77 .2» :III}

1.7,, I7, II JIIIII II‘I‘IIIN'II III‘II

. . "5' 73777777 ”“6; ”qu1
I

.-7 ., ‘
-.. ~-‘ .
h -

u
r 45‘ f.

--

N‘af
.‘.~

\
~
£2

  

-.., -. ..., 3’

71

{I73

   
  
   

IIII'
3.7. I.. 3%: I

7-III III" 7:777}!

III": flIIIIi
I
HI_‘I11'I

., III,
, 77777777777777.32717777?’ ‘. IIW i‘

. “III...

II‘III 'IfiII'u II ‘7'”, ‘
7111" WI

III“;

é“: 7'7":f 7I‘l771'7’7
'l:' r7.. Ir “r177", HI: JIIV'T ‘ 7II }\II ”III 5
Hr"); {07' 1| I" .It ”7 I 7 7 .3 '17 .IIS‘C: "flrl‘ul" 771‘7‘
33.!" . 'III' ' 1"”;3 "WU!" h 'r. ... IA'II'IW "’I.‘;.'
. .- 9w H... 133..
73:}; I "7;'- I?‘ 77.9 43“. R ‘F);“A;Or'.1'i‘¥l' 7‘77. {II 7 4‘77'7
' - I ”I; . ..2 H I' 7.2:. I ‘1‘}! \‘J‘ira‘ Wu". L'I' 0' ;|"’:' \ "l' I7 I
_ I‘¢I I‘J ’gfigfi III“ (”77777171771777.6777 7
I‘ .7“

:2‘
‘
A \“ “V
. K
“‘3 '31"

In,
”II/£77 I7,,' '77'7'7 "III‘III rIIIwI/ 7J7“ ;_.'III7. 77777777

'II'III-H” """I‘5" ‘ ‘- I I
I .‘ 7-,I‘f.",I.‘: '.:'.,‘|- "7‘ I'/‘ ,‘7‘ '17 7"»':.'7. :7I7u7 w . ‘ 77' 7 .
I -. 7 _ .17. , .‘ ':.7 777:“. _ 7 .

“'7“
1"" 17"

-: "IHEIJIV “ “'1'! "If . . .. M '.
IL I '- TZ,“ 7.77.71. ’ I. In'I-‘I‘ Ii"? ' .II-j I I77 \I'.'I7'." " "I‘uv 71;?W“.~;I I‘
,. _. ., -- - .. I

7'}
I; 270', ‘III 7' WE. ‘77” 'I III I -'
7777777777. “777777.77 Q1 a {NIP I (I I I II. '0 .iff,”
71.77 77777 7 77771777777 Ukuu‘ (7'7! 7.777J3'.,777’.777 I73" {I '73: IIIIIIIMM

 

 

I-‘I‘77'

  
 
 
    

I77”;
Iuj’

“EJIIK‘ (to.

 

  
 
 
  
  
   

:g‘
w?”
“fi‘v—so _

A -7 -
-~
ter-n- ..
€E_ ._
‘. ‘
-<-s ‘ .

o.
'.

.
‘ . O ‘—
. ”5“,-

 
 

  
 
 

      

I I . KI. ’ I
I ‘7‘1‘ ” 7777.7 77 77 77 7 7 7
...II . {77%; g

I..', I777 ._ ‘7II77I;7'7., {7.7 .I77 7.7; ,7, .
I. ,7. “7,. ”:17!“ 471.777. I57I77II7I77, 7.77.777. 77777777 .7 77

  

  
 
  

'I'r 7710,77

II

7776 7.77,. :77

   
    
  
 
   

 

 

I ,7. ' .777I7,7II7 (7' IIII 7., . 77. 7 .7 .7 “7 7" f: , 7‘7. 7,7,7. 7‘ 7,77 “7777::
’Ilh‘ '_7‘7";.- 473‘, .7 ‘7'I7I,I7I'.II7"I.‘,".‘7-'.3,712.177II‘I'I' .7. 73' i7.'7I.°, 71;?“ ' I 7 ."f 3'.-'5,‘7‘.’I;IIII'III7,‘IIII7'I-I‘77'1‘;
7o '7'17' 7'7.“ ' . t ‘ ' 1"! I III‘I'IE. 717 "I I: h‘
‘3‘ka 7'7' 1'. I‘" If“ . ‘ .I‘QII.“.;, I j.I'7I:!I ’38:! II {NIH It” "'IIIII I I 1 “ :I:.'."J.:I.II’.":' WI)“ " ‘I‘J" II‘IEIE
u; ' ’I' II}; 7vI ” SI.“ WM IaIQI’IE’n I I'EIII“."'.I‘ Ifftg' .37, .'.u”.’..
777770737, 77.777777 7'777 77' 7 .7. [077‘ .777” 77, ,7'7777. _- 7477,7773 (“I7 . 7,- 777}, "I' 3‘77"“
.SIIw: . '3 ‘III'I I7II-77I'7I ~“' I;7Iit:’;"-.'I‘.:"~(' "r'f'
7 _7.7,7_.77 “77.7777 771:)”: 77 7777777777 (iv: Jr~77l’.{ rI' 7::7‘I’ .77II".777 I7. 7! I .\ .7?-
7.7377777777777777 47‘ II ..N‘I' ,‘IyHI'II.I-‘I'.JI“I"‘:CI"II‘I" [77.4717 7v7-' II
M \‘J‘

 

_ _ - I 7'? I . .
$777,777.77]. 7‘77: ‘77'II7'7‘77'7I'II’I‘7'77737-‘io:'I-.."III J7I7HIJ77I.
II“.-'...I7...I-I IIIN: ‘I "'7 H6577
77 [I 7
I'II.‘ “NH/I I "II In?” ..7' “3ng 'M Ir'7.~"3[5.glIIIIJI7:I .
I» .‘II “I. "‘II'I 7;, 'I ’é‘II’ I"\‘. 'I': I”! .3: I'
‘ ”"I II7I'I‘I I. “ ,1777717‘77 II‘: '9'I.7 ’1: I Iué7 . '
.7717. 7'.'.-.-':‘7r n7!“ .I'JEI777.‘

 
 

U777.

. IIII-«II. ... W ‘:
':-.’. ...,I . :I_. 77" I77.’7." ‘.r
.' .-;- ‘.‘,".."‘ .3 . l- 7I '7"I . 'I M 7,
‘I 'l h (I: ’I""7'J.r . ‘ I I' 'I‘7II “' ‘2. I I7 P..”.:', 1"I .II
I' I" H 7‘": ”‘75 .' """ .' . '>‘I .‘ “". w.“ ‘1 ‘ .‘J‘. .7. .
77,777,777. 7 J II7 .‘I7‘, - 74, 7:7777 «7.77”... .. ,7-II7 .‘7, I7. .77., 7777,77. »:7II7,77,.'7I77I7.7:77{I7-..I77:777 . 7777" 7
I .m, 'I: NIH-5‘ ‘773I'7'7A ‘9 '7‘.’7'7.'x‘.- :3. 35793:“ my. . : ’

,’;‘l', '.77'/ f
77.3.7 7 ‘0' 7973771777 77.771757“ I 7“,, IJ77UI7Iyg7lI‘I-

A'AI‘III WV
.".~‘I-II,“:' I7~'7. ..I In I‘ " H‘In‘w I ‘ I I
’77 7777 7; ~.(77.77(.77. 7773.71ti I] ’7‘ I7} imp-7&7? 7:71:17“ 77:7, 7

. I

I I7": ‘: 7"7I ‘
.7 I]
.~ 7..._7777I‘ " 77 7I7I77I’777177‘n7'777w “€41,777 .7 . 7. 77.7
".""~-'."I-7-I". II III I,7I1l .71.,77-77.
.’ ' ’7 .IIIIQll JIQ’M': aNT [7771' :77,I 3' (If 777,7’777'77; 777.7777 P 77)}: 7.7, 77.72.77

‘4' -_-‘.~.

x 71"“. '7':

  

T]
‘7l7'l I] ”"7” I 7"" .77 7,5727 I777 “‘I‘lfl‘ I7

II
71?»)..7 I37)“;

 
   

   

EFL
7H, 7.; “#74; G
07717 "~I‘..': III!
”I” 7777:7717" ‘flfi’g, '77:? ”Y
1“8077"

III.

”"1“," ”(...-773:5“;

. I: -III7-:us! 7.. '7IIFII‘I;7*7:E'

I I b?“ .7. -, .,
'I‘I’I" ‘.‘.'III:.' IIIII.""t.._. '«II'ITJJIII'I‘J‘ ...:IIIIII ‘ "' III’I'. {IMMIIJinIzsmIz .

       
  
  
  
   
  
     

;. u
u .‘ I
\I-]«;.

I

. ,.'

5.0!! ~,‘
."7
‘>‘ ‘ - I

‘3.
E V
‘ O

WWWWWWWW WWWWWWWWWWWWWWW

31293 01063 2184

I LIBRARY
’ Michigan State
University

 

 

 

 

This is to certify that the

dissertation entitled

FAMILY ENVIRDNMENT (DRREIATES OF AUDIESCENT
DRDIKINGANDWSDRIVII‘E: AS'I'UDYOFTHE
PRIZESSFS OF SEPARATIONAND INDIVIIIJATION

presented by
Cheryl Smith—Winberry

has been accepted towards fulfillment
of the requirements for

Ph 0 Do degree in Psycmlogy

 

%¢fl/ffi:{;/

Major professor/

flJohn Paul McKinney

Date 1/ 23/87

 

MSU is an Alfirman'w- Action/Equal Opportunity Institution O~ 12771

 

 

MSU

LIBRARIES

.

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped below.

 

 

LEA! 0 7 ‘9; I‘

 

 

FAMILY MW COW OF
AmLEmI‘ DRINKIm AND REEKLESS DRIVII‘G:
A STUDY OF THE PRCXIESSES OF SEPARATION AND INDIVIDUATION

By
Cheryl Smith-Winberry

A DISSERTATIG‘J

Submitted to
Michigan State University
in partial fulfil lment of the requirenents
for the degree of

IIXITOR OF PHILDSDPHY

Department of Psychology

1987

ABSTRACT
FAMILY ENVIRONMENT OORRELATES OF

ADOLESCENT DRINKING AND RECKLESS DRIVING:
A STUDY OF THE PROCESSES»OF SEPARATION’AND INDIVIDUATION

By
Cheryl Lyn Smith-Winberry

This study examined the influences of family climate on the
adolescent processes of separation and individuation. It was
hypothesized that adolescents' management of alcohol consumption and
driving practices, both rites of passage in American society, would be
reflective of the manner in which they were separating from their
families. School involvement and social environment were thought to
mediate these relationships. Hypotheses were examined utilizing a
questionnaire methodology with 244 junior and senior high school
licensed drivers. Several variables were examined: (1) family
climate-Family Environment Scale (Moos, 1974), (2) quantity/frequency
drinking index, (3) problem drinking index, (4) transition to drinking
index, (5) risky driving index--e.g. , accidents, moving violations, and
close calls, (6) school comnitment-e.g., GPA, truancy, academic expec-
tations, and (7) social envirorment influences—mg. , peer and parental
support for and modeling of exaggerated drinking and driving practices .

The results indicate that family environment variables are related
to adolescent drinking and driving practices , even after accounting for

peer influences. The stronger relationship between driving practices

and family environment might suggest that family influences are more
important.when initial learning takes place since drinking usually
occurs before driving. Family environment further served to
differentiate between normative and problematic drinking and driving.
Specifically, a constellation of factors termed “family disregard” were
present to a varying degree within the at-risk groups. High rates of
drinking and risky driving were associated with.a social environment
that approved of deviance and with school failure among adolescents.

It was speculated that these families failed to provide their
adolescents with appropriate values, monitoring, and a commitment to
the larger community. In contrast, adolescents who engaged in
normative drinking practices perceived their families to be moderately
cohesive, more individuated, and tolerant of conflict. Abstainers were
closely tied to their families which they perceived as less conflicted,
less individuated, and either very cohesive or very disengaged.

Perhaps moderate cohesion is a prerequisite for the process of
separation; however, more successful adolescents also receive continued
monitoring and guidance. Their families are less approving of
exaggerated drinking and driving, and they have access to other sources
of socially sanctioned adult status--e.g., school competency.

To my husband, Larry, my children, Heather and Gabriel, and my
parents. My family is the core through which all experiences and

accomplishments become meaningful.

iv

AW

I would like to extend my sincere appreciation to my committee
members-QJ.P. McKinney, Ph.D., R.A. Zucker, Ph.D., R. Frankmann, Ph.D.,
Department of Psychology, and D. Smith, Ph.D., Department of Highway
Traffic Safety. Their individual expertise and combined guidance
enriched the quality of the thesis. I am particularly grateful to two
members who have influenced my professional development. Jehn P.
McKinney's commitment to teaching was outstanding. He provided me with
the opportunity to be a part of a researCh team where I was able to
acquire skills from a competent group of individuals. Moreover, I
truly appreciate the many hours we spent discussing issues relevant to
professional and personal development. Robert A. Zucker has provided
excellent clinical and research guidance throughout my graduate
training; however, I am indebted to him for a more significant reason:
he provided me with a role model of competent professionalism. Bob
offered sound guidance in a manner that increased my awareness of
professional ethics, respected my individuality, and supported my
strengths. Additionally, I would like to thank my family and friends
for their support. My husband, Larry Winberry, provided timely support
and abundant tolerance as I progressed through this work. my mother,
Janet.Weathers', unwavering confidence in my abilities allowed me to
persevere during times of personal doubt. Finally, my friends were

always available for consultation and support. Special thanks to JUdy

Larsen, Ph.D., for skillfully providing editorial assistance and Suzy

Pavick for adding the final touches.

vi

TABLE OF CONTENTS

LIST OF THESOOOOOOOO......OOOOOOOOOOOO......OOOOOOOOOOOOO
LIST OF FIGLJRESOOOOCOOOOOOOOOOOOOO............OOOOOOOOIOOOO

WWIWC......-.....OOOOO00.0.00.........OOOOOOOOOOOOO

Theories of Adolescent Socio~emotional Development....
Family Climate Correlates of Adolescent

Separation and Individuation.....................
Adolescent Driving and Family Correlates..............
Adolescent Drinking and Family Correlates.............
Drinking and Reckless Driving: common or

Unique Problems?.................................
Application of NGw’Terminology........................
Significance of This Investigation....................
Hypotheses............................................

WWOOOOOOIOOOOCOOO00.0.0000.............OOOOOOOOOOOOCOO

SllbjectS.O.............OOOOOOOOOIOOOOOOOO00.00.0000...
Prmures.......OOOOOOOOOCOOCOOO......OOOOOOOOOOOOOOO

masureSooooooooocoo-o.000.000.0000....00.000.000.000.

ClaSSification Of S‘ijects.........OOOOOOOOOOOOOOIOOOO

RESULTS.O.............OOOOOOOOOOOOOCOOOO......OOOIOOOOOOOOO

Personal Subject Characteristics of Sample............
Family Environment, Drinking, and Driving
Scale Reliabilities..............................
Testing the Hypothesized Curvilinear Relationships....
Overall Relationship Between Family Environment
and Adolescent Drinking and Risky Driving........
Hypothesis #1 (Family Environment and
Risky'Driving)..............................
Hypothesis #4 (Family Environment and Drinking)..
Hypothesis #7 (Family Environment and the
Transition to Drinking......................
Relative Relationship of Peer and Family
Environment Influences...........................
Hypothesis #2 (Peer versus Family
Environment-Driving).......................

vii

Page

xi

10
37

50
52
53
S6
62
62
62
‘63
7O
73
73

76
80

84

91
93

95
99

99

Hypothesis #5 (Peer versus Family
Environment-Drinking)......................
Hypothesis #8 (Peer versus Family
Environment—-Transition to Drinking)........
School, Peer, and Parental Correlates.................
Hypothesis #3 (Correlates of Adolescent
Risky'Driving)..............................
Hypothesis #6 (Correlates of Adolescent
Drinking)...................................
Hypothesis #9 (Correlates of the Transition to
Drinking)...................................

DIWSSIW.o0000000000000000000000.0000...ooooooooooooooooo
sum.......OOOOOOOOOOOOOOOO0.00000.........OOOCOOOOOOCOOO
”MIX AOOOOOOOOOOOOOOOOOOOO......OOOOOOOOOOOOOOOOOOO...O

WSOIOOOOOO......OOOOOOOOOOOOOO......OOOOOOOOOOCO...

viii

104

108
109

111
111
113
115
135
140
150

IJST OF TABLES

Table Page
1a Descriptive Information on Sample Characteristics......... 74

1b Chi Square Analyses of the Driving Groups by Sex,
COhtrOlla for me.0.0.0.000.........OOOOOOOO0.0.0.0000... 75

1c Chi Square Analyses of the Drinking Groups by Sex,
C0ntr011fi for Age.O.......OOOOOOCOOOOOOO0.0.0.0...0...... 77

1d Percentages of Subjects Within Each Drinking
Category-2A.Comparison to the Zucker and Hartford
(1984) ijjngSOOOOOOOOOO......OOOOOOOOOO......OOOOOOOOOOO 78

2 Internal consistency Scores for Scaled variables on
Measures of Family Environment, Driving, Drinking and
mer supmrt.0.0.000.......OOOOOOOOOOOOOOOOOO0.00.00.00.00 79

3 Correlations of the Curvilinear Association Between
Cohesion, Expressiveness, Control & Organization
and Drinking & Driving After Partialling Out the
linear Associations....................................... 83

4 Intercorrelations of Family Environment Scales............ 86

5 Intercorrelations of the Drinking and Driving

masues.O.........O.................OOOOOOOOOOOOOOOOOO... 87

6 Multiple Regression Analysis of the Family.
mnomnt mlesaIflRiSKy DI-iVi-IEOOOOOOOOOOOOOOOOOO.... 88

7 canonical Analysis for the Family Environment Scales
am tm Drinking masureSOO0.0.............OOOOOOOOOOOOOOO 89

8 Multiple Regression Analysis of Family Environment
mtmTranSitiontODriMing ImeXOOOOOOOOOOOOOOOOOOOOOO 90

9 Correlations Between the Family Environment Scales
WRiSky DriVj-rg.0.0.0.0..........OOOOOOOOOO......OOOOOOO 92

10 Correlations of the Sex x Family Environment Inter~
actions With Driving After the Main Effects Have
w Partia11$ mtOO..........00............OOOOIOOOOOOOO 94

 

ll

 

14

15

l6

17

18

19

20

Table

11

12

13

14

15

16

17

18

19

20

Correlations Between the Family Environment Scales
am ”E Drirlking masureSOOO......OOOOOOOOOOOOOOOI....0...

Correlations of the Sex x Family Environment Inter-
actions With Drinking After the Main Effects Have
m Partialla mtOOOOOOOOOOOOOOOOOI......OOOOOOOOOOOO...

Correlations Between the Family Environment Scales
arfl tm TraDSj-tion to DrinkmgOOOOOOO......OOOOOOOOOOOOOOO

Correlations of the Sex x Family Environment Inter-
actions With Transition to Drinking After the Main
EffeCtS I-Iave m Partialla mtOOOOO......OOOOOOOOOOOOOOO

A Mixed Hierarchical-Stepwise Regression Analysis,
Examining the Correlations of Family Environment

and Driving After Peer Support and Modeling Have

Been Accounted For........................................

A Mixed Hierarchical-Stepwise Regression Analysis,
Examining the Correlations of Family Environment

and Drinking After Peer Support and Modeling Have

Been Accounted For: Entire Sample.........................

A Mixed Hierarchical-Stepwise Regression Analysis,
Examining the Correlations of Family Environment

and Drinking After Peer Support and Modeling Have

Been Accounted For: Cross validation #1...................

A Mixed Hierarchical-Stepwise Regression Analysis,
Examining the Correlations of Family Environment

and Drinking After Peer Support and Modeling Have

Been Accounted For: Cross validation #2...................

A Mixed Hierarchical-Stepwise Regression Analysis,
Examining the Correlations of Family Environment

and the Transition to Drinking Index After Peer

Support and Modeling Have Been Accounted For..............

Correlations Between School, Peer, and Parent Measures
and the Drinking and Driving Measures.....................

Page

96

97

98

100

102

105

106

107

110

112

 

 

10

LIST OF FIGURES

Figure Page

1 Circumplex model: Sixteen types of marital and
fmily SYStaNSooococo-0.0000000cooooooooooooooooooooooooooo 16

2 A heuristic model for changes in influencing
structures affecting behavior over developmental time...... 27

3 An organizational structure for classes of influences
umn drivm WVior-oon.on000.000.000.00oooooooooooooooooo 28

4 An organizational structure for classes of influences upon
drinking mmvj-oroo0.0.0.0.0.00.00........OOOOOOOOOOOOOOOOO 29

5 Defining criteria for adolescent drinking levels,
and percentages of respondents in each categoryb-
by sex and age and for total sample........................ 38

6 variables examined by Jessor & Jessor (1978) in
their study of adolescent transition to alcohol use........ 46

7 Family environment scale subscale descriptions............. 64

8 Internal consistencies, average item-subscale cor-
‘ relations and test-retest reliabilities for FES............ 65

9 Means and standard deviations of FES short form
SlulbscaleSOI0.0...00......0.00.0000.........OOOOOOOOOOOOOOOO 67

10 Illustration of a significant relationship between
Cd‘eSj-on am riSky driVj-rg.0.0.0.0.........OOOOOOOOOOOOOOOO 83

xi

WHEN

The hypothesis that a family's social climate affects the course
of child development and behavioral expression has influenced
theoreticians and researchers since the 18th century. Although the
importance of the family environment is rarely questioned, systematic
attempts to identify, measure, and relate the essential elements of the
environment to specific areas of child development or to particular
behavioral outcomes is a more recent endeavor .

Early attempts at classifying family environment focused on
unitary constructs, such as authoritarianism versus permissiveness in
parental management styles (Lewin, Lippitt & White, 1939; Elder, 1963;
Baumrind, 1967) . Further, many of the studies focused on an age group
in which family influences were thought to be more salient governors of
development--i.e. , infancy through early childhood. One reason for
such a beginning in the area of family research is the realization that
”the family” is a very complex system which changes continuously both
to meet the demands of its individual members, and to find its niche or
homeostasis within the larger social milieu—Le” the community. Any
attarpt at quantification of such an evolving entity is replete with
theoretical and methodological difficulties. Thus, in an effort to
begin to piece together the puzzle, carplexities are simplified, and

extrasystem influences are momentarily ignored .

2

More recently, a number of investigators have attempted to capture
the essence of the family social climate, examining multiple constructs
thought to be important governors of interpersonal behavior and
applying such models to the entire span of the family life cycle (Olson
& McCubbin, 1983). Thus, not only is the family examined at times of
relative closure-i.e., when extrafamilial influences are minimal--but
it is also examined during periods of rapid change and
reformation-i.e., during adolescence when extrafamilial influences
rise as the family prepares for the exiting of one of its members.
Further, various qualities of the family environment are being related
to individual functioning and behavior in many areas of develop-
ment--e.g., social, cognitive, and emotional (Dickinson, Hess, Miyake &
Azuma, 1979; wynne & Cole, 1983; Hess, Azuma, Kashiwag, Dickinson &
Nagano, 1984).

The purpose of this study is to examine the influences of the
family social climate on the normal adolescent processes of separation
and individuation. Adolescents' perceptions of ten constructs of their
family environment-cohesion, conflict, expressiveness, achievement
orientation, intellectual-cultural orientation, independence,
moral-religious orientation, active-recreation, organization, and
control-awhich fall into three general dimensions-~i.e., interpersonal
relationships among family members, directions of personal growth, and
organizational structures-dwere assessed in a nonclinical, high school
population of juniors and seniors (N:244). These perceptions were
related to two areas of behavior thought to be reflective of impending
adult status, behaviors commonly referred to as rites of passage in our

society: (1) adolescent alcohol consumption, and (2) adolescent

3

driving. Further, in an effort to understand the complexity of these
associations, nonfamily influences were assessed and incorporated in
the model. It is anticipated that such a research endeavor will
provide useful information on an important and frequently neglected
aspect of drinking, driving, and the combination of the two; namely,
family influences. However, this is not a model of pathology. It is a
model which deals with the spectrum of possible outcomes in two defined
areas as they relate to various styles of family functioning. The
emphasis here is on understanding the interdependence of the family
process and individual adolescent growth, not in explaining the deviant
expressions of either process.

The need for understanding family influences upon adolescent
functioning appears to be a point of debate within the scientific
community. Developmentally, there seems to be agreement that early
interactions affect many aspects of later functioning. During
adolescence, youngsters are seen as turning from their parents as a
frame of reference to adopt the standards of their peers (Sullivan,
1953). The assumption is that at this point, family influences assume
a subservient role in comparison to peer and other extrafamilial
influences.

In 1963, Brittain subjected this traditional view to empirical
evaluation. The study focused on the issue of the relative influence
of parents and peers on adolescents' choices. Briefly, a large number
of high school girls were asked to respond to a hypothetical situation
after hearing advice from both a peer and a parent. For the
experimental group, the referent was changed between the first and

second administration so that the opinion of a peer at time A was the

4

opinion of the parent at time B. The control group received either
parental or peer opinions during both administrations . The
investigators discovered that the control subjects were more consistent
in accepting the advice of the referent they were presented
with-either parent or peer. However, the experimental subjects were
more likely to vary their opinion depending on the situation. For less
significant problems they would rely on the advise of the peer group.
In contrast, for more intimate problems, they are more likely to turn
to their Regent . Since this study, a large body of literature has
accumulated on the relative influences of the family versus the peer
group during adolescence. An inspection of the empirical findings
leads one to the conclusion that gin systems are important.
Investigations that put forth a polarized view of the issue, suggesting
that the family has little influence or that the family is the only
impactful system, are overlooking the complexity of the adolescent's
social world. Unfortunately, family influences continue to be given a
lower priority in adolescent research, compared to biological , peer and
societal influences. The importance of pursuing such a line of
research becomes clear when one considers the significant impact the
family has on the adolescent's successful accomplishment of
age-appropriate social—emotional tasks (Glueck & Glueck, 1950; Haley,
1980) .

Prior to the presentation of the actual study, a review of the
existing family research dealing with the adolescent's expression of
the age-appropriate need for separation and individuation wil 1 be made .
The focus will be on the aforementioned ten dimensions of family

climate and how these dimensions relate to adolescent emotional and

5

behavioral functioning. As a background for this review, various
theories of adolescent socio—emoticnal development will be presented in
order to provide a foundation for understanding the tasks*with.which a
family with.an adolescent Child is confronted. Finally, the rationale
for choosing drinking and driving as the behavioral areas to examine
further the relationship between family climate and adolescent
development‘will be discussed. What is known about the phenomenology
of drinking and driving will be summarized separately, emphasizing the
family correlates.
Theories of Adolescent Socio—emotional Development

Adolescence has been described as the period in the life cycle
during which the establishment of a personal identity is of paramount
importance (Erikson, 1959). Identity formation has been explored
through research in a variety of areas: ego identity, identity
formation, identity achievement, identity status, and identity crisis.
At the basis of all of these concepts is the idea that the adolescent
strives to develop an inner organization of his/her needs, values, and
attitudes. The task is to come to terms with.one's own uniqueness as a
person. The more solid the development of personal identity, the more
likely the individual is to be able to resist outside temptations to
conform to social pressures. James E. Marcia (1980) describes identity
as an "existential position, to an inner organization of needs,
abilities, and self-perceptions as well as to a sociopolitical stance."
According to his formulation of identity, the adolescent is striving to
come to terms with his/her own individuality, while at the same time
learning to view himself/herself in relation to external social

structures. Failure to resolve these identity issues may lead to

6

"identity diffusion”, which describes a state of feeling empty or
insignificant. This situation has been linked to various conditions
characterized by a withdrawal from reality or by personality
disintegration (Erikson, 1968). In many respects, this theory is the
intrapsychic equivalent of Durkheim's theory of social anomiet

Erik Erikson (1959, 1963, 1968) has been the most influential
writer on identity development. He notes that identity development is
a psychosocial task distinctive of, but not exclusive to, adolescence.
In his theory of psychosocial development, Erikson describes the tasks
at various points of development in terms of crises-~transition points
in which decisions about one's life situation and personal values must
be made. During adolescence, the crisis is identity versus diffusion.
According to the theory, the search for identity does not begin in
adolescence. Rather, it.begins as self-object differentiation at
infancy and reaches its final stage with the self-mankind integration
at old age. What is unique about adolescent effbrts toward identity
integration is that "this is the first time cognitive skills, social
expectations, and physical development coincide to enable young persons
to sort through and synthesize their childhood identifications in
order to construct a viable pathway toward their adulthood" (Marcia,
1980). McKinney et al. (1982) note that it is not the passive
acceptance of ascribed roles, such as American, scholar, Jew or Black
which are earmarks of a successful identity resolution. Rather, the
individual must be active in choosing between these alternatives; "the
individual must be an agent as well as a patient." The resolution of
developmental crises is characterized by a public commitment to a set

of life expectations and personal values.

7
In making Erikson's formulation empirically more explicit, Marcia
(1980) defined four distinct identity types-erole diffusion,
foreclosure , moratorium, and achievement of a mature ego identity—which
he believes reflect possible outcomes of an adolescent's struggle with
identity issues. He postulates two dimensions which distinguish among
the four statuses: the degree of prior struggles with.an identity
crisis and the degree of commitment to a system of values. In role
diffusion, individuals experience no internal anxiety because they
neither challenge ascribed values nor commit themselves to the values
with which they have been presented. Fbreclosure is an adaptation in

which personal struggle is kept to a minimal level by an acceptance of
ascribed values which are presented to the child during development,
usually by a parent. Accordingly, sophisticated cognitive skills which
can assimilate present and past experiences and accommodate one's
belief structures are not employed since the process of an active
search for self is avoided. Moratorium describes individuals who are
struggling to achieve an independent identity; however, the struggle
itself, rather than the final resolution, becomes the focus. SuCh
individuals are unwilling to make a commitment to personal values.
Instead, they seem to rebel against all values which they feel are
being imposed upon them. Finally, a mature ego identity is achieved
when the moratorium crisis is successfully resolved. A personal
commitment to political, religious, vocational, and sexual goals and
values has been made.

Empirical efforts to quantify and study "identity" began with
studies of the validity of the construct. Subsequently, many

personality correlates have been investigated-i.e., anxiety,

8

self-esteem, authoritarianism, moral reasoning and cognition. What
seems clear is that there is abundant evidence that identity issues are
a governing aspect of adolescent development, that youngsters can be
distinguished in terms of their involvement in identity struggles, and
that resolution of identity struggles--e.g., a mature commitment to a
set of values-is a relative concept. That is, a homeostasis is
reached and the individual adopts a new sense of self which continues
to be modified throughout the adult years.

The aforementioned struggles in forming an independent identity
have frequently been interpreted as indicative of a period of
development replete with stress. G. Stanley Hall viewed the adolescent
stage of development in this manner; he adopted the term Sturm und
Drang-storm and stress. This view continues to have its proponents,
particularly among sociologists and psychoanalytically oriented
psychologists (McKinney, Fitzgerald, & Strommen, 1980). Some writers
have gone so far as to characterize all adolescents as possessing a
psychopathology (Goldstein, 1971). Put to the test of scientific
rigor, this view of adolescence has not been supported. In a
well-known longitudinal study of middle-class adolescent males from
ages 14 to 22, Offer and Offer (1975) conclude that over the 8 years of
development, turmoil is not the universal state of adolescence, nor is
it a necessary component of healthy ego development. The study
identified three groups of adolescents: continuous growth group,
surgent growth group, and tumultuous growth group. TWithin each group,
measures of social and emotional functioning were highly correlated.
Thus, there is evidence for a range of ego development among

adolescents that appears to be consistent within groups of adolescents.

9

A natural extension of the pathologyboriented view of adolescence
is the notion that adolescent stress and rebellion will naturally spill
over into the family. As the adolescent struggles to establish.his own
identity, he must rebel against parental values and authority in order
to complete the process. Bandura and.Walters (1959) examined these
hypotheses in a normative sample of adolescents and their parents.

They found that the boys had internalized the parents' values and
standards of behavior to a large degree; consequently, restrictions and
external controls had been lightened as the parents came to believe
that their children could manage their own behavior. Further, these
authors found that emancipation was not a stressful issue, but an issue
that had been in process since preadolescence. Thus, there is no more
evidence to support the view of family dystncti n during adolescence
than there is to support the view of adolescent dysfunction. It seems
logical to assume that families show as much diversity in their ability
to cope with the added demands of adolescence as do the adolescents
themselves.

In summary, the task of forming an independent identity is posited
to be a central component of adolescence. The family counterparts of
this task are the processes of separation from close family ties and
individuation. Inherent in these changes is some normal amount of
stress and conflict. For the majority of families,the adolescents'
needs are accommodated, with the family serving as a reliable
foundation that provides the support needed to make an exit. For other
families, meeting these normal developmental demands is more difficult.
In part, this may be due to a greater degree of conflict within the

adolescent. On the other hand, it could be due to a lack of coping

10

resources within another subsystem of the familyb-i.e., parents,
marital dyad, family, etc. In an effort to sort through these
possiblilties, it.would be helpful to take a closer look at specific
family climate variables as they relate to the range of psychological
and behavioral manifestations of the adolescent's attempt to separate
and individuate.

Family Climate Correlates of Adolescent Separation and Individuaticn

Because the focus of this study is on normal variations in the
social climate of families with adolescents, it seems useful to review
the data on dimensions of family functioning which are hypothesized to
be strengths. To this end, two sets of beliefs will be examined: (1)
clinical interpretation of family strengths, and (2) individual family
members' views of positive family attributes.

Fisher and Sprenkle (1978) believed that a disproportionate amount
of attention had been given to understanding family pathology;
consequently, they developed a study to provide clinical practitioners
‘with information on the theoretical principles other colleagues used in
strengthening family relationships. As a starting point, they reviewed
the literature in three areas of family functioning: cohesion,
adaptability, and commmication. Next, they developed a list of traits
which are used to operationalize these theoretical constructs. An
initial sample of 600 clinicians were asked to rank the seven most
important features of healthy family functioning. Fifty-two percent of
the clinicians responded to the survey. The highest ranked traits, in
descending order, were attentive listening, value sender, flexibility,
differentiation, speak for self, psychological safety, expression of

feelings, support, negotiation, send congruent messages, and attention

11

to affect. The lowest ranked items were: express thoughts, indicate
message is received, spontaneous speech, loyalty, feedback, report
complete thoughts, paraphrase, and physical caretaking. The results of
this survey indicate that clinicians view healthy families as able to
create an environment where family members feel safe, supported, and
valued. Of lesser importance are traits which focus on specialized
interactional or communication techniques .

Another study of family therapists found that cohesion,
adaptability and communication dimensions offered important goals for
treatment. The Group for the Advancement of Psychiatry (1970) asked
family therapists to indicate their primary goals for therapy from a
list of eight goals. Of the 290 respondents, 85% chose improved
communication as primary; 56% chose improved empathy, 56% chose
autonomy and individuation; 34% chose more flexible leadership; 23%
chose reduced conflict. Individual symptomatic improvement was chosen
by 23% and improved individual task performance by 12%. ‘Ihus, the
emphasis is again on creating a supportive climate primarily through
the strengthening of communication skills. Of lesser importance are
system maintenance and personal growth dimensions .

The second primary source of information on healthy family
functioning consists of the opinions of individual family members.
Otto (1963) solicited volunteers from the comrunity to attend ten
one-and-a-half hour family discussion sessions. 'menty-seven couples
from all stages of the life cycle participated in the project. The
sessions were taped and transcribed. The fol lowing list contains the
ten strengths which received the most attention and agreement from the

group:

12

(1) The ability to provide for the physical, emotional, and spiritual
needs of a family.

(2) The ability to be sensitive to the needs of the family members.

(3) The ability to communicate effectively.

(4) The ability to provide support, security and encouragement.

(5) The ability to initiate and maintain growth-producing
relationships and experiences within and without the family.

(6) The capacity to maintain and create constructive and responsible
comunity relationships in the neighborhood, the school, tom and
local government.

(7) The ability to grow with and through children.

(8) An ability for self-help, and the ability to accept help when
appropriate.

(9) An ability to perform family roles flexibly.

(10) Mutual respect for the individuality of family members.

In contrast to clinicians, it would appear that families put more
of an emphasis on interpersonal climate variables. 01 the other hand,
while viewing this dimension as primary, clinicians are more likely to
place greater value on system maintenance aspects than families do.
Further support for this observation cores from a study conducted by
Fisher, Gibbin, and Hoopes (1982). They correlated nonclinical family
members views' about the nature of a healthy family with family
therapists' perceptions. Families ranked items associated with the
dimensions of cohesion higher than therapists. m the other hand,
therapists ranked flexibility and shared leadership-adaptability

items--higher than families did.

13

The findings from such researdh efforts provide valuable data on
"healthy family functioning." The need for establishing criteria for
normalcy and strength becomes clear when one attempts to define and
treat pathology. Hewever, there is a growing concern that broad
definitions of family strengths are inadequate to capture the finer
characteristics that distinguish.families at.different points in the
family life cycle (Duvall, 1962; Rapoport, 1963; Hill, 1964; Olson &
McCubbin, 1983). Thus, attempts are underway to divide the family life
cycle into stages so that phase specific dynamics can be investigated.
"Adolescence" is posited to be one stage in the family life cycle that
requires its own unique skills and strengths in order to meet
adequately the concomitant needs of the developing child and maturing
parents.

The literature on adolescent family development cores from two
primary sources: (1) studies on representative, nonclinical families
‘with an adolescent child who is usually the oldest child, and (2)
studies on clinical populations that attempt to distinguish between the
families of healthy and disturbed adolescents. Of particular interest
is research that isolates family environment variables during different
phases of development, allowing for direct comparisons between phases.
One research team succeeded in such a feat, giving special attention to
the adolescent phase of development. The work of Olson and MoCubbin
(1983), presented in Families, What Makes Them Wbrk?, will be reviewed.
Additionally, selected research focusing on family correlates of
adolescent functioning will be reviewed. The dimensions of
interpersonal relationships, individual growth, and system maintenance

will be emphasized.

l4

Methodological considerations. Before beginning a discussion of
the empirical findings, a brief mention of several methodological
considerations regarding family researdh is needed. First, Jacob
(1975) cogently argues for the need to present separate findings for
each sex. Adolescent males and females exhibit different types of
deviance, suggesting that different family factors may also be
important. Secondly, the adolescent/family literature operationalizes
the concepts of healthy versus unhealthy individual adolescent and
family functioning in a wide variety of ways. This variation makes the
process of direct comparison of studies difficult at best. A related

prdblem concerns the measures which are employed in the studies.
Frequently, the investigators choose to measure family concepts by
developing a new scale or questionnaire. Although such procedures are
not uncommon in the initial stages of researdh, they add to the
complexity of between-study cotparisons. A fourth weakness lies in the
simplistic definitions, albeit necessarily so, of family environment.
As more information accumulates, complex theories of family process can
be devised and tested through the application of sophisticated
multivariate analytic techniques. Finally, sampling biases,
particularly in researCh employing clinical populations, limits the
overall generalizability of the findings.

Based upon the current methodological inadequacies, there exists a
need for research which (1) takes a family systems perspective, (2)
utilizes established—~reliable and valid-instruments, (3) examines a
representative sample of adolescents and their families, (4) employs
several alternative definitions of healthy versus unhealthy

functioning, and (5) considers sex differences when analyzing the data.

15

Empirical findings with normal adolescent families. Olson and
{McCubbin (1983) sought to capture the complexity of marriage and family
life across the life cycle. They developed a theoretical model, the
circumplex model, which classifies families along two
dimensions—~adaptability and cohesion. combining their relative
position on each dimension, families are described as falling into one
of 16 types. Figure 1 illustrates Olson et al.'s model. Balanced
families, represented diagrammatically by the four center areas, are
thought to possess the most resources for dealing with familial stress
and developmental crises. Extreme family types, represented by the
outside corners of the diagram, are characterized by a limited ability
to meet familial and extrafamilial demands. Finally, midrange
families, represented by the middle circle of the diagram, possess
several strengths and resources, yet also manifest certain deficits
which may present difficulties when the families are faced with
specific types of stressors. The family life cycle was divided into
seven stages of development--prechild, young child, school-aged child,
adolescent, launching, empty nest, retirementv-based upon an expansion
of Hill's work (1949).

Olson and McCubbin utilized a survey methodology. A total of 2692
individuals from 31 states responded to the survey. The sample
consisted of 1140 couples, plus 412 adolescents with.equal numbers of
males and females. Each stage of the life cycle was represented by at
least 100 families. The sample included mainly white, working-class
families, with only 15% of parents lacking a high school education.
Indices of family types, family resources, family stress and change,

family coping strategies and marital/family satisfaction were examined.

 

 

<——— Low Hugh —-9

UtSlWCIGIO 5!”le CONNlCHD ileSNlO
CXIAINKABIP» CllhlfifiABIP'JIE
‘ v— V V

. -‘7 ./ ’.//'/;/4///./.’N;7
A

conessou

 

tumour cuuucau r _
unsucactq -'

'LHAOHCALU

/;'7W/;W}W7/

  
  
  
  

CHAONLM l v
tthSnto

   
   

tvf/,

 

 

   
 

Hugh

I

A .o '

D ,. ,

A mum : mmu umau Mum Human

P t‘ OISENGAGIO SIPARMIO CONNECIEO ' thMlSNlo

T W/ / / / _ / I

A W. 7/ I ’/ / ~ //

‘? 77/7f/r 7/////
I

| SlRuCluRlD ,suwciunmv smucwmuu smucwmuu 'smucwmuv

T . DISINGAGIO SIPARAMD CONNECIEO . lNMlSNlO

Y 1

I

,—
c
i

RIGID”
DISINGAGID Si PARA“ CONNKCIEO (WHISHIO

V ;‘<lL/./ [WA/Jr V

QIADRANT III QUADRANT IV
I luuucro ’ 7] mo “not W lumuu

 

 

// /
WWW WWW/fl

 

 

 

 

 

 

 

 

 

 

Figgge l. Circumplex Madel: Sixteen Types of Marital and Family
Systems (Olson & McCubbin, 1983). '

16

17

The following review will focus on the findings of families with
adolescents: (1) perceptions of family environment-cohesicn and
flexibility, (2) stressors specific to adolescence and (3) strengths
and resources.

Related to the family environment, the authors found that levels
of family cohesion were highest among families in the early stages of
the family life cycle. Family cohesion was lower among families with
adolescent children and reached the lowest level with.families in the
launching stage. cohesion rose again during the empty nest phases.
With regard to family adaptability, adolescents viewed their families
as significantly less flexible than did their parents. In general,
parents tended to view their families more similarly than did either
parent and adolescent. Adolescents' perceptions of lower flexibility
and cohesion are explained in terms of the adolescent's need to seek
greater intrapsychic and physical autonomyh-separation and indi-
viduation--from their families. Since they are the ones initiating the
family change, it is reasonable to assume that they will view their
families as less flexible than their parents do. Thus, adolescents
tend to see their families as more extreme and less balanced. (It
should be noted that the use of the word extreme in this context is
relative; these are normal families. Thus, as compared to clinic
families, these families probably would be described as
midrangeebalanced.)

Stress and strain.were evident at all phases of the family life
cycle; however, the type of stress varied as a function of the
developmental level of the family. Overall, families reported more

stress during adolescence than at any other time. Specific stressors

18

include intrafamily strains (30%)--i.e., conflict between children and

parents, more chores that do not get done, etc .-financial strains

 

(60%) and work-family strains (10%) .

 

It was hypothesized that high functioning families at the
adolescent stage would fall into the balanced types. For the most
part, this hypothesis was confirmed. 45% of adolescent families fell
into quadrant III—lower adaptability and cohesion—and about 28% were
in quadrant II—high adaptability and low cohesion. Balanced types
were clearly more functional . They were characterized by the highest
levels of marital and family strengths, the lowest levels of stress and
the highest levels of marital and family satisfaction. Among their
strengths include higher reported levels of family pride and accord.
Further, adolescents in balanced families believed they had good
communication with their parents.

Balanced families also tend to emphasize extrafamilial contacts.
They tended to be more religious, yet spend less hours per week in
active church functions. They reported enjoying friends, relatives and
personal pursuits.

In summary, the findings support the notion that adolescence is a
time of great stress and strain in the family; however, many families
possess strengths and skills which allow them to manage the new
changes. In particular, balanced types-high or low on adaptability
and low on cohesion—appear to function most optimally. Family ties
are allowed to lessen so that individual members can attend to personal
growth activities. In these families, communication is good.

Unfortunately, Olson and McCubbin do not relate their findings to

outcore variables--i.e., school performance, drinking, antisocial

19

behavior, etc. Nor do they discuss differences between the families of
male and female adolescents. However, their findings provide a very
rich understanding of the family environment with adolescent children.

Many other studies address the issue of family correlates of
adolescent functioning, albeit in a less sophisticated manner. Westley
and Epstein (1969) studied 1970 college students and 88 families in
Canada. They focused on problem—solving skills, power, authority, and
roles. They found that families which encouraged autonomy, or a
diversity in roles, had significantly more emotionally healthy
adolescents. Parental marital structure was also related to child
emotional status. Children from mother-dominated and father-dominated
homes had more emotional problems and fewer strengths, whereas children
from father-led and democratic families had few problem and more
strengths.

In a similar study, Balswick and Macrides (1975) asked 417 college
students to describe their adolescence—rebel lion, marital happiness,
control and division of authority. They found that rebellion occurred
in patriarchal-unhappy, patriarchal-very restrictive, and
patriarchal-very permissive family types. Thus, they concluded that a
family's choice of a restrictive or permissive style was less important
than the extent to which it exercised these styles. They also noted
that many respondents reported only mild rebel lion—65% male and 56%
female—dispelling the notion of adolescence as a period of storm and
stress.

Schvanveldt (1973) conducted a study with Mormon families
containing an adolescent member age 12 to 15. He found that the most

common areas of conflict were (1) performing chores, (2) use of time,

20

and (3) expenditure of money. The youngsters expressed a desire to
have more communication with their parents, yet fewer rules and
regulations. The male adolescents did not differ from the females.
Landis (1954) added further support for the notion that adolescents
want a more active role in decision-making. Authoritarian.homes were
associated with.more adolescent-parent conflict than‘were democratic
households. In authoritarian homes, adolescents reported interference
with social life, a lack of voice in decision-making and a lack of
authority as problem areas.

0n the other hand, a certain degree of family control and
regulation is needed to govern adolescent behavior. Jessor and Jessor
(1974) studied the association between maternal ideology (in the areas
of religion, employment and tolerance for deviancy), control, and
affectional ties with adolescent involvement in problem behavior. They
discovered that an inverse relationship between maternal
conventionality and the degree of problem behavior. Further, the
presence of clear rules and regulations were also associated with lower
levels of problem behavior. Affecticnal ties proved to moderate this
relationship for females only.

Empirical findings with clinical samples of adolescent families.
Jacob (1975) reviewed research utilizing the behavioral observation
methodology with disturbed (schizophrenic and non-schizophrenic,
disturbed) families. The purpose of the review“was to test the
hypothesis that variables in the larger social environment of the
family are predictive of individual pathology in adolescent children.
In 1958, Behrens and Goldfarb published the first observational study.

Since that time, Jacob found 57 articles. He divided his review into a

21

discussion of the following family environment variables: conflict,

dominance, affect and communication clarity. Separate results were

presented for schizophrenic (SF) and non-schizophrenic, disturbed (DF)

families. A summary of his findings are as follows:

(1) Conflict

a.

SF: Normal families (NF) interrupted more than SF

in the direction of parent to child. Three studies
indicated greater disagreement in SF, one in NF.

Much to his surprise, Becker (1969) also found that
clinic families exhibit l_e_s_§_ conflict. After
controlling for frequency of speaking, the relationship
was insignificant. Therefore, he urges that this
variable be taken into consideration with direct

observation studies .

b. DF: 23 comparisons yielded very mixed results; ninewere

insignificant and the restwere discrepant.
(2) Dominance

a. SF: 17 comparisons;one study portrays the SF father as
more dominant; 4 others portray the NF father, SF mother
and NF child as dominant. The rest of the results are
mixed.

b. DF: 44 corparisons yieldedtwo trends: (1) NF fathers

have more influence over their children than DF fathers,

and (2) NF are more differentiated.

22

(3) Eat
a. SF: (Mt of 32 comparisons, half were insignificant. The
remaining colparisons were mixed , depending on whether
there is a well child in the family.
b. DF: Out of 33 comparisons, 16 were insignificant. The
remaining studieswere split between viewing problematic
families as having more or less negative affect.

(4) Commmication

 

a. SF: Generally SF produced less clear communication
b. DF: The majority of 44 comparisons vere insignificant.
Where differences emerged, NF fathers and DF mothers
engaged in more simultaneous speech.
Jacob concludes that there is not enough consistent evidence to accept
the popular notion that a disturbed adolescent is merely a reflection
of a disturbed family system. However, he does note that there are
Eny methodological difficulties which make it cumbersote to corpare
studies. SOphisticated analytic procedures, such as meta-analyses, can
be employed to interpret large numbers of findings in a more systematic
manner than mere substantive comparison of findings.

Several studies compared family climate variables in a
nonclinical, yet problematic population of adolescents. Russell (1980)
studied the families of 14 to 17 year old females on the dimensions of
cohesion, adaptability, creativity and support. He used both interview
measures and direct family observations of a game task. During the
game, a conflict situation was introduced. The conflict could only be
resolved if the family allowed the adolescent to develop a solution

(measure of family organization). The findings indicate that girls who

23

have a history of runaway behavior came from families that they
perceive as extreme on the dimesions of cohesion-very high or very
low. High functioning families were more flexible, allowing the
adolescent to assume an active role in decision-making. Further, they
were more supportive and creative. Similarly,‘vanderveen (1976) dis-
covered that disturbed families exhibited fewer coping strategies, less
family integration, and lower family unity than their nondisturbed
counterparts.

Bell and Bell (1982) conducted an ego functioning evaluation of
283 high school girls, ages 15 to 17. They chose the higher and lower
15% to be involved in a subsequent family evaluation. The Family
Environment Scale (FES) was used to assess family climate. The results
indicate that the high functioning adolescents had families who were
more flexible and trusting. They were also less likely to be
triangulated in parental conflicts. like the literature on nonclinical
families, the clinical literature suggests that dysfunctional families
exhibit either excessive or deficient cohesion and organizational
control, and do not allow the adolescent to assume an independent
identity.

An important point to keep in mind is that questionnaire and
interview studies may present different pictures of the family
environment depending on which family member is doing the reporting.

As was previously noted, Olson and McCubbin (1983) found that
adolescents did not perceive their families in the same way that their
parents did. They tended to see less autonomy, more conflict, and less
cohesion. In another study, Nbvak and vanderveen (1970) found that the

normal siblings of disturbed adolescents perceived fewer family

24

problems than the identified patients. In fact, the normal siblings
did not differ in their perceptions of family difficulties from
children from nondisturbed families. However, these siblings did
describe their families as being more task-oriented and less
pleasure-oriented than children from nondisturbed families. Again,
disturbed children described their families as dependent, conflicted,
and less cohesive. Thus, it seems that although the overall picture
may vary between family members, families with dysfunctional
adolescents are described in fewer positive ways.

In comparing the findings from both sets of datae-clinical versus
nonclinical populations-there seems to be evidence for thehypothesis
that families that are not able to accommodate to the autonomous
strivings of their adolescent members will have a negative impact on
the child's overall functioning-reflected in lower self-esteem, more
stress, poorer school performance, a more negative view Of the family
as a source of support, or greater misbehavior. Because studies
reviewed herein are correlational, it is not clear if the adolescent
i overtaxes available family resources by presenting the family with
excessive needs, or if the family is simply unable to meet the gggmel
level of adolescent needs. What is clear is that there is a wide
variation in family envirorments, and that these enviroments can be
related to both successful and unsuccessful mastery of adolescent
age-appropriate tasks. Families who exhibit a moderate degree of
cohesion, organization and control, have lower levels of conflict, and
allow more room for personal growth appear to be associated‘with
adolescents who possess greater ego resources. Further, in our

discussion of adolescent identity formation, it.was argued that higher

25

levels of ego functioning indicate a more successful resolution of
identity crises (Erickson, 1959). It also follows that adolescents who
possess more ego resources will be better equipped to negotiate issues
related to their values and behaviors in areas associated with their

impending adult status.
Statement of the problem. In every society, there are specified

 

activities which are associated with adult status. Margaret Mead
(1949) notes that other cultures provide much clearer guidelines for
when and how an adolescent makes the transition into adulthood.
American culture, on the other hand, tends to have fewer events which
mark this transition-rites of passage-and extends the childhood/
dependent years for a longer period of time. Hewever, there are
certain behaviors, for which there appear to be a social consensus,
that are representative of adult status. These include drinking,
driving, dating, becoming sexually active, cigarette smoking, some
forms of academic achievement, and obtaining a job. The first two
activities, drinking and driving, also carry legal specifications of
who may engage in them. Thus, there is a very clear distinction
between "children" and "adults," even to the extent that under-age
individuals are punished for participation in the activity.
Adolescent drinking and driving also have serious negative
consequences. There is growing concern.about the participation of
adolescents in deviant drinking activities (Biddle, Bank, & Marlin,
1980; Braucht, 1982; Rachel, Maisto, Guess, & Hubbard, 1982). Further,
automobile accidents account for more adolescent fatalities than any
other health problem. When drinking and driving are combined, the

result is frequently a hazardous one: alcohol is thought to be

26

involved in 50% of all highway accidents (Perrine, 1974; Lowman, 1983)
and 60% of fatal accidents among teenagers (Douglass, 1982). Growing
public awareness and concern about these social problems is reflected
in the amount of governmental funds which have recently been
appropriated to investigate and intervene in these areas. Thus, an
understanding of factors influencing these behaviors extend far beyond
academic interest.

The present study examines how one set of variables, family
climate, relates to adolescent drinking and driving behaviors. These
behaviors have been chosen because (1) they can be viewed as behavioral
manifestations of an adolescent's struggle with separation and
individuation issues because of their association with adult status,
(2) they are easier to quantify and measure than other concepts which
have been examined in the literature related to adolescent identity,
and (3) they are significant social issues in their own right.

The investigation of family correlates of adolescent drinking
and/or driving has been suggested by a number of researchers
(Goldstein, 1972; Sobel & Uhderhill, 1976; Zucker, 1984). To date,
there is a paucity of empirical data in these areas, especially related
to driving. At best, studies tend to examine behavior-specific family
correlates, such as parental modeling (Carlson & Klein, 1970; Jessor &
Jessor 1977). Zucker (1979) developed a heuristic model for explaining
changes in influencing structures affecting behavior over developmental
time-see Figure 2. He then applied this model to the acquisition of
drinking—-see Figure 3-and driving behaviors--see Figure 4 (Zucker,
1984). The model calls for the examination of (1) sociocultural and

community influences, (2) family of origin influences, (3) peer and

Soo:....5o< >._m<m:

mmocoooomooo
ago uoe>mnom

.3 30:

m

m>

 

80503.59:

mHUZEfiEmEXV
az< moH><zmm

0) may

 

 

.32; .HoxSN Houmfi 059 3:95398
.850 Hoar—Eon 0:300. HE moHHDQEm Egg 5 8986 wow Hung oflmeuoom e .m 8503

oooxoexo
:53 a :25W

 

   

   

3052.5 .2:

we rll-.l..i_

It...

_ as; .5322 35 _

mmucoa=c_
. fiat—logos:—

_
_

 

 

383:5 3.55600

\ can
35:38.08

 

 

 

 

mooco3:c_
zeta

 

27

 

-Socioeconomic status
of family

~Ethnicity. Religion
influences & Behavior
altematlves

-Neighborhood Value
influences & Behavior
alternatives

I. Soclocultural and
Community Influences

------------ q
-Availabi1ity of vehicles
& values about driving

\

Peers 8- Family
of Procreation

 

 

 

ll & III. Primary Group 8: Family of Origin
Intimate Secondary .
Group Influences Fa Mo Slbs Peers
°Parent Personality -Peer Personality
- lnteractional 8- Interactlonal
Influences

 

 

Influences

-Child Rearing -Peer Socializa-
Pattems tlon Influences

W- ———————— —r —————————

hDrlving Patterns ‘ -Driving Patterns
and Attitudes

and Attitudes

 

 

 

 

 

 

 

l

 

IV. lntra-lndividual -Cognltlve Structure
Influences (including Attitudes
Beliefs. Values)

-Personality Structure
(Motivations. Traits
Temperament)

—-—--—‘———-d

Non-Representational
Elements; including
Genetic & Other

Biochemically Mediated
influences

W

(Eventual) Driving Behavior

 

 

 

Figure 3. An Organizational Structure for Classes of Influences Upon
Driving Behaviors (Zucker, 1984).

28

I. Socioealtural and
Cbmrmmity Influences

II & III. PrimeryGroxpand

Iv. Intra-Individual
Influences

Note. Somewhat modified from Zucker (1979); as refers to Drinking Specific factors, NIB

29

Familj of m
Fa Mo Sibs
-Pareit Personality
actioal

-Inter

Influences IDS Influences
«mm baring

Patterns

 

-30cioeooiomic stems of
family

-Ethnicity, Religious
influences and Behavior
alternatives

-Neighborhood Value
influences and Bel-lavior
alternatives

 

 

-Availabilit:y of alcohol
and values of dr

W of Drinking Settings

 

  

 

 

 

 

-Drinking Patterns,
Attitudes and IS Attihxbs,

Settings

 

 

 

 

 

 

 

 

 

 

a V43!
-Oogmt1ve8tructure

(including Attitudes,
Beliefs, Values)

-Personality Structure

(ibtivatiois, Traits:
Tamerament)

-Nonreprese1tatimal

Elelents; including
Genetic s Other

-Biochemical 1y Mediated

Influences

 

-Drinking Attimdes and
(petitions re: Alcotol

Btiological Actions of
Alcohol

 

(Mattel) Drinking

refers to Norririnking wecific factors.

 

or (Aogrired, Maintained,
Increased, etc.) and Ooiseqlaeioes (Moral/Physiological)

rifle . An Organizational Structure for Classes of Influences Upon trinking Beluvior

30

family of procreation influences, and (4) intraindividual influences.
Within each class of influences, it is posited that there are both
specific and nonspecific variables. Parental modeling of drinking is
an example of a specific, family of origin influence; family climate is
a nonspecific, family of origin influence. Zucker (1984) notes in his
review of youth, alcohol, and driving that "the most obvious and
glaring omission is the comparative lack of attention to peer
influences and family influences in mediating drinking-driving
etiology." Thus, the significance of the proposed study is evident for
both practical and theoretical purposes.

. 1A.thorcugh review of the literature on young drivers and
adolescent drinkers is beyond the scope of this thesis. Because both
areas are of significant social interest, the amount of work devoted to
their study is enormous. However, several factors which are thought to
affect each area will be mentioned in order to present a more complete
‘view of the phenomenology of adolescent drinking and driving. Studies
which address family correlates will be discussed in more detail. It
i is not the intent of this study to suggest that family variables are
the only nor the most important factors related to adolescent drinking
and driving. Hewever, an understanding of their relationship will
assist in clarifying a valuable component of the total picture.
Adolescent Driving and Family Correlates

The problem. Ybung drivers are overrepresented in accidents,
particularly in fatal accidents, considerably beyond their proportion
of the driving population. In 1973, they comprised 21.7% of the
driving population in the United States, yet they were drivers in 36.3%

of all fatal road accidents (National Safety Counsel, 1974).

31

Currently, they are still disproportionately represented among fatal
accident victims (National Institute on Alcohol and Alcoholism, 1982;
Insurance Institute for Highway Safety, 1984). Demographic statistics
suggest that exposure factors--more»driving at hazardous
times--(Goldstein, 1972) and inexperience (Pelz & Schuman, 1971)
mediate this relationship. However, there is mounting evidence that
other, nondriving specific factors may also be involved in the driving
patterns of young drivers. These factors have been referred to as
"extra motives."

Substantive review. Black (cited in Goldstein, 1972) investigated
responses from young drivers who were hypnotized during their
interviews. He reported that many personal-emotional variables emerged
and suggests that these may account for the disproportionate frequency
of road accidents among youths. Although Black did not describe family
relationships, his findings suggest conflict between these youths and
their parents.

Pelz, Schuman and their associates (1971) found a cluster of
extra-motivational factors to be related to the driving
behaviors--accidents and violations--of adolescents. Included in this
list are drives after drinking, speeds on highway, speeds in city,
works excessively on car, races others, drives affected by passengers,
drives to blow off steam, involved in fist fights, and thoughts of
injury while driving. The only parent-related variable was "feels
adult pressure." These authors convincingly argue that teenagers
possess more of these extramotives than their adult counterparts.

Beamish and Malfetti (1962) compared two groups of adolescents

watched on age, education, and miles driven annually. The first group

32

had been referred to the juvenile courts for two or more traffic
violations. The second group was comprised of accident-free
adolescents. Based upon psychological tests, the two groups differed
on several emotional and family variables. compared to controls,
traffic violators: (1) exhibited an impulsive style, avoiding
consideration of the effect of their actions on themselves or others;
(2) tended to be in conflict with.others, including those closest to
themselves-e.g., they perceived their parents as imposing and
restrictive; (3) were rebellious and selfish; (4) were lacking in
self-confidence and exhibited poor self esteem, perhaps contributing to
a need to compensate for these feelings of unworthiness by erratic and
ill-considered actions; (5) reported that their parents and relatives
were inactive in the commnity, producing children with a lessened
sense of civic responsibility (reviewed in Goldstein, 1972).

In a second study by Beamish and Malfetti, remediable traffic
violators were compared to nonremediable violators. The remediables
rated higher on sociability, social and physical activity, dependence
on home--i.e., turning to family for guidance and support--and
appreciation of the arts.

A study by Rommel (1959) provides more direct support for the
hypothesis that youthful violators use driving as an activity in.whi¢h
to prove their adult status, or to compensate for feelings of
unworthiness. They administered the Minnesota Multiphasic Personality
Inventory (MMPI) and the Driver Attitude Inventory to
accident-repeating and accident-free high school drivers. The repeater
group exhibited higher scores on Psychological Deviancy (PD)-disregard

for social mores-and Mania (MA)-excessive activity and enthusiasm.

33

Whth.respect to driving, the repeater groups tended to view'driving as
a.way to display confidence and to prove powerfulness despite danger
indicators.

The antisocial aspect of young violators emerges in several other
studies. Ievonian (1969) studied 1080 driver education students and
found that repeat violators tended to be oriented toward self-benefit
at the expense of others. Further, they noted the rebellious attitudes
presented by this group. Similarly, Goldstein (1972) concluded that
there "are a cluster of personal variables which are correlated with
crashes among young male drivers, labelled by such terms as
anger-rebellion, hostility, argument, distraction, escape and
competitiveness."

While some adolescents appear to be using their automobiles to
release tension which may be due to frustrated independence strivings,
there is also evidence that this group of youngsters is unable to
achieve status from other activities commonly referred to as ”rites of
passage." Kraus, Steele, Ghent and Thompson (1970) gathered data on the
backgrounds of 205 young drivers—-under 21-dwho had recently had
accidents, with 205 matched controls. They found significant
differences in the areas of school achievement, cigarette smoking,
employment and criminal behavior. Specifically, accidentvprone
individuals (1) had failed one or more grades, (2) became cigarette
smokers at or before age 16, (3) secured full-time employment,
exclusive of summer vacation, at or before 17 and before receiving a
driving license, and (4) had been charged with a criminal offense

exclusive of those related to driving.

34

In another study, Harrington (1971) found very similar results.
This research effort represents one of the most extensive efforts to
isolate biographical data which may be related to adolescent traffic
problens. 'Ihree thousand to 8,000 males and 3,000 to 5,800 females
were fol lowed over a four-year period. High accident males—three or
more violations (N=175) -and high accident females—two or more
accidents (N=210) -were compared to accident free subjects. Interview
data were obtained on a nunber of psychosocial variables, including
family relationships. School records provided useful demographic
information on school attendance , achievement, etc . The data were
analyzed separately for males and females. Over 300 variables were
tested and many proved to be significant. The findings most relevant
to the hypothesis of this paper include:
Compared to the low-accident males , high-accident males :
(1) smoked more cigarettes, (2) less frequently were college
students, (3) more frequently wanted to be race car drivers,
(4) began dating at an earlier age, (5) rated their driving
skill lower at ages 16 to 17, (6) completed less education,
(7) played hooky in high school more often, (8) had their own
cars with speed and custom accessories more often at ages 16
to 17, (9) got along less well with their parents, (10)
received less parental approval of their friends at ages 16
to 17, (11) had mothers who lost their tempers more easily
and who babied them more, (12) had consuned more alcohol
before driving, (13) more frequently knew someone who smoked

marijuana, (14) were more frequently in trouble with the law

35

after age 20, (15) more frequently drove when worried, and
(16) acre frequently drove to get away from people.

Compared to low accident females, high accident females]

(I) drove sports cars more frequently, (2) drove more miles
in their lifetimes, (3) belonged to more clubs, (4) felt like
smashing things less frequently, (5) had poorer relations
with their teachers in high school, (6) played hooky acre
often in high school, (7) received less parental approval of
their friends, (8) improved their driving more because they
had been in an accident, (9) attended more car races, (10)
drove acre frequently to get away from other people or to
cool down after an argument with someone, and (11) described
themselves as more conventional, persevering, polished,
self-controlled, friendly, decisive, orderly, sophisticated,
and less frank.

The authors warn.that because of the tremendous number of
comparisons that.were made, the results should be considered
tentatively. At best, one can draw several hypotheses from this study.
The results provide added support for the notion that at-risk males
perceive their families as more conflictual and less able to meet their
needs for separation and individuation. 'With.these needs frustrated,
they may turn to other "rites of passage" in order to assure their
independence-i.e., early dating, drinking, and risky driving. More
legitimate sources of achievement, such as school, are not.availab1e to
them. Hence, these individuals may be at risk for "exaggerating" their
behavior in the driving and drinking realm. For females, the picture

is less clear. There is evidence of some family tension and turning to

36

the peer group, but several of the other variables did not dif-
ferentiate between the at—risk and control groups. Perhaps females
chose activities other than driving in which to assert their indepen-
dence. Or perhaps they do not experience dependency frustrations as
early or as intensely as males, since it is more socially acceptable
for females to remain dependent.

As the preceding review indicates, there is a significant paucity
of data which examines the family environment as a central factor
influencing adolescent driving behaviors. At most, family variables
are considered as a secondary concern. Only one study has made the
family its primary focus. Sobel and Underhill (1976) found that poor
relationships with one's parents, as well as an unstable marriage, are
related to a greater frequency of alcohol-related driving and
automobile accidents (cited in Zucker, 1984). Clearly, there is a need

to address this issue directly.

Methodological considerations. 'In reviewing studies on youthful

 

traffic violators, several problematic methodological issues emerge.

3 First, the definition of ”youthful" varies widely among studies. Some
studies define youthful as 24 and under, while others focus
specifically on high school drivers. From a developmental perspective,
the emotional and social issues vary considerably from 16 to 19, versus
20 to 24 year olds. From a family life spanperspective, the first age
group is in the "adolescent” phase of development, whereas the second
group is in the "launching" stage. Olson and Mccubbin's (1983) work
fbund differences in family climate variables between the two groups.

Thus, more homogeneous group comparisons might be valuable.

37

The second issue deals with sex. A.great deal of traffic research
deals with male subjects. This focus may reflect the reality of who is
involved in more traffic violations and accidents since 85.6% of
drivers of all ages involved in fatal accidents are male (Accident
Facts, 1971). This difference is explained by the fact that males
drive many more miles than females. As female driving hours increase,
it is expected that they will be involved in more accidents. An
alternative explanation is that the deviance process is qualitatively
different between males and females and reckless driving is only one
form of deviant behavior. Some studies fail to report separate
analyses by gender. The ones that do tend to find different profiles
for males and females, suggesting the utility of this approach.

Finally, Klein (1972) points out that violations and crashes are
not highly correlated. He further explains that this finding may be
due to the fact that crashes are low frequency events and that many
violators are not apprehended. However, data resulting from these two
types of dependent variables should not automatically be considered
equivalent.

Adolescent Drinking and Family;Correlates

Theyproblem. In a study of over 900 adolescents who took part in
a national survey (mean age=16 years), it.was determined that 60% of
adolescents are drinkers-infrequent, light, and moderate--and that 6%
are heavier drinkers. Sex and age variables were found to mediate this
relationship. In alacst all age groups, males consume greater
quantities of alcohol than females. Moreover, within each sex, the
percentage of drinkers increases with age. Figure 5 presents the

specific percentages of drinkers by age and sex (reprinted from Zucker

38

 

Nov n6 9: n43 n.mm ”8m: :<
om m.o~ 9.3 a 3 ~.3 oi oi on
«a Q3 Wow H 5 v.3 «.9 «.3 3
no m6 «.3 a 3 v.3 ...: v.3” 3
2. n4 0.5 v.3 v.94. 93 9% mu
3 can b.» .— 3 o ma non v.3 3
2. v4 .4 v a u a 9.5 «.3 mu
2 30s when .. wag as
30388

EH38 "Bomb 8388

and 3.593 09.3“ mod—Sup 3033

3.3300 a 5:9. a 8:0 55 Mom 3:99.!

33:6 “835 89. 8 mm 8388 :95 .36

you 3583 ooh: 9.5230 «33%» a 55... m no.5»

aficoaonflfiu image—=38. cum—3539.0:

5.3800 vim .ud scam—800 18 5:9. a $5» 3.6 mm 53800

$258 9558 T... Ho. .8888 Pagan ”Bax... 8388
Haves you 3393 won 05x53 33%» non 355.5 gnu 33g... .30» a 8:0
356:! 09.3 356.5 .532. you 355.8 325 .588. :53 Bo nod 356.5 :95 55 when

coming confining gaining u§um5§ goon—museum xfiuouo gunned
ammo“ an :58 uuuS an 358 name" on x55 0 coco x55 58... m 8:0 x58 :53 9:8 9:938

 

find 3:6»: flags
Hogan:
\3389.

39.58 3888: 393.5 234 8638 acme—8&5

 

mam—6H0 amt/flu m: «xv—Hun

c.3533: 33.6

39

.253 £9.88 205m ~38. .8 mm:
9.0 xum Elk—00300 3 5 3:00:83: mo unmoufionQ 0:0 .393 050—58 30003001 you 0.2330 05:38 3% E

.ga 3:030: 3:303 :93 3.33 5003.880.“ 035m 30 08.63 3 503333 3:030: 05 «0
5305536 .80 0.3 £0030: 338.600 00: 0000 03.80 :33 «5 00:83 «mega 00.5030 :0 0003 30 803%

.angauufimou .5388 gnu .30 9.5.6 000: no m 0:0 ...-03300 050.53 nun
9.5.5 vu~ ..Bunoooo 9.3.500 you and 002 no 0:0 3 woman 005.9! 032 0:: ...-Son: .283 .98qu300 0.6a: 05 :H

 

.goooggufiouflofifigoiasguSooflgiz .39.

 

 

mun 9m .02 HS 5.3 92 93 .83 3.
no. 3” m... n a 3... 92 fine .89. 3.
3 as ...fi m 3 93 m 3 ....N S
S ...n 32 v p is... o a min 2
S v... ha a a m.n~ o 2 3...” 3
ma «4 ....» a 2 TR N a ...? 2
m... o... ...~ m a ...3 a 2 3H. 3
.... Q. a... o o is .. d 3... 2
z and 2.36 I 09.0.5 09.
393.8 .898: 39.58 8.9.58 3289. 89.5.5 23.: 263.5 age: flgfimfi 3.80

.8ng
\3889.

40

& Harford, 1983). It is clear that drinking, especially among males,
is not a statistically deviant activity. However, from a legal and
social perspective, drinking is an activity which is reserved for
adults and is considered inappropriate for teenagers.

The cause of the recent social concern regarding adolescent
drinking stems from several sources. First, as was noted in the last
review section, alcohol is implicated as a contributing factor in the
nation's number one health prdblem for adolescents-driving fatalities.
Approximately 60% of all fatal motor vehicle accidents involve alcohol
in the 16 to 22-year-old age group. Compared to adults, few teens
report driving while under the influence of alcohol; however, when they
do drink and drive, they are more likely to be involved in a traffic
accident at lower blood alcohol level (Klein, 1972). Secondly,
adolescent problem drinking has been linked to a number of forms of
deviant or antisocial behavior among youth (cahalan, 1970; Ricks &
Berry, 1970). Finally, adolescent problem drinking has been associated
‘with adult alcoholism and prdblem drinking (Rabins, 1966; Blane &
Chafetz, 1979), and also with use of other drugs (Braucht, 1982).

Substantive review. In trying to determine the etiological

 

variables associated with.drinking, particularly problem drinking, a
number of investigators have called for the application of
sophisticated theoretical and methodological designs. Zucker's
developmental model was presented earlier in this review. Jessor and
Jessor, in their triethnic study, developed a similarly complex

formulation of the problem.

41

A thorough review of the adolescent drinking literature is clearly
beyond the scope of this thesis and has been quite adequately presented
elsewhere (Zucker, 1976; Zucker & N011, 1982; Zucker, 1984). Thus, the
present review will focus on family environment.variables as they have
been related to adolescent drinking. Other related variables will be
briefly mentioned, since their consideration provides a conceptual
richness that extends beyond the narrow focus of family environment.
Three sources of data will be reviewed. Two research groups, Zucker
(1975, 1973) and Jessor and Jessor (1968; 1970; 1973; 1975) have been
major activators in the area of adolescent alcohol use. Thus, both of
their studies will be presented separately. The third source of
information comes from several longitudinal studies which associate
adult problem drinking with earlier family, social, and emotional
precursors. Jessor and Jessor emphasize the normal developmental
process of becoming a drinker, while the other studies emphasize the
pathology involved with.problem drinking. Although the focus of this
study lies with the former conceptualization, the problem—oriented
studies are included because they identify family environment
characteristics which are likely to be associated with one aspect of
family functioning; namely, obvious dysfunctional family functioning.
This study‘will not attempt directly to separate alcoholism-prone
adolescents and families from other types of drinkers and drinking
families; the two groups will be studied together. conceptually,
however, it is useful to keep the distinction in mind. It:may be that
families with alcoholism-prone adolescents exhibit a more severe
pathology, and that their failure to meet the developmental needs of

their children is not limited to the adolescent period, but has

42

occurred throughout the family life cycle. Longitudinal work is needed
to test this hypothesis.

Zucker's model proposes a developmental perspective to the
etiology of drinking among adolescents. Initially, the model focused
on family organization and affectional ties in order to explain
adolescent problem drinking. When it becane clear that empirical data
supported multiple causality, three other dimensions were added-peer,
intraindividual, and sociocultural influences. Although the model
attempts to explain a normal developmental phenomenon, its emphasis is
on predicting and understanding prdblem drinking. It is hypothesized
that problem drinking is best understood*within the context of
antisocial behavior. (For a review'of the Zucker model, its advantages
and limitations, see N011, 1980; 1983). Cbnsequently, the majority of
research attempts to isolate variables which could differentiate
prdblem adolescent drinkers from nonproblem adolescent drinkers.

A.major advantage of Zucker's work is that he investigated
drinking related variables separately for males and females. The
I findings will be presented separately. Seventy-five males and 75
females were selected from junior and senior high school students in
order to represent the entire drinking spectrum. Data from in—depth
interviews, psychological testing, and personal diaries was gathered
over a fourdweek period. Two and a half years later, original freshmen
and sophomores-then juniors and seniors-were retested. In addition,
their family members were contacted and asked to participate in the
study. The results described herein are those from this latter tesing

period (Zucker & Barron, 1973).

43

Among males, there was strong evidence for the initial theory
suggesting an association between general antisocial behavior and
problem drinking. For exalnple, on three measures from the California
Psychological Inventory, heavy and problem drinking males differed from
controls. Heavy and problem drinkers reported less hone stability,
optimism, and observance of convention. There was no evidence to
suggest that problem and heavy, nonproblem drinkers should be
conceptual 1y independent categories . Both groups differed from
controls on many antisocial and asocial measures, but they did not
differ from each other.

Several findings emerged with respect to how the parents' own
drinking practices relate to their child's drinking habits. Parental
intake was found to be a more salient predictor of the adolescents '
drinking amount and drinking practices than was parental problem
drinking. Secondly, and somewhat surprisingly, it was discovered that
both maternal drinking and concern about her own drinking practices
were better predictors of adolescent male drinking patterns than were
the father's drinking related behaviors. The authors point out that
little is known about female drinking practices and that this area
warrants further investigation .

Mothers ' attitudes and behaviors regarding discipline and
affectional ties also appeared to be more closely related to child
drinking than the fathers' attitudes and behaviors. With regard to
affective responsitivity, mothers reported themselves as having been
more openly rejecting and ambivalent about expressing concern through
discipline. Sometines they would punish by depriving the child of

human companionship; at other tines they would be overprotective.

44

From.the adolescents' point of view, their behavior was less
related to their mothers' attitudes and values than to their fathers'
behavior. They perceived their fathers as affectively distant and
unrewarding. They also felt their fathers failed to attempt to
socialize their behavior while they were growing up.

The family atmosphere measures indicate that the global family
situation is tense and that it is characterized by a great deal of open
rebellion and resentment.of parental intrusions.

The authors summarize the findings by concluding that there "has
been a failure of the affective environmental supports that provide for
the development of solid positive identification with parental figures.
Thus, by adolescence, the critical factor is this separationrescape,
partly by way of alcohol abuse." It may be that such families
represent the extremes in terms of adolescent separation conflicts and
family climate deficits. It.would be interesting to determine whether
less problematic drinkers also perceive their families to be
affectionally and organizationally deficient, albeit in a less serious
way. If there is a difference in family climate, is the distinction a
qualitative or a quantitative one?

Among females (Zucker & Devoe, 1975), the authors found similar
findings to the pattern noted for males. Many of the indices of
antisocial behavior were related to heavy and problem drinking;
however, the correlations were somewhat lower.

The family environments of the girls also appear to be conflictual
and hostile. The differences lie in the degree of perceived anger and
rejection and the relative influence of parental variables on the

adolescent drinking practices. The mothers of drinking females report

45

more open rejection, nonenurturance, and less companionship than do low
or nondrinkers. FUrther, many mothers report being physically absent
from the home for extended periods of time. Fathers report less
consistent discipline with less affiliative companionship, and more
antisocial behavior. The girls' perceptions of their parents/families
support the accuracy of the parents' reports. In sum, both.ma1e and
female problem drinkers come from dysfunctional households,‘with girls
experiencing more parental rejection and sociopathy than their male
counterparts.

Jessor and Jessor (1968; 1970; 1973) were interested in
understanding the developmental processes which account for changes in
antisocial behavior. Extensive drinking is but one of several deviant
behaviors that they examined. According to their theory, general
antisocial behavior increases during periods of transition. Personal
and sociocultural variables interact to determine the expression of
deviance in any given individual.

To test their model, they identified several variables within
three major explanatory systems--personality, the perceived social
environment and behavior. Figure 6 lists the types of variables which
were examined. Fbur hundred thirty two subjects (188 boys and 244
girls) from grades seven to nine were studied over a three-year period.
The above mentioned variables, as well as drinking habits, were
examined. Subjects were classified into one of five transition groups:
I. abstainers; II. last year transition group; III. second year
transition group; IV. first year transition.group; and‘V. pre-study

transition group.

46

PERSONALITY SYSTEM

Personal Instigations
value on academic achievement
value on independence
Independence-achievement discrepancy
EXpectations for academic achievement

Personal Controls
Tblerance of deviance
Religiosity
Disjunction of drinking functions

PERCEIVED ENVIRONMENTAL SYSTEM

Distal Environment
Parent-friends compatibility
Parent-friends influence

Proximal Environment
Family approval of drinking
Friends' approval of drinking
Friends' drinking pressure
Family models for drinking
Friends' models for drinking
Friends' pressure for marijuana use
Friends' models for marijuana use

BEHAVIOR SYSTEM

Problem Behavior
General deviant behavior
Marijuana behavior involvement

Conventional Behaviors
Church attendance
School performance

DEMOGRAPHIC VARIABLES

Age

Father's education
Mother's education
Family SES

F'gure 6. Jessor and Jessor (1978): variables examined in their study
of adolescent transition to alcdhol use.

47

The results support the hypothesis that there is a constellation
of personality, perceived environment, and behavioral attributes whiCh
exist prior to the onset of drinking and which are correlated with
initiation into drinking. For example, group I scored highest on
achievement-orientation, indicating greater conventionality. Group v
scored the lowest on this dimension, indicating a disinterest in
socially prescribed modes of adolescent status. Groups II to IV fell
in between these two extremes, in the predicted numerical order. A
similar ordering of the groups occurred with several other variables.
Abstainers were described as having the least instigation to prdblem
behavior, the strongest personal controls against transgression, a per-
ceived environment that provides the least approval and opportunity for
drinking and the least amount of general deviance. In contrast, the
group that reported drinking at or before the beginning of the study
exhibited the greatest level of deviance, few controls against
transgression, much opportunity, support and modeling of drinking, and
greater prdblem behavior. The profiles of the other groups fell within
these two extremes, in the predicted order. When all the variables
were combined, they accounted for 37% of the variance between groups
among girls and 47% of the variance among boys.

Demone (1975) found similar results utilizing a cross-sectional
methodology. Compared to abstainers and nonmative drinkers,
pathological drinkers were more likely to come fromidivorced families
with a parent who has or has had a drinking problem, or a parent who is
ardently opposed to drinking. These adolescents were more likely to
spend time with and rely on peers, engage in antisocial behavior, date,
work, skip school and repeat grades. Further, they reported drinking

48

at an early age--i.e., between ten and 12 years of age. Demone
concluded that this group of adolescents seemed to be rejecting adult
sanctioned forms of adolescent behavior and adopting adult roles and
privileges.

The major strength of Jessor et al.'s work lies in its longi-
tudinal focus in accounting for "normal" drinking transitions between
adolescents. The family variables that.were studied relate more to
drinking specific variables, as opposed to the Zucker study, which
focused more on family climate variables. Jessor et al.'s study
suggests that parental modeling of drinking behaviors and approval of
adolescent drinking are related to the earlier experimentation with
these behaviors. In another study, Biddle, Bank and Marlin (1980)
found that parental drinking patterns had little influence on specific
adolescent drinking practices. Rather, parents affected their
adolescents through normative standards and peers through modeling
behavior, indicating the different kinds of relationships an adolescent
has with his parents and peers. What is clear is that adolescents who
1 drink excessively (Zucker) and/or earlier (Jessor) have ample access to
alcoholic beverages and modeling of drinking behaviors in their
environments. Perhaps such factors are necessary but not sufficient in
understanding the phenomenology of adolescent drinking.

Several longitudinal studies shed light on family variables from a
more pathology-oriented framework (McCord & McCord, 1960; 1962; Robins
et al., 1962; Robins, 1966; J0nes, 1971; Berry, 1967; Ricks & Berry,
1970). Subjects were initially contacted as children and then followed
up a number of years later. Although there is great variability among

studies--i.e., subject selection, subject characteristics, definition

49

of "problem drinkers," instrument chosen, method of fol low up,
etc.--there are several similarities which allow for tentative
generalizations: (1) similar categories of the family environment were
examined—affectional interaction, parent reward structure , parental
modeling; (2) control groups of nondrinking, disturbed children were
present (in addition to several normal controls); and (3) identified
adults were diagnosed as alcoholics. These studies are reviewed by
Zucker (1976; 1979; 1982).

With respect to family environments, the longitudinal studies
indicate that there are distinct deficits in the affectional ties
between family members, particularly within the marital dyad. Often
such difficulties are apparent within the parent—child relationship as
well. mereas the father is portrayed as more consistently absent, the
mother alternates between overinvolvenent and open rejection. Not
surprisingly, many mothers were found to have emotional disturbances
and to resent their role within the family. Although the conflict
within these families appears to be great--i.e. , high rates of verbal
and physical assault—the discipline practices are more adequately
described as inconsistent than as harsh. Combined with inadequate
discipline, the parents exhibit greater sociopathy and fewer
ideological controls-i.e. , adherence to religious or moral beliefs.
Thus, although these studies are describing the family characteristics
of one group of drinking adolescents, those who becone adult
alcoholics, the findings provide tentative support for the proposed
model of dysfunctional family interactions. In order to account for
the spectrum of normal to excessive drinking among adolescents, greater

specificity with a representative population is needed.

50
Drinking and Reckless Driving: Common or Unique Problems?

According to the National Highway Traffic Safety Administration's
National Center for Statistics and Analysis, teenagers constitute about
eight percent of the driver population and account for approximately
six percent of the vehicle miles travelled; however, statistics
indicate that they are not only disproportionately represented among
accident-involved drivers , but that tley are also more likely to be
involved in drinking related accidents. Specifically, in 1982 15% of
all alcohol-involved drivers in accidents were teenagers (statistics
summarized in Vegega, 1984). men involved in an alcohol-related
accident, teenagers are generally found to have lower blood alcohol
levels than their adult counterparts. Thus, the combination of exces—
sive drinking and careless driving appears to be a special problem for
teenagers. The question remains as to whether these two areas share
enough common etiological components be more accurately considered a
single phenomenon. An inspection of the literature suggests that the
answer may be both yes and no. That is, within a subset of the
population, there are aspects of excessive drinking and reckless
driving which could be grouped into a single category. cm the other
hand, in the remainder of the population, drinking and driving related
factors are clearly separate phenomena.

Specific examination of the factors which can distinguish problem
drinkers and drinking-drivers during adolescence is lacking. The
majority of evidence appears to come from the adult literature.
Selzer, Vinokur, and Wilson (1977) conducted a psychosocial comparison
of drunken drivers and alcoholics. They found that compared to

controls, alcoholics were significantly different on virtually every

51

variable. Specifically, they drank more frequently and greater amounts
per drinking occasion, drank more to relieve tension, experienced more
troublesome and uncomfortable effects from drinking, had more stress in
their personal and professional lives, and participated in fewer
non-drinking social events. Men arrested for drunken driving were
distinguishable from both controls and alcoholics on the
above-mentioned variables. They appeared to fall somewhere between the
control subjects and the alcoholics. Based upon these results, the
authors concluded that alcoholics and drunken drivers are not
homogeneous groups, but may share features which distinguish them from
the larger population, differing from each other primarily in degree of
pathology.

In a similar study, Clay (1974) provided evidence which supports
the notion that problemidrinking is largely a separate phenomenon from
problem drinking and driving. She compared alcoholics and DWI
offenders to controls and to each other. Although alcoholics were
involved in more drinking-related accidents, they were less risky
drivers. That is, the number and severity of their driving offenses
were less significant than the offenses of the DWIs. In fact, the
majority of the DWI group's offenses had no alcohol involvement.

Further examination of the DWI group led Clay to conclude that
there may be a section of the population whidh is characterized by an
aggressive irresponsibility which permeates many behavioral realms,
including drinking, driving, property management, interpersonal skills,
and compliance to societal norms and rules. To a much greater extent

than alcoholics, EMU offenders had been processed by the legal system

52

for violations which ranged from destruction of property to criminally
assaultive behavior.

Other authors have proposed theories of deviant behavior which
claim that there is a single dimension which underlies various
manifestations of antisocial behavior. For example, Donovan and Jessor
(1985) conducted a factor analysis on a diverse groupiof problem
behaviors-problem drinking, illicit drug use, delinquent-type
behaviors, and precocious sexual intercourse-which.were assessed on
both an adolescent and a young adult sample. Their results support the
notion that these behaviors may be accounted for by a single syndrome.
Certainly, reckless driving, though not included in this study, could
be considered a form of antisocial behavior in light of the aggressive
qualities noted by many authors.

Though the questions regarding the commonality of these two
prdblematic behaviors are far from resolved, future work should keep in
mind that there are likely to be similarities among a subset of the
population. The task which remains is to find out which factors
subsume the commonality and how these factors differ in the population.
In this study, it is hypothesized that the spectrum«of drinking and
driving behaviors are linked because of their value as a rite of
passage, and that family system processes operate to differentiate the
successful management of these behaviors among adolescents-i.e., to
distinguish between abstainers, normative, and problematic drinkers,
and between risky and normative drivers.

Application of a New Terminology
Examining the link between family influences and adolescent

drinking and driving practices is an area.which has not been given much

53

attention in the past, as is apparent from the preceding literature
review. However, in addition to the general lack of attention, there
is another problem which is preventing advancetent in this area. When
family influences are examined, they are not well conceptualized. The
result of this practice is a collection of results which seem isolated
from each other and somewhat trivial. What is needed is the
application of a systematic theory for understanding family influences.
A systems perspective provides such a theory.

As a function of utilizing such a theory, investigators would have
a new terminology for discussing family influences, which includes
concepts such as system relationships, individual growth, and system
maintenance. Instead of addressing seemingly isolated behaviors, one
can look at a host of behaviors which are conceptually related to a
construct lmown to operate in family systems. In this study, the
Family Environment Scale (Moos, 1974) has bee) selected as the tool
which will measure three dimensions of family functioning-quality of
interpersonal relationships, individual growth pursuits, and system
maintenance techniques. The application of these concepts and this
terminology to adolescent drinking and driving represents a vast
improverent over the current state of affairs.
Significance of This Investigation

In determining the relative significance of any investigation, the
contribution of the findings to three areas needs to be assessed: a)
methodological significance, b) theoretical significance, and c)
practical significance. This study has important implicatiom for all

three areas.

54

Theoretically, this study proposes a model which might account for
some of the variation Observed in adolescents' struggles with identity
issues. It dealS‘with.both healthy and dysfunctional manifestations.
The use of a family systems perspective will not only contribute useful
information to the understanding of family process during the
adolescent phase of development, but also provides a new framework for
conceptualizing adolescent drinking and reckless driving.

The methodology of this study avoids a number of limitations noted
with prior research. First, family climate is operationalized along
several dimensions, using a standardized tool--the Family Environment
Scale (FES) (Moos, 1974). Secondly, multiple measures of drinking and
driving behaviors, and alternate definitions of what constitutes
problems in these areas are incorporated. These two advantages allow
for more direct comparability between this study and past studies.
Third, by examining mediating variables-such as GPA, academic
expectations, truancies, peer modeling, peer support, family support,
and accessibilityh-the study will be able to make more precise
conclusions about the relationship of family influences to other
influences that the literature shows to be important. In this sense,
the methodology is ecologically oriented, rather than arbitrarily
focused on unitary constructs. Another methodological advantage is the
use of sophisticated, multivariate techniques which allow for the
interrelating of a greater number of variables while reducing the
capitalization on chance findings. Finally, the design employs a
normative population rather than a clinical population, allowing for

greater generalizability.

55

Practical 1y, by operationalizing the processes of separation and
individuation into drinking and driving related variables, this study
will provide useful information for researchers and clinicians who are
interested in these major social problems. Moreover, by focusing on
the family environment, the belief in the individual pathology of
adolescents might be lessened. Such an effort will be worthwhile in

providing practitioners with alternative avenues for intervention.

56

Hypotheses

The general premise of this study is that adolescents who perceive
their families as less functional will be more likely to force their
way into adulthood by exaggerating behaviors considered to be rites of
passage, such as drinking and driving. Within the context of this
study, exaggerated drinking refers to individuals who consume large
quantities of alcohol on a more frequent basis, resulting in more
problems associated with their drinking. Exaggerated driving refers to
individuals who have been involved in any automobile accident as
drivers or who have received frequent moving violations. The findings
of Olson and McCubbin (1983) suggest some characteristics which
distinguish functional from dysfunctional families during adolescence.
Specifically, less functional families appear to be at the extremes on
cohesion and adaptability. Moreover,'Olson and Mccubbin found that
less functional families express greater overall conflict and less
commitment to or involvement in individual growthrproducing activities.

In the present study, the Family Environment Scale (FES) was
utilized. Although it cannot be assumed that the FES and Olson and
McCubbin's instrument,FACES, measure the same constructs and hence will
produce identical results, several of their scales appear to be
theoretically similar. For example, the items on the cohesion scale of
the FACES seem to be similar to the cohesion and expressiveness scales
of the FES. Further, the FACES adaptability scale appears to be
similar to the organization and control scales on the FES. To measure
conflict and individual growth pursuits, Olson and Mccubbin developed
their own items and did not report any psychometric data on these

items. An advantage to using the FES is that it contains scales that

57
measure conflict and five areas of individual growth pursuits, and that
these scales have been standardized on a normative population.

Extrapolating from this analysis, it was hypothesized that
families which fall at the extremes of cohesion, expressiveness,
organization, and control, and that exhibit greater conflict and fewer
individual growth pursuits for its members would have adolescents who
experienced difficulties in managing the normal adolescent tasks of
separation and individuation.

The aforementioned literature review presented several research
findings which support parts of this theory; however, the majority of
the studies did not use a family systems approach, nor did they regard
family influences as their main area of interest. consequently, it is
difficult to integrate the findings in such a way that precise
relationships can be detailed and specific hypotheses formulated. At
best, the extrapolations from studies and theories regarding family
functioning during adolescence, and the empirical literature on
adolescent drinking and driving can be combined to outline some
tentative hypotheses which are more exploratory in nature than
confirmatory.

Based upon the stated objectives of the proposed study and the
implications from the theoretical and empirical literature presented
herein, the following hypotheses were developed:

Comparison of variations in Adolescent Driving Styles

Hypothesis #1 (family functioning):
It was hypothesized that families which fell at the extremes
on the dimensions of expressiveness, cohesion, organization,

and control, are high in conflict, and are low in the five

58

areas of individual growth wmfld have adolescents who were
riskier drivers—defined in terms of self reported (1)
automobile accidents, (2) automobile near misses, and (3)
seat belt disregard.

Hypothesis #2 (peer versus family influences):
It was anticipated that peer modeling of and support for
reckless driving would be more highly associated with a
riskier driving style (Vegega, 1984; Pelz & Schuman, 1971;
Goldstein, 1972; Harrington, 1971). However, an understanding
of the family climate would add significant information to
the understanding of adolescent driving styles.

Hypothesis #3 (school, peer and parental correlates):
Earlier research suggests that there are a whole host of
extramotivational factors that inf luenoe tre driving patterns
of adolescents (Pelz & Schuman, 1971) . Included in this
category are school, peer and family variables. Thus, it was
hypothesized that adolescents who were riskier drivers would
be (1) less committed to school achieverent and activities,
(2) more influenced by peer modeling of and support for risky
driving behaviors, and (3) more likely to perceive their
parents as tolerant of risky driving behaviors .

Comparison of Adolescents with Normative and Excessive Drinking Styles

Hypothesis #4 (family functioning):
It was hypothesized that families that fell at the extreres
on the dimensions of expressiveness, cohesion, organization,
and control, are high in conflict, and are low in the five

areas of individual growth would have adolescents who

59

exhibited exaggerated drinking-defined in terms of self
reported (1) frequency of drinking, (2) quantity of alcdhol
consumed per drinking occasion, (3) problems associated with
drinking, and (4) frequency of drunkenness in the last year.

Hypothesis #5 (peer versus family influences):
It was anticipated that.peer modeling of and support for
alcohol use would be more highly associated with.an
exaggerated drinking style (Wbdarski, 1984; Braudht, 1982;
Lener & Spanier, 1980). However, an understanding of the
family climate would add significant information to the
understanding of adolescent drinking styles.

Hypothesis #6 (school, peer and parental correlates):
It has been suggested that a thorough model of adolescent
drinking include both.drinking specific and nonspecific
influences in a number of domains--i.e., peer, school, and
family (Zucker, 1982). Family climate can be viewed as a
family of origin, nonspecific factor. In order to get a
broader view of the phenomenon, several other factors were
included. Specifically, it was hypothesized that adolescents
who were more exaggerated drinkers would be (1) less
committed to school achievement and activities, (2) more
influenced by peer modeling of and support for exaggerated
drinking behaviors, and (3) more likely to perceive their
parents as tolerant of exaggerated drinking behaviors.

Comparison of Abstainers and NOrmative Drinkers

Hypotheses 7 to 9 address the family functioning correlates, peer

versus family influences, and school, peer, and parental mediators

60

associated with normative adolescent drinking and abstinence. However,
the data on this comparison group is so lacking that it is difficult to
formulate specific hypotheses. Jessor and Jessor (1975) found the
abstainers to be more committed to conventional behavior, more involved
in traditional academic pursuits, more dependent upon their families,
and less involved in antisocial activities. According to their theory,
antisocial behavior increases during times of transition. It may be
that this group is not yet ready to make the transitions associated
with the processes of separation and individuation. They may be either
slower in their development or unable to make the transition. In the
former case, they may resemble the normative drinkers in the next few
years. In the latter case, they may be caught up in a dysfunctional
family system. The exact nature of the system is unclear.

Demone (1972) believes that abstainers are a statistically deviant
group of adolescents. Whereas pathological drinkers reject most formal
adolescent activities and adult-sanctioned standards of behavior,
abstainers overreact in the opposite direction. They tend to adopt
unquestionably adult—prescribed forms of behavior, such as academic
excellence, family dependence, and religious involvement. Moreover,
they rarely if ever engage in antisocial behaviors or other normative
behaviors considered to be rites of passage-i.e., dating or
maintaining employment during the school year.

In the area of family relationships, Demone noted that abstainers
tend to come from larger, intact families, to have good communication
with their parents, to participate with parents in discussions
regarding values and standards of behavior-inc1uding drinking-and to

feel strongly obligated to their families. Although these areas are

61
not systematic measures of the perceived quality of family environment,
they do suggest that abstainers, as compared to pathological drinkers,
may perceive their families as more cohesive, expressive, religious,
intellectual ly-cultural 1y oriented and less conflicted.

Cathiningthefamilydatawithtleschoolandpeerdata, itmaybe
that the abstainers have not begun tl'e process of adolescent
separation. Such a conclusion would be in line with Jessor and
Jessor's (1975) speculations. Detone's sample of abstainers were among
the youngest group of subjects—seventh and eighth graders. Thus, with
time, it is likely that many of them will resemble normative drinkers;
however, it is unclear what this group will be like if they remain
abstainers by the latter high school years.

Based upon such information, it was hypothesized that abstainers
would differ from normative drinkers in the ways in which they perceive
their families, but the exact relationships would be unclear. Finally,
it was hypothesized that their families would be more influential than
their peers in determining drinking behaviors, and that they would be
more committed to school achieverent and activities .

Sex as a Moderator Variable

The existence of sex differences in the drinking and driving
behaviors of adolescents has been well documented. Thus, it was
hypothesized that males and females might perceive their families
differently. Hypotheses #1, #4, and #7 were tested for sex differences

according to the method specified by Cohen and Cohen, 1975.

Subjects

The subjects for this study were high school junior and senior
licensed drivers (N=244) who were involved in a school based driver's
awareness program--Promoting Responsible YOung Drivers through
Education (PRYDE)-which is a project sponsored by the Department of
Transportation under the direction of Jehn Paul McKinney, Ph.D. The
school was located in a rural community in a midwestern state. The
entire school was tested twice a year-—fall and spring-as part of an
effort to develop and evaluate the PRYDE Program. The questions for
this study were included in the fall, 1984 administration. Subjects
were informed that their participation was strictly voluntary. See the
results section for a more detailed description of subject
characteristics.

Procedures

 

The questionnaire was administered by high school teachers during
the first.two hours of school. The students had 55 minutes or one
class period in which to complete the questionnaire. A.minimum of two
project staff persons--i.e., nonschool personnel-dwere located in the
school in order to answer questions and/or helping students discuss

their reactions to the test material. The students were instructed to

62

 

63

complete their questionnaires independently. See Appendix A for the
questionnaire.

Measures

mrsonal subject variables. Information on age, sex, grade, race,

 

 

and parental marital status was collected. The information was used to
describe characteristics of the sample.
Family climate. The instrument that was used to measure family

climate is tl'e short form of the Family Environment Scale (FES) (Moos,

 

1974) . The FES is designed to measure and describe interpersonal
relationships among family members, directions of personal growth that
are emphasized in the family, and the basic organizational structures
of the family. The instrument is comprised of ten independent scales.
The description of each scale appears in Figure 7.

The 90—item, full scale was developed by Moos (1974) from the
responses of over 1,000 individuals in 285 families. Families were
representative of a variety of religious orientations, minority group
membership, and psychiatric involverent. The 90 items were chosen on
the basis of (1) 50—50 item split in way of responding, (2) high
intrascale correlations, (3) low interscale correlations, and (4)
maximal discrimination among families. The short form is a 40-item
instrument-4 items per scale-which correlated .80-.96 with the long
form. It allows for a more rapid assessment of family functioning and
was designed partly for the purpose of group administrations.

Figure 8 displays the internal consistency coefficients obtained
by Moos on a sample of 285 families. The short form has very similar
values--not specifically reported in Moos (1974) . Finally, the

average scores for the 285 families on each of the subscales are

1.

2.

3.

10.

Figure 7.

cohesion

Expressiveness

Conflict

Independence

Achievement
Orientation

Intellectual
Cultural
Orientation

Active
Recreational
Orientation
Moral-
Religious

Emphasis

Organization

Control

64
Relationship Dimensions

The extent to which family members are
concerned and committed to the family and the
degree to which family members are helpful and
supportive of each other.

The extent to which family members are allowed
and encouraged to act openly and to express
their feelings directly.

The extent to which the open expression of
anger and aggression and generally conflictual
interactions are characteristic of the family.

Personal Growth.Dimensions

The extent to which family members are
encouraged to be assertive, self-sufficient,
to make their own decisions and to think
things out for themselves.

The extent to which different types of
activities (i.e., school and work) are cast
into an achievement oriented or competitive
framework.

The extent to which the family is concerned
about political, social, intellectual
cultural activities.

The extent to which the family participates
actively in various kinds of recreational and
sporting activities.

The extent to which the family actively
discusses and emphasizes ethical and religious
issues and values.

System Maintenance Dimensions

Measures how important order and organization
is in the family in terms of structuring the
family activities, financial planning, and
explicitness and clarity in regard to family
rules and responsibilities.

Assesses the extent to which the family is
organized in a hierardhical manner, the
rigidity of family rules and procedures and
the extent to which family members order each
other around.

Family Environment Scale Subscale Descriptions.

65

Average Eight.week

Internal Inter—scale Testrretest
consistencies correlations Reliability
Subscales (N2814) (N2814) (N247)
Cohesion .78 .58 .86
Expressiveness .71 .48 .73
Conflict .75 .56 .85
Independence .64 .45 .68
Achievement Orientation .65 .49 .74
Intellectual-Cultural
Orientation .78 .54 .82
Active Recreational
Orientation .68 .48 .77
Moral Religious Emphasis .79 .55 .80
Organization .78 .52 .76
Control .70 .51 .77

Figuge 8. Internal Consistencies, Average Item—Subscale Correlations
and TestrRetest Reliabilities for RESP-Normative sample (Moos, 1974).

66

presented in Figure 9. These norms will be useful in comparing the
sample in this study with the sample on which the instrument was
developed.

Questions of scale validity were addressed by the authors in the
following way. First, there were no consistent sex differences in
perceptions of family social environments. However, there was enough
variability among families to suggest that such differences may occur
‘within specific families. Secondly, adolescents tended to see less
emphasis on cohesion and expressiveness and somewhat more emphasis on
conflict than did parents. Adolescents also saw less independence,
intellectual-cultural orientation and moral-religious emphasis, but
somewhat more achievement orientation and active recreational
orientation. The above two findings are similar to the findings
reported by Olson and McCUbbin using a different family environment
instrument. Third, cohesion and expressiveness decreased and conflict
increased in larger families. Fourth, clinic families scored lower on
cohesion, intellectual—cultural orientation and active-recreational
orientation. They obtained higher scores on both conflict and control.
The clinic families also obtained lower scores on expressiveness, and
independence and higher scores on achievement orientation (Moos, 1974).
The last area involved an analysis of high versus low drinking
families. Although the parents not the adolescents are the drinkers, I
will discuss the findings in greater detail.

Families with parents who drank alcohol more frequently reported
less moral-religious emphasis. This was the only scale that
differentiated high versus low drinking families. Between members,

there were other differences in perception. Children in low drinking

67

Subscales (Ns285 families)
Mean S.D.
Cohesion 2.91 0.86
Expressiveness 2.25 0.82
conflict 2.15 1.09
Independence 2.76 0.77
Achievement-Orientation 2.26 0.93
Intellectual-Cultural Orientation 2.71 0.94
Active-Recreational Orientation 2.74 0.93
Moral-Religious Emphasis 1.94 1.30
Organization 2.37 1.02
Control 2.15 0.95

**Each subscale has 4 items

Figure 9. Means and Standard Deviations of FES Short For
Subscales--N0rmative sample (Moos, 1974).

68

families perceived more emphasis on cohesion, expressiveness,
achievement orientation and organization than did the children in the
high drinking families. On the other hand, these latter children
perceived more emphasis on intellectual-cultural orientation. The
mothers in high drinking families shared their children's perceptions
of their families. HOwever, the fathers perceived fewer difficulties.
The authors concluded that the results are consistent with the clinical
impression that mothers and children are affected more than the father,
who is usually doing the drinking.

In a second study, the authors discovered that families tended to
fall into different family typologies based upon their scores on the 10
areas of social environment (Moos & Moos, 1976).

Drinking behaviors. The items in this area were adapted from the

 

questions used by Jessor and Jessor in their triethnic study, and items
suggested by Zucker and N011 (personal communications, 1984) . Items
included: (1) frequency of alcohol was consumption in the past 6
months, (2) frequency of drunkenness in the past.year, (3) amount of
alcohol consumed in a 3-hour social occasion, and (4) number of
problems which have occurred as a result of drinking-trcub1 ‘with
teacher or principal, trouble with friends, trouble with police,
trouble with parents, having driven after drinking and been.criticized
by a date. These measures allowed for an evaluation of fgeggeggy_and

intensity of drinking patterns, as well as negative consequences.

 

Specifically, two indices of drinking were developed: (1) a
quantity/frequency index (a product of the standardized responses to #1
and #3 above), and (2) a problem drinking index (a summation of the

number of problem areas, including drunkenness, weighted by the

69

frequency of occurrences within each areas). These two indices were
used to analyze the data for the adolescents in the sample who have
made the transition to drinking.

A final area of interest.was the understanding of perceived family
environment differences between abstainers and nonproblem drinkers.
The variable that measured this difference could be thought of as a
transition to drinking index. In the following section a criteria for
classifying subjects into one of three drinking categories-abstainers,
normative, and excessive drinkers--is outlined. Those subjects who
fell into the abstainer category were coded 0 and those subjects who
fell into the normative drinker category were coded 1. This simple
dichotomous variable was preferable to the quantity/frequency and
problem indices for these comparisons since any variation of the latter
two variables would be due to the drinkers--abstainers always receive a
score of 0 on these variables because they do not drink. Thus, by
creating a dichotomous variable, the differences that emerged were due
to between-group rather than within-group variation.

Drivingibehaviors. Based upon the traffic safety research
reviewed herein, the following items were chosen to measure driving:
(1) number of accidents as a driver, (2) number of accidents as a
motorcyclist, (3) number of near misses as a driver, and (4) seat.belt
disregard. Responses on each of the four items were standardized and
summed to form one "risky driving" index.

School, peer, and parental correlates. In order to get a more
complete picture of adolescent drinking and risky driving, three sets
of variables were included in the study: school, peer, and parental

mediators. The specific items included: (1) grade point average, (2)

70

academic expectations, (3) school truancy, (4) peer support of
adolescent drinking, (5) peer support for risky driving, (6) peer
modeling of drinking, (7) peer involvement in automobile accidents, (8)
peer involvement in moving violations, (9) perceived parental support
of adolescent drinking, and (10) perceived parental support of
adolescent risky driving.
Classification of Subjects

Drinking categories. The adolescent drinking literature suggests
that abstainers and problem or excessive drinkers may be qualitatively
different from normative drinkers. Thus, this study compared
abstainers to normative drinkers and excessive/problem drinkers to
normative drinkers. In order to accomplish this task, definitions of
normative and problem/excessive drinking were developed.

cabalan, Cisin and Crossley (1969) argued for the need to use
several definitions of "problem drinking“ when conducting alcohol
related research. Jessor and Jessor (1975) defined drinking in terms
of drunkenness in the past year, plus problems associated with
drinking. Specifically, they classified individuals as problem
drinkers if they reported either (1) drunkenness at least 6 times in
the past year, or (2) negative consequences two or more times in the
past year in at least three out of five areas. Using this definition,
29.3% of 7481 moderate drinkers were classified as "problem drinkers."
The authors noted the arbitrary nature of this, or any, definition.
Consequently, they developed two other definitions based on (1)
frequency of drunkenness, or (2) prdblems associated with.drinking.
The second definition of problem drinking was drunkenness twice a month

or more in the past year. Definition III was based on the negative

71

consequences of drinking: having experienced, over the past year,
drinking-related negative consequences at lease twice in any single
area, with at least one experience in an additional area.
Examining personality, perceived social environment, and
behavioral variables, they found that with all three definitions,
problem drinkers coild be distinguished from nonproblem drinkers.
Moreover, the pattern of findings remained essentially invariant over
the three different definitions. The findings demonstrate the utility
of considering both frequency and problem indices of drinking in a
definition.
Zucker (1980) defined a continuum of drinking groups from
abstainers to heavy drinkers (see Figure 5). In his work, be combined
this criteria with an alcohol-related prdblem index in order to
identify problem drinkers.
Based on the above literature, three levels of drinking were
differentiated in the present study:
(1) Abstainers-~nondrinkers
(2) ‘NOrmative Drinkers-(1) infrequent, light, and moderate
drinkers (Zucker, 1980) with no more than four occurances of
one or more alcohol related problems, and noimore than ten
occurrences of drunkenness in the past year (Jessor & Jessor,
1972)

(3) Excessive Drinkers-(1) moderate heavy or heavy drinkers
(Zucker, 1973; 1975), or (2) one or more occurances of
alcohol-related problems in each of the six areas, or two or

more prOblems in at least three out of the six areas, or 10

72

or more problems in a single area combined with at least one
problem in a second area.
Because the focus of the study is on the entire spectrum of drinking,
and not on identifying the extreme:group, the definition is somewhat

less restrictive than either the Jessor et a1. or Zucker definitions.

RESULTS

Personal Subject Characteristics of the Sample

Descriptive statistics were obtained for all of the subjects on
the following variables: age, grade, sex, race, and parental marital
status. The results appear in Table 1a. The majority of subjects are
16 (38%) and 17 (55%) years of age and are enrolled in their junior and
senior years of high school. The sample contains slightly more males
(10%) than females, with 97% of the subjects being Caucasian. Finally,
approximately 25% of the adolescents come from families in which the
parents are currently divorced.

In order to get an estimate of what percentage of the sample could
be considered "at risk," a classification scheme was developed. A
categorical variable, driver status, was developed from a combination
of the number of accidents and the number of close call. At-risk
drivers included those subjects who were involved in at least.one
driving accident and/or who reported a near miss at least once per
month. Table 1b presents the results of a chi-square analysis of the
difference between male and female subjects within each driving
classification, controlling for age. No significant differences were
found (p>.05). An average of 35% of the subjects fell within the risky

driving category.

73

74
Table 1a

Descriptive Information on Sample Characteristics

 

 

Characteristic Mean Mode Group Percentages (%)
Age 16.70 17.00 16 yrs. (38)
17 yrs. (55)
18 yrs. (07)
20 yrs. (00)
Grade --- 12.00 11th (44)
12th (56)
Sex ——- Males Males (55)
Females (45)
Race —-- Caucasian Caucasian (97)
Hispanic (01)
Black (00)
Native American (02)
Asian (00)
Parents ---- Married thrried (76)
Marital Divorced (24)

Status

 

75

 

mo. . m 58. u
063 numxm 98:3,...

 

8. Am ...a H fie.
.no. u chosen who

mo. A m ...E H fins
.86 u «page... do

 

 

 

 

 

 

mm.mm wm.eo wo.ov mw.>m wo.he am.vm
H m x mm m 3 @836 baa
mm.mo wm.mm wo.om mm.mo wo.mm wm.mm
m m on we mm on mum>euo ommm
moneom onmz mofimeom moan: mofimeom mofimz
me u own now be u own now me u own new

08. 8: 83:8

no... @520ch

son 9.: 35:8

 

 

woe mom ooHHonucoo .xmm an mmsouo.m:e>euo new no mwmwamc< cocoon ecu

0: 63mg

76

Drinking status was determined by theoretical classification
schemes suggested by Zucker (1980) and Jessor et a1. (1974).
Definitions of four categories of drinkers-i.e., abstainers, norma-
tive-light, normative—moderate , and exaggerated—were based upon two
criteria--quantity/frequency of drinking and problems associated with
drinking. Table 1c presents the results of a chi square analysis of
the drinking groups by sex and age. controlling for age, the
percentage of males and females in each drinking category did not
significantly differ (p.> .05). Table 1d demonstrates that the
percentage of drinkers within each category (collapsed over age and
sex) is not significantly different from the percentages found by
Zucker and Hartford (1983) for 16- to 18-year-olds, with the exception
that the females in this sample tend to be heavier drinkers than the
females in the Zucker and Hartford sample. In all subsequent analyses,
the light-normative and moderate-normative categories will be combined
into one normative drinking category.

Family Environment, Drinking, and Driving Scale Reliabilities

Because the relationship scales are composite scores-~e.g., the
sum of four items per scale--the internal consistency values for each
scale were computed. The results of this analysis are presented in
Table 2, and indicate that cohesion, conflict, culture, religion,
organization, and control are measured with adequate reliability;
however, the internal consistency values for expressiveness,
individuation, achievement orientation and recreation reflect low
reliabilities. Thus, the ability of these latter scales to account for

a significant amount of the variance among adolescents on the drinking

77

 

me. A m..ee m cue:
.~H~.m u mumsam ecu

mexmeemfin. me.xm.eemfie3
.Hee.m u unease ecu .eee.m u mumeem one

 

 

 

 

 

wees em.em em.em ee.em ee.em ee.e~
m m mm mm as me eoumumammxm
we ee.e ee.me em.me em.¢ e~.me
e H e on m e ounnmeoz
we me.e me.em ee.ee ee.~e ee.em
e H es me em he semen
we e~.ee ee.e~ ee.e~ ee.m~ ee.em
e e no em as ea emcemumna
mmHmeom mon: moneom moan: mofimeom moan:
me u mom

Hoe eceeeouucou

 

 

om< MOM ocfiaouucou .xom an monouo mcwxceuo ecu mo womaamce phenom ecu

on menus

78

 

 

emem whee emem 3.2 Eflfizufieh .m
em.emue.ee e~.meue.e em.emum.em ee.mmue.em Hugues .a
"emcee new» eeueee whomenem sea .HHH
eeeem {me $03. emeém €885,135 .m
ee.-nm.ee em.meue.e em.emum.em em.mmne.e~ segues .<
“$30 Home mated modmeom .HH
eeeem £03 Seem Seem bugging .m
em.~mue.em ee.eene.ee ee.mmneum~ ee.emue.em umxose .a
"Ameeo Hams wanes. mmem: .H
ooumuooomenm Bmumerlmfiugoz unmefilgeumfioz mumfimumoe 53m

 

 

mange see: 8085

 

ocm noxoom 93 cu :Omeumfimeou «Inbommumo mfixfiuo comm cote: muooflosm mo mommucooumm

pa manna.

79
Table 2
Internal Consistency Scores for Scaled variables on Measures of Family

Environment, Driving. Drinking and Peer Support

 

Scale Alpha

 

Family'Environment:

Cbhesion .64
Expressiveness .38
Conflict .70
Individuaticn .14
Achievement orientation .18
Culture .41
Recreation .27
Religion .60
Organization .52
control .41
Driving:

Risky Driving Index .38
Drinking:

Problem Drinking Index .69
Quantity/Frequency Index .72

Peer support for:

Drinking .30
Driving .40

 

80

and driving measures is attenuated. Any relationships that emerge on
these scales are likely to be very rdbust.

Among the drinking and driving dependent measures, the problem
drinking index and the quantity/frequency index were measured*with
considerable accuracy, while the peer support scales and the risky’
driving index.were measured with moderate reliability.

Testing the gypothesized Curvilinear Relationships

Hypotheses #1, #4, and #7 specify that families which function on
the extremes of cohesion, expressiveness, control, and organization
will have adolescent members who are more likely to act out in the
areas of drinking and driving. Such a relationship is nonlinear and
hence inappropriate for linear regression analysis. Hewever, Cbhen
(1978) provides a cogent argument for the ability of linear regression
analysis to determine if nonlinear relationships exist. By stepping
the linear and curvilinear forms of a variable, in that order, into a
hierarchical analysis, one can determine if there is a significant
curvilinear relationship after the linear relationship has been
accounted for.

In the present study, two relationship scales-cohesion and
expressiveness-and two maintenance scales-organization and
control--are posited to relate in a curvilinear fashion to adolescent
drinking and risky driving. In order to test these hypotheses, fOur
hierarchical analyses were conducted for each dependent variable,
stepping in the linear form of the family environment variables first
(Y=bX1 + a) and then the curvilinear form (YébXI + le*X1 + a). The
null hypotheses for the partial, semipartial, beta and B weights all

test whether there is a unique, significant curvilinear relationship.

81

Table 3 presents the significance tests for the partial correlations of
the curvilinear forms of the relationship and the maintenance scales
with the drinking and driving measures. Alpha has been set at .1 so
that curvilinear relationships are more readily identified.

The results of this analysis indicate that significant curvilinear
relationships exist for cohesion with three of the dependent
measures—quantity/ frequency index (2 < .05) , transition to drinking
index (2 < .10), and risky driving index (p_< .01), as well as a
significant linear relationship with the risky driving index
(L< .001) . Similarly, organization has a significant curvilinear
relationship with two of the drinking measures—quantity/ frequency
index (p_< .05) and transition to drinking index (25 .10)-and
expressiveness has a significant curvilinear relationship with one of
the measures—quantity/frequency index (p_< .01) . Finally, control is
not significantly related to drinking or driving in its curvilinear
form (2.) .05) .

Figure 10 further illustrates the curvilinear phenomenon.

Although the relationship between cohesion and the risky driving index
is presented, the schematic representation of the curvilinear
relationship between any of the other family environment variables and
the dependent measures would be similar. A significant curvilinear
relationship suggests that at both low and high values of variable A,
in this case cohesion, the values of variable B--risky driving—remain
relatively constant at either a high or a low level. At moderate
values of variable A, the value of variable B is in the reverse
direction of its value at the extremes of A. Further analyses would be

needed to determine whether the actual function is av or a I \

82
Table 3
Correlations of the Curvilinear Association between Cohesion,
Expressiveness, control & Organization and Drinking and Driving

After Partialling Out the Linear Associations

 

 

 

Partial Prdb. of
variables Correlations F Sig. linear R
I. Licensed drivers:
A. Risky driving index
Expressiveness -.03 .26 -- .05
Control .08 1.39 -- --
Organization -.04 .46 -- .01

II. NOrmative and problem drinkers:

A. Quantity/frequency index

Cohesion -.17 5.13 .05 --
Expressiveness -.19 6.39 .01 --
control .01 .33 -- --
Organization -.15 3.93 .05 --
B. Problem drinking index
Cohesion -.06 .73 -- --
Expressiveness -.08 1.01 -- ---
Control .07 .74 -- --
Organization -.03 .16 --— --

III. Abstainers and nonrproblem drinkers:

A. Transition to drinking index

Cohesion -.13 2.91 .09 --
Expressiveness .02 .75 -- --
Control -.00 .75 -- .05

Organization -.13 2.87 .09 --

83

Risky

Driving
Status

 

Nbrmative

 

 

Low High

cohesion

Figure 10. Illustration of a significant curvilinear relationship
between cohesion and risky driving.

84

shape. Thus, as illustrated in Figure 10, a significant curvilinear
relationship between cohesion and risky driving indicates that the
precise relationship will be either a K. ,I (a) or an inverted m (b)
function.

In the following analyses, the appropriate form‘of the family
environment variables will be selected for each dependent variable.
The specified curvilinear form of the significant family environment
scales is a U—shaped function. Thus, subjects scoring on the
extremes--i.e., scale values of 0 and 4-dwill be put into one group
(coded as 1) and subjects scoring in the middle range--i.e., scale
values of 1 to 3-dwill be put into a second group (coded as 0).

Overall Relationship Between Family Environment and Adolescent
Drinking and Risky Driving

Hypotheses #1, #4, and #7 require ten separate regression
equations for each dependent variable-three drinking variables and one
driving variable. The resultant number of regression equations would
equal 40. ‘With alpha set at .05, it.wculd be expected that two of
these equations would yield a significant outcome strictly by chance.
In order to reduce the probability of type I error, it has been
suggested that.when the independent and dependent variables are
multiple measures of the same or similar constructs--i.e., are
significantly intercorrelated-a conservative first step*wculd be a
multiple regression analysis, or a canonical analysis, depending on the
number of dependent variables. Ruth.such an analysis, one would
determine whether any relationship exists between the two sets of

variables. If a relationship does exist, the exact nature of the

85

relationship can then be teased out by conducting further analyses on
each dependent variable individually (Cohen S. Cohen, 1975) .

Table 4 presents the intercorrelations for the ten areas of family
functioning. Table 5 presents the intercorrelations of the three
measures of drinking and the one measure of risky driving. It should
be noted that the drinking measures are based upon different samples.
The quantity/frequency and problem drinking indices are based on data
from subjects who have made the transition to drinking, while the
transition to drinking index is based upon data from abstainers and
normative drinkers. Finally, the risky driving index is based upon
data from all subjects-licensed drivers.

It is evident that the family enviromnent dinensions-r_ela:

tionships: cohesion, expressiveness, and conflict; growth:

 

achievement-orientation, individuation, intel lectual-cultural

orientation, recreation, and religion; maintenance: organization and

 

control--are significantly intercorrelated (p < .05-.001). Similarly,
the two drinking measures--quantity/ frequency and problem drinking
indices--the one transition to drinking measure, and the one risky
driving measure are significantly intercorrelated (pf .05) . Thus, the
first step in analyzing the relationships between family envirorment
and adolescent drinking and risky driving-hypotheses #1, #4, and
#7—wi11 be a multiple regression. Since this is the first step in an
exploratory analysis, alpha will be set at .1 so that overall
relationships are more readily identified.

The multiple regression equations for the family envirorment

scales and the four dependent variables are presented in Tables 6-8.

 

*a

86
Table 4

Intercorrelations of Family Enviromnent Scales

 

Cohesion (1)
Cohsq (2) -.09

Express (3) .59 -.15

*** **

Expsq (4) -.49 .30 -.46

*** *** ***

*

*** * ***

*** *** *** *** ***

*** *** *** *** *** ***

Recreat (8) .67 -.20 .56 -.45 .28 .54 .52

*** *** *** *** *** *** ***

Religion (9) .53 -.15 .35 -.33 .25 .47 .38 .54

*** ** *** *** *** *** *** ***

Culture (10) .62 -.19 .52 -.39 .20 .49 .45 .64 .51

*** ** *** *** *** *** *** *** ***

Orgarliz (11) 074 -021 050 -050 013 060 046 064 .56 059

*** *** *** *** * *** *** *** *** ***

Orgsq (12) -028 038 -029 027 -028 -035 -047 -033 -018 -031 -020

*** *** *** *** *** *** *** *** ** *** **

COHtIOl (13) .37 -026 .19 -031 052 043 040 .39 038 028 040 -025

*** *** *** *** *** *** *** *** *** *** ** ***

 

 

F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12

*** (B < .001).
** (p_ < .01 ).
* (p < .05 )

87
Table 5

Intercorrelations of the Drinkingyand Driving Measures

 

Risky driving (D1)

Quantity/ (132) .57
frequency ***

Problem (D3) .78 .41
drinking *** ***

Transition (D4) .08 .40 .38
to drinking -- *** ***

 

 

D1 D2 D3 D4

*** (E < .001).
** (P. < .01 ).
* (E < .05 ).

88

 

ooo.

”63ch
33:00
:oflumuwcmmuo
mag 16
coflmem
coaumwuomm
"bogged/HUGH
ucQ=w>mEU<
mmwcgwwmwumxm
vmsoo

av.m mmN\HH omo. mma. 5mm. scammcou

855% Ema

 

883353

.m 393.5 mm mm counspé menswm m m mamauflg 833.53,

 

 

awe/ED mwmam cam mwdmum ungoufiém 32mm 93 mo naming conmmummm EMBED:

o OHQMB

89

 

 

 

boa. msmfinoum
mmm. awue\:mso
m mHQMflum> amoflcocmo
th.I vmmxm
mmm.| omeoo
mmo.u aoamaflmm
com. mummuomm
mos. 6H>HacH
ham. m>waau<
3?: $30
mva. muficmmso
moo. Houucoo
H mfinmaum> Hmoacocmo
mam. m mm.m mm. mm. mo. N
omo. ma mm.~m mm. mm. ma. H
@353 cofimamuuoo
mocmoauaamflm mo mumsvmnaeu m xfies Hmoacocmo msHm>cmmam uwgssz

 

 

woman—mm: @3qu 05 can mmHmom DCQEOHH>5 338m can now mflmfig Hmoacccmu

N. wanna

90

 

Houucoo

whose

musuaco
882$
coaumouomm
cOwumsca>accH
acmem>wflco<
Dowfimcoo
mmmcm>wwmwumxm

2:. SA 823 m8. «8. 3m. 8:8
853% 8. 83355.

 

868383 a 2865 an «m 8832 88.8. m m 6333: 833.3,

 

 

xwocH mcaxcflpo cu cofiuamcmua can can ucwecoufi>cm wwwemm mo wflmxamcm cowmwmummm mamwuasz

m manna

 

(n

m

m

1.5!:

91
The results of the analysis for the risky driving index demonstrate an
overall relationship between these two sets of variables (R:.36,
ELF .01). Similarly, the transition to drinking index is also
significantly related to the set of ten family environment dimensions,
although the overall relationship is not as strong (R:.30,Jp_< .1).

Table 7 presents the canonical equation for the ten relationship
scales and the two measures of adolescent drinking. The results
indicate that a significant first.order association‘exists (p_< .05).
After the first order relationship is accounted for, there is no
significant linear relationship between the two sets of variables
(Wilks Lambda, p_> .05). Thus, hypotheses will be analyzed
individually for each of the ten family environment areas.

After determining scale reliabilities, appropriate scale
measurements-i.e., linear versus curvilinear--and the overall
relationship between the independent and dependent sets of variables,
we can turn our attention to the examination of the specific
hypotheses.

Hypothesis #1 (Family Environment and Risky Driving)

Regression analyses were used to test the hypothesis that family
environment is related to adolescent risky driving behaviors. The
previous analysis found that the prediction of curvilinearity emerged
for cohesion, but not for expressiveness, organization, or control.
Thus, in this set of regression analyses, the linear forms of all ten
FES scales will be utilized, as well as the curvilinear form«of
cohesion.

The Pearson product moment correlations in Table 9 indicate that

compared to safe drivers, risky drivers perceive their families as

Table 9

Correlations between the Family Environment Scales and Risky;Driving

Cohesion

Cohsq

Expressiveness

Conflict

Achievement
Orientation

Individuaticn

Recreational
Orientation

Religion

Intellectual-
Cultural

Organization

Control

*** (P. < .001).
** (B < .01).
* (P. < .05).

 

92

-026

***

-ol4

**

-015

*‘k

-002

—.10

-013

-.21

***

-.21

***

—.26

***

-023

***

-012

 

 

th

..m we

enmmm

 

 

 

Or.

93

moderately cohesive (p’< .01); however, there is also a significant
linear trend (2 < .01) such that within the moderate range, the riskier
drivers perceive their families as less cohesive. Moreover, the risky
drivers perceive their families as less involved in four areas of
individual growth-religion (p < .001) , individuation (p < .05) ,
recreation (p < .001) , and intellectual-cultural involvement (2 <
.001)--1ess organized UE.< .001), less controlling (p’< .05) and less
expressive (p_< .01).

Table 10 presents the findings from the analysis of sex as a
moderator variable. The literature suggests that the driving profile
of males differs from the driving profile of females. If this pattern
is true, then the interaction of sex and the family environment scales
should account for significant.variance above and beyond the main
effects of sex and family relationships. HOwever, the hierarchical
regression analysis does not support the hypothesis that sex is a
moderator variable. The partial correlations of the interaction
variables were not significant at the .05 level. Thus, sex will not be
utilized in understanding the association between family environment
and adolescent driving in any further analyses.

Hypothesis #4 (Family Environment and Drinking)

It was hypothesized that families at the extremes of cohesion,
expressiveness, organization and control, high in conflict, and IOW'in
the five areas of individual growth would have adolescent members who
consume more alcohol and.who have more problems associated with their
drinking. Among the drinkers, it was found that the predictions of
curvilinearity were substantiated for cohesion, expressiveness, and

organization on the quantity/frequency index; however, none of the

94

Table 10

Correlations of the Sex x Family Environment Interactions With Driving

After the Main Effects Have Been Partialled Out

 

 

Partial Partial
variables correlations F Sig.

 

1. Licensed Drivers:

A. Risky Driving Index

Organize -.05 .63
Control .02 .10

Sex * Cohesion .00 .22 --
Sex * Cohsq -.01 .30 --
Sex * Express -.01 .26 --
Sex * Conflict -.01 .46 --
Sex * Achieve -.06 .88 --
Sex * Individ .00 .28 --
Sex * Recreate .02 .14 -—-
Sex * Religion .04 .42 --
Sex * Culture -.01 .20 --

* .....—

* —_

 

95

curvilinear relationships emerged on the problem drinking index. Table
11 presents the matrix of the correlations between the ten family
enviroment scales and the two measures of drinking. The results
indicate that adolescents who consume more alcohol on a frequent basis
perceive their families as moderately cohesive (p < .01) , moderately
expressive (p < .01), more conflicted (p < .05), more individuated (p <
.01) , and more recreationally oriented (p < .01) . Further, adolescents
who report having more problems with their drinking habits perceive
their families as less involved in intellectual-cultural activities (p
< .05) and less organized (p < .05).

The analyses of sex as a moderator variable are presented in Table
12. The interaction of sex and the relationship scales of conflict and
control are significant for the problem drinking index (2 < .05) .
Specifically, perceived control and conflict are unrelated to drinking
problems among males, and positively related to drinking problems among
females.

Hypothesis #7 (Family Environment and the Transition to Drinking)

Of the four hypothesized curvilinear relationships between family
environment and drinking status, only cohesion and organization proved
to be significant. Thus, in the following regression equations, the
curvilinear forms of these two scales and the linear fonts of the other
eight scales will be utilized. The product moment regression
correlations between family enviroment and the transition to drinking
index are presented in Table 13.

Compared to adolescents who have made the transition to drinking,
abstainers perceive their families as lying on the extremes of cohesion

(p < .01) and organization (p < .05), less conflicted (p < .001), less

Table 11
Correlations Between the

Measures

Cohesion

Cbhsq

Expressiveness

EXpsq

Conflict

Achievement
Orientation

Individuaticn
Recreational
Orientation
Religion
Intellectual-
Cultural
Organization

Organizesq

Control

*** (p_ < .001). **

 

96

Family Environment Scales and the Drinking

 

(E < .01). *

.— -008
-.20 -"
** ....-
-- -.03
-020 -..
** -—_
013 -001
* ..—
.09 .07
.19 -.02
** ...-
.17 .04
** —-—
'_011 -009
-001 -013
.... *
—- -012
....— *
-011 -.-
005 -002
Quantity/ Problem
Frequency Drinking
Index Index

(2 < .05) .

97

Table 12
correlations of the Sex x Family Environment Interactions With Drinking

After the Main Effects Have Been Partialled Out

 

 

Partial
variables Correlations F Sig.
I. NOrmative and prdblem.drinkers
A. Quantity/Frequency Index

* Cohsq —.08 1.00 --

* Expsq .08 1.11 --

* Conflict .07 .79 --

Sex * Achieve -.07 .79 --

Sex * Individ -.06 .53 --

Sex * Recreate -.05 .42 --

Sex * Religion -.02 .65 --

Sex * Culture —.05 .39 --

Sex * Orgsq -.03 .16 --

Sex * Control .02 .51 --

B. Problem Drinking Index

Sex * Cohesion .09 1.35 --

Sex * Express .01 .34 --

Sex * Conflict .16 4.58 .034'

Sex * Achieve .11 2.06 --

Sex * Individ .14 3.41 --

* Recreate .05 .40 --

* Religion .04 .25 --

* Culture .05 .46 --

* Organize .01 .48 --

* Control .17 5.10 .025

 

98

Table 13

Correlations Between the Family Environment Scales and the Transition

 

to Drinking
Cohsq .13
*
Expressiveness .07
Conflict .23
*'k*
Achievement .16
Orientation *
Individuaticn .11
*
Recreational .08
Orientation --
Religion .01
Intellectual- .07
Cultural --
Organizesq -.12
control .16
*

 

 

Transition to

Drinking Index

*** (p < .001).
** (p < .01).
* (p < .05).

 

H

 

In

99
achievement oriented (p,< .001), less individuated (p_< .05), and less
controlling (p < .01).

Table 14 presents the results of the analyses of the family
environment by sex interactions. Only one scale emerged.as significant
QB < .05). Specifically, achievement orientation was unrelated to the
drinking status of females (2 > .05) while it was positively associated
with the drinking status of males (2 < .05) . Males who have made the
transition to normative drinking perceive their families as more
achievement oriented.

Relative Relationship of Peer and Familnynvironment Influences
Hypothesis #2 (Peer versus Family Environment-—Driving)

It.was predicted that peer variables would be significantly
related to the driving measure, but that an understanding of family
environment would account for a significant percent of variance after
peer variables had been accounted for. Three items were chosen to
measure peer influences: (1) the amount of perceived support
adolescents receive from their peers for risk taking while driving, (2)
the number of peers who have been involved in traffic accidents, and
(3) the number of peers who have received a moving violation.

In order to test this hypothesis a mixed hierarChical-stepwise
regression equation was computed. The first three peer variables were
entered as a set. As predicted, they accounted for a significant
amount of variance for the risky driving measure (Rs.31,¢p_< .001).

The second step of the analysis was to add the remaining ten family
environment scales into the equation. Since there were no prior
hypotheses regarding the ordering of the scales, they were entered in a

stepwise fashion. If a family environment variable accounted for at

100

Table 14
Correlations of the Sex x FamilyiEnvironment Interactions With

Transition to Drinking After the Main Effects Have Been Partialled Out

 

Partial

variables Correlations F Sig.

 

I. Abstainers and non-problem drinkers

A. Transition to Drinking Index

Sex
Sex
Sex

*tfififiifl-fifl'}

Cohsq
Express
Conflict
Achieve
Individ
Recreate
Religion
Culture
Orgsq
Control

.13
—.00
-.04
-.15
-.03
-.08
-.09
-.07

.11
-.09

2.74
.32
.21

3.95
.15

1.03

1.52
.74

2.10

1.15

o
O
b
(n

 

101
least .01% of the variance and the total regression equation remained
significant, the scale was retained in the equation. The results of
this analysis are presented in Table 15.

The findings support the general hypothesis that family
environment contributes significant understanding to the phenomenon of
adolescent risky driving. With the entire sample, three out of the ten
scales were significantly related to risky driving. Specifically,
risky driving was associated with less cohesiveness within the moderate
range and less intel lectual-cultural involvement (R increased from .31
to .44, p < .001).

During the stepwise regression part of the analysis, the ten
family environment scales were ordered in a way which was determined
strictly by mathematical computations. Shach a procedure was
appropriate given that there were no hypotheses regarding a theoretical
ordering; however,chance factors were nuch more influential in
determining the final rank order. Thus, it was decided to conduct a
cross validation analysis in order to determine to what degree the
ordering is due to chance variations . After the mixed
hierarchical-stepwise analysis was performed on all the subjects' data,
the sample was randomly divided into two groups and the regression
equations were repeated on each group. The results of these analyses
appear in the middle and lower portions of Table 15.

The regression equations for the two cross validation samples
support the conclusion that family environment is significantly related
to risky driving, even after peer influences have been accounted for.
In both samples, at least five FES scales were significantly related to

risky driving. The results from the cross validation analysis also

102

Amosfiuooo oHnfi.

 

ooo. oo.o oHH\o HNH.u ooo. ooo. ooo. on. uoHHucoo o
ooo. oo.o HHH\o ooo.: MHo. moo. omo. oom. muomuoom o
ooo. Ho.m ~HH\o oH~.: ooo. ooH. ooo. ooo. oH>HocH m
ooo. oo.m oHH\o ooo.- oHo. ooH. moo. moo. m>mHaua o
ooo. m~.o oHH\m ooo.: oHo. HoH. mHo. ooo. muouHoo o
ooo. oH.o mHH\o ooo.: ooo. moH. ooo. ooo. :onmaoo o
ooo. mo.o oHH\o oHo. ooo. ooo. HHH. ooo. uoquu Hmooz
ooo. ooo. moo. How. uuomosm Hmooz
moo. ooo. ooo. moo. uncoooo Home H
”Ho goHuocHHom> ooouo .HH
ooo. oo.o oo~\o HoH.u oHo. ooH. ooH. ooo. omaoo o
ooo. oo.oH oo~\m ooo.: mHo. HoH. ooH. ooo. mHouHoo o
ooo. mo.HH oo~\v Hoo.n moo. ooH. ooH. ooo. :onmapo o
ooo. ~o.o oom\o ooo. ooH. ooo. ooo. «Ho. meoHu Hmooz
mom. omo. moo. mmm. unmoHoom Home:
ooo. ooo. ooo. ooo. uncooom ammo H
“mHmeom mHHucm .H
.m m N m m
.mHm 2890 on 29on 85.6 mm Bygone 988m o deHnHoz mHnoHHms doom

 

 

Mom coucoooo< comm o>mm mcHHmooz com Duommom Moon Houmm mca>fiuo can

 

Domecoufl>cm mHoemm mo mcoHumHoHuoo may mcacoemxm .mommfimcm COmmmonom omozmoumlHMOmcoumuowm ooxflz <

ma ofinma

103

 

 

ooo. mv.v vHH\m ooo.: oao. «ma. mmm. mwv. om>mocH o
ooo. mm.v mHH\> NHH.I mHo. HmH. own. one. Houucoo m
ooo. mm.m mHH\o hmm.| mHo. moH. mHm. mow. ousuHoo v
ooo. mn.m haa\m chH.I omo. on. mmH. mvv. m>mfinu< m
ooo. mm.v mHH\v vhm.l hmo. MHH. NeH. ohm. omeoo m
ooo. mm.m maaxm mom. avo. moo. moo. Ham. umxoau Home:
mmH. mmo. eve. cam. ucooooom Hwooz
ooo. ooo. ooo. ooo. homoso Howe H
"No coflumoflg moouo .HHH
m m mm m
.ofiw 3890 mo 39on @986 mm coughed onoom m 333:: mHanHSw doom

 

104

demonstrate the operation of chance factors in determining which scales
appear in the equation, and in what order they appear. Thus, one trust
be cautious in over-interpreting these results. The only scale which
appeared in all three equations was intellectual-cultural involveuent.
(he or both forms of cohesion also energed. Thus, it is likely that
these two family dimensions have the strongest association with
adolescent driving with this sample of adolescents. Beyond this
statement, all that can be said with any degree of confidence is that
family environrrent adds significant information to the understanding of
adolescent risky driving after peer influences have been accounted for.
Hypothesis #5 (Peer Versus Family Ehviroment--Drinking)

Similar to hypothesis #2 , it was anticipated that (1) peer support
for adolescent drinking, and (2) modeling of drinking behaviors would
account for a significant amount of the variance associated with
adolescent drinking. No precise ordering of the family enviroment
scales was predicted. Thus, tre sane mixed hierardlical-stepwise
equation was employed, and the sane criterion for retaining a scale was
used. Cross validation analyses were again conducted.

Tables 16 to 18 present the analyses for the two drinking
variables: (1) quantity/frequency index, and (2) problem drinking
index. The hypothesis that peer influences would be significantly
related to adolescent problem drinking, with family environment
variables contributing additional neaningful information was confirmed
(R increased from .24 to .39, p_ < .01) . Three family enviroment
scales were related to problem drinking with the full sample and the
two cross validation samples. Specifically, greater problems

associated with drinking were correlated with higher achievement

105

 

 

ooo. ooH\o ooo.: oHo. mHH. HmH. ooo. oH>HocH o
ooo. ooH\m ooo. ooo. oHH. ooH. moo. muomuooo m
Hoo. ooH\m ooH.: Hmo. ooo. oHH. ooo. moHamogo o
ooo. HoH\v ooo. mHo. moo. ooo. HHo. m>mHau< o
moo. moH\m omH.: moo. ooo. ooo. ooo. muouHoo m
ooo. ooH\o ooH. ooo. ooo. omo. ooo. ocHHmooe Homo
Hoo. ooo. ooo. Hoo. unomooo Home H
.xwocH oconHuoosmHnoHoo ”mHoEoo mHHuom .HH
ooo. oo.m ooH\o HoH. oHo. ooH. ooH. ooo. oH>HocH o
ooo. m~.o ooH\o oHo.: oHo. mHH. omH. ooo. muouHoo o
ooo. ~m.o ooH\o ooH. ooo. ooH. ooH. ooo. moomuomm m
Hoo. Ho.o ooH\m ooH.: ooo. ooo. oHH. moo. ooHoHHmo o
ooo. oo.o HoH\o ooo.: oHo. ooo. moo. Hoo. omoxm o
ooo. oH.o moH\m ooH.: ooo. Hmo. ooo. ooo. vogue o
ooo. oo.~ ooH\~ ooH. oHo. oHo. ooo. ooH. ocHHmooe Homo
HoH. oHo. oHo. HmH. uponaom Homo H
.xoocH mocmoooum\muwucmoov umfloemm muflucm .H
.m M mm m
.on 2890 on $9on 885 mm cough? 386m o mHoHHHoz 63393 imam

 

 

Hoo ooucoooua comm m>ommmcHHmooz oco buommom Homo Hanna mcchHHo oco

 

Homecoufl>cm wHoemm mo mcoflumHoHuoo ecu moocflemxm .mmmmHmcm COammonom owfizmoumlamowooumuwmm ooxflz <

ma ofinme

106

 

 

HHo. oo.o oo\o ooo.: oHo. ooH. oHo. ooo. moHooopo o
ooo. oo.o oo\o Hoo. HHo. ooH. ooo. omo. momumxm o
ooo. oo.o oo\o ooo. Hoo. ooH. ooH. ooo. moomuumm m
ooo. oo.o Ho\m ooo.: ooo. ooH. ooH. ooo. oH>oocH o
oHo. Ho.o oo\o ooo. ooo. HoH. ooH. ooo. mowonuo o
oHo. mo.o oo\m ooo.- ooo. ooo. oHH. moo. :oHomaoo o
oHo. oo.o oo\o ooo. Hoo. ooo. ooo. oHo. oconoos ammo
ooo. moo. moo. ooo. uncooso Homo H
Axooco mcoxcouo eooooumv "Ho ecouoooom>ommouo .HH
mHo. oo.o oo\o ooH.- ooo. moH. ooo. moo. omeoo o
HHo. oo.o oo\o ooH. oHo. ooH. oHo. moo. ostau< o
ooo. mo.o oo\o ooo. ooo. ooH. ooo. mmo. muouHso o
ooo. oo.o oo\o ooo. oHo. ooH. ooH. moo. HoHocoo m
ooo. om.m Ho\m ooo.: ooo. ooH. HoH. moo. ommxm o
ooo. oo.o oo\o ooo.- moo. ooH. HmH. ooo. :oHoonm o
moo. oo.o oo\o Hoo. ooo. ooo. ooH. moo. muomuoom o
oHo. oH.H oo\o ooH. oHo. ooo. ooo. moH. ocHHmooe Homo
HoH. oHo. oHo. HoH. uncooom Homo H
AxoocH oocoowooo\ouoncoooo "Ho :oHuooon> ooono .H
o o oo o
.oHo 2395 on mHoeHo 866 o o Emoofi “Wazoo o mHmDHoz 280e,. mono

 

 

How coucoooo< zoom o>om mcoooooz com Duommom noon oopm<mmcoxcona coo

 

Dcoecouo>cm mooeom mo mGOHDmHoHHou on» mcocoeoxm .womemcm QOommoomom omozmoumlHMOHoouoooom ooxoz fl

NH oanme

107

 

 

oHo. mo.o Ho\o oHo. oHo. oHH. ooH. ooo. m>mHau< o
oHo. oo.o oo\o ooo. ooo. oHH. ooH. Hoo. ooHHocoo m
ooo. mo.o oo\m moo.u ooo. oHH. ooH. oHo. muopHoo o
HHo. oo.o oo\o oHo. Hoo. HoH. HoH. ooo. mommuomo o
oHo. oo.o mo\o ooH.: ooo. ooo. ooH. ooo. ooocoogo o
ooo. mo.o mo\o mHo. ooo. ooo. ooo. moo. ocHHoooe Homo
oHH. oHo. oHo. oHH. uuommso Homo H
ixwocH ocHxaHuo ewHoooo. uoo ooHuooHHm> ooooo .HH
Hmo. mo.H oo\oH ooH.: ooo. ooo. ooo. ooo. omoxm o
ooo. mo.o moo ooo.: HHo. ooH. ooH. ooo. H895 o
moo. Ho.o oo\o omo. mHo. ooo. HoH. moo. muomuomm o
moo. oo.o Ho\o ooo.- oHo. ooo. ooH. ooo. muouHoo o
ooo. oo.o oo\o mHH.- oHo. Hoo. omH. Hoo. Houucoo m
ooo. om.o oo\m ooH. ooo. ooo. ooH. ooo. oH>HocH o
ooo. oo.o oo\o ooH.: ooo. ooo. HHH. ooo. coooHHmm o
ooo. oo.o mo\o ooo.: Hmo. ooo. Hoo. ooo. omnoo o
ooo. oo.H oo\o ooo. ooo. ooo. ooo. HoH. ooHHmooe Homo
omH. ooo. ooo. omH. uuomoom Homo H
Axoooo mocooooHM\>uoucooOo "N* c0oumooom> moouo .H
...m mm mm m
.on 2.890 on 39.8 866 oo 88:on 89$ m 333:: 282$ mono

 

 

How coucooood zoom o>mz mcoooooz coo unommom Hoom Hound mcoxcono ocm

 

ucoecouo>cm mHoeom mo mCOopoHouuoo ooo.mchoeoxm .mommoocm aeommowmom omosmoumIHMOHLUHmuoom ooxoz <

ma oHnoB

108

orientation, greater participation in recreational activities, and less
family organization.

The pattern of relationships between peer/ family influences and
alcohol consumption is different from that pattern found with alcohol
problems. Alcohol consumption was not significantly related to peer
influences. In this case, family environnent influences appear to be
nore important than peer influences . Five cannon family environnent
scales entered into the full sample regression equation and the two
cross validation equations. Increased consumption was significant
related to moderate cohesion and expressiveness, less religious
orientation, greater participation in recreational activities, and less
intellectual-cultural involvenent.

The aforenentioned cautions regarding the family environnent
scales which appeared in these regression equations, as well as the
ordering of the scales, apply to the results in this section.
Hypothesis #8 (Peer Versus Family Environrrent--Transition to Drinking)

Specific hypotheses regarding the relationship between peer/ family
variables and the transition to drinker status were limited to the
speculation that peers may play a less significant role in determining
adolescent drinking behaviors and attitudes than family enviroment.
The lack of more specific hypotheses led to the utilization of analyses
which were exploratory in nature.

In the subsequent analyses, the sane procedures were fol lowed as
outlined in hypotheses #2 and #5 above. The peer variables include:
(1) peer support for drinking, and (2) peer modeling of drinking
behaviors.

109

Huth.the full sample (Table 19), peer influences account for a
significant amount of variance in predicting abstainer/normative
drinker status (R:.301,'p.< .01). However, the data in the middle and
lower portions of Table 19 demonstrate that in one cross validation
equation, peer influences are not significantly related to the
dependent variable (Rs.308,'p_> .05) whereas in the other cross
validation sample, the relationship is significant (R=.312, p < .05) .
The differences in the magnitude of R.between the three samples are
small and the associated statistical significance varies as a function
of the degrees of freedom—dwhich are larger in the full sample--and
random error-dwhich is greater in one cross validation sample than the
other. Hence, it can be concluded that peer influences are probably
marginally associated with.drinking status.

After peer influences have been accounted for, family environment
accounted for a significant amount of variance for the full sample and
for one of the cross validation samples. Out of the 2 to 3 scales that
stepped into the two significant equations, conflict.was the only scale
that entered into both equations. It appears that normative drinkers
perceive their families as more conflicted than do abstainers.

School, Peer, and Parental Correlates

In order to gain a more complete picture of the phenomenon.of
risky driving during adolescence, three additional areas were examined:
school commitment, peer influences, and perceived parental attitudes
towards adolescent driving behaviors. The results of a regression
analysis of adolescent risky driving, alcohol consumption, problem
drinking, and transition to drinking with these three areas appear in

Table 20, in the form of Pearson product moment correlations.

110

 

 

omo. mo.o moo\o ooo. ooo. ooo. ooo. ooo. ouoooso m
ooH. oH.o ooH\m. ooo... oHo. omH. ooH. ooo. 88ng o
ooo. oo.m moo\o ooo. moo. omo. moo. mom. mmoooxm m
ooo. mo.o ooo\m moo. omo. moo. omo. Nmm. aboomcoo N
omo. oo.o ooo\~ oom. ooo. ooo. ooo. mom. mcoooooe noon
ooH. Hoo. Hoo. ooH. floooso Homo H
”No ecouoooom> moouu .ooo
moo. oo.m ooo\m moo. omo. ooo. omo. omo. oumouoom o
ooo. oo.o moo\o mom. moo. omo. moo. ~oo. o>oono¢ m
moo. no.0 ooo\m omo. omo. omo. ooo. mom. uooomooo m
moo. mo.o ooo\~ mmm. moo. moo. moo. oom. mcoooooe noon
omo. mmo. mmo. omm. uuoomom noon o
uoo acouoooom> moonu .oo
omo. oo.m ooo\m ooo.: moo. moo. mmo. omm. :Oomooom o
moo. o~.o moo\o omo. ooo. omo. moo. oom. o>oooo< m
ooo. mo.m ooo\m omo. ooo. mmo. omo. oom. uooomcoo o
ooo. oo.o ooo\m ooo. omo. ooo. omo. oom. ocoooooe Hoom
moo. omo. ooo. moo. uncooom ooom o
"oooemm oooucm .o
m m mm m
.oom oomoooo mm 393m omega mm ooowomoa ooooom m oomouoo: oonmooo> mono

 

 

mom ooucoooom zoom o>mmxmcoooooz ocm uuommom Hoom uoumm xooco mcoxcooo cu economcmoa one ocm

 

ucoecooo>cm mooeom mo oncoumooouou one moocoeoxm .momhomcm COommoomom omozmoumloMOonoomooom ooxoz.<

mo oonma

lll

Hypothesis #3 (Correlates of Adolescent Risky Driving)

 

Adolescents who perceive themselves to be riskier drivers report
lower grade point averages (p < .001) , greater absenteeism (p < .01) ,
and fewer academic expectations (2 < .05) . In relationship to their
peers, they see themselves as having more friends who have received
noving violations (2 < .001) and/or been involved in traffic accidents
(2 < .001) . Further, more of their friends are frequent drinkers (p <
.01) . Surprisingly, they do not perceive their friends as encouraging
risk-taking while driving, or excessive drinking, to a greater extent
than do their safety- conscious counterparts. Finally, riskier drivers
believe that their parents are more tolerant of adolescent drinking (2
< .05) and risky driving (2 < .05).

Typothesis #6 (Correlates of Adolescent Drinking)

 

The correlations that appear in Table 20 under the
quantity/frequency and problem drinking indices are based on the data
from subjects who have already made the transition to drinking. The
results indicate that heavier drinkers report lower grade point
averages (p < .001); however, they do not differ from their lighter
drinking counterparts on absenteeism or academic expectations. In
terms of their peer group, heavier drinkers perceive their friends to
be more supportive of adolescent drinking (2 < .05) and risky driving
(2 < .05) . However, in terms of the number of friends who drink,
heavier drinkers do not differ from lighter drinkers (p > .05) . The
friends of the heavy drinkers appear to exhibit less caution in their
driving practices, as is evident from the fact that they have been

cited for more moving violations (2 < .05) . Finally, heavier drinkers

112

Table 20

Correlations Between School, Peer, and Parent Measures and the Drinking

 

and Driving Measures

 

Peer support .06 .13 .22 .20
for drinking -- * ** **
Peer support .09 .16 .25 .10
for driving -—- * *** --
Friends with .30 .15 .15 .15
moving violations *** * * *
Friends who .24 .10 .09 .11
have accidents *** -- —- --
Friends who .15 .12 .10 .27
drink ** .... _—_ ***
Parent approval .11 .14 .26 .18
of drinking * * *** **
Parent approval .13 .17 .24 .08
of driving a- * mu ....
Grade point -.32 -.24 -.32 .02
average *** *** *** -——
Absenteeism .17 .06 .31 .03
** .... *** ..—
expectations * -- -- --

 

 

Risky Quantity/ Problem Transition

Driving Frequency Drinking to Drinking
Index Index Index Index

*** (p'< .001).
** (p_ < .01).
* (p < .05).

113

believe that their parents are more tolerant of adolescent drinking (2
< .05) and risky driving (2 < .05).

The results for the problem drinking index are similar to the
results for the quantity/ frequency index with the exception that the
magnitudes of the correlations are generally greater. Problem
drinkers, in comparison to non-problem drinkers, report lower grade
point averages (p < .001) and greater absenteeism (p < .001) but no
difference in their academic expectations. Thus, they may not have
given up all academic pursuits, but their behavior indicates that
excellence in this area may be difficult or unrealistic. Similar to
heavy drinkers, problem drinkers report receiving more peer (p < .01)
and parental (p < .001) support for adolescent drinking and risky
driving. Moreover, problem drinkers report lmowing more friends who
drink (2 < .05) and who have received moving violations (2 < .05) .
These facts might indicate that they either choose or sonehow find
themselves involved in a social climate which supports adolescent
drinking.

Typothesis #9 (Correlates of the Transition to Drinking)

The correlations for this neasure are based upon the data from
abstainers and non-problem drinkers. The results presented in Table 20
indicate that compared to normative drinkers, abstainers perceive
differences in the amount of peer support that they receive for not
drinking and in their parents' beliefs regarding drinking, but not in
their degree of school commitnent. Specifically, grade point average,
absenteeism, and academic expectations were not related to drinking
status (p > .05) . However, non-drinking adolescents perceive their

friends and parents to be less supportive of adolescent drinking (2 <

114

.01) . Moreover, abstainers report having fewer friends who drink (p <
.001) and fewer friends who have been cited for a noving violation (p <
.05) .

A final point regarding the data presented in Table 20 deserves
nention. The magnitude of the correlations between the drinking and
driving neasures, and the school , peer and parent neasures differ.
Problem drinking has the strongest relationship with the school
neasures, followed by problem driving; moreover, problem drinking is
also strongly related to the peer and parental support measures of both
drinking and driving. Problem drivers appear to have more friends who
would also be classified as problem drivers and drinkers. Thus, the
teenagers who are having problems with their drinking and/or driving
might be thought of as less committed to traditional or adult
prescribed rites of passage-such as academic excel lence-and more
committed to demonstrating their independence in the areas of drinking
and driving. Despite the fact that their behavior in these areas tends
to take on an exaggerated, perhaps even antisocial quality, they still
perceive their friends and parents as supporting their efforts to
assert their adulthood through these two channels.

Compared to non-problem drinkers, abstainers receive less support
for and have fewer drinking role models. Despite the fact that they do
not differ from normative drinkers on academic neasures, they are able
to maintain better grade point averages than adolescents who drink more
alcohol on a frequent basis. Thus, both abstainers and light,
non-problem drinking adolescents appear to be equally conmitted to
school; however, as alcohol consumption increases or as problems

associated with drinking increases, conmitnent to school decreases.

DISCUSSION

The findings from this study lend support to the general premise
that perceived family environment is related to the transition to
drinking and the management of drinking and driving practices among
adolescents. However, the findings also suggest some modifications of
the specific hypotheses regarding what constitutes a "functional"
family system during the adolescent phase of development. It is
evident that not all ten aspects of family environment, as measured by
the Family Environment Scale, are equally important in their
association with the target behaviors, nor do they manifest identical
patterns of relationships to the two target areas. In the following
chapter, the relationships between family environment and adolescent
risky driving, alcohol consumption, problematic drinking, and the tran-
sition to drinking will be discussed, integrating the findings from
this study with the existing literature in each respective area. Next,
the question of how adolescent driving practices are linked to the .
transition to drinking and the management of alcohol consumption‘will
be addressed by examining the empirical evidence from this study and
other researCh and comparing the findings to existing theories. It was
the premise of this study that adolescent drinking and driving are
linked because of their status as rites of passage in American society.
They may also share other commonalities which present theoretician with

the task of specifying and integrating the associations between them.

115

116
A third area which deserves mention is the methodological questions and
problems which arose during the execution of this study. Finally,
directions for future work in the area of family environment correlates
of adolescent drinking and risky driving‘will be suggested.

Adolescent risky driving. Initially, it was hypothesized that
cohesion, expressiveness, organization, and control would be related to
risky driving in a curvilinear fashion, such that adolescents who
perceived their families as excessively high or IOW'On these qualities
would be more likely to engage in risky driving practices. In fact,
the curvilinear relationship emerged for only one scale, cohesion, and
the form of the relationship was in the opposite direction from what
was expected. That is, riskier drivers perceived their families as
moderately cohesive. However, the presence of a significant linear
relationship suggests that adolescents at or above the median on
cohesion report a lower incidence of risk taking during driving, while
adolescents below the median present a mixed picture. ‘With less
perceived family cohesion, some of these adolescents engage in more
risk taking, while at the lowest level of cohesion, many adolescents
report few incidents of risk taking. These findings present two
questions. First, why did the curvilinear relationship emerge on only
one scale? Second, why was the relationship in the opposite direction
from what.was predicted?

The basis for positing the aforementioned curvilinear
relationships rests on the assumption that these four family
environment scales are similar to the cohesion—adaptability constructs
proposed by Olson and Mccubbin (1983). The stated definitions of the

scales appear to be similar; however, their psychometric properties may

117

be different enough to render them separate but related neasurenents of
family functioning. The exception to this conclusion appears to be
cohesion since it.was curvilinearly related to three out of the four
dependent measures.

.A second possibility centers around the scope of the circumplex
model. Olson and McCUbbin's curvilinear model of family functioning is
a broad based model which was designed to relate to other criteria of
healthy family functioning. In this study, the model is being applied
to individual behaviors which are not inherent qualities of family
functioning. Thus, the concept of a balanced family system may not be
associated with individual behavior in the same way in which it is
associated with other measures of family functioning; the associations
are likely to vary in magnitude and direction. Such a pattern emerged
with the findings from this study. The issue of the connection.between
system—level and individual level variables will be discussed in more
detail later.

The direction of the association between cohesion and risky
driving suggests a modification of the initial hypothesis. One of the
main advantages of moderate family cohesion during the adolescent phase
of development is that it allows the adolescent member some distance to
try out new roles and to engage in behaviors which the family might not
approve of. Thus, it was hypothesized that moderate cohesion would be
associated with "normal" amounts of experimentation in the area.of
driving, and that exaggerated driving practices would be associated
with.either high or low levels of cohesion. HOwever, it appears that
risk taking, whether normal or exaggerated, is associated with a

moderate level of cohesion. Perhaps such an environment is necessary

118

to engage in transitional behaviors, but is not sufficient to account
for the differences be’oween normal and exaggerated risk taking while
learning to drive.

The existence of a significant linear relationship suggests that
sone distinction can be made between nornal and risky drivers with
regard to their perceived family cohesion. Within the moderate range,
adolescents who perceive their families as less cohesive also report
more risk taking. Perhaps at very high and low levels of cohesion,
they lack the support , encouragenent , and motivation to begin the
processes of separation and individuation. On the other hand, at
noderate levels they are able to engage actively in these processes
although they still require more, rather than less family support in
order to manage the tasks successfully. A study by Hotch (1979) lends
support for this interpretation.

Hotch examined the style of hone-leaving among a group of older
adolescents. She found that an agent or active style was associated
with moderate levels of family relatedness, whereas a patient or
passive style was associated with both high and low levels of
relatedness . Moreover, agentic adolescents reported higher career
aspirations , anticipation for success, and personal motivation. Thus,
Hotch concluded that anong this group, the challenges involved in the
transition from a dependent to an independent status were viewed as a
challenge which could be mastered, and that the family served as a
support structure for the adolescents. However, in families with high
or low relatedness, the hone environnent was either too attractive or

too lax to instill in the adolescents with the motivation to leave.

119

If sone of the tasks and challenges inherent in this transition
include the acquisition and managenent of adult-like behaviors, such as
drinking and driving, then Hotch's interpretations fit nicely with the
findings from this study. The difference lies in the ability to
distinguish between at-risk drinkers and drivers. Along with
variations on the dinension of cohesion, several of the other family
enviroment scales suggest possible discriminators between at-risk and
normative drivers .

The hypothesis that exaggerated risk taking would be associated
with fewer avenues for individual growth was largely supported. Risky
drivers report less of an emphasis on religion, recreation, and
intellectual-cultural orientation; moreover, they report feeling less
individuated from their families. Perhaps then, as Ronnel (1959)
concluded, reckless driving is used as a means to achieve self-worth
and adult status. Support for this conclusion cones from the finding
that risky drivers report greater school failure, and are thereby
isolated from a socially approved form of individual recognition.

The literature further suggests that peer pressure and a need to
impress one's peers may be prine contributors to the risks that sone
adolescents take while learning to drive. Indeed, this study found
that risky drivers report having more friends who might be classified
as risky drivers. However, even after the effects of peer influence
have been accounted for, perceived family enviroment still plays a
major role. Although similar findings have been reported in the
literature (Harrington, 1971; Sobel & Underhill, 1976), these studies
do not give adequate attention to the impact of family influences as

they try to explain the phenonenon of adolescent risky driving.

120

Of interest in the present study is the lack of a significant
relationship between family conflict and risky driving. The
predominant finding in the driving literature which includes family
variables is that adolescents report high levels of family anomie and
discord. This apparent discrepancy may be explained by the manner in
which conflict was neasured in this study. All four questions refer to
frequent verbally or physically expressed anger. However, the conflict
in the families of at-risk drivers within this sample may be largely
nonverbal, since these adolescents perceive their families as less
expressive than the families of safer drivers. It may be that the
conflict the adolescents experience is expressed thrwgh their driving
behaviors or is acted out in sone other way. Alternatively,
risk-taking adolescents perceive their parents as more tolerant of
adolescent drinking and risky driving perhaps, leading to less conflict
over their behavior.

A final point regarding the findings on risk taking while driving
is the energence of a pattern of what will be terned "family
disregard." Riskier drivers report perceiving their families as less
organized, less controlling, and less involved in intellectual-cultural
activities (a neasure of societal connectedness or conservatism).
Within such an unstructured enviroment, it would seem that the
standards for appropriate behavior would be difficult to teach and
enforce. With the simultaneous presence of friends who are engaged in
risky driving practices, it is likely that these adolescents have both
the influence and the opportunity to becone antisocial in their driving
practices. The antisocial quality of repeat traffic violators has been

noted by many writers (Ievonian, 1969; Pelz & Schuman, 1971; Goldstein,

121

1972). Moreover, Beamish and Malfetti (1962) noted the co-occurance of
adolescent rebelliousness and a lack of involvement of parents in
connunity and societal affairs among a groip of repeat traffic
offenders. Since cohesionr-in its linear formr-and intellectual-
cultural orientation have the strongest associationS'with.risky
driving, it might be concluded that.giving the adolescent a feeling of
connectedness and responsibility to others, whether it be family or
society, is an important way to promote safer driving practices.
Adolescent alcohol consumption. In explaining the findings
regarding adolescent drinking practices, it is useful to take a closer
look at the differences between the two indices. Prdblem drinking
refers to conflicts that the adolescent has experienced with
significant others-family, school personnel, peers, and law
enforcement authorities-as a direct result of his/her drinking. The
only reference to quantity of alcohol consumption is a general measure
of frequency of drunkenness. On the other hand, the quantity/frequency
index is a pure measure of alcohol consumption, without any '
consideration for the effects that the consumption may have on other
aspects of the adolescent's life. Given these two measures, it is
possible that two individuals can report equal alcohol consumption, yet
experience very different degrees of associated problems. Hence, in
the context of defining exaggerated drinking as a form of risk or
problem potential, the problemldrinking index is a muCh more pure
measure of this concept. The quantity/frequency index may include
connotations of prdblems, especially for the heavier drinkers, but the
implication is at best tentative; this index can be conservatively

thought of as a measure of the range of normative drinking, with the

122

extrene representing a risk potential groip. With this distinction in
mind, the results from this study provide sone neaningful information
regarding the family environnent correlates of adolescent drinking.

The expression of the need to achieve adult status through
drinking is most accurately represented as a spectrum of possible
outcones ranging from abstinence on one extrene, to problematic
drinking on the other. Both extrenes are statistically deviant groups
with the majority of the spectrum consisting of adolescents who engage
in normative drinking practices. To appreciate the extrenes, one mist
understand the normative process. Thus, our discussion will begin with
an examination of the findings regarding alcohol consumption.

Contrary to expectations , adolescents who consune greater amounts
of alcohol more frequently perceive their families as moderately
cohesive and expressive . The curvilinear relationship was
insignificant for the system maintenance scales of organization and
control. However, these results are in the sane direction as the
relationships found for the risky driving index. Employing the sane
explanation, perhaps the adolescents who are most agentic in expressing
their need to individuate are the ones who are motivated to explore
actively new adult-like behaviors. Hotch (1979) found this style of
hone-leaving to be associated with noderate degrees of family
relatedness, a construct which is probably closely related to the
cohesion and expressiveness variables in this study. In their quest
for independence agentic adolescents are likely to over-indulge during
this transitional period. Fron the work of Jessor and Jessor, we might
also expect that they will be engaged in other forms of norm-violating
behavior as well (1972) .

123

The findings of moderate degrees of expressiveness and greater
conflict in families in which adolescents take risks in drinking and
driving suggest that these adolescents are actively engaging their
families in the process of separation, and that this process involves
some degree of turmoil. Olson and McCubbin's results support the
conclusion that these characteristics are present in normal,
well-functioning families. The adolescents who report limited and
infrequent alcohol consumption perceive less conflict in their families
and high or low levels of expressiveness. 'Within highly expressive
families, the parents may be clearly and forcefully transmitting their
values against alcohol resulting in compliance on the part of the adol-
escents and less conflict. Sudh a process may also indicate a slower
rate of adolescent separation. Within less expressive families, the
conflict may be curtailed due to the lack of emphasis on verbally
expressing feelings. In either case, teenagers who consume less
alcohol perceive their parents and friends as less approving of
adolescent alcohol consumption, although the groups do not differ in
the number of friends who drink.

Further support for the notion that the adolescents who are
actively experimenting with alcohol are also more agentic in their
style of home-leaving comes from the finding that they perceive
themselves as more highly individuated. They view their families to be
tolerant of the individuality of each member and flexible enough to
allow'personal growth pursuits. Moreover, these adolescents probably
have more opportunity to drink since their families encourage

recreational activities.

124

Only one area suggests that individuals who are heavier drinkers
may be at risk for problems. Specifically, heavier drinkers report
lower grade point averages, although their rate of truancies and their
academic expectations do not.vary from their lighter drinking
counterparts. Since the group of heavier drinkers also contains a
larger number of problem drinkers, it may be that the finding of lower
grade point average is largely due to the difficulties experienced by
this latter group. In comparing the correlations between academic
failure and each of the drinking indices, the larger correlation
between academic failure and problem drinking (prdblem drinking index =
-.32; quantity/frequency index = -.24) suggests that this conclusion
has some merit.

The group of adolescents above the median on the
quantity/frequency index probably contains a fair number of adolescents
WhO‘Will experience difficulty‘with.their drinking practices, and the
aforementioned family representation provides no specifications for
distinguishing between these two groups of drinkers. For such an
understanding, we must turn to an examination of the problem drinking
index.

Adolescents who report problems with their drinking manifest
perceived family environnents that are characterized by sone of the
qualities of parental disregard. Similar to riskier drivers,
problematic drinkers report less organization and less
intellectual/cultural orientation. combined with greater peer and
parental approval of exaggerated drinking and risky driving, there
would appear to be a paucity of internal and external controls against

the deviant or antisocial management of alcohol.

125

Viewing an admission of problematic drinking as an index of
potential difficulty is further supported by the school experiences of
adolescents in this groip. They report the lowest grade point averages
and the highest truancy rates. However, their academic expectations do
not differ from their nonproblematic drinking counterparts. The future
may present sone harsh realities for these adolescents as they realize
that their academic expectations will be difficult if not impossible to
attain given their current academic performance.

The conflict that was predicted to exist in the families of
problematic drinkers proved to be a significant discriminator only
among females. Moreover, females who report problems with their
drinking perceive their families to be more controlling, yet less
organized. Within such an environnent, the presence of demands with an
accompanying lack of structure might lead to greater ambivalence on the
part of the adolescent, and hence, greater conflict. The difference
between the sexes might also be related to the donle standards in our
society. Independence and careless behavior are tolerated , perhaps
V even encouraged—"boys will be boys"--among males to a greater extent
than they are among females. The lack of difference in the drinking
patterns between females and males—a finding which has been
contradicted by other research (Zucker & Hartford, 1980)-conbined with
the double standard suggests that females may be more susceptible to
criticism for their behavior.

Finally, it is of interest that none of the affective qualities of
the family environnent proved to differentiate this groip of drinkers
from the group of non-problematic drinkers. Nor did four of the

personal growth scales energe as significant. Both normative and

126

problematic drinkers appear to perceive their families as possessing
environmental qualities which.would encourage separation and
individuationr-i.e., cohesiveness, expressiveness, moderate conflict,
and an opportunity to engage in individual growth pursuits. HOwever,
problematic drinkers report an absence of some essential socializing
elements. Perhaps the family‘s greatest contribution in influencing
adolescents' drinking practices lies in the areas of transmission of
values (regarding all forms of normrviolating behaviors) and
organization.

The transition to drinking3 An alternative approach to addressing
the issue of family environment correlates of adolescent drinking is to
examine the perceptions of abstainers and normative drinkers. Jessor
and Jessor's (1975) work on the transition from nonrdrinker to drinker
status has already been reviewed. For the purposes of this discussion,
it is important to keep in mind their conclusion that drinking is a
normal developmental phenomenon. Hence, abstainers are a statistically
deviant group. According to the theory proposed herein, these
adolescents have not begun to engage in the process of separation from
family ties. If this is so, then they are likely to differ in their
perceptions of their family enviroment. The findings from this study
suggest that such differences do exist.

Rather than viewing their families as moderately cohesive, a
quality which has been associated with normal family functioning during
the adolescent phase of development, abstainers perceive their families
as possessing either lOW'Or high cohesion. Perhaps the home is either
too rewarding or too frustrating to allow'themnto focus their energies

on separation issues.

127

Evidence for the hypothesis that abstainers are less involved in
the processes of separation and individuation comes from the
adolescents' perceptions of several other dimensions of family
functioning. First, this group reports that there is less conflict in
their families than the conflict which is reported in the families of
their normative drinking counterparts. Since a higher level of family
conflict is a normal part of this phase of the life cycle, the lack of
stress might indicate a tranquility'which precedes the onset of
separation stressors. Second, abstainers believe that their families
put less of an emphasis on achievement strivings. This finding appears
to be largely due to the priority that normative drinking males put on
achievement. Here again, the cultural gender biases emerge; by
adolescence, males are expected to accept their adult mission.of
financial and occupational success, as measured by the FES achievement
orientation scale. These pressures are not as strong for females and
younger males. In this case, the difference between the males may have
more to do with.deve10pmental rather than chronological maturity.
Perceived individuation further differentiates between abstainers and
normative drinkers, with the normative drinkers reporting higher levels
of individuation. Finally, the abstainers perceived their families to
be less controlling. In the presence of less conflict and fewer
challenges to the central value structure of the family, parents can
perhaps afford to be less controlling.

The values and models presented to abstaining youth tend to
de-emphasize drinking. Both parents and peers are less approving of
adolescent drinking; moreover, abstainers find themselves involved in a

peer network where abstinence is the norm. The research evidence would

128

suggest that over the next few'years, the majority of this group‘will
enter into the normative drinking group. It.would be interesting to
note whether those who do not make the transition experience greater
difficulty in exiting the family.

The negative relationships between drinking and school per-
fbrmance, noted with both the quantity/frequency and problemldrinking
indices, do not appear to play a part in separating normative and
abstaining youth. Both.grcups report good sdhool performance, high
connitnent, and high achievenent expectations.

Overlap between drinking and driving status. The pattern of
findings for the drinking measures are different enough from the
pattern for driving measure to call into question the feasibility of
viewing the two behavioral areas as outcones of similar family
dynamics. Moreover, only 52% of the exaggerated drinkers are risky
drivers, although the majority (75%) of abstainers are safe drivers.
The correlation.between the two status groups is .25 (p_< .001)
indicating that they may share common etiologies; however, the search
for explanatory factors should go’beyond an examination of family
dynamics.

Both the drinking and driving literature indicate that individual
and social factors --i.e., antisocial behavior, school failure, and a
desire to take risks-ma 'be related to an exaggerated expression of
drinking and driving (Zucker, 1971; Zucker, 1975; Jessor & Jessor,
1971; Goldstein, 1971; Harrington, 1971; Levonian, 1969). In this
study, risky driving and prdblematic drinking were found to have
several similarities. In terms of family environment, a pattern of

relative family disregard emerged, resulting in less of a social

129

conscience and fewer internal and extenal restraints. Further,
adolescents in these groups perceived greater peer support or modeling
fer engaging in these behaviors. Finally, the acquisition of risky
behaviors was associated with.greater school isolation and failure.

Jessor and Jessor (1985) examined the behaviors associated with
this transitional phase in great detail. They found a cluster of
behaviors which seemed to be a manifestation of a general antisocial
tendency in a group of their adolescents. For example, adolescents who
were problem drinkers were also more likely to get into fights with
peers, parents and teachers, to be involved in illegal activities, and
to have experimented with other narcotics. Although risky driving was
not one of the behaviors that they examined, the findings from this
study and other research suggest that it may be an expression of
antisocial tendencies in a circumscribed group of adolescents.

It is clear that the phenomena of adolescent drinking andidriving
are complex, involving some degree of overlap. It is likely that an
explanation of the overlap would require several constructs, perhaps
including the ideas of rites of passage and antisocial behavior.

Methodological considerations. Several methodological issues
developed during the course of this study. First, it is clear that the
kinds of interpretations one can make regarding drinking or driving
will vary depending on how these phenomena are measured. At this time,
it is not clear how alcohol consumption relates to problem drinking,
or, for that matter, how problem drinking is to be defined. Thus, the
best measures we have are comprised of multiple items which render

themselves amenable to a summation index. Analyzed individually, the

130

results from the separate items are likely to be different and
difficult to integrate. .

A second methodological consideration revolves around the power of
studies to detect relationships between distal factors-such as family
functioningb-and criterion variables. This issue appears to be related
to the question of level of measurement and has been addressed in some
areas of psychology, such as organizational behavior (Schneider, 1983).
At the heart of the issue is the determination of how'different levels
of a system interact, how'each level can be conceptualized and
measured, and how the various levels can be connected to one another.
Studies in applied psychology, and in the areas of drinking and driving
in particular, tend to show stronger relationships between behavioral
predictor and behavioral criterion variables. For example, in this
study, peer modeling of moving violations, a behavior, was the best
predictor of adolescent risky driving, also a composite of behaviors.
The question becomes one of associating group level variables, such as
family environment-also referred to as distal variables-dwith
individual level variables, such as behavior.

The procedures for measuring a construct that is part of a dynamic
process present many complex problems. FUrthermore, if a behavior
derived from a construct is not directly observable, the error
associated with its measurement increases. In conducting family
research, it is difficult to obtain significant results because the
hypothesized family constructs tend to be dynamic and not directly
observable-referred to frequently as distal constructs. Hence it is

difficult to operationalize and measure the constructs. The greater

131

measurement error is naturally associated with a general reduction in
the power of the study.

Beyond the issue of intralevel measurement is the issue of
defining connections between links. The associations between
interlevel constructs will tend to be weaker than the links between
intralevel constructs. Again, the key to increasing the power of the
study to identify the interlevel links is to identifying the major
connections and to measure them with as much accuracy as is possible.

Once a problem within a systems framework has been
operationalized, it is necessary to make the generalizations appro-
priate to the levels that are being measured. In this study, family
environment was defined in terms of the adolescent's perceptions.
Hence, the generalizations need to speak to the individual's
reconstruction of system level phenomena. In order to obtain a more
objective view of the family environment, family interactions would
have needed to be coded by an observer. Alternatively, individual
perceptions could have been pooled across several family members,
yielding a systems-level perception rather than an individual
perception.

From the aforementioned discussion, it is clear that the kinds of
obstacles involved in doing family research are numerous, and the
relationships which do emerge can be considered robust-given the
general lack of power in these kinds of studies. However, the results
from this study should be interpreted and generalized with caution
since it is intended to be explanatory in nature and it is based on

individual perceptions of a dynamic process involving many interacting

132

constructs. Further studies utilizing sophisticated methodologies are
needed.

A related methodological issue concerns specific features which
affected the power of this study. The internal consistency scores of
the scaled items were IOW'tO moderate, reducing the probability of
detecting actual relationships. It may be that the internal
consistencies are lower than expected because of the nature of the
testing situation. These questions were included in the second
administration of a series of questionnaires designed to develop and
evaluate the effectiveness of a drivers education program. Some
students expressed resentment with the length of the questionnaire, as
well as concern that their responses would not be kept confidential.
These factors probably contributed to the lowered scale reliabilities,
as well as other systematic error. The result was a reduction in
statistical power. However, an examination of the adjusted R
statistics suggests that the shrinkage due to error is not great.
Thus, the results that.did appear are probably fairly robust. If the
sources of measurement error had been reduced, more associations might
have been identified, yielding a more complete and accurate picture of
the phenomena addressed in this study.

Finally, the ability of this study to detect sex differences was
also reduced due to the lowered statistical power and the number of
variables compared to the total number of subjects. The psychological
literature is replete with studies which report sex differences based
upon separate analyses of the data for males and females. SuCh a
method is fallacious given that (1) the error terms are not

independent, (2) the number of subjects differ for each analysis,

133

yielding different degrees of freedom for the tests of significance,
and (3) the value of the error terms vary. If the conclusions made
from interpreting the results Obtained by analyzing the data separately
fer each sex go beyond genderespecific descriptions to a comparison of
the two profiles, the conclusions should be considered suspect. In
order to make direct comparisons, an analytic technique which includes
a sex by predictor interaction term should be employed. In this study,
such an analysis was conducted; however, after the main effects were
entered into the regression equations, the degrees of freedom
associated with the interaction.terms were probably not great enough to
detect the differences which might exist. Since many studies in the
area of adolescent drinking and driving have failed to consider these
obstacles, the number of sex differences which are reported may be
inflated.

Future directions. Based upon the results of this study, a number
of suggestions can be made regarding future work in the area. First,
the overall power of the research can be enhanced by being aware of
systems issues in designing and conducting research; utilizing
multivariate, correlational data analytic techniques; sampling from
large, representative populations; and limiting the number of variables
relative to the sample size.

Secondly, the theory presented herein considered that aspects of
ego functioning may be related to the links between family system
characteristics and individual behavior. It may be helpful to treasure
ego functioning-i.e., Marcia's four statuses of identity

resolution-directly.

134

Third, it would be useful to explore the relationship between
normal and deviant expressions of adolescent drinking and driving. An
example of such a model has been discussed in the work of Jessor and
Jessor (1985).

Finally, longitudinal or cross sequential research is needed in
order to assess directly the course that different adolescents and
their families follow as they struggle with the processes of separation
and individuation. Moreover, such research could relate changes in the
adolescent's drinking and driving habits, as well as other behaviors
considered to be rites of passage, to changes in the family's ability
to cepe with the fluctuating needs of their independence-bound
adolescents. The suggestion of such a link, at this point, is merely

tentative.

 

This study sought to examine the influences of the family social
climate on the normal adolescent processes of separation and
individuation. It was hypothesized that adolescents' management of
alcohol consumption and driving practices, two areas of behavior
considered rites of passage in American society, would be influenced by
the manner in which they were separating from their families of origin
and consolidating their sense of individual identity. Thus, an
examination of drinking and driving practices, together with
information on their performance and involvement in school (another
source of socially approved status for adolescents) would have
implications for the amount of turmoil adolescents experience as they
attempt to master an important developmental task: establishment of a
sense of self that integrates childhood/family experiences and adult
role expectations. Furthermore, perceptions of family
environment--e.g., relationships, personal growth.options, and system
maintenance features-dwere hypothesized to influence these processes by
eitherequipping or failing to equip adolescents with.assential ego
resources and coping skills.

These hypotheses were examined utilizing a questionnaire
methodology with 244 high school junior and senior licensed drivers who

were involved in a school-based driver's awareness program--Prcmoting

135

136

Responsible Young Drivers through Education. Several variables were
examined: (1) family climate—Family Environment Scale (Moos, 1974),
(2) quantity/frequency drinking index, (3) problem drinking index, (4)
transition to drinking index, (5) risky driving index-e.g., accidents,
moving violations, and close calls, (6) school comitment—-e.g., GPA,
truancy , academic expectations , and (7) social environment inf lu-
ences-e.g. , peer and parental support for and modeling of exaggerated
drinking and driving practices.

The results were detailed, providing some support for the
hypotheses , and suggesting further modifications . Integrating the
specific results yielded several important findings. First, family
environment variables were indeed related to adolescent drinking and
driving practices, even after accounting for peer influences.
Developmental 1y, initiation into drinking occurs prior to initiation
into driving (Zucker, personal communications). Thus, the fact that
more family variables were related to driving practices than to
drinking practices suggested that family influences may be most
apparent during the period of transition, and then become less obvious
as other factors-e.g. , peer influences and societal norms-~increase in
salience.

A second important finding was the role that family environment
played in differentiating "normal experimentation" from exaggerated and
problematic involvement. Risky drivers and problematic drinkers
perceived their families to be less organized and less connected with
the community. Additionally, risky drivers viewed their families as
less control ling. This constellation of factors was termed ”familial

disregard" and indicated that these youngsters may be ill-equipped to

137

manage appropriately the transitional demands of adolescence.
Further, it was speculated that their supply of ego resources may be
attenuated by a history of failure to provide adequately for personal
needs .

Related to the issue of differentiating between normative and
exaggerated drinking and driving practices was the finding that social
and academic factors changed as one moved up the continuum from
abstinence to drinking, or safe driving to risky driving.

Specifically, as the evidence for ineffective management of drinking
and driving increased, status derived from socially sanctioned
activities, such as school, decreased. Moreover, the social
environment appeared to change such that deviant behavior was supported
by peer and parental attitudes and behavior. These findings were
consistent with the literature on the emergence of antisocial behavior,
and suggested that family environment may play a role in the

deve 10pment of such patterns .

Finally, it appeared that certain aspects of family functioning
were associated with behavior reflective of normative experimentation
and accompanying struggle. In order to begin the process of
separation, the family in the adolescent phase of development may need
to provide a moderate degree of cohesion. Too much cohesion may make
the nest too appealing to leave, while not enoigh cohesion may cause
undue separation anxiety. Moderate expressiveness allows the family to
transmit important values while also allowing for the expression of
disagreement and conflict. Youth in this type of family environment
perceived themselves to be more individuated, while other family

environment configurations were associated with fewer personal growth

138

options. These patterns were consistent with the findings of Olson and
McCubbin (1983) and Koch (1979).

Extrapolating from these findings, one can develop a profile of
adolescents who may be at risk for experiencing difficulty in
consolidating a sense of identity manifest in problematic drinking
and/or careless driving. These adolescents are likely to begin
drinking at an earlier age, and to be engaged in other forms of
antisocial behavior (Zucker & Hartford, 1980; Jessor & Jessor, 1972) .
They are likely to perceive the family environment as disengaged,
failing to provide adequate affectional ties as well as socialized
values. Consequently, the adolescents may turn to peers with similar
values and behavior, further consolidating tl'e deviant practices. Sore
of these young people may come to feel increasingly cut off from the
larger community that could be used as an alternative for obtaining
guidance and support. As energy is devoted toward binding anxiety
associated with premature or ill-equipped independence, there is less
energy and motivation available for mastery of other socially
appropriate avenues of adult stat'us--e.g. , academic competency.

The theoretical literature on adolescent development provides a
framework for these observations. Marcia (1980) describes sote of
these adolescents as reflecting role diffusion, though many of them may
appear to be at the moratorium stams. Normative drinkers and drivers
could also be described as being at the moratorium status. The
difference between these two groups is that problematic drinkers and/or
risky drivers are probably at much greater risk for failure to make an
adequate transition into adulthood, and for further manifestation of

deviant behavior. longitudinal data of both a behavioral and emotional

139

nature is needed in order to clarify these issues. Moreover, family
life span research, which takes a developmental focus, can further
delineate the ways in which family environment influence tlese

processes .

APIENDICES

APPENDIX A

STUDENT QUESTIONNAIRE

140

APPENDIX A

STUIENT QUESTIONNAIRE

This is a questionnaire that deals with a person's attitudes and
behaviors towards both driving and drinking. Most people find tress
questions both interesting and easy to answer.

Do not begin until you are told to do so. For sole questions you will
check one or more answers and/or write your answer. Some of the
questions might not apply to you; leave them blank. If you have any
questions, please raise your hand and someone will help you.

THIS QUESTIONNAIRE IS COMPIETELY CONFIDENTIAL. No one will know what
your answers are, so please be as precise as possible.

NAME

 

TODAY ' S DATE

 

PARTICIPATION IN THIS STUDY IS MY VOLUNTARY. YOUR C(MPIEI‘ION
OF THE QUESTIONNAIRE CONSTITUTES YOUR ASSENT TO PARTICIPATE.

141

PLEASE ANSWER THE FOLLOWING ABOUT YOURSELF.

1. My age is:
A. younger than 13 years

B. 13
C. 14
D. 15
E. 16
F. 17
G. 18
H. 19
1. older than 19 years

2. I am:
A. Male
8. Female

3. My grade in school:

A. 9th
8. 10th
C. 11th
D. 12th

4. My racial identity is:
A. White (not Hispanic)
8. Hispanic
C. Black (not Hispanic)
0. Native American
E. Asian

5. My parents are divorced or separated:
A. Yes
8. No

o. My grades in school are mostly:

A. A’s
8. 8’5
C. C’s
D. D’s
E. F’s

7. How many non-classroom activities do you participate in (for
example, student council, sports, clubs, church/community
groups, etc.)? ______

8. The number of days I have been absent from school this year is:
A. 5 days or less
8. 6 to 10 days
C. 11 to 15 days
D. 15 to 19 days
E. 80 or more days

9.

10.

11.

12.

13.

14.

142

I have been involved in an automobile accident (either as a
driver, passenger or pedestrian) in which alcohol was a factor.
(Circle as many as apply to you)

A. I was the driver

8. I was the passenger

C. I was the pedestrian

If a high school student wanted an alcoholic beverage:

A. it would be easy to get a hold of

B. it wouldn’t be easy, but if a person really wanted it, they
could get it

C. it would be nearly impossible to get a hold of

I have a:

A. drivers license

B. learners permit

C. neither a drivers license or a permit

If you answered "C“ to the last question (#11), skip to
question #15.

when I want a car to drive somewhere:

A. I have my own car

B. I do not own my own car, but it is pretty easy to get a
hold of one

C. it wouldn’t be easy, but if I really wanted a car, I could
get a hold of one

D. it would be nearly impossible to get a hold of one

How many auto accidents have you been involved in:
____ as a driver?

as a passenger?
as a motorcyclist?

How many "near misses" or "close calls" have you been involved in
as a driver:

A. nearly every day

8. almost once a week

C. almost once a month

D. only rarely

E. never

I wear a seat belt:
A. always

8. often

C. .sometimes

D. rarely

E. never

143

BELOW IS A SET OF STATEMENTS ABOUT DRINKING AND ABOUT DRIVING AFTER
DRINKING. WE ARE INTERESTED IN YOUR PERSONAL REACTION TO EACH

STATEMENT. PLEASE INDICATE THE DEGREE TO WHICH YOU PERSONALLY AGREE
OR DISAGREE WITH EACH STATEMENT BY CIRCLING THE LETTER IN FRONT OF THE
ANSWER WHICH BEST DESCRIBES HOW YOU FEEL ABOUT THE STATEMENT.
REMEMBER, THERE ARE NO RIGHT OR WRONG ANSWERS; WE ARE INTERESTED ONLY
IN YOUR PERSONAL REACTIONS.

16.

17.

18.

19.

80.

21.

I would not be accepted by my friends unless I drank.
A.~ Strongly Agree

8. Agree

C. Disagree

D. Strongly Disagree

When I am with my friends, I feel uncomfortable asking for
non-alcoholic beverages at occasions where alcoholic beverages
for being served.

A. Strongly Agree

8. Agree

C. Disagree

D. Strongly Disagree

I feel uncomfortable around friends who do not drink if I
am drinking.

A. Strongly Agree

8. Agree

C. Disagree

D. Strongly Disagree

At social occasions with my friends, drinking alcoholic
beverages is a fairly common activity.

A. Strongly Agree

8. Agree

C. Disagree

D. Strongly Disagree

As far as I know, none of my friends are too concerned about
wearing seat belts.

A. Strongly Agree

8. Agree

C. Disagree

D. Strongly Disagree

I would feel uncomfortable asking a friend to slow down if
I felt he/she was driving too fast or carelessly.

A. Strongly Agree

8. Agree

C. DiSagree

D. Strongly Disagree

144

88. Driving around, or "cruising“ is an 1.nportant activity among
my friends. We do it whenever we can.
A. Strongly Agree
B. Agree
C. Disagree
D. Strongly Disagree

83. I would not be accepted by my friends if I payed attention to
all traffic regulations--i.e., if I didn’t speed or roll a
stop once in a while. -

A. Strongly Agree

B. Agree

C. Disagree

D. Strongly Disagree

84. I would not be accepted by my friends if I tried to prevent
them from driving after drinking too much.
A. Strongly Agree
8. Agree
C. Disagree
D. Strongly Disagree

25. I would ask a friend to drive me home if I felt I had too much
to drink and drive safely.
A. Strongly Agree
B. Agree
C. Disagree
D. Strongly Disagree

86. My friends would understand if I refused to get in the car with
a friend I thought had too much to drink and drive safely.
A. Strongly Agree
B. Agree
C. Disagree
D. Strongly Disagree

27. Driving after drinking is fairly accepted among my friends.
A. Strongly Agree
B. Agree
C. Disagree
D. Strongly Disagree

PLEASE ANSWER THE FOLLOWING QUESTIONS ABOUT YOURSELF, YOUR OWN USE AND
NON-USE OF ALCOHOL: AND YOUR DRIVING BEHAVIOR.

88. In the past six months, how often would you say you drank
any alcoholic beverage? (Circle one)
A. Never drank in the past 6 months
B. Less than once a month
C. About once a month

D. Two or three times a month
E. About once a week
F. Two-six times a week

G. At least once a day

89.

3C).

31.

145

How many times would you estimate that you have gotten drunk or
pretty high on alcohol during the gas: zgags?

A. Once a week or more

B. Twice a month

C. Once a month

9. 6-10 times during the year

E. 4-5 times during the year

F. 2-3 times during the year

6. 1 time during the year

H. None

Over the past year, how many times would you say you
gggxg a 9;; after you had had too much to drink?

A. Once a week or more

B. Twice a month

C. Once a month

D. 6-10 times during the year

E. 4-5 times during the year

F. 2-3 times during the year

G. 1 time during the year

H. None

Over the past xgag, how many times were you a passenger in
a gag when the driver had had too much to drink?

A. Once a week or more

B. Twice a month

C. Once a month

D. 6-10 times during the year

E. 4-5 times during the year

F. 8-3 times during the year

G. 1 time during the year

H. None

146

32. During the gag; ygggs how many times have each of the following
happened to you? (Mark gag oval for each question or row.)
40 or
8-3 4-5 . 6-9 more
BgLs 999s Iimsa Iimss limes Iimss

( ) ( ) ( ) ( ) ( ) ( ) You’ve gotten into trouble
with your teachers or
principal because of your
drinking.

( ) ( ) ( ) ( ) ( ) ( ) You’ve gotten into
difficulties of any kind
with your friends because
of your drinking.

( ) ( ) ( ) ( ) ( ) ( ) You’ve gotten into trouble
with the police because of
your drinking.

( ) ( ) ( ) ( ) ( ) ( ) You’ve driven when you’ve
had a good bit to drink.

( ) ( ) ( ) ( ) ( ) ( ) You’ve been criticized by
someone you were dating
because of your drinking.

( ) ( ) ( ) ( ) ( ) ( ) You’ve gotten into trouble
with your parents because
of your drinking.

33. During a 3 hour social occasion, how many total drinks
would you have?
A. None--I don’t drink
B. 1-3 drinks

. 4-6 drinks

. 7—9 drinks

. 10-18 drinks

. 13 or more drinks

TINUU

NOW WE WOULD LIKE TO ASK YOU SOME QUESTIONS ABOUT YOUR FRIENDS'
DRINKING AND DRIVING BEHAVIORS.

34. As far as you know, about how many of your friends drink
alcohol at least sometimes?
A. None of them drink alcohol
B. 1-8
C. Several
D. Most of them
E. All of them

 

147

[1.)
(.0

As far as you know, about how many of your friends have
received a traffic ticket for a moving violation?

A. None of them have received a ticket
B. 1-2
C. Several

D. Most of them
E. All of them

36. As far as you know, about how many of your friends have been
involved in a traffic accident as a ggiggg?
A. None of them have been in an accident
B. 1-2
C. Several
D. Most of them
E. All of them

37. As far as you know, about how many of your friends have driven
a car after drinking too much?
A. None of them drink and drive
B. 1-8
C. Several
D. Most of them
E. All of them

NOW WE WOULD LIKE TO ASK YOU SOME QUESTIONS ABOUT YOUR FAMILY’S
FEELINGS RELATED TO DRINKING AND DRIVING.

38. How do you think your parents (or your family) feel about people
your age drinking?
A. Strongly approve
B. Approve
C. Don’t care one way or the other
D. Disapprove
E. Strongly disapprove
F. I don’t know

39. How do you think your parents (or your family) feel about
people your age driving after they have been drinking?
A. Strongly approve
B. Approve
C. Don’t care one way or the other
D. Disapprove
E. Strongly disapprove
F. I don’t know

40. How do you think your parents (or your family) feel about
people your age ignoring traffic laws from time to time?
A. Strongly approve
B. Approve
C. Don’t care one way or the other
D. Disapprove
E. Strongly disapprove
F. I don’t know

148

THE NEXT SET OF QUESTIONS ARE STATEMENTS ABOUT FAMILIES. YOU ARE TO
DECIDE WHICH or THESE STATEMENTS ARE TRUE OF YOUR FAMILY AND WHICH ARE
FALSE. IF YOU THINK THE STATEMENT IS IBQE 95 EQSILX Iggg OF YOUR

FAMILY, CIRCLE THE LETTER "T". IF YOU THINK THE STATEMENT IS FALSE 0R

-————— --——-.— -—---- --- —————— ---

41. Family members really help and support one another. T F
48. Family members often keep their feelings to themselves. T F
43. We fight a lot in our family. ' T F
44. We don’t do things on our own very often in our family. T F
45. We feel it is important to be the best at whatever we do. T F
46. We often talk about political and social problems. T F
47. We spend most weekends and evenings at home. T F

48. Family members attend church, synagogue, or Sunday

School fairly often. T F
49. Activities in our family are pretty carefully planned. T F
50. Family members are rarely ordered around. T F
51. We often seem to be killing time at home. T F
58. We say anything we want to around home. T F
53. Family members rarely become openly angry. T F
54. In our family, we are strongly encouraged to be

independent. T F
55. Getting ahead in life is very important to our family. T F
56. We rarely go to lectures, plays or concerts. T F
57. Friends often come over for dinner or to visit. T F
58. We don’t say prayers in our family. T F
59. We are generally very neat and orderly. T F
60. There are very few rules to follow in our family. T F
61. We put a lot of energy into what we do at home. T F

68. It’s hard to "blow off steam“ at home without
upsetting somebody. T F

63. Family members sometimes get so angry they throw things. T F

64. We think things out for ourselves in our family. T F

66.

67.

68.

69.

70.

71.

78.

73.

74.

76.

77.

78.

79.

80.

149

How much money a person makes is not very important
to us.

Learning about new and different things is very important
in our family.

Nobody in our family is active in sports, Little League,
bowling, etc.

We often talk about the religious meaning of Christmas,
Passover, or other holidays.

It’s often hard to find things when you need them in
our household.

There is one family member who makes most of the
decisions.

There is a feeling of togetherness in our family.

We tell each other about our personal problems.
Family members hardly ever lose their tempers.

We come and go as we want to in our family.

We believe in competition and “may the best man win.“
We are not that interested in cultural activities.

We often go to movies, sports, events, camping, etc.
We don’t believe in heaven or hell.

Being on time is very important in our family.

There are set ways of doing things at home.

44-4-4

.4

REFERENCES

WW8

Zucker, R. A., and N011, R. Personal commnications, Michigan
State University, October, 1984.

150

Accident Facts, 1965-1971 Eds. National Safety counsel, Chicago, Ill.

Adelson, J. (1980). Handbook of Adolescent Psychology: JOhn‘Wiley &
Sons, NEW YOrk.

 

Balswick, J. 0., & Macrides, C. (1975). Parental stimulus for
adolescent rebellion. Adolescence, 19, 253-266.

 

Baumrind, D. (1967). Childcare practices anteceding three patterns of
preschool behavior. Genetic Psychology Monographs, Z§, 43-88.

 

Beamish, J. J., & Malfetti, J. I" (1962). A.psychologica1 comparison
of violator and nonviolator automdbile drivers in 16-19 year age
group, Traffic Safety Research Review, §(1), 12-15.

 

Behrens, M., & Goldfarb,‘W; (1958). A study of patterns of
interaction of families of schizophrenic children in residential
treatment. American JOurnal of Orthgpsychiatryy gg, 300-312.

 

Bell, L. G., & Bell, D. C. (1982). Family climate and the role of the
female adolescent: Determinants of adolescent functioning. Family

Biddle, B. J., Bank, B. J., & Marlin, M. (1980). Social determinants
of adolescent drinking-dwhat they think, what they do, and what I
think and do. JOurnal of Studies on Alcohol, 41(3).

 

Blane, H. T., & Chafetz, M. E. (Eds.) (1979). YOuth, Alcohol and
Social Policy. NEW York: Plenum.

 

Bdb, S. (1968). An investigation of the relationship between identity
status, cognitive style, and stress. Unpublished doctoral
dissertation, SUNY at B.

 

 

Braudht, G. N. (1982). Prdblem drinking among adolescents: a review
analysis of psychosocial research. In Alcohol and Health
Monographs, No.4: Special Population Issues. Rockville, MD:
National Insitute on Alcohol Abuse & Alcoholism.

 

 

151

152

Breuer, H. (1973). Ego identity status in late adolescent college
males as measured by a group administered incomplete sentence
blank and related to inferred stance toward authority.
unpublished doctoral dissertation, New‘York University.

Brittain, C. V. (1963). Adolescent choices and parent-peer
cross-pressure. Sociological Review, gg, 385-391.

Cahalan, D. (1970). Problem Drinkers. San Francisco: JOssebeass.

 

cabalan, D., Cisin, I. H., & Crossley, H. M. (1969). American Drinking
Practices: A.National Study of Drinking Behavior and_Attitudes.
NeW'BrunswiCk, N.J.: Publications Division, Rutgers Center of
Alcohol Studies.

 

carlson,‘w. L., & Klein, D. (1970). Familial versus institutional
socialization of the young traffic offénder. JOurnal of Safety

 

Clay, M. L. (1974, December). Characteristics of high risk drivers,
alcoholic and otherwise. Paper presented at the Nbrth American
congress on Alcohol and Drug Problems, San Francisco.

Cohen, J. (1978). Partialed products are interactions; partialed
powers are curve components. Psychological Bulletin, §§, No. 4,
858-866.

 

Cbhen, J., & Cohen, P. (1975). Applied multiple
regression/correlation analysis for the behavioral science, JOhn
Wiley & Sons.

Davidson, S. (1978). The level of differentiation of ego functioning
in the identity:achieved and identity-foreclosed status. Un-
published doctoral dissertation, Boston University.

Demone, H. W. (1972). The nonuse and abuse of alcohol by the male
adolescent. Proceedings of the Second Annual Alcoholism
Conference of the National Institute on Alcohol Abuse and
Alcoholism, JUne 1-2.

 

Dickson,'wu, Hess, R., Miyake, N., & Azuma, H. (1979). Referential
cxnmunication accuracy between mother and child as predictor
of development in the United States and Japan. Child
Development, Q9, 53-59.

Donovan, D. M & Marlatt, G. A. (1982). Personality subtypes among
driving-while-intoxicted offenders: Relationship to drinking
behavior and driving risk. JCurnal of consulting and Clinical
Psychology, 59(2), 241-249.

Douglass, R. In (1982). Ybuth, alcohol and traffic accidents. in
Alcohol Health Monographs, No. 4: Special Population Issues.
Rockville, MD: National Institute on.Alcdhol Abuse and
Alcoholism.

153

Donovan, J. E. & Jessor, R. (1978). Adolescent problem drinking:
psychosocial correlates in a national sample study. Journal of
Studies on Alcohol, 19(9) , 1506-1524.

 

Douvan, & Adelson, J. (1966). The adolescent experience. New York:
Wiley.

 

Divall, E. M. (1957, 1962). Family Develgament. Chicago: Lippinoott.

Elder, G. H. (1963) . Parental power legitimation and its effect on the
adolescent. Sociometry, 36, 50—65.

 

Erickson, E. H. (1959). Identityand the lifecyle; Psycholgical
Issues. New York: International U—niversity Press, I.

Erickson, E. H. (1963). Childhood and Society. 2nd ed. New York:
W.W. Norton.

 

Erickson, E. H. (1968). Identity: Youth and Crisis. New York, W.W.
Norton.

 

Fisher, B. L., Gibbin, P. R., & Hoopes, M. H. (1982). Journal of
Marital and Family Therapy, 8, 273-282.

 

Fisher, B. L., & Sprenkle, D. H. (1978). Therapists' perceptions of
healthy family functioning. International Journal of Family
Counselipg, 6, 9-18.

 

 

Gleuck, S., & Gleuck, E. (1950) . Unraveling juvenile delinquency.
Cambridge, MA: Harvard University Pracs.

 

Goldstein, L. (1971) . Youthful drivers as a special safety problem.
In The young driver: Reckless or unprepared. North
Carolina Symposia on Highway Safety, _§_.

Group for the Advancement of Psychiatry (GAP). (1970) . Treatment of
families in conflict. New York: Science House.

 

 

Haley, J. (1980). Leaving Home. New York: McGraw-Hill.

Hall, G. S. (1904). Adolescence, 2 vols. New York: Appleton.

 

Harrington, D. M. (1971, Septenber) . The youngdriver follow-up
study: An evaluation of the role of human factors in the first 4
years of driving. California Department of Motor Vehicles, Rept.
No. 30.

 

Hess, R., Szuma, H., Kashiwagi, K., Dickinson, W. Nagano, S.,
Halloway, S., Mayake, N., & Price, G. (1984). Family influences
on school readiness and achievement in Japan and the United
States: An overview of a longitudinal survey. In H. Stevens
H. Azuma, & K. Hakuta (Eds) , Child Development and Education
in Japafl. New York: Freeman Press.

 

154

Hill, R. (1949). Families under stress. New York: Harper.

Hill, R. (1964) . Methodological issues in family development
research. Family Process, 3(1), 186-206.

 

Hogan, R. (1973) . Moral conduct and moral character: A psychological
perspective. Psychological Bulletin, _7_9_, 217-232.

Hotch, D. F. (1979) . Separating fran the family: A study of
perceptions of home-leaving in late adolescence. Unpublished
doctoral Tissermdon , Michigan State University.

 

Insurance Institute for Highway Safety. (1984, April). Highway loss
reduction status report, 1_9_ (7) .

 

 

Jacob, T. (1975) . Family interaction in disturbed and normal
families: A methodological and substantive Review. Psychological
Bulletin, 82(1), 33-65.

 

Jessor, R., Graves, T. D, Hanson, P. C., et a1. (1968). Society,
personality, and deviant behavior: A study of a triethnic
commmity. New York: Holt, Rinehart and Winston.

 

Jessor, R., & Jessor, S. L. (1973). Problem drinking in youth:
Personality, social and behavioral antecedents and correlates.
Proceedings of the Second Annual Alcoholism Conference of the
National Institute on Alcohol Abuse and Alcoholism, June, 1972.
Washington D.C. DHIEW.

 

Jessor, S. L., & Jessor, R. (1974). Maternal ideology and adolescent
problem behavior. Developmental Psychology, IQ, 246-254.

 

Jessor, R., & Jessor, S. L. (1975). Adolescent development and the
onset of drinking: A longitudinal study. ery Journal of
Studies on Alcohol, g, 27-51.

 

 

Jessor, R., & Jessor, S. L. (1977) . Problem behavior and psychosocial
development . Iondon: Academic Press .

 

Jessor, R., Young, B. H., Young, E. B., & Tesi, G. (1970). Perceived
opportunity, alienation and drinking behavior among Italian and
American youth. Journal of Personality and Social Psychology, _1_5_,
215-222.

 

Jones, M. C. (1971) . Personality antecedent and correlates of
drinking patterns in women. Journal of Consulting and Clinical
Psychology, §_6_(1), 61-69.

 

Klein, D. (1971) . Adolescent driving as deviant behavior. In The
Yourg Driver, Reckless or Ungepared, North Carolina Symposifi on
Highway Safety, _5_.

155

Kraus, A. 8., Steele, R., Gent, W., & Thompson, M. (1970, Jure).
Pre-driving identification of young drivers with a high risk of
accidents. Journal of Safety Research, _2_(2), 55-66.

Landis, P. H. (1954) . The ordering and forbidding techniques and
teenage adjustment. School & Society, _82, 105-106.

 

Ierner R. M., & Spanier, G. B. (1980) . Adolescent development: A life
span perspective. New York: McGraw-Hill Book Company.

 

 

Devonian, E. (1969, July). Personality characteristics of juvenile
driving violators. Accident Analysis and Prevention, 1(1) , 9-16.

 

Lewin, K., Lippitt, R., & White, R. K. (1939) . Patterns of aggressive
behavior in experimentally created "social climates." Journal of
Social Psychology, 19, 271-299.

 

 

Iowan, C. (1982/1983) . Drinking and driving among youth. Alcohol
Health and Research, 1(2) , 41-49.

 

Mahler, C. (1969) . The assessment and evaluation of the cgping styles
of two ego identity status group: Moratorium and Foreclosure, two
Hentity conflict-arousing stimuli. Unpublished masters thesis,
SUNY at B.

 

Marcia, J. E. (1966) . Developtrent and validation of ego identity
status. Journal of Personality and Social Psychology, §(5) ,
551-558.

 

Marcia, J. E. (1967) . Ego identity status: Relationship to change in
self-esteem, "general maladjustxrent," and autlnritarianism.
Journal of Personality, 15(1) , 119-133.

 

Marcia, J. E., & Friedman, M. L. (1970). Ego identity status in
college women. Journal of Personality, £9) , 249-263.

 

Marcia, J. E. (1980). Development of ego-identity status. In Muuss,
Adolescent Behavior and Society, Random House.

 

Mattson, D. R. (1974) . Alienation versus expgaration and commitment:
Personality and family correlates of adolescent identity statuses.
Report from the Project for Youth Research. Copenhagen: Royal
Danish School of Educational Studies.

McCord, W., & McCord, J. (1960) . Origins of alcoholism, Stanford, CA:
Stanford University Press .

 

McCord, W., & McCord, J. A. (1962). longitudinal study of the
personality of alcoholics. In Pittman, D. J. 5. Snyder, C. R.
(Eds), Society, culture and drinking patterns, New York: John
Wiley.

156

McKinney, J. P., Fitzgerald, H. E., & Strcnnen, E. (1982).
Developmental psychology: The adolescent and my adult, (2nd
ed), Dorsey Press.

bead, M. (1949). Coming of age in Samoa. New York: Mentor Books.

Moos, R. H. (1974). Familyenvironment scale. Palo Alto: Consulting
Psychologist.

Moos, R. H., & Moos, B. (1976). A typology of family social
environnents. Family Process, 1_5_, 357-371.

 

muss, R. E. (1980) . Adolesc_e_nt behavior and society: A book of
read1_ng' 3, Third edition, NY: Random House.

National Transportation Safety Board. (1974, July). Special study:
Youth and Traffic Safety Ed. Rept. No. NI‘SB-STS—71-3, Washington,
D.C.

Neuber, K. A., & Genthner, R. W. (1977). ‘The relationship between ego
identity, personal response, and facilitative camunication.
Journal of Psychology, _9_5_, 45-49.

National Institute on Alcohol and Alcoholism: Traffic Safety Statistics
of Young Drivers, 1982.

N011, R. B. (1980) . Early assessment of child and family
vulnerability in alcoholic families: A proposed nethodology.
Unpublished masters thesis, Michigan State University.

Noll, R. B. (1983). Young male offspring of alcoholic fathers: Early
development and cognitive differences from the MSU Vulnerability
Study . Unpublisred doctoral dissertation, Michigan State
University.

Novak, A. L., & VanderVeen, F. (1970). Family concepts and emotional
disturbance in the families of disturbed adolescents with normal
siblings. Family Process, _9_, 157-171.

Offer, D., 8. Offer, J. B. (1976). Fran teenage toyouananhood. New
York: Basic Books.

Olson, D., & McCubbin, H. (1983). Families: What makes them work?
Sage Publication.

Orlofsky, J. L, Marcia, J. E., & lesser, I. M. (1973). Ego identity
status and the intimacy versus isolation crisis of young
adulthood. Journal of Personality and Social Psychology, 31(2) ,
211-219.

Otto, H. A. (1963). Criteria for assessing family strengths. Family
Process, _2_(2), 329-337.

157

Pelz, D. C., & Schuman, S. H. (1971, Jure). Are yang drivers really
more dangerais after control ling for exposure and experience?
Journal of Safety Research, §(2), 68-79.

 

Pelz, D. C., & Schuman, S. H. (1971, April). Exposure factors in
accidents and violations of yang drivers . University of
Michigan.

 

Pelz, D. C., & Schuman, S. H. (1971). Motivational factors in
crashes and violationg of young drivers. Paper presented at
meetings of American Public Health Association, Mineapolis.

 

 

Perrine, M. W. (1974) . The Spectrum of Drinking Drivers, Wiley Press.

Podd, M. H., Marcia, J. E., & Rubin, B. M. (1970). The effects of ego
identity and partner perception on a prisoner's dilemna game.
Jairnal of Social Psychology, 3, 117-126.

Rachel, J. V., Maisto, S. A., Gues, L. L, & Hubbard, R. L. (1982).
Alcohol use amorg youth. In J. DeLuca (Eds), Alcohol and health
mongraphs, no.1: Alcohol consumption and related problems.
Rockville, MD: National Institute on Alcohol Abuse and Alcoholism.

 

Rappoport, R. (1963). Normal crises, family structure, and mental
health. Family Process, 2, 68-79.

 

Ricks, D., & Berry, J. C. (1970). Family and symptan patterns that
precede schizophrenia. In Roff, M. s. Ricks, D.F. (Eds), Life
History Research in Psychopathology, Minneapolis: University of
Minnesota Press.

 

Robins, L. N. (1966) . Deviant Children Grown up, Baltimore: Williams
& Wilkins.

 

Ronmel, R. C. (1959, March). Personality characteristics and
attitudes of youthful accident-repeating drivers . Traffic Safety

 

 

Rosenfeld, R. U. (1972) . The relationship of ego identity to
similarity among self, ideal self, and probable occupational role
concept among college males. Unpublished doctoral dissertation,
University of Maryiand.

 

Russell, C. S. (1980) . A nethodological study of family cohesion and
adaptability. Journal of Marriage and Family Camseling, _6_(4) ,
459-470.

 

Schuster, D. (1971) . 'Ihe yourg problem driver. In The Young Driver:
Reckless or Unprepared? North Carolina Symposium on Highway
Safety, 5.

 

Schvaneveldt, J. (1973) . Mormon adolescents: Likes and dislikes
toward parents and hone. Adolescence, _8_, 171-178.

 

158

Selzer, M. L.,‘Vinokur, A., & Wilson, T. D. (1977). A.psychosocial
comparison of drunken drivers and alcoholics. JOurnal of Studies

 

Sobel, R., & Underhill, R. (1976). Family disorganization and
teenage auto accidents. JOurnal of Safety Research, g, 8-18.

 

Sullivan, H. S. (1953). The integpersonal theory ofgpsychiatry.
New York: W.W. Norton.

 

Sundberg, N., Schrma,‘V., Wbdtli, T., & Rohila, P. (1969). Family
cohesiveness and autonomy of adolescents in India and the United
States. Jburnal of Marriage and the Family, 403-407.

 

vanderVEen, F. (1976). content dimensions of the family concgpt test
and their relation to childhood disturbance. Unpublished
manuscript. Institute for JUvenile Research, Chicago.

 

 

VEgega, M. E. (1984). Deterring drinking-driving among youth: Some
research needs. Paper presented at the American Psychological
Association, Toronto, Canada.

 

 

VErdonik. F., & Sherod, In (1984). An inventory;of longitudinal
research on childhood and adolescence, Social Science Research
council, New York.

 

 

WEstley,'W. A., & Epstein, N. B. (1969). Silent.minority: Families
of emotionally healthy college students. San Francisco:
Jessey-Bass.

 

 

Wbdarski, J. S., & Hoffman, S. D. (1984). Alcohol education for
adolecents. Social Wbrk in Education, 69-92.

 

wynne, L. C., & Cole, R. E. (1982). The Rochester Risk Research
Program: A new look at parental diagnoses and family
relationships. In H. Stierlin, L.C. Wynne, & M. Wirsching (Eds),
Psychosocial intervention in schizophrenia: An international
view, Berlin: Springer-VErlag, 25-48.

 

Zucker, R. A., & Riester, A. E. (1968, December) Adolescent social
structure and drinking behavior. Personnel and Guidance JOurnal,
304-312.

 

Zucker, R. A. (1976). Parental influences on the drinking patterns of
their children. In Greenblatt & SchuCkit (Eds.), Alcdholism
Problems in women and Children.

 

Zucker, R. A. (1979). Developmental aspects of drinking throughout
the young adult years. In Blane, H. T. & Chafetz, M. E. (Eds.),
Ycuth, Alcohol, and Social Poligy, Plenum Press.

Zucker, R. A., & Harford, T. C. (1983). National study of the
demography of adolescent drinking practices in 1980. Jaurnal of
Studies on Alcohol, 44, 974-985.

 

159

Zucker, R. A. (1984). Youth, alcohol, and driving: A developnental
overview. Paper presented at American Psychological Association.

Zucker, R. A., & Barron, F. H. (1973). Parental behaviors associated
with problem drinking and antisocial behavior amorg adolescent
males. Proceedings of the First Annual Conference on the National
Institute of Alcohol Abuse and Alcoholism, June 1971, Washington
D.C., DHEW.

Zucker, R. A., & Devoe, C. (1975) . Life history characteristics
associated with problem drinking and antisocial behavior in
adolescent girls: A comparison with male findings. In Wirt,
Winokur, & Merrill (Eds.) , Life history research in
psychopatmlogyo

 

Zucker, R. A., & Noll, R. B. (1982). Precursors and developnental
influences on drinking and alcoholism Etiology from a
longitudinal perspective. In J. DeLuca (Eds), Alcohol and Health
Monogrgahs, No. 1: Alcohol Consumption and Related Problems.
Rockville, MD: National Institute on Alcohol Abuse and Alcoholism.