delinquents, as compared with one-third of the parents of
nondelinquents. Few of the parents of the delinquents were
inclined to reason with their sons. Also, these parents
were found most frequently to be mentally retarded, emo
tionally disturbed, alcoholic, criminalistic, dependent,
irresponsible, or ailing in health. Significant correla
tions existed especially with the occurrence of unfavorable
traits in the mothers' families and, as the Gluecks ob
served, the mother assumes major responsibility in the
early phases of child-rearing in our society.
The primary role of intra-family relationships in
the genesis of delinquency has come under increasingly
minute scrutiny. A classic in this field was the research
conducted by Healy and Bronner (1936). Theirs was the
first action research designed to study family relation
ships rather than the individual delinquent. The attempt
was to explain why so often one child in a family became
delinquent while another did not. Their focus was upon
whether the parent-child relationship might have a differ
ent meaning for each sibling. The findings indicated that
16
the only way in which the delinquent differed from his
nondelinquent sibling was in the nature of their relation
ship to their parents. In 91 per cent of the cases the
delinquent child felt thwarted and rejected from early in
life, even though in many instances the parents were un
aware either of their own role in the delinquent's concept
of himself or of his feelings toward his family.
The role of family relationships in delinquency was
examined by Craig and Glick (1963), also, under the aus
pices of the New York Youth' Board. This study involved
gathering extensive information on family relations, by
means of interviews by case workers, for a sample of 244
boys. Using this information, the boys were rated on the
basis of three social criteria: supervision of boy by
mother, discipline of boy by mother, and cohesiveness of
family. On the basis of these ratings, predictions were
made as to whether or not the boys would show delinquent
behavior. Their cases were followed from the time they
entered elementary school till they reached the age of 17.
Of the 27 boys predicted delinquent, 23 were serious or
persistent offenders. Of the 193 predicted nondelinquents,
186 were, in fact, nondelinquents. Of the 19 boys who were
predicted as having an almost even chance of becoming de
linquent or remaining nondelinquent, almost half were delin
quent and the other half nondelinquent. The authors con
cluded that the Modified Glueck Prediction Table is a
17
method for sensitizing us to certain specific factors in
family life which, if allowed to persist, will tend to
produce delinquency.
These results, obtained by the Youth Board, are not
unique. They are supported by similar findings of the
Maximum Benefits Project in Washington, D.C., in which the
Social Prediction Table was applied to 179 children from
high delinquency areas (Craig and Glick, 1963).
Despite the fact that the Glueck Social Prediction
Table is yielding significant data on the etiology of delin
quency and its predictive validity is being established,
it nonetheless suffers from two shortcomings. First, con
siderable time, expense, and professional skill are neces
sary to make accurate predictions. Considerable family
information must be compiled by skilled investigators, and
highly trained raters must then evaluate the data and come
to agreement on the ratings, a total process which makes
large-scale application of the table impracticable. Sec
ondly, there is reason to believe that accuracy of predic
tion would fall off sharply if the same methods were
applied to predicting outcomes for children from so-called
low delinquency areas. It is apparent that parents of
"better" families are more adept at presenting a socially
acceptable facade, are less subject to overt turmoil, and
are perhaps more subtle in giving their children reason to
reject parental authority.
18
Needed at the present time is not only a predictor
of potential delinquents but an indicator of the extent to
which a given young offender is attitudinally committed to
delinquency, thereby yielding a prediction of his amena
bility to rehabilitative efforts. The following formula
tion focusing on defective intrafamily situations in terms
of authority relationships suggests such an approach. It
has been derived primarily from the investigator's experi
ence with delinquent boys.
Authority Relationships and Delinquency
If the American ideal of parent-child relations
can be summed up in a slogan, it is as stated by Smart and
Smart (1953, p. 72), "Parents protect and children are
dependent." Children who have parents upon whose authority
they can rely for security are rewarded with love and
affection for so doing, and, thus, surrender part of their
potential independence for close parental supervision. The
feeling of dependence and the love that sustains it cement
the authority of parents over children. Moreover, this
feeling of dependence allows the youngster to define him
self as a "child." He can afford to feel irresponsible
and protected in ways which most adults cannot.
The child who has become dependent upon his par
ents, however, is something of a captive in that he can be
made frightened and insecure when love and affection are
19
withdrawn. The essence of socialization, thus, resides in
the differential use of parental power. Not only do most
parents define some kinds of behavior as good and others
bad; they also act upon these value judgments by subjecting
their children to positive and negative sanctions. The
temporary withdrawal of love is more or less implicit in
the use of negative sanctions. This withdrawal is usually
frightening to the child; therefore, he is subtly coerced
into seeking approval by conforming to parental expecta
tions .
The authority that adults, other than parents, have
over children rests on different foundations. It is not
the threat to withdraw love and affection that makes chil
dren obey teachers, police, or neighbors. It is, rather,
the fact that parents have made any adult a legitimate
authority figure in the eyes of the child by means of their
relationship with the latter. Unless parents effectively
endow such people with a temporary title to rule, children
will not accept such authority, and later, they will reject
society’s role as the ultimate authority. To help insure
that the authority of other adults is made legitimate, boys
and girls are made aware, sometimes painfully, that such
people can and do inform parents of their misadventures.
Since parents typically have the ultimate power, they be
come the center of the communications network for other
adult authorities. Whether the child is at school or on
20
the streets, he is made to feel that none of his behavior
can be hidden from his parents.
The youngsters who receive the necessary emotional
props to define themselves as dependent do not mind being
the subjects of this frequently unintentional and unrecog
nized conspiracy. They, in fact, would feel very uncom
fortable without the security of such limits. Most
American children feel compelled to give their parents
self-incriminating information even when there is reason
able assurance that parents will not receive this informa
tion from other sources. The children feel the need to
confess in order to relieve themselves of the responsi
bility of hiding information.
The authority relationships which become the
building blocks for socialization do not, however, develop
automatically. The capacity of children to be dependent
must be cultivated and made secure before it can be used
to sanction and teach. The agents of socialization must
prepare their subjects properly in order to rule them.
Indeed, the story of childhood is not really about chil
dren. It is the story of the way parents, primarily, and
other significant adults rear children. This can be seen
quite clearly in cases where, for a variety of reasons,
these authority relations never develop. The children in
volved are left to socialize themselves by dealing with
their own problems as they arise. In fact, it is the
21
investigator's observation from working with nearly 3,000
delinquents, age 12 to 16, that youngsters who have this
task thrust upon them no longer consider themselves chil
dren- -they do not define themselves in the same way as do
ordinary young people.
Some of the family difficulties that can stunt the
growth of respect for authority in the eyes of the child
have been discussed earlier in this paper. Actually, there
are a myriad of psychological, structural, and interper
sonal factors that underlie failures of parental control.
Descriptive of these factors are studies by Andry (1960),
Ball (1962), Bandura (1959), Glueck and Glueck (1962),
Hartelius (1965), McCord and McCord (1964), Medinnus
(1965), Reiss (1951), Silver (1966), and Stott (1965).
While any or all such conditions as parental: alcoholism,
emotional disturbance, permissiveness, overstrictness,
culture-conflict, physical handicap, mental insufficiency,
poverty, neglect, indifference, separation, and marital
discord could be considered causes of delinquency, it is
more precise and appropriate to consider the impact these
conditions have upon the child's acceptance-rejection of
adults as meaningful, important, and protective symbols of
society.
Regardless of what accounts for specific failures
of parental authority, the net result is more or less uni
form. The child's relationship with his parents becomes
22
extremely problematic. Either feelings of dependence are
not cultivated and the source of parental control is lost,
or the parental power that does exist is not effectively
and systematically used. Thus, the child loses whatever
benefit is inherent in being responsive to parental con
trol. Further, since the parents cannot, or do not wish
to, assume responsibility for the problems of their chil
dren, the latter learn to hide these problems from them.
The childhood "confessionals" become impossible, and the
burden of responsibility for behavior, thus, cannot be
lightened. The parents who provide material support for,
but fail to lead their children, must be manipulated by
them so that the latter's difficulties in other settings
are concealed. Rather than confide in parents, it becomes
easier for the child to keep his problems to himself, even
though concealment is a man-sized job. And, since the
child must shoulder the responsibilities that his parents
do not, his burdens begin to approximate those shouldered
by adults.
Moreover, children who cannot afford to feel depen
dent upon their parents begin to view themselves as adults. ,
They become reluctant to define themselves as "children"
because they cannot face the strains of being dependent
without the supporting cast that is crucial to a stable
definition of this role.
This state of premature autonomy results in certain
23
behavioral, as well as subjective, manifestations. Given
the failure to develop respect for the authority of par
ents, an absence of communications between parents and
other authority figures, and the development of a prema
turely adult self-definition, autonomous children are
likely to make most of their own decisions. Someone must
assume the responsibility for decision-making about routine
behavior. If parents do not manage to win this responsi
bility from their children, then the latter will assume
(with a certain telling logic) that the responsibility is
theirs. These are the children, then, who make such deci
sions as: Not to attend school on nice days when they have
something "better to do," or when they do not want to face
the teacher with a lack of homework. This type of
decision-making combined with a disregard for society's
authority, as embodied in its laws, leads to increasingly
more serious offenses as the person emerges into adoles
cence and young adulthood; and, the more effective the
surrounding neighborhood is in nurturing budding del in-
*
quency, the more rapid and striking is this process.
In addition to the loss of restraints against com
mitting illegal acts, one of the most important consequen
ces of premature autonomy shows up in encounters with adult
authorities. Most children leave pre-adolescence with a
relatively firm predisposition toward accepting adult
authority in those settings where it traditionally has
24
official jurisdiction. They are inclined, a priori, to
view the authority of teachers and policemen as legitimate,
and therefore, willingly deliver the behavior that these
authorities come to expect. However, the children de
scribed previously emerge into adolescence with no such
attitudes toward adult authority. Since parents cannot
pass on the "title to rule,” the authority of officials is
not made legitimate; and the relationships between delin
quents and other authority figures are left to develop sui
generis. Each institution that comes in contact with these
youngsters must attempt to win legitimacy for its claims to
authority without help. As might be predicted, the results
of these encounters are extremely problematic. Rather than
accept the authority of school teachers and policemen, etc.,
these youngsters develop a complex set of criteria for
accepting or rejecting the authority of these officials--
a set of criteria that the adults involved frequently can
not meet.
Another major difficulty with the prematurely adult
stance is that it cannot be dropped when inappropriate. As
a self-definition, it is precariously rooted; thus, it must
be defended at all costs. The youths demand that the
adults and children with whom they interact reinforce their
precarious posture. Obviously, this facade, along with the
bulk of the delinquent's behavior, is something that teach
ers, probation officers, policemen, and adults in general,
25
as well as nondelinquent peers, are usually unable to
honor.
Perhaps, the basic feature of true premature auto
nomy, however, is a display of total indifference to the
expectations of parents, as well as to their disapproval.
This indifference is accompanied by a general withdrawal
from family life as a whole. These children become board
ers with their own families. They may eat and sleep in the
home but they remain only partially informed and almost
completely disinterested in family activities and problems.
In line with this is the investigator's experience that
the delinquent boy possesses relatively little factual in
formation about members of his family. Observations sug
gested, also, that this is the case regardless of ethnic
derivation, socioeconomic class, intelligence, personality
pattern, or offense record. Moreover, individual vari
ability in the amount of family information possessed by a
given boy seemed to correlate inversely with his degree of
commitment to delinquent values, as demonstrated by him in
the treatment setting. That is to say, the less informa
tion a boy can give about his family, the more he can be
predicted to be a difficult subject for rehabilitation.
This holds true in spite of any temporary anti-authority
attitudes, suspicion, or negativism that might be posited
to account for the delinquent child's failure to provide
such information. For, even after close rapport has been
26
established with individual boys, this writer, as well as
the other treatment personnel, have encountered the same
phenomenon.
In short, then, it appears that any one or a number
of factors exist within the delinquent child's family which
impair his ability to accept and be subservient to author
ity, in any or all of its forms, including society and its
laws. Upon the severity and extent of these factors de
pends his degree of commitment to delinquent values; atti-
tudinal distance from the family; and his factual knowledge
about family members. Also, it is this writer's belief
that influences outside the home will not determine which
child becomes delinquent. Rather, their impact appears to
be upon the rate at which a given child progresses toward
delinquency, the nature of his offenses and their fre
quency. Therefore, it is hypothesized that any of the
following four conditions, if measured, would serve as a
quantitative indicator of any of the other three.
(a) Severity and extent of authority-impairing
influences within the home.
(b) A youngster's degree of delinquent commitment.
(c) His attitudinal distance from the family.
(d) The amount of his factual knowledge about
family members.
In view of the practical needs of working with
large numbers of children, any measuring instrument to be
27
used for predicting future delinquents or for evaluating
current offenders must be: quick and easy to administer;
independent of reading-comprehension skill; nonthreatening
to the subjects; quick and easy to administer and score;
objective in its results; easily understandable to the non
professional user; and capable of providing well defined
norms for delinquents and nondelinquents as well as their
subgroups. The present research, using a youngster's
knowledge of facts about his family as a primary variable
related to delinquency, was designed to produce just such a
measuring instrument; one based upon the investigator's
theoretical approach to the etiology of delinquency. All
hypotheses tested in the course of the investigation are
discussed below, in a separate section, for the purpose of
clarity. They were derived from the investigator's theo
retical formulations outlined in the preceding paragraphs,
and do not depend upon other constructs or assumptions.
II. HYPOTHESES TESTED
The following hypotheses were derived from the
assumption that delinquency and lack of family information
are directly related to defective intrafamily relation
ships .
Delinquents and Nondelinquents Compared
Hypothesis One. Delinquents possess less knowledge
about their families than nondelinquents. Therefore, (a)
delinquent boys' scores on a Family Information Test (FIT)
will differ significantly from those of nondelinquents,
and (b) FIT Scores of nondelinquent siblings of delinquent
boys will be intermediary to, and significantly different
from, those of the other two groups.
FIT Scores and Possible Equivalents
Hypothesis Two. The more delinquent the boy, the
less he knows about his family. Thus, (a) FIT Scores will
differentiate significantly between subgroups of delinquent
boys classified by treatment workers along a continuum of
delinquency, and (b) the higher the FIT Score, the greater
the probability of a boy being classified into the most
delinquent subgroup.
29
Hypothesis Three. The more delinquent the boy, the
less susceptible to treatment he is. Therefore, (a) FIT
Scores of subgroups of boys classified according to treat
ability will differ significantly and (b) the higher the
FIT Score, the less treatable the boy.
Predictions to Rehabilitation Outcome
Hypothesis Four. FIT Scores, Treatment Prognosis
Ratings, and Delinquency Classification Ratings are alter
native methods of predicting to outcome in a given rehabil
itation program. Therefore, FIT Scores will predict to
outcome as well as either, or both, and may be substituted
for these expert judgments in situations where the latter
are not feasible.
III. PROCEDURE
The Experimental Groups
A total of 312 delinquent boys were included in the
experiment. They were chosen on the basis that the Supe
rior Court of Los Angeles County had adjudicated them as
delinquent and placed them in the Los Angeles County Pro
bation Department Junior Camp Program. The only selection
criteria for assignment to a given camp are: age 13 to 16
years, maintenance of ethnic population balance within a
given camp, and available bed space in the camp.
The first 100 experimental subjects were those boys
who comprised the total population of Camp Joe Scott,
Saugus, California, during a two-week period. Later, 224
consecutive boys arriving at the camp were added. The
total group was divided into two parts by assigning the
first 162 cases to Experimental Group I, and the second 162
cases to Experimental Group II. (This procedure made it
possible to work out predictive procedures on the basis of
Group I results and to check their efficiency by applying
them to Group II.) A loss of 12 cases was sustained
through circumstances beyond the experimenter's control.
30
31
These were boys who did not complete the program due to
the following reasons: (a) They were returned home by the
court after a rehearing. (b) They were returned home after
developing severe physical illness. (c) They ran away
prior to being given the FIT, and were not returned to the
camp by the court. As a result of these losses, the final
number of boys was reduced to 154 in Group I and 158 in
Group II.
The Control Groups
The investigator obtained, from school personnel in
the same districts from which the delinquent boys came,
lists of boys who were considered to be good citizens of
the academic community, and who were not achieving at above
the average academic level. These lists were then screened
to eliminate boys who had prior police contacts. From the
boys remaining on the lists, individuals were selected to
be included in the control group on the basis of Age,
Ethnic Group, and IQ. These control group factors were
equated to their counterparts in the experimental group,
according to the percentages shown in Appendix D. This
process produced a control group of 100 boys that differed
from the experimental group by absence in the former of
known delinquent behavior. Also, all of the controls
lived in high delinquency areas and had been classified by
school personnel as belonging to the lower socioeconomic
32
class. As Appendix D shows, this was not true of all the
delinquency boys. Any effect of this lack of matching
would be expected to decrease the amount of difference in
FIT Scores between the control group and the experimental
group rather than to increase it.
The lists referred to above gave the names of 23
boys who met the criteria for inclusion in the control
group and who were brothers of boys in the experimental
group. It was possible to interview 16 of these boys.
They were included in the above control group, and they
were also constituted as a separate control group to test
Hypothesis IB.
The Family Information Test
A measurement of each boy's amount of family knowl
edge was needed in order to test the stated hypotheses.
The investigator, on the basis of his own experience,
formulated 32 questions about family members which could be
expected to differentiate between delinquent and nondelin
quent boys. The results of a pilot study conducted by the
investigator in 1964 revealed that nondelinquents from a
low delinquency area answered four times as many of the
questions as the 100 delinquent boys mentioned above. The
FIT and instructions for its administration are given in
Appendix B. An important aspect of the test administration
was that due to the adverse reactions of most of these boys
33
to tests of any type, the FIT was camouflaged by making
the questions part of a usual and expected process of veri
fying with the boy certain necessary facts such as ad
dresses, phone numbers, the correct spelling of names, and
the correct names and residences of separated parents.
Approach
At Camp Joe Scott, a staff member of more than five
years' experience in the rehabilitation of delinquent boys
is assigned as Intake Worker. He has the responsibility to
take each boy through the intake procedure, including:
dressing in, individual interviews, group sessions, and
assignment to caseload. This worker was asked to provide
the investigator with the following information on each of
the 224 consecutive boys prior to administering the FIT to
the latter: Age, Delinquency Classification, Treatment
Prognosis, Gang Membership, Ethnic Group Membership, Neigh
borhood Delinquency, Rate, and Socioeconomic Class. To
enable the worker to obtain this information in a reliable
fashion, he was given a set of standard instructions.
These are reproduced in Appendix A. The data for the first
100 experimental cases was obtained by each of the 13 line
staff members, assigned to the camp at that time, process
ing each boy on his caseload according to the above pro
cedure. As a result of including these 100 cases in G I,
it was expected that the data and results from this group
34
would be more variable than those from the 158 boys in
G II. The latter had all been rated and tested by one
person, the Intake Worker.
Neither the test results, the classification data,
nor the theoretical assumptions underlying the project
were communicated to the staff members at the camp. The
experimental group's protocols were not scored until after
the given boys had departed from the camp. These precau
tions effectively minimized any unconscious bias which
would have spuriously affected treatment approaches and
outcome.
The approach used by the investigator to standard
ize control group interviews, and gain the boys' coopera
tion, is given in Appendix E. The FIT was given without
introduction, as part of the interview context. Sociomet
ric data were obtained from the schools by the investiga
tor. Each subject was given a code number, and all lists
of names were destroyed so as to protect the identity and
privacy of each individual and family.
It was ascertained that the California Test of
Mental Maturity was routinely administered in the school
districts and camp school. Thus, the I.Q. scores of all
boys included in the experiment are those obtained by this
test.
All FIT protocols were scored by the investigator,
according to the scoring procedure described in Appendix C.
35
The final score for each test was simply the total number
of questions that a boy was not able to answer factually,
to the nearest whole number.
The assessment of treatment outcome was obtained by
using the results of the classification system in operation
at the camp when this research program was initiated there.
Each boy, upon leaving camp, was classified by staff con
census into one of the following four categories:
1. Graduates: Those boys who benefitted maximally
from the treatment program, and whose attitudes and adjust
ment indicated a better than even chance of nonrecidivism.
2. Releasees: Those boys who, though they remained
in the program until recommended by the camp staff for re
lease, benefitted minimally from treatment. They were
viewed as having a very doubtful chance of nonrecidivism,
based upon their attitudes and adjustment.
3. Removals: Those boys who, due to their acting
out behavior and recalcitrant attitudes, could not be re
tained in an open treatment setting such as a camp. They
were returned to court for a change of treatment plan, and
were usually forwarded to a more restrictive setting such
as the California Youth Authority.
4. AWOLs: Those boys who ran away from the camp,
and who were not returned after further adjudication by the
court. If a boy was returned, the incident was considered
as part of the overall picture and treatment was continued
36
with the boy.
As a result of the above procedures, the following
data were obtained and recorded on a 3" x 5" index card
for each of the 312 boys in the experimental group:
(1)
Name
(2)
Age
(3)
FIT Score
(4)
Delinquency Classification
(5)
Treatment Prognosis Rating
(6)
Gang Membership
(7)
Ethnic Group Membership
(8)
Delinquency Area Rating
(9)
Socioeconomic Status
(10) I. Q. Score
(11)
Treatment Outcome
IV. RESULTS AND DISCUSSION
The data from the study lend themselves to logical
division into the following sections: Normalizing of FIT
Scores, FIT Reliability, Comparison of Experimental and
Control Groups, Analysis of Data from Group I, Predictions
to Group II, FIT Validity, and Further Findings.
Normalizing of FIT Scores
The raw FIT scores obtained from the 312 experi
mental subjects had a skewed distribution. (This is shown
in Appendix F.) Data in this form made parametric ap
proaches such as product-moment correlation unwarranted.
The FIT distribution was normalized by a modified square
root transformation, a permissible procedure with data of
at least the ordinal scale of measurement (Guilford, 19S6).
New FIT Score values were obtained by substituting the fol
lowing quantity for each score: ^10(X+.5), where X ■ the
original score. The transformed scores provided the more
normal distribution shown in Appendix F. The table in
Appendix H gives the transformed score for each obtainable
score on the FIT. All values for measures of central ten
dency and dispersion reported in this paper are given in
37
38
transformed scores. Their raw score equivalents may be
obtained by consulting the table.
FIT Reliability
The reliability of the test was determined by use
of the split-half method. The FIT is composed of four main
sections, each dealing with specific family members, and
containing an odd number of questions. Separation into two
halves unavoidably distributes questions from each category
unequally. Thus the specific subject matter in each half
differs. For example, one half of the items might include
three of five "Sibling" items, two of the three "Parents"
questions, six of the 11 "Father" items, and five of the
11 "Mother" questions. Because of this, two estimates of
the test's reliability were obtained, each using a differ
ent division of the items.
Each time, a table of random numbers was used to
obtain a new serial order of the test questions from 1 to
32. A separation of odd from even numbered items gave the
two halves. Using the first division of items, the relia
bility of the test was determined for the total experimen
tal group of 312 cases. Using the second division of items,
the reliability was determined for the last 101 cases in
the experimental group. In both instances, a split-half
coefficient of .59 was obtained. Application of the
Spearman-Brown Prophecy Formula (Edwards, 1954, p. 176)
39
gave a full-scale estimate of the test's reliability of
.75. Because of the necessary difference in subject matter
of the two halves, whatever division of items is made, it
may be that this is a lower estimate of reliability than is
inherent in the instrument.
Comparison of Delinquents
and Nondelinquents
Hypothesis One had two parts: (a) Delinquent boys'
scores on a Family Information Test (FIT) will differ sig
nificantly from those of nondelinquents, and (b) FIT scores
of nondelinquent siblings of delinquent boys will be in
termediary to, and significantly different from, those of
the two other groups.
Test of the hypothesis of no difference in the FIT
scores of delinquents and nondelinquents was carried out
by means of the Kruskal-Wallis One-Way Analysis of Variance
(Siegel, 1956). An obtained H of 101.58 made the null hy
pothesis untenable at far beyond the .001 level (H of
10.83 required at .001 with df = 1). Thus, the experi
mental hypothesis that FIT scores of delinquents and non
delinquents differed significantly was accepted.
Table 1 gives the FIT means of the delinquent group
and the nondelinquent group, as well as those of their
respective subgroups. The FIT mean of the nondelinquents
(7.19) was lower than that of the delinquent boys (10.30).
Since the scoring system on FIT is such that a lower score
indicates more information, the direction of the difference
in means revealed that the nondelinquents possessed more
information about their families than did the delinquents.
This direction was in accord with the interpretation made
of the expected relation of family information and delin
quency.
Table 1.--Summary Chart of Experimental and Control Groups
and Their Subgroups
Subj ects N
I.Q.
Mean
Age
Mean
Fit
Mean
All Delinquent Boys 312 86.0 14.5 10. 30
Delinquent Siblings 16 86. 75 14.9 10.42
All Nondelinquent Boys 100 89.88 14 .5 7.19
All Nonsiblings 84 89.95 14.0 6.91
Nondelinquent Siblings 16 90. 31 14.1 9. 33
Matched Nonsiblings 16 87.63 14.1 6.81
In dealing with the second part of the hypothesis,
comparison was first made of the FIT scores of delinquents
and those of their nondelinquent brothers. Sixteen pairs
of scores, with one tie, were involved in this comparison.
The Wilcoxon Matched-Pairs Signed Ranks test (Siegel, 1956)
41
was used. The obtained T of 16.5 made the null hypothesis
untenable at the .02 level (T of 20 required in a two-
tailed test, with N ■ 15). Therefore the experimental hy
pothesis that FIT scores of delinquents and their nondelin
quent brothers differed significantly was accepted. These
results were substantiated by a t test of the difference
between means of paired observations (Walker and Lev,
1953). The obtained t, with df = 15, of 2.818 exceeded
the two-tailed value of t * 2.602, for p ® .02.
To obtain a similar situation for comparison of
nondelinquent siblings and nondelinquent nonsiblings, the
control data were then used to select, from the 84 nonsib
lings, 16 who individually matched the 16 siblings in
relation to Age, I.Q., Ethnic Group, Socioeconomic Class,
and Delinquency Area. For those instances in which equal
I.Q. scores were not available, the nonsibling with the
most similar I.Q. to that of the given sibling was chosen.
As shown in Table 1, the mean FIT score of the 16 nondelin
quent siblings was 9.33 and that of the 16 matched nonde
linquent nonsiblings was 6.81. The hypothesis of no dif
ference was tested by the Wilcoxon approach. The obtained
T value of 3 permitted rejection of the null hypothesis
at beyond the .01 level (two-tailed T of 20, with N • 16).
Based on these two comparisons and on the size of
the means of the three groups involved, it was accepted
that the FIT scores of nondelinquent siblings were
42
intermediate to, and significantly different from, those
of their delinquent brothers and those of nondelinquent non
siblings. The delinquents showed less information about
their families than did their nondelinquent siblings, and
the latter in turn showed less information about their
families than did nondelinquent nonsiblings. This finding
gave support to the idea that factors associated with
family life in the homes of the delinquents had some dele
terious effect, also, upon the nondelinquent siblings.
Analysis of Data from Group I
In making the analyses necessary for testing the
remaining hypotheses in the study, all of which were con
cerned with relationships of subgroups of delinquents, the
procedure used was to deal separately with Group I and
Group II. The analyses were run with the Group I data.
This made it possible to check the stability of results, by
comparing them with the results for Group II. It also per
mitted making predictions to Group II on the basis of Group
I results and provided a check on the efficiency of the
prediction procedure.
It was predicted, in Hypothesis Two, that (a) FIT
scores would differentiate significantly between subgroups
of delinquent boys classified by treatment workers along a
continuum of delinquency, and (b) the higher the FIT Score,
the greater the probability of a boy being classified into
43
the most delinquent subgroup. Seven delinquency categories
were originally used by the camp workers in classifying the
boys. In analyzing the results, these were combined to
form two categories. This was done partly because of the
small number of cases in some categories and partly to
bring the number of categories into accord with those for
other variables, to permit the computation of similar
statistics for all. Those boys classed as Average Delin
quent, Disturbed Strongly-Delinquent, and Hard Core Delin
quent were combined into the Most Delinquent Group. Those
classed as Situational Delinquent, Fringe Delinquent, Dis
turbed Mildly-Delinquent, and Disturbed Moderately-
Delinquent were combined into the Least Delinquent Group.
These consolidations were logically consistent with the
descriptive criteria given for classification of the boys
into the seven categories.
A two-fold classification of FIT scores was also
made, by dichotomizing at the median FIT score of 11.2.
With this two-fold classification of both variables, Delin
quency Classification and FIT score, a Chi-Square test was
made of the hypothesis of no difference between FIT scores
of Most Delinquent and Least Delinquent subgroups. The
obtained Chi Square of 43.52 called for rejection of the
null hypothesis at beyond the .001 level (Chi Square of
10.83 required at .001 with df « 1).
The question of what kind of correlation
44
coefficient to use to express the degree of relationship
between variables was a difficult one in this study because
of the range of variables involved. For some pairs of
variables (e.g., FIT scores and I.Q.) it was possible to
compute Pearson r, and if that pair of variables were to
be dealt with in isolation, that correlation coefficient
would be the preferred one. For other pairs of variables,
either a biserial r or a point biserial r would be indica
ted. With still other variables, the only coefficients
that could be computed were those involving a 2 x 2 table:
tetrachoric r and phi. Tetrachoric r involves more assump
tions regarding the nature of the dichotomized variables
than does phi. The latter is subject to the disadvantage
that the maximal size of correlation possible varies ac
cording to the way in which the dichotomizing was done.
(If, for example, one variable is dichotomized so that half
the cases fall in each category, a correlation of 1.00 is
possible only if the other variable also has half the cases
in each category. The greater the divergence from a 50-50
division on the second variable, the lower the maximal
possible phi.) Since the nature of the data made it im
possible to obtain similar divisions on all variables, phi
coefficients computed for them would not be comparable
because the maximum possible phi would not be the same for
all. In view of this it was decided to use tetrachoric r
rather than phi. The assumptions underlying tetrachoric r
45
are that the two variables be continuous, normally distrib
uted, and linearly related. Though it could not be demon
strated that these requirements had been met, tetrachoric
r was accepted on the basis that a given correlation coef
ficient's magnitude be interpreted with reservations and
with regard to its level of significance. The point of
view taken in connection with this choice was that ex
pressed by Guilford (1956, p. 306): "As in most statisti
cal operations where the true form of the distribution is
unknown, we can here remember that we have taken the chance
of faulty assumptions and interpret results with the requi
site reservation." The significance levels were those of
Chi-Square, obtained as with any contingency coefficient,
and are given in Table 2.
The tetrachoric r computed for the variables of
FIT score and delinquency classification was .779. Thus,
the evidence supported the hypothesis that the higher the
FIT score, the greater the^probability of a boy being
classified into the most delinquent subgroup. The corre
lation coefficient and the level of significance suggested
that a boy's FIT score was a relatively efficient predic
tor of his commitment to delinquency, as rated by treatment
staff.
Hypothesis Three. It was hypothesized that (a) FIT
scores of subgroups of boys classified according to treat
ability would differ significantly and, (b) the higher the
46
Table 2.--Summary Chart of G I FIT Scores (N * 154) Corre
lated with All of the Variables Studied
(The coefficients given are tetrachoric correlations. The
significance levels are those obtained by Chi-Square Tests,
df - 1.)
Variable
rt
Significance
Level
I.Q. 84 and Below
vs.
I.Q. 85 and Above . 203
Not
Significant
Age 13 and 14
vs.
Age 15 and 16 .167
Not
Significant
Most Delinquent
vs.
Least Delinquent .779 .001
Fair and Poor Prognosis
vs.
Excellent and Good .653 .001
Formal Gang Members
vs.
Peripheral § Nonmembers .534 .001
Negro-Americans vs.
Mexican-Americans and
Anglo-Caucasians .124
Not
Significant
High Delinquency Area
vs.
Moderate and Low .208
Not
Significant
LowerClass
vs.
Middle and Upper .191
Not
Significant
47
FIT score, the less treatable the boy.
In testing this hypothesis, the Treatment Prog
nosis Ratings of Poor and Fair were combined, as were those
of Good and Excellent. As in the preceding analysis, the
FIT scores were dichotomized at the median for Group I,
11.2. Computation of Chi Square for the resulting 2 x 2
table gave a value of 24.70, significant at beyond the .001
level. The null hypothesis was, consequently, rejected.
This evidence, and the correlation of .653 of FIT scores
with Treatment Prognosis Ratings supported the experimental
hypothesis.
Hypothesis Four. This hypothesis is concerned
with prediction to outcome in a rehabilitative setting.
More specifically, it was hypothesized that FIT scores,
Treatment Prognosis Ratings, and Delinquency Classification
Ratings are alternative methods of predicting to outcome
and that FIT scores will predict to outcome as well as
either, or both. Analysis was first made of the relation
ship between FIT scores and outcome results.
The Kruskal-Wallis One-Way Analysis of Variance was
applied to the four groups of boys classified as Graduates,
Releasees, Removals, and Awols to test the hypothesis of
no differences in FIT scores between these groups. On the
basis of an obtained H value of 58.30, the null hypothesis
was rejected at beyond the .001 level (H value of 10.83
required with df * 3 and p ■ .001).
48
Examination of the four distributions seemed to suggest a
natural dichotomizing of the outcome variable, with Gradu
ates and Awols grouped together and with Releasees and Re
movals grouped together. The Mann-Whitney U Test (Siegel,
1956) was used to test separately the hypothesis of no
difference between FIT Scores of Graduates and Awols and
the hypothesis of no difference in FIT Scores of Removals
and Releasees. The null hypothesis was accepted in both
instances. In the former case, the obtained Z of -.1403
gave a probability of .30 that the obtained difference
between Graduates and Awols could have occurred by chance.
In the latter instance, the obtained Z of -.2891 gave a
probability of .78 for chance occurrence of the obtained
difference in FIT Scores between Releasees and Removals.
This finding with respect to the FIT Scores of
Graduates and Awols is in accord with the general impres
sion of the investigator that the majority of boys who ran
away were more similar to Graduates than to the boys in
either of the other two groups. Most of the runaways ap
peared to be the result of emotional upset (homesickness,
feelings of failure, etc.) rather than delinquent reaction
to authority. This evaluation is supported by the fact
that 14 of the 19 boys who ran away had been given a Delin
quency Classification Rating that would put them in the
Least Delinquent group. Just the reverse was the case with
the Treatment Prognosis Ratings, with 13 of the 19 boys
49
receiving ratings of Fair or Poor and only 6 being rated
Good or Excellent. A possible explanation of this diver
gence is that, in the Treatment Prognosis Ratings, emotion
al upset was considered in making ratings. The problem of
predicting Awols will be dealt with in the section on
future research. Suffice it to say at this point that FIT
scores permit no differentiation between this group and
Graduates and that a supplementary procedure, preferably
objective in nature, is needed.
Since FIT score results, Delinquency Classification
Ratings, and general impression were in accord with group
ing Awols with Graduates, rather than with Releasees and
Removals, the outcome variable was dichotomized in this
way. The Kruskal-Wallis One-Way Analysis of Variance was
applied to test the hypothesis of no difference in the FIT
scores of the two groups so formed. The obtained value of
57.83, as compared to a required H of 10.83 (with df « 1,
for p * .001) resulted in rejection of the null hypothesis.
Tetrachoric correlations between FIT scores and
outcome, Delinquency Classification Ratings and outcome,
and Treatment Prognosis Ratings and outcome,were, respec
tively, .636, .539, and .692. Bearing in mind that the as
sumptions made in computing tetrachoric r may have been
faulty, probably the most that can be said in interpreta
tion is that all three indicate at least a moderate degree
of relationship to outcome and that they are fairly similar
so
in size. This is far from definitive evidence, but is in
general accord with the experimental hypothesis that these
three procedures may be viewed as alternative predictors to
outcome and that the FIT scores predict about as well as
the others and may be substituted for them in situations
where expert judges are not available to make the ratings.
In analyzing the results of Group I, tetrachoric
correlation coefficients were computed between outcome and
all other available variables. In addition to the three
just discussed, these variables included I.Q., age, gang
membership, ethnic category, delinquency area, and socio
economic status. All these correlations are presented in
Table 3. The 2 x 2 tables used in obtaining these coeffi
cients were all set up in such a way that they showed the
tendency for favorable position on one variable to go with
favorable position on the other (e.g., high values on I.Q.
with low scores on FIT). Determination of the significance
of the relationships in the tables was made by use of Chi
Square (Siegel, 1956).
All nine of the variables were significantly
related to outcome, gang membership, membership in such
minority groups as Negro and Mexican-American, residence
in a high delinquency area, and lower-class socioeconomic
position are rather generally considered as contributing to
delinquency proneness. The finding that they were associ
ated with outcome in a rehabilitative setting is, thus, in
SI
Table 3.--Summary Chart of the Variables Studied Correlated
with Group I (N * 154) Rehabilitation Outcome
(The correlations given are tetrachoric coefficients. The
significance levels were obtained by Chi-Square tests,
df » 1.)
Variable
rt
Significance
Level
I.Q. 84 and Below
vs.
I.Q. 85 and Above .381 .01
Age 13 and 14
vs.
Age 15 and 16 .389 .01
Most Delinquent
vs.
Least Delinquent .539 .001
Fair and Poor Prognosis
vs.
Excellent and Good .692 .001
FIT Score 11.2 and Above
vs.
Below the 11.2 Median .636 .001
Formal Gang Members
vs.
Peripheral and Nonmembers .428 .01
Negro-Americans vs.
Mexican-Americans and
Anglo-Caucasians . 264 .05
High Delinquency Area
vs.
Moderate and Low .360 .01
Lower Class
vs.
Middle and Upper .485 .01
52
line with what would be expected. The comparative size of
the various correlation coefficients indicates, however,
that none of them would be as useful for predicting to out
come as the three variables of FIT score, Delinquency Clas
sification Rating, and Treatment Prognosis Rating. The
same was the case with I.Q. and age, which were signifi
cantly related to outcome but which had correlation coeffi
cients lower than the variables just mentioned.
In analyzing the results of Group I, correlation
coefficients were also computed between FIT scores and all
other variables, as was shown in Table 2. This was done
primarily for the purpose of investigating the possibility
of using more than one variable to predict to outcome, but
some of them are relevant in another connection. If FIT
scores are valid indicators of delinquency proneness, it
might be expected that they would show some relationship to
other variables that are regarded as contributing to such
proneness: gang membership, ethnic group membership, delin
quency area, and socioeconomic status. The generality of
these findings is limited by the fact that the boys studied
/
were not representative of the general population with
respect to distributions on these variables. The signifi
cant relation between gang membership and FIT scores, with
a correlation of.534, was in line with the expectation.
For the other three variables, the Chi-Square values were
not large enough to permit rejection of the null
53
hypothesis. Thus, within this restricted group, these re
sults were not in accord with what had been anticipated.
The final step in the analysis of the Group I data
involved making a statistical prediction of outcome from
FIT test scores. The procedure involved obtaining a bi
serial correlation coefficient between FIT scores and out
come and using it to determine the point on the scale of
FIT measures that would give maximal separation in the two
categories of the outcome variable (Guilford, 1965).
This was done first with the outcome variable di
chotomized into the two groups combining Graduates and
Awols, on the one hand, and Releasees and Removals on the
other. Such division permits making predictions of prac
tical value in terms of official action taken by camp
authorities to eliminate boys from the camp. The Awol
group would not come under the heading of eliminees in this
sense. The procedure was also used with a second division
of the outcome groups to form Graduates as one category
and the combination of Awols, Releasees, and Removals as
the second. In this instance, the validity of the test
was assessed in terms of its ability to predict to
graduation.
Without knowledge of FIT scores, the best predic
tion to outcome that could be made would be that of the
modal outcome category, Graduate/Awol in the case of this
subgroup. The outcome predicted for all 154 boys would,
54
thus, be that of Graduate/Awol. It would be correct in 89
of the 154 cases. Using the critical division point on
FIT, the prediction of Graduate/Awol was made for the 83
boys who had a score of 11.2 and below and of Release/
Removal for the 71 who had a score of 11.6 and above. The
results of using this procedure are presented in Table 5.
It shows that the total number of correct predictions was
114. This is an excess of 25 over the number obtained by
using the modal category. Its forecasting efficiency would
be stated as 25/89 or 28 per cent.
When the dichotomizing was done in the second way,
there was a total of 109 successful predictions out of 154.
This was an excess of 25 over the 84 in the modal category,
and it involved a predictive efficiency of 30 per cent.
Thus, the particular method of dichotomizing the outcome
variable had little influence on the results.
Predictions to Group II; FIT Validity
The Group II correlations of all variables with FIT
Scores and outcome are given in Table 4. Because the total
group of 312 delinquent boys was divided into two sub
groups, it was possible in this study to go beyond the kind
of prediction dealt with in the last section. In such a
situation the entire set of data is used as a basis for
forecasting. To the extent that the data involve aspects
particular to the sample, the procedures may be expected
55
Table 4.--Summary Chart of Variables Studied Correlated
with G II (N * 158) FIT Scores and Outcome
(Tetrachoric coefficients and Chi-Square significance lev
els are given.)
Variable
With
FIT
Scores
Sig.
Lev.
With
Out
come
Sig.
Lev.
I.Q. 84 § Below
vs.
I.Q. 85 ^ Above .399 .01 .302 .05
Age 13 $ 14
V S •
Age 15 $ 16 .227
Not
Sig. .215
Not
Sig,
Most Delinquent
vs.
Least Delinquent .865 .001 .673 .001
Fair and Poor Prog,
vs.
Excellent and Good .804 .001 .681 .001
Releases § Removals
vs.
Graduates § AWOLS .856 .001
_ _ - _
FIT Score 11.2 § Above
vs.
Below the 11.2 Median
- . _ •
.856 .001
Formal Gang Members
vs.
Peripheral § Nonmemb. .600 .001 .556 ,001
Negro-Americans vs.
Mexican-Americans £
Anglo-Caucasians .319 .05 .297 .05
High Delinquency Area
vs.
Moderate 5 Low . 324 ~ .05 .167 .05
Lower Class
VS
Middle § Upper .415 .01 .187 .05
56
to have less efficiency when applied to other groups. A
more rigorous assessment of the validity of the test is
made when the procedures based on one sample are applied
without modification to another group.
It was anticipated that, when predictions of out
come for the Group II data were made with the Group I Cri
terion Measure, the prediction efficiency would not remain
the same. This expectation was based on a preliminary
analysis of variance that was done of the eight outcome
groups (Graduates, Releasees, Removals, and Awols from
Group I and Group II, respectively). This test was made
at the outset in order to provide an overall assessment of
whether there was a significant difference anywhere among
the eight groups.
The Bartlett Test for Homogeneity of Variance
(Walker 5 Lev, 1953) gave a value of 14.133, which was only
fractionally below the one needed for rejection of the null
hypothesis at the .05 level. Consequently, it was decided
to apply a nonparametric technique. The Kruskal-Wallis
One-Way Analysis of Variance of the eight groups yielded
an H of 86.0. This far exceeded the 24.32 needed for sig
nificance with p a .001 and df * 7. This evidence of the
presence of differences among the FIT distributions sug
gested the possibility that there might be a significant
difference between Group I and Group II. This comparison,
made with the same test, gave an H of 13.85 ( with a value
of 10.83 required for significance at the .001 level with
df * 1). The hypothesis that the two sets came from the
same population of FIT scores was, thus, rejected. If the
FIT could be considered a valid indicator of delinquency
proneness, the direction of the difference between the two
distributions indicated that the boys in Group II were less
delinquent than those in Group I. Other information re
lating to this point was obtained when the Group II results
were analyzed, and it is presented in a subsequent section.
The difference found in the initial overview of the
two groups made it seem probable that techniques developed
with Group I would not have the same efficiency when used
with Group II. The results obtained with this assessment
of FIT validity are presented in Table 5. With the output
categories of Graduate/Awols and Releasees/Removals, ap
plication of the cutting score from Group I resulted in
133 correct predictions of Group II outcome. This involved
an excess of 31 successful predictions over the number ob
tained with a modal-category prediction, a predictive
efficiency of 301. When the other form of dichotomizing
the outcome variable was used, a predictive efficiency of
35% was obtained, this figure being higher than those ob
tained in predicting to outcome in Group I. These predic
tions to Group II are of particular interest because of
the similarity of this situation to what occurs in a prac
tical prediction setting once a cutting score has been
Table 5.--Results of Predictions to Outcome in Group I and Group II with Use of
FIT Criterion Measure from Group I
Predictions to Group I
Predictions to Group II
Dichotomy -
Awols with Graduates
Awols
Dichotomy -
with Graduates
Outcome
Categories
Rel 5 Grad §
Rem Awol Total
Rel $
Rem
Grad §
Awol Total
Successful
Predictions 48 66 114 35 98 133
Unsuccessful
Predictions 17 23 40 21 4 25
Total 65 89 154 56 102 158
Predictive
Efficiency*: 28%
i
30%
Dichotomy
Eliminees with Awols
I
1 Dichotomy -
Eliminees with Awols
Outcome
Categories
Rel, Rem, '
§ Awol Grad Total
Rel, Rem,
$ Awol Grad Total
Successful
Predictions 67 42 109 53 67 120
Unsuccessful
Predictions 17 28 45 16 22 38
Total 84 70 154 69 89 158
Predictive
Efficiency*: 30% 35%
*Predictive efficiency is determined by finding the excess of successful
predictions with the cutting score over the successful predictions without it
(modal category of outcome) and dividing the former by the latter.
59
developed. Had the cutting score for Group I been avail
able prior to the time when the Group II boys came to camp,
it could have provided help in determining whether or not
a boy would be sent to camp.
For purposes of comparison the same procedure that
was used in determining the cutting score for Group I was
applied to the Group II data. The score was somewhat lower
than that obtained with the Group I data, a finding in ac
cord with the significant difference in central tendency
between the two groups. (The mean of Group I was 10.82 and
of Group II, 9.79.) Since the previous analyses had indi
cated that the method of dichotomizing the output variable
did not markedly influence the results, this section of the
work was done with only one set of categories. Graduates
and Awols were combined in one group; and Releasees and
Removals formed the other. Table 6 presents the predictive
results obtained when the Group II cutting score was used
with the Group II data and also with the Group I data. The
predictive efficiency in these two cases is somewhat lower
than that found when working with the Group I cutting
score; but in the case of predicting to Group I outcome
with the Group II cutting score, the predictive efficiency
(27%) is close to those found with the other comparison.
The FIT scores gave an increase in prediction efficiency
with the use of both cutting scores on both groups, despite
the fact that there was a significant difference in the
Table 6.--Results of Predictions to Outcome in Group II and Group I with Use of
FIT Criterion Measure from Group II
Predictions to Group II Predictions to Group I
Outcome
Categories <
Rel $
v Rem
Grad ^
Awol Total
Rel §
Rem
Grad §
Awol Total
Successful
Predictions 48 45 123 56 57 113
Unsuccessful
Predictions 8 27 35 9 32 41
Total 56 102 158 65 89 154
Predictive
Efficiency 21% 27%
61
distribution of test scores between the two groups.
Analysis of other information on the two groups
provided another means of assessing FIT. If FIT was a
valid indicator of delinquency proneness, the direction of
the difference between the means for Group I and Group II
indicated that the boys in Group II were less delinquent
than those in Group I. If this were the case, it would be
expected that other variables associated with delinquency
would differ from Group I to Group II. More specifically,
it would be hypothesized that Group II would have a smaller
proportion of boys in each of the following categories:
Formal Gang Member, Negro-Americans, High Delinquency Area,
and Lower Socioeconomic Class. The results are summarized
in Table 7. All four differences were in the expected
direction, but only the decreases in Formal Gang Members
and Lower Socioeconomic status boys were significant (at
the .05 level, as determined by a Chi-Aquare test).
Further Findings
In addition to the variables already discussed,
analysis was also made of the data on age and I.Q.,for
both groups. The correlations of these variables with FIT
score and with outcome are presented in Tables 2, 3, and 4.
Both of them have lower correlations with outcome in both
groups than FIT score, Delinquency Category Rating, and
Treatment Prognosis. Although it would be desirable to
62
supplement the predictive efficiency of FIT by using other
objective variables with it in a multiple regression equa
tion, preliminary try-out of these two indicated that they
provided negligible help and that it probably would be more
profitable to focus attention on other variables.
Table 7.--Summary Chart of Predictions from G I to G II,
Independent of FIT Scores, Based Upon Objective Factors
Related to Delinquency
Predictions of Change
~X2
in Proportions from Sig.
G I to G II df a 1 Level
Decrease in Formal
Gang Members 4.20 .05
Decrease in Not
Negro-Americans .26 Sig.
Decrease in High
Delinquency Area Boys .95 Not
Sig.
Decrease in
Lower Class Boys 4.98 .05
Comparison of the results obtained with age and I.Q.
for Group I and Group II is of relevance in connection with
the question of differences between the two groups. Both
variables showed lower correlations with outcome in Group
II than in Group I. From Group I to Group II there was a
change in the age-composition of the boys in the camp. A
63
Chi-Square test of the significance of the decrease from
Group I to Group II in the number of 15- and 16-year-old
boys yielded a value of 4.77, significant at the .05 level
(which requires a value of 3.84, with df * 1). No such
change occurred with respect to I.Q. during the same peri
od. A nonsignificant Chi-Square of .12 was obtained for
I.Q. differences from Group I to Group II, with I.Q. di
chotomized at the median. It is suggested that the drop
in correlation between these two variables and outcome may
have resulted from changes in the camp treatment program
introduced to help work with the younger boys. A program
of verbal counseling has traditionally been part of the
camp program. It is generally accepted by staff in the
camp program that, other factors being equal, the older and
more intelligent a given boy, the easier it is to communi
cate verbally with him and the more he will benefit from
such verbal counseling. With the increase in proportion
of younger boys during the past year (the period during
which the 158 boys in Group II were studied), the staff has
made changes in the program to gear it to younger boys.
In place of verbal counseling, there has been increased
focus on treatment-centered activities. It might be ex
pected that these would not only meet the needs of the
younger boys, but also those of the less intelligent chil
dren. A drop in correlation between these two variables
and outcome might, thus, occur. The change, however, was
64
not large and may be simply a matter of sampling varia
bility. This probably is the case with a number of the
other variables.
However, in connection with Area Delinquency Rate
and Socioeconomic Class, another factor may have been oper
ative. Several of the 13 staff members who processed the
first 100 boys were not fully experienced in working with
delinquents. Their classification on these two social fac
tors may have included a degree of inaccuracy. The boys in
Group II were processed by a single experienced worker, so
the ratings for this group would be expected to be both
more consistent and more accurate. The finding of a sig
nificant correlation between FIT scores and these variables
is in line with the interpretation that FIT scores provide
a measure of delinquency proneness and that these two
social variables (ADR and SEC) are related to such prone
ness .
In regard to the correlations with Outcome, boys
from the Lower Class and High Delinquency Areas tend to
deal less in verbal concepts than in overt behavior. Thus,
in addition to making more probable the success of younger
boys in the camp, the activity-oriented treatment program
may also have enabled more of the Lower Class and High
Delinquency Area boys to succeed.
In general, the overall outcome picture for Group
II was better than for Group I. The former included 19
65
more graduates, 6 fewer Awols, 6 fewer Releasees, and 5
fewer Removals. This change is in the direction that would
be predicted on the basis of their lower FIT scores, though
it is not statistically significant.
One additional finding may be mentioned in connec
tion with relationship of FIT scores to delinquency prone
ness. At the camp, those boys who are positive in attitude
and demonstrate willingness and ability to be helpful to
other boys are given recognition by appointment to the
"Boy Council." Of the 312 boys in the experimental sample,
91 achieved this status. Their mean FIT score of 8.89 was
midway between that of the remaining boys in the experi
mental group (10.6) and the 100 boys in the control group
(7.1).
Future Research
Further work on the theory of family genesis of
delinquency and considerable study of FIT are necessary
before either of them can be regarded as adequately estab
lished. Because of the restricted samples investigated by
the present research, cross-sectional samples of other non
delinquent and delinquent populations will be necessary to
determine the generality of these findings. The predictive
efficiency of FIT remained essentially the same in this
study from Group I to Group II, despite the fact that there
were changes both in the characteristics of the boys and in
66
camp procedures. However, until the test is administered
in camps with markedly different treatment policies, they
may be considered applicable only in other settings similar
to this camp. If the results are supported by broader
sampling, the indefinite results discussed under Further
Findings may be clarified. Also, it may be possible to
develop norms by which to identify predelinquents. Such
is the experimenter's ultimate goal.
Study needs also to be made of the test's item
reliability and validity. It is probable that some ques
tions will have to be discarded and more efficient ones
formulated.
Finally, work needs to be done to develop supple
mentary objective instruments that, used with FIT in a
multiple regression equation, will increase predictive
efficiency. In this study it was found that discrimination
between Awols and Graduates could not be made on the basis
of FIT scores. Personal observation suggested that some
kind of measure of emotional disturbance would be helpful
in making this differentiation. Currently, ratings by
skilled personnel are the means of bringing this factor
into the picture. It would be highly desirable to have an
objective instrument to provide information with respect
to this variable.
V. SUMMARY AND CONCLUSIONS
The purposes of this research were: (a) to test
the interpretation of delinquency as a function of defec
tive intrafamily relationships, and (b) to tentatively
establish the validity and reliability of an objective
delinquency proneness test developed from the above inter
pretation .
Observations with delinquent boys in a treatment
setting suggested that these boys were characterized by:
premature autonomy, attitudinal distance from the family,
and lack of factual knowledge about family members. The
more outstanding were these manifestations of emancipation
from the family and parental authority, the greater ap
peared to be a given boy's commitment to delinquent values,
and the less his response to treatment. These observations
indicated that a measure of family knowledge might be an
efficient indicator of commitment to delinquency and of
potential for treatment.
A Family Information Test (FIT) containing 32 fac
tual questions about family members was empirically devel
oped, and administered to 312 experimental subjects and
100 controls. The experimental group consisted of boys
67
68
adjudicated as delinquent and ordered placed in the Los
Angeles County Probation Department Junior Camp Program.
Boys with clear police records, good school citizenship,
and not better than average academic achievement constitu
ted the control group, which was matched to the experi
mental group in relation to Age, I.Q., Ethnic Group Member
ship, Socioeconomic Class, and Area Delinquency Rate.
Prior to administering the FIT to the experimental
subjects, camp treatment personnel rated each boy according
to a standardized system of Delinquency Classifications and
Treatment Prognosis. When a boy departed from the camp,
the type of departure, as determined by treatment person
nel, was recorded- This gave an estimate of Treatment
Outcome.
The major findings from the research were:
(1) Delinquent boys, in placement at Camp Joe
Scott, Saugus, California, possessed significantly less
family information than the nondelinquent controls.
(2) 16 nondelinquent brothers of experimental sub
jects possessed significantly more family information than
the latter, and significantly less than 16 matched non
delinquent nonsiblings.
(3) Of all the variables studied, FIT Scores,
Delinquency Classifications, and Treatment Prognoses demon
strated the highest correlations with Treatment Outcome.
(4) The equivalence of the above correlations
69
indicated FIT Scores were acceptable substitutes for the
expert ratings of delinquent boys by treatment personnel.
(5) The FIT reliability of .75 was established by
the split-half method.
(6) The FIT's validity was tentatively established
on the basis of (a) independent evidence with regard to
objective factors associated with delinquency, and (b)
significant results from Criterion Measure predictions
independently made from one half of the experimental group
to the other.
Generalization from the results to populations not
studied was unwarranted. Broader based, cross-sectional
sampling of delinquents and nondelinquents, and item analy
ses of the FIT, would have to be carried out before the
theory and the test could be accepted as workable.
REFERENCES
70
REFERENCES
71
Andry, R. G. Delinquency and parental pathology. Spring
field, 111.: C. C. Thomas, 1960.
Ball, J. C. Social deviancy and adolescent personality:
an analytic study with the MMPI. Louisville:
University of Kentucky Press, 1962.
Bandura, A., § Walters, R. Adolescent aggression. New
York: The Ronald Press, 1959.
Bloch, H. A., 6 Neiderhoffer, A. The gang: a study in
adolescent behavior. New York: Philosophical
Library, 1958.
Briar, S., 5 Piliavan, I. Delinquency, situational induce
ments, and commitment to conformity. Social
Problems, 1965, 13 (1), 34-45.
Cavan, Ruth S. Readings in juvenile delinquency. New
York: J. E. Lippencott Co., 1964.
Cloward, R. A. Illegitimate means, anomie, and deviant
behavior. American Sociological Review, 1959,
24 (2), 164-176.
Cloward, R. A., 5 Ohlin, L. W. New perspectives on juve
nile delinquency. New York: Columbia School of
Social Work, 1959.
Cohen, A. K. Delinquent boys, the culture of the gang.
Glencoe, 111.: Free Press, 1955.
Cohen, A. K., § Short, J. F. Research in delinquent
72
subcultures. Journal of Social Issues, 1958, 14
(3), 20-37.
Cowden, J. E. Predicting institutional adjustment and
recidivism in delinquent boys. Journal of Criminal
Law, Criminal and Police Science, 1966, 57 (1),
39-44.
Craig, Maude M., 5 Glick, Selma J. Ten years experience
with the Glueck social prediction table. Crime
and Delinquency, 1963, 9 (3), 249-261.
Edwards, A. L. Statistical methods for the behavioral
sciences. New York: Rinehart, 1954.
Glueck, S., 6 Glueck, Eleanor T. Unraveling juvenile
delinquency. Cambridge: Harvard University Press,
1950.
Glueck, S., § Glueck, Eleanor T. Family environment and
delinquency. London: Rutledge § Kagan Paul, 1962.
Guilford, J. P. Psychometric methods (2nd ed.). New York:
McGraw-Hill, 1954.
Hartelius, H. A study of male juvenile delinquents. Acta
Scandinavia, 1965, 182, 7-138.
Healy, W., § Bronner, Augusta F. New light on delinquency
and its treatment. New Haven, Conn.: Yale Univer
sity Press , 1936.
Hsu, F. L. Americans and Chinese: two ways of life. New
York: Henry Schuman, 1953.
Jacks, I. Psychological testing in institutional
73
rehabilitative programs - a gentle critique. Crime
and Delinquency, 1962, 8 (1), 34-39.
McCord, W., 6 McCord, Joan. The effects of parental role
model on criminality. In Ruth S. Cavan (ed.),
Readings in juvenile delinquency. New York: J. B.
Lippincott, 1964.
Medinnus, G. R. Delinquents' perception of their parents.
Journal of Consulting Psychology, 1965, 29 (6),
592-593.
Merton, R. K. Social theory and social structure.
Glencoe, 111.: Free Press, 1957.
Monohan, T. P. Family status and the delinquent child: a
reappraisal and some new findings. Social Forces,
1957, 35, 250-259.
Montagu, A. M. F. The biologist looks at crime. The
Annals of the American Academy of Political and
Social Science, 1941, 217, 46-57.
Nye, F. I. Family relationships and delinquent behavior.
New York: Wiley, 1958.
Perlman, I. R. Delinquency prevention: the size of the
problem. The Annals of the American Academy of
Political and Social Science, 1959, 322, 1-9.
Peters, C. C., 5 Van Voorhis, W. R. Statistical procedures
and their mathematical bases. New York: McGraw-
Hill, 1940.
Podolsky, E. The chemical brew of criminal behavior.
74
Journal of Criminal Law, Criminal and Police Sci
ence , 1955, 45, 675-679.
Reckless, W. C., Dinitz, S., § Kay, Barbara. The "good"
boy in a high delinquency area. Journal of Crimi
nal Law, Criminal and Police Science, 1959, 48,
18-25.
v
Reiss, A. J. The accuracy, efficiency, and validity of a
prediction instrument. American Journal of Soci
ology, 1951, 61, 552-561.
Robison, Sophia. Juvenile delinquency: its nature and
control. New York: Holt, Rinehart 5 Winston,
1961.
Scarpitti, F. Delinquent and nondelinquent perceptions of
self, values and opportunity. Social Problems,
1965, 13 (1), 60-65.
Schuessler, K. F., 8 Cressey, D. R. Personality and
characteristics of offenders. American Journal of
Sociology, 1950, 55, 476-485.
Selye, H. The stress of life. New York: McGraw-Hill,
1950.
Shaw, C. R., 8 McKay, H. D. Juvenile delinquency and urban
areas. Chicago: University of Chicago Press,
1942.
Sheldon, W. H., Hartl, E. M., § McDermott, E. Varieties of
delinquent youth. New York: Harper, 1949.
Siegel, S. Nonparametric statistics for the behavioral
75
sciences. New York: McGraw-Hill, 1956.
Silver, A. W., § Derr, J. A comparison of selected per
sonality variables between parents of delinquent
and nondelinquent adolescents. Journal of Clinical
Psychology, 1966, 22 (1), 49-50.
Smart, R., § Smart, Mollie. An introduction to family
relationships. Philadelphia: W. B. Saunders,
1953.
Sutherland, E. H. White collar crime. New York: Dryden,
1949.
Spiller, B. Delinquency and social class. Journal of
Criminal Law, Criminal and Police Science, 1965,
56 (4), 463-478.
Stott, D. H. Family situations conducive to behavior dis
turbance in delinquents. Social Work, 1965, 10,
14-17.
Tefferteller, Ruth S. Delinquency prevention through re
vitalizing parent-child relationships. The Annals
of the American Academy of Political and Social
Science, 1959, 322, 69-78.
Walker, Helen M., § Lev, J. Statistical inference. New
York: Holt, 1953.
Woods, Frances J. Cultural values of American ethnic
groups. New York: Harper, 1956.
APPENDICES
76
APPENDIX A
77
INSTRUCTIONS TO THE INTAKE WORKER
DELINQUENCY CLASSIFICATION
"Use your own casework judgment, based upon the two
descriptions below, to decide whether the given boy’s de
linquent behavior is primarily the result of his value
system (his commitment to delinquent values) or to disturbed
emotions. Though, both may be found to varying degrees
in all boys, decide which is the most dominant within the
given boy. Then, use one of the seven categories which
follow to identify him more specifically.”
MANIFESTATIONS OF A SOLID DELINQUENT COMMITMENT: This boy
knows what he does, before he does it; he is in conscious
control of his behavior. He knows that society frowns on
what he does, but he frowns on society; what he does is
right, as long as he can get away with it. He usually gets
along well with others, especially of his own ethnic group
and delinquent persuasion. He even gets along well with
adults, if they leave him alone; and he conveys this to
them by attitude, if not by word. He couldn't care less
about adults, though he is extremely wary of the authority
they wield.
He feels completely justified in dealing out retri
bution to anyone who ’finks," acts like a "punk," or who
78
aid the authorities. In short, he has his code, his be
liefs, his well defined behavior patterns, and his solid
in-group, and he is extremely loyal to all of them. He
even takes consequences stoically, because he sees himself
as losing a battle bravely fought against overwhelming
odds.
Very few boys fit this pattern totally, but the
degree to which this type of value system underlies a
given boy's behavior is the degree to which he is delin
quent.
MANIFESTATIONS OF EMOTIONAL DISTURBANCE: The forms and
disguises of underlying emotional disturbance are many and
complex, and sometimes it is a bit too easy to classify a
hard to handle boy with the label ''Emotionally Disturbed."
However, the following are some indicators which should
make one strongly suspect the existence of emotional dis
order within a boy:
Bedwetting, Facial Tics, Severe Nail Biting, Stam
mering, Stuttering, Compulsive Masturbation, Obviously
Effeminate Mannerisms, Frequent Nightmares, Painful Shy
ness, Habitual Withdrawal Into Solitude, Compulsive Steal
ing (even when he knows he will be caught), Habitual
Hyperactivity, Frequent Peer Conflicts Over Trifles, Inap
propriate Rage Reactions, Habitual Faulty Judgment--sincere
inability to gauge the consequences of acts, Chronic Worry,
79
Continual Nervousness, An Opinion of Himself as Worthless,
and Repeated Acts of Arson or Solo Burglary.
The picture is one of his inability to do away with
the things about him that make him worried, unhappy,
ashamed, or that get him into trouble. No matter how hard
he tries, they persist. As a matter of fact, the harder
he tries to improve, the more the above symptoms may worsen
until he is helped to eliminate their causes.
Quite often, he is very dependent and suggestible.
He may strive mightily for approval from adults even though
he may be in constant rebellion against them. Basically,
he accepts what he knows of society's rules and regula
tions, but he is unable to live up to the standards deman
ded of him. Generally, he is without any well formed
loyalties, codes, beliefs, or behavior patterns. In their
place, lie confusion and anxiety as the major determinants
of his behavior.
CLASSIFICATIONS: (Assign each boy to one of the categories
below)
1. Hard Core Delinquent: His commitment to the
delinquent values indicated above is very solid, and he
organizes most of his life around them. Usually, an orga
nized gang has completely displaced his family as the focal
point of his existence. More often than not, he will be
either: (a) Surly (but not openly defiant unless the
80
authority figure is week), sophisticated (in terms of past
experiences and present attempts to get what he wants) ,
and sly (in calculating the odds and taking advantage of
unguarded situations). Or, (b) Polite, but impassive, in
an attempt to mask his true feelings about the restraints
imposed upon him by authority. In this case, he attempts
to keep his sophistication and deviousness hidden. In both
cases, the only time true anxiety is shown is when attempts
to play the angles and "get by” are thwarted by a powerful
authority structure, and he is exposed for what he really
is--a dyed in the wool delinquent. As was said before,
these boys are relatively rare.
2. Average Delinquent: This is the type of boy
most people have in mind as a stereotype when the word
"delinquent” is used. Much of his overt behavior resembles
that of the Hard Core delinquent, but he is not as com
mitted to delinquent values. He is usually a gang member,
but not as solidly as the Hard Core boy. Frequently, he
"belongs” to more than one gang--his loyalties have not
solidified. Adults still retain some meaning for him aside
from the mantle of authority which they wear. He cannot be
classified Hard Core, but neither is he a Fringe Delin
quent .
3. Fringe Delinquent: This boy seldom belongs to
an organized gang, though he may be peripheral to, or iden
tify with, one without actually being a member. His
81
commitment to delinquent values is weak, but may still be
noticeable as a facade. He is chiefly striving for an in
dependent identity and is quick to go off half cocked or
to follow the lead of pseudo-masculine, sophisticated de
linquents in his attempts to achieve it. He is game for
anything, especially if it will enhance his status in the
eyes of his peers, or it looks like fun. Serious confront
ation by the authorities usually makes him wonder why he
was such a fool, though he typically gets into trouble the
same way all over again after a period of time. Adults
are quite important to him, and he may become quite attached
to them if they act consistently and fairly in aiding him
toward wholesome adult manhood. When this happens, he
becomes a positive leader who counteracts the delinquent
influence of the sophisticates whom he previously followed.
4. Situational Delinquent: This boy is never a
gang member, though he may have recently become a part of a
wild crowd. He has no commitment to delinquent values;
they are alien to his way of thinking. He is essentially
normal in every respect, though he may be very upset over
the bind he got himself into or about some feature of his
immediate life situation. His outstanding characteristics
are that he was displaying at least an average adjustment
to all phases of his life until approximately one year ago.
Then, SOMETHING CRUCIAL OCCURRED THAT DISRUPTED HIS LIFE.
For example: His father died suddenly; his mother suffered
82
a nervous breakdown; his parents decided rather suddenly
upon a divorce; etc. Whatever the shattering situation
was, IT IS THE BEFORE AND AFTER NATURE OF HIS TOTAL ADJUST
MENT that must be critically assessed. Behaviorally and
attitudinally, he is conspicuous because of his wholesome
ness in spite of the fact that he may be in turmoil because
of his present predicament. Typically, he gravitates to
the adults in a placement setting rather than to his peers
with whom he has little in common. He may even be galled
by a program geared to rehabilitation because he sincerely
feels he is not in need of it, and basically he is correct.
Needless to say, the Situational Delinquent is rare.
5. Disturbed, Strongly Delinquent: Though this
boy may, in many ways, resemble the Hard Core or Average
Delinquent, his emotional disturbance is severe enough to
be seen as the PRIMARY UNDERLYING BASIS FOR HIS DELIN
QUENCY. His delinquent approach to life is more of a shell
(in the nature of a defense mechanism) around his person
ality rather than being part of his core. The indicators
which distinguish him from the Hard Core and Average Delin
quents are: Chronic Worry and/or Tension and Nervousness;
periods of depression with or without instances of aggres
sive anger. Typically, he will display both. He is actu
ally dissatisfied with himself and his life, but he has
been able to make only a delinquent adjustment, which has
alleviated some of his tensions while creating others. His
83
behavior runs hand in hand with his mood: When he is feel
ing cheerful and optimistic about himself, he is dependable
and likable; when he is upset or feeling pessimistic, he
is disagreeable and untrustworthy. He is capable of loy
alty and close friendship--usually only along delinquent
lines until his emotional conflicts are alleviated.
6. Disturbed, Moderately Delinquent: This boy
shows more clearly the symptoms of emotional disturbance,
while moderate delinquent commitment is also displayed.
There is little doubt that the former is primary to his
delinquencies. Typically, there is upheaval, turmoil, or
disorganization within the family, and has been for some
time. It is not unusual for this boy to have been in a
prior placement. His behavior is unpredictable, from high
ly responsible to quite sly. He has seldom been more than
peripheral to gang activities. His peer relationships are
loose; he has a friend one day who is his enemy the next--
and conflictual. He displays ability in resolving these
conflicts, but resolutions are only temporary. He is char
acteristically dependent upon adults and is by turns mature
or childish and demanding in his relationships with them.
7. Disturbed, Mildly Delinquent: This very imma
ture, little (though he may be physically big) boy does
not know what delinquent values are. He is too confused
to have any consistent values. His family is chaotic, so
cially, economically, and psychologically. He frequently
84
has been in several prior placements. His delinquent be
havior is merely one part of his frantic, acting out, con
fusion. He can be expected to manifest more than one of
the indicators of emotional disturbance listed above. He
is typically the lowest on the peer pecking order, and is
usually the butt of their teasing and cruel name-calling.
They would be happy to get rid of him. He behaves errati
cally, and his social skills leave much to be desired.
He is extremely dependent upon adults and their approval,
but the very intensity of this need defeats him. He is so
childishly impulsive in his demands, complaints, or at
tempts at manipulation that he continually forces adults
to censure his behavior.
TREATABILITY
The concept of treatability is a difficult one.
The most efficient way to go about estimating it is to
limit your assessment to the boy himself and omit the
family's prognosis or the condition of his neighborhood.
Think of it in terms of, ''Other things being equal, what is
the prognosis for this boy within this program?" To do
this, divide the problem into two parts in order to arrive
at a final solution: (1) What type of adjustment can he
be expected to make to and within the program? For exam
ple: The boy who cannot or will not contain himself has
poor possibilities of completing the program; the boy
85
whose desire it is to conform to "what they want me to do”
may not have the resiliency to withstand the pressures of
the program unless he has better than average impulse con
trol; the boy who enthusiastically accepts the program from
the basis that he knows he needs some help can be expected
to make a good adjustment within the program. (2) What
is this boy's potential for benefit from this program? To
what degree is he capable of change, growth, improvement?
Of crucial importance here are: His evaluation of himself;
his view of his life; the severity of his problems; and his
motivation to improve his lot. The boy who is dissatisfied
with his life, uncomfortable with himself, motivated to
improve, and whose problems are relatively mild, is an ex
cellent candidate for rehabilitation. The boy who is
basically satisfied with himself and his existence and who
is unmotivated to put a stop to contacts with the law is a
poor candidate. Naturally, each boy shows the four factors
in varying combinations and seldom in as clear-cut a form
as in the above examples. Then, too, motivation is often
the key factor. A strongly motivated boy can overcome
other handicaps; an unmotivated boy will get nowhere even
if the other factors are optimal.
Use the combined estimate of his ability to adjust
within and benefit from the program to arrive at his over
all "Treatability.” Of the following four categories,
choose the one which seems to most clearly fit the boy's
86
over-all prognosis. Rate him as: Excellent, Good, Fair,
or Poor.
GANG MEMBERSHIP
If a boy is an established member of an organized
gang, indicate him as: Formal. If his activities have
been peripheral to an organized gang, or if he is a member
of a loosely organized group of boys, indicate him as:
Peripheral. Include under the peripheral category the boy
who "belongs" to several gangs but who is a loyal member
of none. For the boy who is none of these--he has partic
ipated in no gang activities--indicate: None.
SOCIOECONOMIC CLASS
Objective data about the family's socioeconomic
status are not usually available. Therefore, use all
available information from the case file and that gathered
from your interviews with a boy to classify his Socio
economic Class, according to the following broad catego
ries : Lower, Middle, and Upper.
NEIGHBORHOOD DELINQUENCY RATE
Using your own on-the-job experience, indicate the
boy's neighborhood rate of delinquency as High, Moderate,
or Low.
87
APPENDIX B
DIRECTIONS FOR THE FAMILY INFORMATION TEST
1) Before interviewing a boy, obtain a description of the
family constellation from his casefile. Indicate on the
answer sheet the absence, and its duration, of any immedi
ate family member from the home. For example: "Father
dead 10 years--stepfather in home 7 years"; for such a case,
in which the parent substitute has been in the home more
than one year, the questions about "father" should be ap
plied to the stepfather. Indicate the whereabouts of any
sibling who has been out of the home for more than a year.
Any facts of this nature that are not made clear in the
casefile, may be checked with the boy, AFTER ALL QUESTIONS
HAVE BEEN ASKED.
2) Interview each boy PRIVATELY, as part of an intake in
terview for face-sheet data. DO NOT PROMPT HIM TO CORRECT
HIS ANSWERS. Ask each question in a neutral, non-probing
manner. However, establish some degree of rapport with the
boy so as to insure his cooperation in trying to answer the
questions correctly.
3) Make every effort to determine whether each answer
given by a boy is a guess (an approximation) or a fact of
which he is certain.
4) Use the following recording system:
FACT * an answer given as a certainty.
D.K. * An "I don't know" response to a question.
GUESS ■ An uncertain response (I think maybe...I'm not
sure...About...Around...I guess).
Indicate the response to EACH PART of a multi-part
question.
Verbatim answers may, also, be recorded if you wish.
Their content is often significant from a casework
standpoint.
QUESTIONS
SIBLINGS: DO NOT APPLY TO SIBLINGS DECEASED OR LIVING OUT
OF THE HOME MORE THAN ONE YEAR.
1. What is the full name of each of your brothers and sis
ters? (THE MIDDLE NAME IS THE ITEM OF IMPORTANCE,
HERE.)
2. How old is each of your brothers and sisters?
88
3. What is the birthday of each of your brothers and sis
ters?
4. What does each one like to do in his spare time?
5. What important sicknesses or accidents has each one
had?
PARENTS:
6. In what city did your parents meet each other?
7. How long have they been together?
8. When is their wedding anniversary? (explain if neces
sary)
FATHER: APPLY THE QUESTIONS TO STEPFATHER, ETC., IF APPRO-
PRIATE.
9. What is your father's first and MIDDLE name?
10. How old is he?
11. In what city was he born?
12. What is his birthday?
*13. What kind of job does he do? (Get job description.)
*14. Who does he work for? (The name of the employer.)
*15. On what street is his job located?
*16. What does he do in his spare time?
*17. What important sicknesses or accidents has he had?
18. In what city did he go to school?
19. What grade did he finish in school?
MOTHER: USE STEPMOTHER, ETC., IF APPROPRIATE.
20. What is your mother's first, MIDDLE and MAIDEN name?
21. How old is she?
22. In what city was she born?
23. What is her birthday?
89
24. What kind of work does she do?
*25. Who does she work for? (The name of the employer.)
*26. On what street is her job located?
*27. What does she do in her spare time?
*28. What important sicknesses or accidents has she had?
*29. In what city did she go to school?
30. What grade did she finish in school?
31. If anyone in your home has died, what did he die of?
32. In what city (or town) were you born?
90
SCORING
Each "D.K." or "Guess" response was scored as a
failure, as was any multipart answer which gave definite
responses to less than half of the parts.
For those cases in which a stepparent or guardian
had replaced a natural parent more than one year prior to
the test, the questions pertinent to that parent were
applied to the parent substitute rather than to the natural
parent. In the case of a parent living outside the home
for more than one year, with no stepparent in the home,
each "D.K." or "Guess" response to certain questions (13
through 17 for father, 25 through 29 for mother) was scored
as one-fourth of a point wrong. When a parent had been
deceased more than one year, the above questions applying
to that parent were omitted, and the "D.K.” and "Guess"
responses to the remainder of the questions pertaining to
that parent were scored as one-fourth point wrong. The
final score for each test was simply the total number of
questions wrong, to the nearest whole number.
91
APPENDIX C
DATA SHEET
DATE____________
NAME_____________________________________________
AGE______________
ETHNIC I.D._____________________________________
GANG I.D._______________________________________
SOC. EC. GRP.__________________________________
DEL. CLASS._____________________________________ FIT_SCORE______
TREAT. PROG.____________________________________
I .Q._____________
FINAL DISPOSITION_________________________ (To what degree did
he succeed or fail in the camp program?______________________
WEEKS IN CAMP______________________________
INTAKE COMMENTS (PSYCHOLOGICAL DATA, DELINQUENT HISTORY,
PRIOR PLACEMENTS, AND FAMILY CONSTELLATION $ HISTORY
Anecdotal Camp Record (List major items such as: Mayor,
Boy Council Member, Removal-Return, Awol-Return, etc.)
FIT ANSWER SHEET
1 ____________________________ 17
2 18
3 ____________________________ 19
4 ____________________________ 20
5 ______________________________ 21
6 22
7 ____________________________ 2 3
8 ____________________________ 24
9 25
10
11
12
13
14
26
27
28
29
30
15 31
16 32
No. of Questions Missed
93
APPENDIX D
Comparison of Experimental and Controls
on All Variables
Variables
Delinquents
(N= 312 )
Nondelinquents
( N = 10 0)
I.Q. 86.0 89.88
Age 13 42=13.47% 14=14%
Age 14 106=33.97% 32=32%
Age 15 146=46.79% 49=49%
Age 16 18=5.77% 5 = 5%
Mex.-Amer. 88=28.21% 29=29%
Neg.-Amer. 116=37.18% 37=37%
Anglo-Cauc. 108=34.61% 34=34%
Lower Class 223=71.47% 80=80%
Middle Class 84=26.93% 20=20%
Upper Class 5=1.60% NONE
High Del. Area 193=61.86% 100=100%
Mod. Del. Area 99=31.73% NONE
Low Del. Area 20=6.41% NONE
APPENDIX E
94
Control Group Interview Format
"We are interviewing many young people from this
area in an attempt to get a picture of community life. We
would like your help. Would you be willing to answer a few
questions for us?" If the subject responded in the nega
tive, the interview was gracefully terminated at this
point. After receiving an affirmative answer, the follow
ing questions were asked in order to obtain background
data:
"What is your age?" "On what street do you live?"
"Who do you live with?” In cases of parents absent
from the household, then, "How long has your
(father, mother) been (away, deceased)?" "How long
has your (stepfather, stepmother, grandparent) been
with you?" "How many brothers and sisters do you
have?"
Any additional questions made necessary by the in
dividual boy's family constellation were asked at this
point. The FIT was then given without introduction.
APPENDIX F
Distribution of Test Scores of Groups I and II
(N = 312) Using Numbers of Items Missed
IA
>s
o
t-
o
X
6
3
Z
3 0
2 8
2 6
22
2 0
1 8
1 6
1 k
1 2
1 0
Frequency Obtained by
Arithmetic Smoothing
Original Frequency for
Each FIT Score
8
6
2
0
I 2 3 4 5 6 7 8 9 10 1 1 12 B 14 15 16 17 1 8 1 9 20 21 22 2 3 2k 25 26 27 28 2 9 30
C n
Fit Test Scores
Number of Boys
APPENDIX G
Distribution of FIT Scores of Groups I $ II (N = 312) Normalized by
a Monotonic Transformation, using 10(X+.5) as the new score. The
Method of Best Fit (Guilford, 1954) produced the curve.
3 0
2 8
2 6
Best Fitting Fre
quency for each
Transformed Score
24
22
2 0
1 8
Original Frequency
for each Transformed
Score
1 6
1 4
l 2
l o
8
6
4
2
0
16 1 7 1 9 15 18 1 0 13 7 9 11 1 2 5 6 8 3 1 2
Transformed FIT Scores
97
APPENDIX H
Transformed Score Values for
All Obtained FIT Scores
X X'p X xT
28 16.9 14 12.0
27 16.6 13 11.6
26 16.3 12 11.2
25 16.0 11 10. 7
24 15. 7 10 10.2
23 15.3
9 9.7
22 15.0
8 9.2
21 14.7
7 8.7
20 14. 3
6 8.1
19 14.0 5 7.4
18 13.6
4 6.7
17 13.2
3 5.9
16 12.8
2 5.0
15 12.4
1 3.9

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Behavioral Seriousness And Impulse-Control Balance In Delinquency

PDF

Prediction Of Therapeutic And Intellectual Potential In Mentally Retardedchildren

PDF

Improvement Rate In Tuberculosis As A Function Of Accessibility To Repressed Emotional Disturbance

PDF

Ego Diffusion In Women With Behavioral Disorders And The Integrating Effects Of Psychodrama In Identity Consolidation

PDF

Perceptual Defense And Somatization: A Comparison Of The Perceptual Thresholds Of Obese And Peptic Ulcer Patients

PDF

An Investigation Into The Communication Style Of Suicidal Individuals

PDF

Self-Worth, Future Goals, Mood And Time Perception

PDF

An Experimental Study Of Aggression Habit Patterns And Sex Role

PDF

A Semantic Differential Investigation Of Critical Factors Related To Achievement And Underachievement Of High School Students

PDF

Effects Of Different Treatment Procedures On Reading Ability And Anxiety Level In Children With Learning Difficulties

PDF

Personality Variables And Intellectual Abilities As Determinants Of Concept Learning

PDF

The Relationship Of Dependency To Verbal Learning Without Awareness

PDF

Personality Variables Associated With Narcotic Addiction As Measured By The Minnesota Multiphasic Personality Inventory

PDF

Acculturation And Value Change

PDF

Measuring Thought Process As An Ego Function In Schizophrenic, Mentally Retarded And Normal Adolescents By Means Of The Rorschach

PDF

The Concept Of Sexual Identity In Normals And Transvestites: Its Relationship To The Body-Image, Self-Concept And Parental Identification

PDF

Sex-Role Preferences Of Early Adolescents In Relation To Adjustment

PDF

First Steps In An Attempt To Construct An Objective Test Of Character Structure

PDF

Psychological Test Changes In Schizophrenic Patients Under Brief Stimuluselectroconvulsive Therapy

PDF

Psychological Variables In The Mother Related To Infant Feeding Patterns

Asset Metadata

Creator Venezia, Peter Salvatore (author)

Core Title Delinquency As A Function Of Intrafamily Relationships

Degree Doctor of Philosophy

Degree Program Psychology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,Psychology, clinical

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Metfessel, Milton (committee chair), Jacobs, Alfred (committee member), Lovell, Constance (committee member), Schrag, Clarence C. (committee member), Seward, Georgene H. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-149617

Unique identifier UC11360206

Identifier 6713764.pdf (filename),usctheses-c18-149617 (legacy record id)

Legacy Identifier 6713764.pdf

Dmrecord 149617

Document Type Dissertation

Rights Venezia, Peter Salvatore

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA