THE EFFECTIVENESS OF THE MANN INVENTORY IN I
CLASSIFYING YOUNG DRIVERS INTO BEHAVIORAL
CATEGORIES AND ITS RELATIONSHIP TO
SUBSEQUENT DRIVER PERFORMANCE

Thesis for the Degree of Ed. D. .
MICHIGAN STATE UNIVERSITY
FRANCIS CARL KENEI.
1967

  
    
    

LIBRA R Y
Michigan State
University

This is to certify that the
thesis entitled
THE EFFECTIVENESS OF THE MANN INVENTORY IN CLASSIFYING

YOUNG DRIVERS INTO BEHAVIORAL CATEGORIES AND ITS RELATIONSHIP
TO SUBSEQUENT DRIVER PERFORMANCE

presented by

Francis C. Kenel

has been accepted towards fulfillment
of the requirements for

JILL.— degree in_Ed_\LC_a_LiOH

Major professor

 

Date October 27, 1967

0‘169

  

 

 

 

 

 

 

 

 

 

 

ABSTRACT

THE EFFECTIVENESS OF THE MANN INVENTORY IN CLASSIFYING
YOUNG DRIVERS INTO BEHAVIORAL CATEGORIES AND ITS
RELATIONSHIP TO SUBSEQUENT DRIVER PERFORMANCE

by Francis Carl Kenel

Previous studies investigating the causes of motor
vehicle accidents almost invariably demonstrate that most
are attributable to human failures of various kinds. In
recent years the area of personality factors, attitudes,
and the quality of personal adjustment, "We drive as we
live," has been emphasized as important to safe driving.

The primary purpose of this study was to investigate
the validity and reliability of the Mann Inventory.

The Mann Inventory consists of 63 items that appear
to reflect an individual's feelings toward himself,
others, and established social mores. Reaction to
items in the Inventory are expressed by checking
one of five responses——always, usually, sometimes,
rarely or never.
An effort was made to ascertain whether or not persons
classified into one of six behavioral categories on the
basis of observed behavior would respond differently to
items in the Mann Inventory.

The secondary purpose of the study was to ascertain

whether or not the observed behavior which identified an

individual within a behavioral category would be reflected

in subsequent driving record.

 

Francis Carl Kenel

Sexton High School, Lansing, Michigan, was selected
for this investigation on the basis of:

l. A student body of approximately 2000, representing

a broad cross—section of the pOpulation.
2. A Driver Education program reaching several
hundred students each year.
The sample population of this study consists of .mr
1057 subjects, 523 males and 534 females who
were enrolled in, and had satisfactorily completed
Driver Education at Sexton High School, Lansing,
Michigan, between September, 1958 and June, 1960. -EL

3. Availability of comprehensive and professional

police Traffic Record Files.

All students enrolled in the Driver Education program
completed the Mann Inventory during the last week of the
Driver Education course. On the basis of rater observation,
conducted throughout the semester, each rater placed the
individual Inventory response sheet into one of the six
appropriate behavioral categories.

The hypotheses were tested by means of Chi Square,

The Cell. Square Contingency Method and Product—Moment
Correlation.

Statistical analysis of the data reveals:

1. When students were grouped on the basis of
observed behavior, response to items in the Mann Inventory
were significantly different for each behavioral category.

Sixty of the 63 items were significant at or above the .10

level of confidence for the sample population.

 

 

 

 

Francis Carl Kenel

The response to items with the sample population ‘
divided by sex reveals 37 items that differentiate signifi- i;
cantly for females and A7 items for males at or above the .10 A
level of confidence. Response to 20 to 39 specific items were
identified with any one of the six behavioral categories.

2. The reliability of the Mann Inventory was deter-
mined by application of Product—Moment coefficient of
correlation to two separate administrations of the instrument.
Correlation values of .697 — .986 were derived. With 63
degrees of freedom, these values indicate a high to very high
correlation with marked or dependable relationships.

3. The relationship between the behavioral categories
and driving records was of the same direction and magnitude
as was the relationship between the Mann Inventory and
observed behavior. As the individuals behavior deviated
from category 1 (well adjusted), toward category 6 (marked
evidence of problems of adjustment in school), his driving

record of convictions and/or collisions increased.

 

 

 

 

THE EFFECTIVENESS OF THE MANN INVENTORY IN CLASSIFYING

YOUNG DRIVERS INTO BEHAVIORAL CATEGORIES AND ITS

RELATIONSHIP TO SUBSEQUENT DRIVER PERFORMANCE

By

Francis Carl Kenel

A THESIS

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

DOCTOR OF EDUCATION

Department of
Secondary Education: Curriculum

1967

 

 

 

ACKNOWLEDGMENTS

To Dr. William Mann, my major advisor, friend and
chairman of my guidance committee.

To the other members of my guidance committee,

Dr. John Fuzak, Dr. Walter Johnson, and Professor
Gordon Sheehe.

To Dr. Charles Porter, Dean College of Applied
Science and Technology, Illinois State University, for
technical and statistical assistance in computer program-
ming.

To my wife, Betty J. Kenel and the children for

their many sacrifices.

11

PM? ‘5; ‘

- o c‘ -. -

.“v.9.:‘vh‘;‘ri .213"; ’“T‘l , a ‘ OI - .- J.-(J ’ I it...
‘~ ~oo',s-|~<I.V..4 4 3......

Q ' ‘ 1 . l g a I - . '. I- --

 

 

 

TABLE OF CONTENTS

ACKNOWLEDGMENTS

LIST OF TABLES . .

LIST OF APPENDICES

Chapter

I.

II.

III.

IV.

THE PROBLEM .

Statement of the Problem.
Importance of the Problem

Basic Assumptions
The Hypotheses

Definition of Terms

Organization of the Study

REVIEW OF LITERATURE ON THE RELATION-
SHIP OF PERSONALITY AND PERSONAL
SOCIAL ADJUSTMENT TO ACCIDENT IN-

VOLVEMENT.

Summary

DESIGN AND METHODOLOGY

The Test Instrument

Sample Selection
The Raters . .
The Data . . .

0

Procedure of Analysis.

Null Hypothesis.
Summary .

STATISTICAL ANALYSIS

Validity of the Instrument:

Differ-

ence Between Response to Items in
the Mann Inventory by Behavioral

Categories

iii

vii

l

\O\OCDCDU'TU'I

A2

I
..
r
..

.

p

'l
.’.4
a.
--=*.
I»;
o o
+—
-~
0
'.'

, . . .
,, fry; I an. H- «II H14
'. I‘JEE4.*.“"‘?1 rarelM'Gr.”

i‘
. I.
., .

 

9 ’9‘ A‘

.—

 

 

 

 

 

Chapter Page

Reliability of the Mann Inventory . . A7
Prediction of Driving Behavior . . . A9
Conviction-Collision Experience by

Behavioral Categories, Combined

Female and Male Populations . . . 51
Frequency of Conviction for Moving
Violations. . . 5“

Comparison of Collision Involvement

by Behavioral Categories, Combined

Female and Male Populations . . 61
Chargeable and Non-chargeable Colli-

sion Experience by Behavioral

Category . . . . . . . . . 65

Summary. . . . . . . . . . . 69

V. SUMMARY AND RECOMMENDATIONS . . . . . 73
Conclusions . . . . . . . . . 7A
Discussion . . . . . . . . . 77
Recommendations . . . . . . . . 8O
BIBLIOGRAPHY . . . . . . . . . . . . 82
APPENDICES . . . . . . . . . . . . . 90

iv

 

Table
A.l.

4.2.

14030

A.A.

l4050

A.6.

“.7.

“.8.

“.9.

 

 

LIST OF TABLES

Inter-Group Correlations for Response
to Items in the M. I.-Successive Test
Administrations--Females . . . .

Inter-Group Correlations for Response
to Items in the M. I.-Successive Test
Administrations--Males. . . . .

Cumulative Driving Record, by Behavioral
Category for the 26 Months Immediately
Following Satisfactory Completion of
Driver Education. . . . .

Conviction-Collision Experience by Be-
havioral Categories, Combined Male
and Female Populations. .

Frequency of Conviction for Moving Vio-
lations by Behavioral Category, Com-
bined Female and Male Populations .

Conviction Experience for Moving Vio-
lations by Behavioral Category--
Female Subjects ,

Conviction Experience for Moving Vio-
lations by Behavioral Categories
--Male Subjects . . . .

Collision Involvement by Behavioral
Categories--Combined Female and Male
Populations . . . .

Collision Involvement by Behavioral
Categories--Female Subjects

Collision Involvement be Behavioral
Categories--Male Subjects.

 

A8

A8

50

52

55

58

59

62

6A

66

 

 

 

Table

4.11. Chargeable and Non-Chargeable Collision
Involvement by Behavioral Category--
Female and Male Populations Com—
bined . . . . . . . .

A.l2. Chargeable and Non-Chargeable Collision
Involvement by Behavioral Category-—
Female Population . . . . . .

4.13. Chargeable and Non-Chargeable Collision

Involvement by Behavioral Category-—
Male Population. . . . . . .

vi

 

Page

67

70

71

 

 

 

 

 

LIST OF APPENDICES

 

Appendix Page
A. The Mann Inventory and Response Sheet . 91
B. Inventory Items Found to be Signifi-
cantly Related to Behavioral
Categories . . . . . . . . . 97

vii

 

 

CHAPTER I

THE PROBLEM

 

Identification of the unique characteristics of
the problem driver has been the subject of numerous
investigations employing a variety of approaches. Re-
search efforts have encompassed categories of sensory
and motor skills, age, intelligence, personal credit
rating, social adjustment, emotional stability and
personality factors, or combinations of these and
several other variables.

In reference to physical characteristics, the
Eno study,l Cobb,2 and Fletcher,3 have shown slight but

positive relationship between static visual acuity and

 

 

 

l"Personal Characteristics of Traffic—Accident
Repeaters," Eno Foundation for Highway Traffic Control,
Saugatuck, Conn., 19A8, pp. 6A.

2P. W. Cobb, "Automobile Driver Tests Administered
to 3,663 Persons in Connecticut, 1936-37, and the Re-
lation of the Test Scores to the Accident Sustained,"
Unpublished Report to Highway Research Board, Washington,
D. C., July, 1939.

3E. D. Fletcher, "Visual Acuity and Safe Driving
Ability," Journal of the American Optometric Association,

XX. 7 (19A9). pp- “39—AA2.

 

 

 

 

 

 

 

 

 

driving record. Burg,“ in an investigation of dynamic
visual acuity, static visual acuity, and driving record
reports a somewhat stronger relationship between dynamic
visual acuity and driving record than between static
visual acuity and driving record.

Finesilver5

reports no apparent relationship be-
tween auditory sensitivity and subsequent or prior

driving record.

6 7

Cobb and Lauer report low correlations between
motor skills and driving record. Lauer8 and Marshg demon-
strate a positive relationship between age, (identified

as under 25 and over 65), as a characteristic related to
disproportionate accident involvement. However, other

research indicates that the age differential is incidental

 

“Albert Burg, "An Investigation of Some Relation-
ships Between Dynamic Visual Acuity, Static Visual
Acuity and Driving Record," University of California,
Department of Engineering, April, 1964.

5S. G. Finesilver, "National Symposium of the
Deaf; Driving and Employability," University of Denver,
College of Law, February 12-1A, 1962.

6Cobb, op. cit.

7A. R. Lauer et al., "Aptitude Tests for Army
Motor Vehicle Operators," Report 981, Personnel Research
Section, TAGO, Department of Army, Washington, D. C.,

1952.

8A. R. Lauer, "Age and Sex in Relation to Acci-
dents," Bulletin 60, Highway Research Board, 1952.

9B. W. Marsh, "Aging and Driving," Traffic
Engineering Reprint, November, 1960.

 

 

MW' 1"“ W.'V.fliz"~5-T’""’ “ - ’. - . - ' :-"-..-.:t.4'.:m~.l. - : ifl.-*:r.z.-'*- ""'-r""' Jm-E'lr: “4*"- - ~ ,

 

 

 

 

and that the personality and behavior patterns of these

age groups are the real variables.

While no direct relationship between intelligence
and accident involvement can be readily identified,
Kremple,lo reports that persons with ten or more years
of formal education tend to be less susceptible to
accidents than those with fewer years of education.

Birnbachll

in using intelligence as one criterion, re-
ports no difference in the average intelligence rating
of either the accident repeater or the accident-free
group. The accident—free group, however, appears to
possess greater knowledge and understanding of safety
concepts than do the repeaters.

Tillman and Hobbsl2 using residence, age and sex
as control factors, submitted the names and addresses of

96 male drivers who had experienced four or more colli-

sions and a control group of 100 accident—free drivers to

 

10R. J. Kremple, "A Study of Application Blank
Items to Discover Those Differentiating Accident-Free
from Accident—Prone Truck Drivers" (unpublished Master's
dissertation, Western Reserve University, Cleveland,
Ohio, 1951).

11s. B. Birnbach, "A Comparative Study of Accident
Repeater and Accident-Free Pupils" (unpublished Doctoral
dissertation, New York University, 19N7).

12W. A. Tillman and G. E. Hobbs, ”The Accident
Prone Automobile Driver," A Study of Psychiatric and
Social Background, American Journal of Psychiatry,
CVI. 5 (19A9), pp. 321-331.

 

 

juvenile court, adult court for records other than auto—

mobile accidents, public health agencies, venereal disease
clinics, local credit bureau and three other social agen—
cies. Their findings revealed that 66 per cent of the
high accident group were known to one or more agencies;
while among the control group, only nine were known to

any one of the agencies and in no instance was one of
these persons known to more than one agency.

Heathl3 in a comparative study employing personal
data and the Thurston Temperment Schedule reports that of
763 traffic offenders who had a record of at least three
offenses and a control group of 195 subjects who had a
record of no accidents or violations that the repeaters
scored much higher in the area of impulsiveness and on
the social scales. He concludes, therefore, that
emotional factors were the major determinants of one's
driving record. In summary, Brodylu indicates that safe,
lawful, courteous driving is primarily a problem of
emotional makeup and social adequacy.

It appears then that the greatest potential for the

identification or prediction of problem drivers or accident

 

13E. D. Heath, "The Relationships Between Driving

Records, Selected Personality Characteristics, and Bio-
graphical Data of Traffic Offenders and Nonoffenders,"
The Traffic Safety Library, Illinois State University.

 

l“Leon Brody, "Personal Characteristics of Chronic
Violators and Accident Repeaters," Highway Research
Board, Bulletin No. 152, National Academy of Sciences,
National Research Council.

 

 

 

 

repeaters in general rests in the investigation of factors

of emotional stability and personality variables.

Statement of the Problem

 

The primary purpose of this study is to investigate
the validity and reliability of the Mann Inventory, (here-
after referred to as the M. I. or the "Inventory,") as
an objective measure of selected behavioral categories.

The secondary purpose of the investigation is to
determine the relationship, if any, between the several
categories and subsequent accrued driving record--

collisions and/or convictions for moving violations.

Importance of the Problem

Death and injuries resulting from motor vehicle
mishaps have risen steadily since the late 1950's. In
spite of improvements in highway and vehicle design,
increased enforcement and educational effort, and con—
tinuous exhortation in both newsprint and other mass media
to drive carefully, approximately 50,000 fatalities and
1 3/4 million injuries are experienced annually. This
fact becomes even more significant when one realizes that
the death rate per 100 million miles driven has increased
from 5.2 in 1961 to 5.7 in 1964, and the rate per 100,000

population has risen from 21.3 in 1958 to 2u.9 in 19614.15

 

15Accident Facts, 1965 edition, National Safety
Council, Statistics Division, p. 9.

 

 

 

Studies investigating the causes of motor vehicle
accidents almost invariably demonstrate that most are
attributable to human failures of various kinds. In
recent years the area of personality variables, atti—
tudes, and the quality of personal adjustment has been
emphasized as important to safe driving. An especially
promising approach can be summarized by, "We drive as
we live, and high accident rates may reflect our patterns
of adjustment to modern life. Also, the attitudes and
values held by accident repeaters appear to differ from
those of safe drivers."16

If we accept as significant indicators those per-
sonality characteristics most frequently identified among
accident repeaters, it appears that they are basically
emotionally and socially maladjusted and resentful of
authority. To a somewhat lesser degree they lack self-
discipline, have an irregular job history, are immature
in their thinking and planning for the future, and have
had an over-authoritative mother or father or parent
figure. Possibly another indication of social-emotional
maladjustment is evidence of an uncooperative attitude.

Assuming that one or a combination of these charac-
teristics identifies the potential traffic violator——

accident involvee/or' accident repeater, then a major

 

16Ross A. McFarland, The Major Strategies in the

Prevention of Highwayglnjuries and Fatalities, Harvard
School of Public Health, pp. 3-A.

IA‘.. _ _'-

__AL

- ~54 d
'lv‘:’:‘“";f"":‘.:a «r yrtyo . v V O. [cf-#1 ' 0..
. . 4 _ .

1 .‘t:..'l‘ .{45 ' -I‘- .‘

I
l O. o- a...
0 0'0-"
(*A

”4-1'

 

 

 

consideration appears to be the need for the development

of some method to identify him before an accident pattern
develops. Credence is given to the possibility of such

17 when he reports that "signifi-

a procedure by Brazell,
cantly higher percentages of students who received low
attitude grades in driver education courses had accidents
and moving violations than did students who received high
attitude grades." (The above attitudinal grade was based
on teacher evaluation of student behavior.)

Further evidence tends to indicate that it is
possible to identify a certain percentage of the extreme
accident repeaters or traffic violators out of a total
test group by using various paper and pencil instruments,

plus interview techniques and/or case studies. However,

with the possible exception of the Psychological Inventory

 

developed by Dr. Charles F. Hanner,18 there does not ap-
pear to be any single instrument or method that can be
administered to an individual that will justify conclud-
ing that he will be involved in accidents, although it
may be stated that he possesses many characteristics that

indicate a propensity toward accident susceptibility.

 

17Robert E. Brazell, A Comparison of Various Behind-
the-Wheel Training Methods, Safety and Traffic Division,
Automobile Club of Michigan, 1962, p. 37.

18Charles F. Hanner, "Use of Psychological Inven-
tory in Writing Insurance for Youthful Male Drivers,"
Traffic Safety Research Review, National Safety Council,
VII, 1 TMarch, 1963), pp. 5-9.

 

 

 

Basic Assumptions

The investigation of this problem is based on the
following assumptions:

An understanding of behavior is essential for
identifying a majority of the underlying causes of
traffic collisions and violations.

Raters with an academic background and professional
experiences in the behavioral sciences can predict a
youthful driver's likelihood of accruing a traffic record
of violations and/or collisions on the basis of behavior
observed over a time period of at least 18 weeks.

Behavioral categories for predicting driver perform-
ance can be identified by response to items incorporated

in the "Inventory."

The Hypotheses

 

The hypotheses to be tested in this study are:

Hol.-—There is a significant relationship between
the rater's classification of subjects on the basis of
observed behavior and the subject's response to items in
the M. I.

H02.--There is a significant positive relationship
between response to items in the Mann Inventory, by be-

 

havioral categories, when the test is administered to
two sample groups drawn from the same population.
H03.--There is a significant positive relationship

between the rater's classification of subjects on the

 

 

~t-.~T- .‘~

" .' -".'
‘e- "’ .

 

 

 

basis of observed behavior and the subjects subsequent
driving record during the 26 months immediately follow-

ing satisfactory completion of driver education.

Definition of Terms

 

Conviction.--Legal action arising out of the
issuance and disposition of a citation for a moving Vio-
lation other than one received at the time of a collision.
(A conviction arising out of a collision will be listed
under the category--chargeable accident.)

Chargeable Accident.--An accident involvement for

 

which the driver record file indicates the issuance of a
citation and conviction of a moving violation arising out
of the involvement.

Non-Chargeable Accident.-—An accident involvement

 

for which no enforcement action was taken against the

individual involved.

Organization of the Study

 

The general plan of the study is to present in
Chapter Two, the review of literature on the relation-
ship of personality and personal/social adjustment to
driving performance. The third Chapter is an account
of the methodology used in collecting, organizing and
tabulating the data, and the techniques employed in
analyzing them. The results of the analysis are reported
in Chapter Four. The summary, conclusions and impli-

cations for further study appear in Chapter Five.

 

 

 

 

CHAPTER II

REVIEW OF LITERATURE ON THE RELATIONSHIP
OF PERSONALITY AND PERSONAL SOCIAL

ADJUSTMENT TO ACCIDENT INVOLVEMENT

A review of literature reveals that man has at-
tempted for many centuries to explain apparent unequal
vulnerability to misfortune, (accidents). In fact, con-
cern for this phenomenon is found in both religion and
methology.l

The search for an explanation of accident occur—
rence has progressed from the concept of "Act of God"
to simple data collection of observable physical factors
to the identification of unsafe behavior as a primary
factor in accident causation.

From the early 1920's2’3 until the late 1950's“

it was believed by many investigators that the accident

 

 

 

lWilliam Haddon, Edward Suchman and David Klein,
Accident Research Methods and Approaches (New York:
Harper and Row, 196H), p. 386.

 

2Major Greenwood and Hilda Woods, "The Incidence
of Industrial Accidents with Special Reference to Multi-
ple Accidents," Medical Research Committee, Industrial
Fatigue Research Board, Report No. A, 1919.

 

 

3Major Greenwood and G. Yule, "An Inquiry Into
the Nature of Frequency Distributions Representative of

IO

 

 

 

W III: ' "H ' WWW- ". 'T
m r't .

 

 

 

 

 

ll

proneness concept could explain man's involvement in
accident producing situations.

The work of Adelstein,5 however, demonstrates the
weakness of the accident proneness concept when the fac-
tors of time and population at risk are controlled. His
findings, with these controls, demonstrated that when
time and exposures to risk were controlled, the accident
rate for subsequent years actually increased when those
men who had experienced the greatest number of accidents
during the first year of a three-year study were removed
from second and third year computations.

Miller,6 demonstrates statistically that repeated
accident involvement is an inadequate criteria for acci-
dent proneness. He illustrates mathematically that,

among a group of 1,000 employees experiencing 1,500

 

Multiple Happenings with Particular Reference to the
Occurrence of Multiple Attacks of Disease or of Repeated
Accidents," Journal of the Royal Statistical Society,
LXXXIII (1920), pp. 255-279.

“Tillman and Hobbs, op. cit., pp. 321-331.

 

5A. M. Adelstein, "Accident Proneness: A Criti—
cism of the Concept Based Upon An Analysis of Shunters'
Accidents," Journal of the Royal Statistical Society,
Series A, CXV (19527: pp. 35H-AOO.

6Gene Miller, "Accident Repeaters May Not Be
Accident-Prone," National Safety News, LXVII (March,

1953), pp- 3-5-

 

 

 

l2

accidents within a given time period that 223 persons

might be accident free. Conversely on the basis of chance,
one person could be involved in as many as seven mishaps,
and four others in six mishaps.

On the basis of the research reviewed it cannot be
stated conclusively that accident proneness does not
exist, however, it appears that proneness has received
undue attention as an explanation of accident causation.

If one accepts accidents as a form of human behavior,
a more promising conceptualization of the accident pheno—
menon appears to lie in the behavioral or adjustive ap-
proach involving both psychological and social factors.

. . . We begin by assuming that the accident did

not simply happen by chance, but that it developed

through the same stages as any form of human be—

havior and was characterized by the same kind of
social and psychological forces as determine other
types of social action. Once we view it in this
fashion, we may analyze the accident situation for
the kinds of decision-making processes involved,

the motivational forces, the cultural influences,

the role of social interaction, etc. In other

words, we can now approach accidents as behavior

and apply the same explanatory models that the be-

havioral sciences have developed for understanding

human behavior in general.

If we discount the fact that much of the early work
in the field of accident prevention was conducted to prove
or disprove the accident proneness concept and concen-

trate instead on psychological characteristics and social

adjustment, we find upon examination that the research

 

7Herbert H. Jacobs and others, Behavioral Approaches
to Accident Research (New York: Association for the Aid
of Crippled Children, 1961), p. N3.

 

 

 

 

 

 

l3

 

reveals a consistent pattern of digression from accepted
norms of behavior among accident involvees. Further evi—
dence tends to demonstrate identifiable personality vari-
ables. Dunbar8 describes accident repeaters as impul-
sive, impetuous people who concentrate on immediate
pleasures and satisfactions. They are resentful of
authority, essentially rebellious and cannot tolerate
even self-discipline. Their attitudes toward sex and
family are cavalier; they desire excitement and adven-
ture, do not like to plan for the future, and are apt to
act on the spur of the moment. They are sociable and
happy-go-lucky, concerned over their own health and ex-
press guilt over the lack of achievement. Dunbar further
states that:

. . patients who are accident prone typically do
not finish the educational courses which they under
take, have an unstable work record, experience many
changes in their Jobs coupled with ups and downs of
their income, are spontaneous and casual in their
social relationships and have superficially good
sexual adjustment, but are irresponsible toward
their sexual partners and families.9
Similar findings by Alexander indicate that:

The correlation between a tendency toward impul-
sive action and a lack of tolerance for external

or internal discipline on the one hand and acci-
dent proneness on the other appears to constitute

 

8F. Dunbar, §ynopsis of Psychosomatic Diagnosis

and Treatment as quoted by L. LeShan, "Dynamics in
Accident-Prone Behavior," Psychiatry, 1952, p. 73.

9F. Dunbar, "Medical Aspects of Accidents and Mis-
takes in the Industrial Army and in the Armed Forces,"
War Medical, IV, 1943, p. 167.

 

 

 

 

 

 

I!

 

14

a definite casual relationship. It is easily
understandable that an impulsive person, full of
hostility and guilt feelings, will be inclined to
have accidents. His actions are real and he has
at the same time also a tendency towards self-
punishment and suffering. He is incautious and
at the same time is inclined to pay a price in
the form of physical injury for his aggression.

Wong and Hobbsll in investigating personality

lO

characteristics among patients hospitalized as a result
of accident involvement found the same characteristics

in high accident groups in industry. Seven personality
traits described as being characteristic of patients with
multiple accidents were studied. They reported that the
high and low accident individuals differed markedly as a
group in their personality traits and social adjustments.
The high accident worker was frequently from a broken
home, showed evidence of conflict with authority in both
childhood and adulthood, experienced contact with the
Juvenile court, had an irregular work history, and a his-
tory of being fired. Further characteristics were marital
discord and contact with social agencies. The low acci-
dent worker rarely showed difficulty of this nature in
his life history. The authors concluded that aggressive-

ness, which lead to difficulties in their personal lives,

3?}.
.. .
. ‘
- -. .59

v . 1

O " $
.417. -..
Odludpo-

«I
0.6..-

" syn-c
”'3'.
a“. A

00v .‘ ‘,
'°".‘!"."-w.nl~

. . ...
111‘.
. I
0. ‘ " f ‘
J ' ‘fi-hfl- ' i-——-‘--

 

10F. Alexander, Psychosomatic Medicine as quoted
by L. LeShan, "Dynamics in Accident-Prone Behavior,"
Psychiatry, 1952, p. 73,

 

 

11w. A. Wong and G. E. Hobbs, "Personal Factors in

Industrial Accidents," Industrial Medicine and Sur er ,
XVIII (July, l9u9), pp. 291-295.

 

 

 

was the same factor that produced the high accident

record at work.

Tillman and Hobbsl2 report similar findings for
a group of 40 taxi—cab drivers representing the high and
low accident groups in the company they worked for. A
portion of their study consisted of obtaining the per—
sonal history of each driver, including his parental
background, childhood and adolescent history and subse-
quent adult adjustment. Additional information was ob-
tained from police files. This information was then com-
pared with the individual's driving records. The group
being tested though small produced highly significant
results. The high and low accident groups differed
markedly in their personality characteristics. The high
accident group showed marked intolerance for and aggres-
sion against any authority dating from early childhood
experiences in an unstable home background. In many in—
stances this personality feature was evident in anti-social
behavior. The high accident record itself was one mani-
festation of these personality characteristics. Among the
high accident taxi drivers the following characteristics
were evident:

l. A high incidence of parental divorce rate

predominated. One or both parents were de-

scribed as excessively strict. The father

 

l2Tillman and Hobbs, op. cit.

 

 

16

was often a poor provider with a record of
heavy drinking.

In the area of childhood adjustment ll of the
repeaters exhibited instability of an aggres-
sive type, such as temper tantrums, frequent
fights, bully characteristics, leader of a
gang and frequent appearances at juvenile
court. An equal number displayed complete
opposite behavior manifested by regressive
tendencies.

Adjustment appeared to be average in terms of
academic standing, but frequent truancy and
disciplinary infractions were evidenced.
Social adjustment was lacking to a marked
degree. They appeared to have many acquain-
tances but few friends. Emotional attachments
were superficial and varied from day to day as
their impulsiveness dictated. In a group

they always attempted to be the center of
attention.

Behavioral patterns indicated immaturity, show-
ing little concern over their problems, and

personal distress tended to be accepted.

Since it was felt that taxi—cab drivers might not

be representative of the general driving public, the
names and addresses of two paired groups: (1) 96 male

drivers who had suffered four or more automobile

 

 

 

accidents, and (2) 100 drivers who had never experi-

enced an accident. The names of both groups were sub-
mitted to juvenile court officials, the adult court for
records other than automobile accidents, public health
agencies, venereal disease clinics, local credit bureau,
and to three social service agencies. In reply 66 per
cent of the high accident group were known to one or more
agencies. Among the 100 accident-free drivers only 11
per cent (9) were known to any of these agencies, and in
no instance was one of these persons known to more than
one agency.

In the high accident group, two persons were known
to all sources. Three were known to four sources, nine
to three sources, 60 to two sources, and 32 to one source.
In the high accident group, the credit bureau had con-
tacted 3A.3 per cent, social agencies 17.7 per cent, public
health venereal disease clinics 1u.u per cent, and the
adult court 3A.3 per cent exclusive of traffic charges,
and the juvenile court 16.6 per cent.

In the low accident group, the credit bureau knew
six individuals, the social agencies, adult court and
juvenile court each had contact with one individual.

This report indicates that social maladjustment is com-
mon among both the general driving public and taxi-cab

drivers who have experienced high accident frequency.l3

 

l3Ibid.

*—

 

 

Schulzinger states:

The likelihood of an accident increases fur—
ther with inexperience; with excessive speed; the
partaking of alcoholic beverages; physiological or
psychological fatigue; with certain specific
psychological conflicts; and with the degree or
intensity of such non—specific emotional factors
as anxiety, hostility, frustration, and guilt.

The chances for an accident are greatest if the
young man is irresponsible and maladjusted; grew

up in a maladjusted home; and was exposed in child—
hood to violence and strife, to over-authoritative
parents or parent figures, to separation or loss

of parents, and to accidents to himself, parents

or siblings. The probability of an accident also
grows with any increase in the duration or inten—
sity of exposure to environmental hazards.

An accident is least likely to occur in a
healthy well-adjusted young girl, age thirteen,
born and reared in a normal, secure and loving
home, in which both parents are well adjusted and
if she has not experienced or witnessed major or
frequent accidents in the past. The likelihood
of an accident decreases further with the hours
of night, the month of February, the mid-week,
the home environment and a minimal exposure to
hazards. Even with increased exposure and in-
creased physical, physiological or psychological
stress or strain, an accident is relatively fin—
1ikely to occur in this type of individual.1

In a study comparing accident repeaters and acci—

dent-free junior high school students, Birnbach15 employed

The Bell Adjustment Inventory, National Safety Education

Test, the New York State Physical Fitness Test for Boys,

and The Rogers Strength and Physical Capacity Test. He

reports that:

 

l“Morris S. Schulzinger, The Accident Syndrome;

 

The Genesis of Accidental Injury (Springfield, Illinois:

 

Charles C. Thomas Publishers, 1956), p. 213.

lSBirnbach, op. cit.

 

 

 

 

19

The repeater tends to be maladjusted in several
areas, comes from a home where there is a high
incidence of divorce and/or one in which at
least one of the parents is overly strict.

As reported by teachers, the repeater is care—
less in preparing his school work and lacks
industriousness. He seeks to avoid school as
much as possible and capitalizes on his in-
juries and illnessess to stay away. Truancy

is higher in this group than among accident
free groups. Teachers generally rate these
students poor in cooperativeness.

The repeater tends to be aggressive and attempts
to dominate by physical means. In sports and
games he is considered a poor sport due to his
non-cooperative attitude.

Physical adventure and participation in danger-
ous and strenuous undertakings are regularly
sought and enjoyed. Recklessness and eagerness
to take chances and impulsive actions seems to
best characterize the repeater. Due to his
poor emotional adjustment and home and family
relationships, it is difficult, he finds, to
meet distress placed upon him by family and
school.

Among repeaters, nail bitting is prevalent. He

is superstitious and given over to ritualism.

20

When frustrated, thwarted, or subjected to
emotional strain, the repeater becomes re-
bellious. He is rebellious of authority and
hates to be commanded. This resentment is
often shown by bringing harm to himself and
others. He shows no inhabitions, signs of
remorse or fear of punishment, and attempts
to regain prestige by physical powers, bully-
ing, acts of bravado, and by flaunting

authority.

The Eno study indicates a number of personality

characteristics differentiating violation/accident re—

peater groups from violation/accident free groups. The

outstanding differences were as follows:

1.

Repeaters are not as well informed regarding
safe driving practices and regulations as are
free. The lack of information tends to in-
crease among those with high accident fre-
quencies.

Repeaters tend to have more personality mal-
adjustment than free, and this condition tends
to increase among the more serious repeaters.
They have poor motor control under both normal
and frustrating conditions, and are more easily

upset by frustrating situations.

 

“WIT

vw" .

1. ?w~‘

w

J

'03 .‘E: -

 

 

 

 

21

 

3. Attitudes toward certain aspects of driving
and law enforcement are significantly poorer
among repeaters.

Brody, in summarizing the research project "Per-
sonal Characteristics of Chronic Violators and Accident
Repeaters," states:

Three general conclusions .

l. The problem of safe, lawful, and courteous
driving is primarily a problem of emotional
makeup and social adequacy. So-called psycho—
physical functions (reaction time, glare re—
covery time, etc.) do not, per se, differen-
tiate between good and bad drivers. The
latter may excel in these functions in many
instances, while the former may occasionally
be inferior without jeopardy to their driving
records.

2. With regard to the psychological (as distin-
guished from the psychophysical) factors noted
above, other research studies indicate that the
following specific characteristics tend to be
evidenced by chronic violators and accident
repeaters: they are apt to be aggressive and
intolerant of others. They tend to resent
authority. They are inclined to have an
exaggerated opinion of their importance and
their abilities. They are likely to be lacking
in responsibility and often act impulsively,
on the spur of the moment. The basis for such
characteristics is likely to be obscure. Just
as eight-ninths of an iceberg lies below the
surface of the water, most of the factors and
forces that shape an individual's personality
are hidden in his background, often in early
childhood experience.

3. Obviously, here is an extremely difficult and
complicated problem. It is not surprising,
therefore, that work at the New Jersey Clinic
and similar work elsewhere have not produced
simple formulas for detection or correction
of problem drivers. While the general impor-
tance of personal adjustments and personality
trends are indicated, it cannot be said with

 

 

 

 

 

 

 

 

22

assurance: use this or that test in screen-
ing drivers for licensing purposes or in driver
renexamination. But the development of such
tests remains one of the prime needs and ob-
jectives. Exoeriments toward that end will

be continued.

17

Schuster and Gilford in a three—part study com-

pared: (1) the better than average driver (a person

 

who had a clear driving record for at least three years
with no traffic citations or accident involvements),
(2) chronic violators (persons who were currently under
revocation for extremely poor driving records), and (3)
accident repeaters (persons who had been involved in
three or more accident involvements during the preceding
three years). They found that when compared with selected,
better than average drivers, the accident repeaters were
significantly: more self-assertive, more active, more
adventuresome, more sociable, more ambitious, younger,
and less self-reliant. The observed differences probably
reflect true differences in personality traits and atti-
tudes of accident repeaters.

Compared with the selected better than average
drivers, the chronic violators were significantly: less
tolerant of others, less objective (hyper—sensitive),

more adventuresome, less self-reliant, less friendly,

 

l6Brody, op. cit., pp. 1-2.

17B. H. Schuster and J. P. Gilford, An Analysis
of Accident Repeater and Chronic Violator Drivers, A
paper presented at the 47th National Safety Congress and
Exposition, October, 1959, Chicago, Illinois.

 

 

 

_
l

 

 

 

 

 

 

23

and less stable emotionally. They were further charac-
terized by fewer years of formal education and tended
to drive more during hours of darkness.

West18 in 1946, diagnosed accidents of teenagers as
lack of social responsibility. His observations include
the incorporating in the curriculum, preparation for life,
such as good citizenship and safety education necessary
to function in a modern technological society. It ap-
pears then that a function of education in this sense
must include effective psychiatric diagnosing and counsel-
ing for the accident repeaters.

19’20 reports a highly successful experience

Hanner
when using a psychological inventory developed by himself
and assistants to categorize male automobile insurees.
Subsequent driving experience substantiated that it is
possible to classify drivers on the basis of such an in—

ventory. Additional cross validation using the Personal

History Data indicates that personal history data likewise

 

may be sued successfully, it does not, however, have as

 

18Roscoe L. West, "The Philosophy of the Accident
Problem," The Journal of Educational Sociology, October,
19u6: pp- 7fig77-

l9

 

Hanner, "Use of the Psychological Inventory . . .,"
op. cit., pp. 5-9.
20

Charles F. Hanner, "Use of Personal Data In
Underwriting Setting," Traffic Safety Research Review,
VII, 1 (March, 1963).

 

“ms“ ~

 

 

 

24

great a potential use as the inventory, except that when
the two are used together they may give even greater in-
sight into the potential driving behavior of insurees.
Perhaps the most ambitious research project under-
taken to date to determine the relationship between per-
sonality, behavioral factors and social factors, is that
reported by Lynnete Shaw in "The Practical Use of Pro-
jective Personality Tests as Accident Predictors,"21
employing the TAT and a counter-part developed by the
Industrial Psychological Services and Operational Re-

search Bureau, the SRT, Social Relations Test.

 

This project represents an effort to develop a
screening device for prospective bus drivers for PUTCO
(Public Utility Transportation Corporation, Johannesburg,
South Africa). It was conducted by the Industrial Psycho-
logical Services and the Operational Research Bureau,
Johannesburg, South Africa. The results of this ten year
study demonstrate a very strong relationship between re-
sponse to items in the Thematic Apperception Test--TAT--

and Social Responsibility Test--SRT--and driving record

 

when the tests are modified specifically for urban Afri-
can natives. The study further dramatically illustrates
the relationship between a good driving record, as

measured by company records of convictions for traffic

 

21Lynnete Shaw, "The Practical Use of Projective
Personality Tests as Accident Predictors," Traffic
Safety Research Review, IX, 2 (June, 1965), 35.

 

 

 

m w“ e .

 

/ \ 1' l

O */

|
.p."

. f

   
    

”i

d
It . _' ..

 

 

25

violations and collision involvement, and personal ad-
justment and behavior.

One of the illustrations consists of a group of
281 subjects, 51 of whom had had repeated accident in-
volvements over short periods of time. On the basis of

the SRT and TAT test scores it was found that 96 per cent

 

of these accident repeaters could have been accurately
predicted.
From 1956 to 1960, 120 drivers were; involved in ’“r
one accident and ten drivers in more than one accident
due to driving in a reckless or dangerous manner; 51 of
these drivers had, for one reason or another, been given
the previously identified personality tests. Upon exami-
nation of test scores it was found that 80 per cent of
this dangerous risk group had failed the TAT and SRT tests.
Miss Shaw categorizes accident risk into five broad
classifications described, in part, as follows:

The Potentially Bad Accident Risk

 

The badly integrated or maladjusted person.

The person who is emotionally unstable and ex-
tremistic.

The person with a distorted apperception of life
and a distorted sense of values.

The over-confident, self—assertive person.

The person who harbors grudges, grievances and
resentments.

The blame-avoidant person who is always ready
with excuses.

The intolerant and impatient person.

The inadequate person with a driving need to prove
himself.

The helpless and inadequate person who is con-
stantly in need of guidance and support.

 

 

26

The Potentially Poor Accident Risk

 

The person who displays little energy, stamina
or interest.

The person who exhibits the personality faults
of the bad accident risk, but in a less pro-
nounced manner.

Also included in this group is the person whose
faults or weaknesses are such that they could
possibly improve in time, but the learning
process is likely to be so expensive that he
is really a poor risk--especia11y as a
professional driver.

The Potentially_Borderline Accident Risk

The weak person who could be easily influenced.

The person who is nervous and lacking in self-
confidence.

The person who has unresolved conflicts, but who
gives evidence of a constructive, but as yet
unsuccessful, effort to make adequate adjust-
ments to these conflicts.

The person with a not-too-pronounced anxiety
neurosis. (This may make him ultracautious
at the moment but he must be regarded as a
doubtful risk, as he may deteriorate.)

The Potentiallijair Accident Risk

 

The person whose intentions are good, even if his
capabilities are not always commensurate with
his intentions.

The person who has certain weaknesses or unre-
solved conflicts which, although they will not
unduly upset the balance of his personality,
will nevertheless always impose a certain
strain on him and slightly impair his effici-
ency.

Also included in this group is the person whose
faults and weaknesses are such that they will
probably improve in time but who at the moment
cannot rise above them and who is currently
therefore only a fair risk.

The Potentially_Cood Accident Risk

 

The balanced and mature and well-controlled person
with a healthy and realistic outlook, satis-
factory interpersonal relations, a kindly and
tolerant attitude to others, a well developed
social and civic conscience and an ingrained
sense of responsibility.

 

 

 

 

 

 

 

27

The person who, as yet, cannot be said to be quite
mature but whose motivations are sound and who
demonstrates an ability to learn quickly by
experience and profit by his mistakes. (This
sort of person will undoubtedly have a learn-
ing period before his record stabilizes it-
self at a good level.)

The contented person who is in no way outstanding
but who is friendly, cheerful, adaptable and
accepting--provided he is reasonably intelli-
gent, realistic and mature.

The rather withdrawn introvert-—provided that he
is not too maladjusted [underlining added]

 

 

 

It would, in any case, be virtually impossible
to prepare a guide that would cover all contin-
gencies, for, one must stress again, it is the
totalgpersonality pattern that matters and parti-
cularly the balance and integration of that pattern.
This whole study has emphasized that, if this is
sound, and the person functions on a fairly even
keel, is well adjusted to his circumstances and
has learned to live with himself and his environ-
ment, then the prognosis is good. If there is a
pronounced imbalance of any kind, if the person is
poorly integrated, if his adjustments are adequate
or unrealistic, then the prognosis is bad.

 

The writer feels that, although there is no
doubt that the Good accident class does contain
some strange personality patterns, the study re-
ported here has lent confirmation to the fact that
the Good group nevertheless contains a preponder-
ance of individuals who are representative of the
average good citizen, in that, although they may
have their weaknesses, their limitations and their
faults, they are nevertheless reasonably well
balanced and well integrated people who have
developed an ethical attitude toward life and a
socially responsible code of behavior.22

Summary

Early studies generally attempted to demonstrate

accident proneness as the major factor in the occurrence

of injurious mishaps.

 

22Ibid., pp. an, 65, 69.

.4

 

 

 

Within the past two decades other theoretical
models and research have demonstrated the weaknesses of
the accident proneness concept.

A more promising approach, involving both psycho—
logical and social factors, appears to lie in the be-

havioral or adjustive model of accident prevention.

 

 

 

CHAPTER III
DESIGN AND METHODOLOGY
The Test Instrument

The M. 1. consists of 63 items that appear to re-
flect an individual's feelings toward himself, others,
and established social mores. Reaction to items in the
"Inventory" are expressed by checking one of five re-
sponses--always, usually, sometimes, rarely, or never
(see Appendix A for copies of the M. I. and the response
sheet).

The M. I. in its original form consisted of 100
items selected on the basis of face validity by Dr.
William Mann of the Highway Traffic Safety Center at
Michigan State University, East Lansing, Michigan. The
selected items represented a compilation of the feelings
expressed by 100 Michigan high school students toward
the police, school, cars, family, peers, personal ex-
pectations-~desires--habits, and society. Intensive case
studies including personal interviews had been conducted
with each of these 100 students subsequent to identifi-
cation by driver education teachers as the worst drivers

in their respective schools. The persons involved in the

29

 

 

30

gathering of this data were at the time enrolled in the
course, "Personality Factors in Traffic Safety," taught
by Dr. Mann and were pursuing advanced degrees in the
field of Traffic and Safety Education.

A pilot study to ascertain potentially selective
questions from among the list of 100 items was then con-
ducted. Twenty high school teachers of Driver Education
in central Michigan were requested to evaluate students
enrolled in their classes. Prior to initiating this
evaluative process the teachers were brought together
and criteria established for the identification of three
behavioral categories: (1) very aggressive, (2) very re-
served, (3) average. The criteria for evaluation were
as follows:

1. Very Aggressive: Any student who, in the
opinion of the driver education instructor
through personal observation in the classroom
and/or during practice driving instruction,
displays behavior that is exceedingly aggres—
sive, is a show-off, is extremely egotistical
or temperamental.

2. Very Reserved: Any student who, in the opinion
of the driver education instructor through per-
sonal observation in the classroom and/or during
practice driving instruction, displays behavior
that is exceedingly cautious and timid.

3. Average: All students who do not fall into
either of the other classifications.1

 

1John G. Schaff, "Personal Attitude Survey" (unpub—
lished Master's dissertation, Michigan State University,
1957).

 

 

31

The "Inventory" was then administered to a sample
population of 451 students. (Figures are not available
as to the numbers of male and female students involved.)
As a result of observations employing the established
criteria, 80 students were identified as very aggressive,
86 as very reserved, and 285 as average.

The response to each of the items were tabulated
for each of the three identified categories, and signifi-
cant differences were determined by chi square analysis.
The results of this analysis disclosed that 18 items
appeared to discriminate between the three categories:
four at the .10 level, eight at the .05 level, one at
the .02 level, four at the .01 level, and one at the .001
level of significance. Thirty—seven items were deleted
from the "Inventory" since nearly all students gave identi-
cal responses.

An adjustment scale was established by determining
the response(s) of 85 per cent of the Average population
to each of the remaining 63 items. Analysis revealed
that the response of 85 per cent of the Average popu-
lation deviated from this response pattern by 7—19 points.

The remaining 63 items were edited on the basis of
the pilot study. These items make up the instrument now
identified as the M. I.

The M. I. was subsequently administered to three

successive classes-—a total of 42 persons-—who had been

 

 

 

 

32

referred to the Ingham County Driver Safety School,
Lansing, Michigan, after having been cited and found
guilty or responsible, of/for, at least two moving vio-
lations and/or collisions within the preceding twelve
months. Using the criteria previously established, 15
persons were identified as very reserved, 21 as very
aggressive, and four as average. The remaining two indi-

viduals scored four and five of six lie items incor-

 

rectly and deviated by 46 points each on the adjustment
scale. Their response to significant items vacillated
from marked aggression to very reserved.

On the basis of the above research and other work
undertaken by the author, it appears that the M. I. may
be a helpful instrument for use in the identification of

basic behavior patterns and for predicting driving be-

 

havior.

Observation tends to indicate, however, that greater
discrimination of behavior is required than that employed
in the initial efforts, namely Very Aggressive, Very Re-
served and Average. As a result, the sample population
in this study shall be classified according to the follow—
ing six categories of observable behavior.

1. Behavior characterized by well adjusted inter-

action with persons and consistent with the
norms of the society in which the individual

lives.

 

 

33

2. Behavior generally characterized by satis-
factory interaction with persons and society,
but with periodic withdrawal from contact with
people.

3. Behavior generally characterized by satis-
factory interaction with persons and society,
but with periodic efforts toward assertive
action.

4. Behavior characterized by forceful, outgoing
action or vigorous efforts to assert oneself
over others.

5. Behavior characterized by withdrawal from con-
tact with other persons.

6. Behavior characterized by a pendulum effect,
vacillating between extremes of aggression and

withdrawal.

Sample Selection

 

The sample population of this study consists of
1057 subjects, 523 males and 534 females who were enrolled
in driver education at Sexton High School, Lansing, Michi-
gan between September, 1958 and June, 1960.

All of the sample population had: (1) enrolled in

and satisfactorily completed driver education during the

 

periods of September, 1958, to June, 1959, or September,
1959, to June, 1960, (2) attained a minimum age of 16

years or a maximum age of 18 years by time of course

 

 

 

 

   

r' I /J ,/

_.' ;!
/
. /
(m4: ‘

     

J

 

_.___.u———

 

/
1'
34

completion, (3) enrolled in grade nine through twelve
or had drOpped out of school upon reaching his six-
teenth birthday after having completed the eighth grade.

With an enrollment of approximately 2,000 students,
Sexton High School affords a comprehensive cross-section
of the ethnic and socio—economic groups of both the city
of Lansing and the surrounding rural areas.

Unlike many schools of similar size, the policy
in driver education at Sexton is to assign students to a
single instructor for all phases of the program. In—
structional phases——classroom, multiple car method of
instruction, and on-street experience--take place con-
currently over an 18-week semester. This system provides
the instructor greater opportunity to become more closely
acquainted with students and to evaluate their behavior
than is typically true in a program of comparable size.
This type of scheduling also helps to overcome one of the
potential weaknesses of the initial evaluation of the
M. I. in that the evaluation of student behavior is limited
to two raters, rather than to the possible variations in
interpretation and classification of the 20 persons COOper-
ating in the previously mentioned pilot study.

The sample population is atypical in that a large
percentage of both dropouts and graduates, tend to remain
in the Lansing area. This is due to the presence of

Michigan State University, the Lansing Community College,

 

 

 

 

 

 

35

the state capitol and associated governmental agencies,
and an extensive industrial complex. However, as a re-
sult of this tendency to remain in the community, there
is also greater probability of accumulating a record of
either convictions or collisions within the city.
If a relationship between observed behavior and
a record of moving violations and/or collisions is to r
be identified, an effective enforcement and records
system is essential. The driver record file of the “
Lansing Police Department; which were made available
for this research project, have been acclaimed nationally

for their effectiveness.

The Raters

 

Each of the raters had several years of teaching
experience, including work at the junior high, senior
high and college levels. During the investigation both
raters were employed as full—time teachers of Driver
Education at Sexton High School.

Both raters had earned Master of Science degrees
in Traffic and Safety Education with emphasis in the be-
havioral sciences.

Both raters were pursuing doctoral degrees at
Michigan State University. One was enrolled in Traffic
Safety-Teacher Education with a cognate area in Counsel-

ing and Guidance, and the second was admitted to a

 

 

 

36

 

program in Special Education—Educational Deprivation-
Emotionally Disturbed.

The raters had conducted Driver Improvement
sessions for the Ingham County Driver Improvement pro-
gram for at least one year prior to the investigation.
Persons were directed to this program by the local
Traffic Court Judges and the Driver Licensing Division,
Secretary of State's Office after having accrued an un-
satisfactory driving record.

As a result of professional preparation and experi-
ence both raters were sensitive to the kinds of behavior
that are manifested by persons likely to become involved

in traffic mishaps.

The Data
Two types of data were gathered on each student:
(1) response to items in the M. I. by behavioral category,
and (2) individual driving records.

Response to Items in the
M. I. by Behavioral

Category

The behavior of all students was observed through-

 

out the semester by the raters. These observations took
place while in the driver education classroom, during
instruction on the off-street driving facility, while
receiving on—street driving instruction, or anytime

that circumstances brought the student and rater

 

 

A

 

 

 

 

  
  
   

4 r ' ‘*
.ff ” A // '
'7
III.
I i
U
. ‘l

/
I
\,

-‘k-‘m~ .u 4-

 

A

 

37

together. Where conflicting information or observations
led to a question of the student's basic behavior pat-
terns, consultation was held with the school counselors
and other members of the teaching staff and/or reference
was made to the student's permanent record file in an
effort to gain additional data.

The M. I. was administered to each student during
the last week of a semester course in driver education.
Twenty minutes were allocated for completion of the
"Inventory." Students were requested to answer all items
as honestly as possible with assurance given that the re—
sults would not be used for grading purposes.

After weighing all available information, each
individual response sheet was assigned to one of the six
behavioral sub-groups; each sub-group was further classi-
fied as to male and female.

Individual Driver License
Record

The individual driving records--convictions and/or
collisions--were obtained from the driver record file of
the Lansing Police Department. The decision to use the
driver record file of the Lansing Police Department was
made in light of three factors:

1. A high enforcement index, approximately 25-35,

in any given year, and a police force and re-

cord file acclaimed for their effectiveness.

 

 

38

Since a majority of students enrolled in
driver education at Sexton High School are
between the ages of 15 years, 9 months, and
16 years, 6 months, at the time they com-
plete driver education, it is assumed that a
major portion of their driving experience
during the first two years will take place
within the city limits of Lansing, Michigan.
At the time of the study a paradox existed
which since has been rectified. Persons be-
tween 16 and 17 years of age could be charged
with moving traffic violations; however, such
cases were heard in probate court. Upon con—
viction for a traffic offense final disposi-
tion was posted in the police department driver
record files. Such a record of conviction
would not be found in the driver record file
of the Secretary of State's office, due to a
Michigan statute which precluded posting such
records to the file. Hence, a more accurate
summary of an individual's driving record
during the first two years could be obtained

from the local police department.

Procedure of Analysis

 

Three procedures were utilized in the analysis of

the data:

(1) chi square test of significance, (2) the

  

  
 

.. I w 7

.’ . .
I I .
| .

.. 4': ‘.

I l 'ri
-v. .‘.

' .0 :
I " 1|. -.. x

r I '..

v- " 1‘, .

,/

,/ x
/
/

 

 

 

39

cell square contingency method, and (3) product-moment

coefficient of correlation.

Chi Square Test of
Significance2

 

 

The chi square test of significance was employed
to determine the difference between the responses of the
six behavioral categories to the 63 items in the M. I. E—

The Cell Square E
Contingency3

 

 

The cell square contingency was employed to deter-
mine which response or responses, by behavioral category,
contributed most significantly to the total chi square
value for each "Inventory" item.

Product-momentyCoefficient
of Correlation‘l

 

The product—moment coefficient of correlation was
employed to determine the reliability of the M. I. Appli-
cation of the product—moment coefficient correlation is
in effect the corrollary of the rational of the Spearman-

Brown split half reliability coefficient. Two separate

 

2Sidney Siegel, Non-Parametric Statistics (New York:
McGraw-Hill Book Company, Inc., 1956), pp. 174-179.

 

3J. P. Guilford, Fundamental Statistics in ngchology
and Education (New York: McGraw-Hill Book Company, Inc.,

1950), pp. 368-369.
“Allen L. Edwards, Statistical Methods for the Be-

havioral Sciences (New York: Rinehart and Company, Inc.,
1954), pp. 142—177.

 

 

 

 

 

 

 

40

administrations of the same test are compared rather than

split half or odd—even items.

Null Hypothesis

The following are restatements of the hypothesis
in null form, i.e., the hypothesis that there is not a
true difference between the six behavioral categories as
they relate to the criterion.
There is no significant positive relationship
between the rater's classification of subjects
on the basis of observed behavior and the sub-
ject's response to items in the M. 1.
Ho There is no significant positive relationship
between response to items in the M. I. by be-
havioral category, when the test is adminis-
tered to two sample groups drawn from the same
population.
Ho There is no significant positive relationship
between the rater's classification of subjects
by behavioral category and subsequent driving
record accrued during the 26 months immedi-

ately following satisfactory completion of
driver education.

Summary

The sample population was drawn from Sexton High
School, Lansing, Michigan. Response to items in the M. I.
were collected upon completion of Driver Education during
the school years of 1958-59 and 1959-60.

Driver license data for the sample was collected
from the driver record file of the Lansing Police Depart—
ment for a period of 26 months immediately following

satisfactory completion of Driver Education.

.V V‘-‘

I“ '
I / .
‘hv.
.““
V.
\‘ t '.
\ ‘ r._.
\.. s ‘ 2
I ‘ ‘_
‘.
.
,
. ' 1.. _ - 1—
.___‘| H I § '- . .,l ._ .

   

/ 7

.

-._

as.

1'
’7‘ ' - A
‘.

   

 

3 -— [mi —Jl - — -_ E‘I’ . .

41

Chi square analysis was employed to determine
significant differences in reSponse to each of the 63
items. The .10 level of significance shall be employed
to determine the acceptance or rejection of responses
by behavioral category to each of the 63 items.

The cell square contingency was employed to deter-
mine the most significant response or responses by be-
havioral category to each of the "Inventory" items. The
.05 level of significance shall be employed to determine
specific response or responses to items by behavioral
category.

The product—moment coefficient of correlation was
employed to determine the reliability of the instrument
when comparing one administration of the test to a second
administration. A correlation of .50 shall be required
for the acceptance of reliability coefficient. (For 60
degrees of freedom, Guilford indicates that a correlation

of .250 is significant at the .05 level.)5

 

5Guilford, op. cit., p. 610.

 

 

 

 

 

 

 

CHAPTER IV

STATISTICAL ANALYSIS

In this chapter are presented the analysis of:

(l) the relationship between the rater's classification

_ ”"1

of subjects on the basis of observed behavior and the

P.—
I

subject's response to items in the M. I., (2) the relia-
bility of response to items in the M. I. by behavioral
category when comparing one test administration to an-
other, (3) the relationship between the rater's classifi-
cation of subjects by category and subsequent accrued
driving record during the period of 26 months immediately
following satisfactory completion of Driver Education.
Validity of the Instrument: Difference

Between Response to Items in the
M. I. by Behavioral Categories

 

 

 

The null hypothesis tested for each item in the
M. I. was:

Hol There is no significant positive relationship
between the rater's classification of subjects
on the basis of observed behavior and the sub-
ject's response to items in the M. I.

When the response of the combined male and female

populations are analyzed by behavioral category, chi

42

 

 

 

 

 

 

 

 

 

 

square values for 60 of the 63 items are found to be

significant at or above the .10 level.

Response by behavioral category to items 2, 30,
and 60 are not significantly different than zero at the
.10 level; thus, the null hypothesis of no significant
difference between response to items in the M. I. by
behavioral category must be accepted for these three
items.

For the remaining 60 items the null hypothesis
must be rejected since response to items are signifi-
cantly different than zero at the following levels:

1. Response by behavioral category to items 10,

21, 23, 26, and 31 are significantly different
than zero at the .10 level.

2. Response by behavioral categories to items 46
and 58 are significantly different than zero
at the .05 level.

3. Response by behavioral categories to items 19
and 63 are significantly different than zero
at the .02 level.

4. Response by behavioral categories to items 7,
28, 34, and 38 are significantly different than
zero at the .01 level.

5. Response by behavioral categories to the re-
maining 47 items is significantly different

than zero at or beyond the .001 level.

 

 

 

 

 

 

44

Employing chi square analysis with the behavioral
categories further divided according to sex identifies
37 items that are significant at or above the .10 level
for the female population. Behavioral categories 4 and
6 do not appear in the analysis of data for the female
population. In terms of course evaluation——pass or fail
--both raters appear to have exhibited an identical bias
in that females who were classified into either of the
aforementioned categories failed in such disportionate
numbers that an insufficient female population remained
for applicable statistical treatment.

Response of the female population by behavioral
category to 26 items, numbers: 2, 6, 10, 17, 18, 19, 20,
22, 23, 26, 27, 3o, 31, 34, 35, 39, 40, 42, 50, 51, 53.
56, 58, 60, 62, and 63 are not significantly different
than zero; thus, the null hypothesis of no significant
difference between response to items in the M. I. by
category must be accepted for these 26 items.

For the remaining 37 items the null hypothesis must
be rejected since response to items are significantly
different than zero at the following levels:

1. Response by behavioral categories to items 4,

7, 15, 25, 28, 37, 38, 46, and 49 are signifi-
cantly different than zero at the .10 level.

2. Response by behavioral categories to items 8,

12, 16, 21, 41, and 59 are significantly

different than zero at the .05 level.

 

 

 

 

 

45

3. Response by behavioral categories ot items ll,
14, and 57 are significantly different than
zero at the .02 level.

4. Response by behavioral categories to items 3,
5, 13, 36, and 47 are significantly different
than zero at the .01 level.

5. Response by behavioral categories to items 1,
9, 24, 29, 32, 33, 43, 44, 45, 48, 52, 54,

55, and 61 are significantly different than 2-

 

zero at the .001 level or above.

Analysis of the 63 items for the male population
identifies 47 items significant at or above the .10 level.

Response of the male population by behavioral cate-
gory to 16 items, numbers: 2, 15, 21, 23, 25, 26, 28, 29,
30, 31, 36, 38, 46, 55, 58, and 60 are not significantly
different than zero at the .10 level; thus, the null hy-
pothesis of no significant difference between response to
items in the M. I. by behavioral category must be accepted
for these 16 items.

For the remaining 47 items the null hypothesis must
be rejected since response to items are significantly
different than zero at the following levels:

1. Response by behavioral category to items 5, 10,

13, 19, 35, and 39 are significantly different

than zero at the .10 level.

 

 

 

46

2. Response by behavioral category to items ll,

12, 14, 32, 34, 39, 45, 51, and 52 are signifi-
cantly different than zero at the .05 level.

3. Response by behavioral category to items 20

and 24 are significantly different than zero
at the .02 level.

4. Response by behavioral category to items 4, 6,
8, 9, 17, 22, 27, 37, 41, 53, 56, 61, and 62
are significantly different than zero at the
.01 level.

5. Response be behavioral categories to items 1,

3, 16, 18, 33, 4o, 42, 43, 44, 47, 48, 49, 5o,
54, 47, and 59 are significantly different than
zero at the .001 level or above.

Through the use of the cell square contingency matrix
it is possible to identify a specific response or responses
that contribute most significantly and positively to the
total chi square value for each of the significant items.

Investigation reveals that 20 to 39 items are signifi-
cantly related to the various behavioral categories. The
number of related items by behavioral category and sex is

shown below.1

 

1For specific items significantly related to each
behavioral category, see Appendix B.

 

 

 

 

 

 

47

For females:

Category 1 = 28 items
Category 2 = 25 items
Category 3 = 36 items
Category 4 = 20 items
For males:
Category 1 = 25 items
Category 2 = 36 items
Category 3 = 31 items
Category 4 = 39 items
Category 5 = 39 items
Category 6 = 20 items

Reliability of the M. I.

 

Ho There is no significant positive relationship
between response to items in the M. I. by be-
havioral category, when the test is adminis-
tered to different groups drawn from the same
population.

At 63 degrees of freedom, with coefficients of
correlation not less than .755 and .697 for females and
males, respectively, reliability coefficients are signifi-
cant above the .01 level; thus, the null hypothesis must
be rejected since the correlation coefficients are signifi-
cantly different than zero above the .05 level criterion
value of .250.

Inter-Group Correlations for Females, Table 4.1,
and Inter—Group Correlations for Males, Table 4.2, demon-
strate the level of correlation derived by use of the

product-moment correlation coefficient, when test adminis-

trations are compared for successive years.

 

 

48

TABLE 4.1.-—Inter-group correlations for response to items
in the M. I.-successive test administrations--females.

 

Category Always Usually Sometimes Rarely Never

 

.986 .976 .968 .961 .985
.970 .901 .887 .893 .965
.916 .785 .771 .828 .861
.906 .755 .810 .768 .896

 

TABLE 4.2.—-Inter-group correlations for response to items
in the M. I.-successive test administrations-—males.

 

Category Always Usually Sometimes Rarely Never

 

.t

O\U'l

.965 .925 .942 .944 .964
.951 .872 .850 .809 .907
.934 .859 .858 .898 .887
.860 .764 .805 .710 .849
.921 .754 .697 .808 .887
.952 .879 .848 .808 .901

 

Guilford states:

Taking correlations just at large, without particular
regard to their use and as a general orientation, we
may say that the strength of relationship can be
described roughly as follows for various r's:

Less than .20 . . . Slight; almost negligible

relationship
.20 - .40 . . . Low correlation; definite
but small relationship
.40 - .70 . . . Moderate correlation;

substantial relationship

-w_.5__— .‘72'.

--._. .
.‘rfi

 

 

 

 

 

149

.70 - .90 . . . High correlation; marked
relationship

.90 - 1.00 . . . Very high correlation;
very dependable relation-
ship2

Application of Guilford's constraints to the data
in Tables 4.1 and 4.2 demonstrates a high or very high

correlation in all categories.

Prediction of Driving Behavior
Ho There is no significant positive relationship
between the rater's classification of subjects
by behavioral category and subsequent driving
record accrued during the 26 months immedi-
ately following satisfactory completion of
driver education.

If an individual's adjustment provides an index to
his future behavior as a driver, a relationship should
exist between his behavior and his driving record as mea—
sured by collision involvement and/or convictions for
moving violations.

Investigation of H03 is an attempt to ascertain
whether or not the identification of persons by behavioral
category would be of value in predicting an individual's
driving record for some identified period of time.

In Table 4.3 is contained a summary of the driving
record of the sample population according to behavioral

category and sex for a period of 26 months immediately

following satisfactory completion of Driver Education as

 

2Guilford, op. cit., p. 165.

 

 

 

 

50

.pcmEo>Ho>cfi coamfiaaoo HmqulnmconHHHOo memmmnmnoncoc pcm mapmwwsmco neon go Upcooc m cue: mcomsmd go accede u a

muccEm>Ho>cH coamfiaaoo Hmuouuuzaco mconHHHoo mammowgmnolcoc do UHOocL m peas wcomsmq no access u c mucoEm>Ho>cH COHwHHHoo HmHOpII

maco mCOHmHHHoo manmmMmeo mo Uhoowp m cufiz mcompoa no Lopez: u z mpcmEo>Ho>cH :onHHHoo proullmconHHHOQ oLoE so Amv no chooop
- m spa: wcowhoa mo popes: u z mucmEo>Ho>cH conHHHoo AHV do Ucoomg m zpfiz mcompod mo LonEsc u q mucoEc>Ho>cfi :onHHHoo Hmpou
IImCOHmHHHOo CH cm>Ho>CH mCOmme mo Lopez: Hopou n x mmcofip0fi>coo do Lopszc HmpOpllmcofiu0H>coQ no GLOOoL m spa: mCOmLmQ no coped:
Haqu u h mucoEo>Ho>cH conHHHoo HauoullmCOHmHHHoo mLoE Lo AHV pcm mCOapoH>c00 oLoE Lo Amv mo Upooop m cpfiz wcomhoa no LopESC u H
mmcofiuoa>200 no 9096:: Hmuoullmcofimfiaaoo on pcm mcofipoH>COO once so Amv do Upcoos m Spas mcompmq mo Looszc u z mmcofipoa>
1:00 Mo amass: Hmpoullmcofip0H>COo oboe Lo Amv mo ULOomL m cufiz mCOmLmQ mo LooESC n o mpcme>Ho>cH coamfiaaoo HmHOpII.wconfiHHoo
mace Lo AHV 0cm coauo«>coo Adv no Upoooh m cufix wcomsoa mo pecan: n m mmcofimfiaaoo on .coHu0H>coo AHV go Upcomp m npfiz mcomhon
mo amass: n m mcofip0H>coo AHV mo Upooop m cue: mcowpmq mo Lopfizc n D mmCOHmHHHoo oLoE Lo AHV p59 .mCOHpoH>coo on spas mCOmLmQ

mo Lopez: u o meOHmfiHHoo on new mcofiuofi>coo on spas mCOmLmQ mo LopEsc HmpOp u m mmgomopwo CH mcomaoa mo Lopezc Hmpou u <

.onE u E .mHMEwm u m

 

 

eHue MH-OH MHINH NNIOH mH o:-mm 20Huoe em-mH smumH oaumm HHum m AH m-m s em 2
sum HnH mum elm s was m-m mIN NIH cum HIH H m mum o m a o Hsowooso
MHum mus one eHum mH smumH manm mH-m sHue mmueH mum H m 3-: mH mm s m stomoeso
muH mum mum sum HH mH-sH NH-OH mum o 2-“ o-e m m s-e HH em s
mH-m HH-OH ale mmum :H omumm msumm .mmneH Hmum oelmm mum w MH . mum s m: z m
o o mum o m m-m mu: 0 muH m-H mum H m H-H m s m 3 see some
m-H OH-m OHnm mus sH mmuwH seuHm mum THIH monoH HH-o NH Hm sue me em 2 w
o mum HIH sum .m on: mHum NIH an: HH-m muH m s mum Hm m: ..s m Hto memo
muH mu: mum :.m m MHIHH smuHm m-m sum HHum HIH mH oH OHuw mm :m z m stowooso
muH v-0 mum m-H m eHum HH-HH o o o elm m HH e-e om so a
o Hmuom mHumH mnH em mm|mm ssumm m-m mmueH sm-eH mum mm ms mmumm MMH 0mm 2 H stowocso
muH mHumH OH-OH sum am Hmumm meuos o OH-m OHum mum mm mm mmusm mmm mmm s
, . m
a o z 2 H x e H z o s m o o m < sow mMMMmsmmMm

 

.cofipwospm co>HHU
mo coapoHQEoo atonemmmfiumm mafionHom aflopMHpoEEH mcucoe em on» sou mzowmpmo Hmpofi>mzon an .pLOOmH mafi>fipp m>Hpmasesoll.m.z mqm<e

 

 

 

 

 

51

identified by the Driver Record Files of the Lansing,
Michigan, Police Department.

Empirical evidence appears to indicate a difference
between the driving records, convictions and/or collisions,
of those individuals identified with behavioral categories
1, 2, and 3,

Category 1. Behavior characterized by well adjusted
interaction with persons and consistent with the
norms of the society in which the individual lives.
Category 2. Behavior generally characterized by
satisfactory interaction with persons and society,
but with periodic withdrawal from contact with
people.

Category 3. Behavior generally characterized by
satisfactory interaction with persons and society,
but with periodic efforts toward assertive action.

and those persons whose observed behavior placed them in
behavioral categories 4, 5, and 6.

Category 4. Behavior characterized by forceful,
outgoing action or vigorous efforts to assert one—
self over others.

Category 5. Behavior characterized by withdrawal
from contact with other persons.

Category 6. Behavior characterized by a pendulum
effect, vacillating between extremes of aggression
and withdrawal.

Conviction-Collision Experience by Behavioral
Categories, Combined Femgle and

Male Populations

In Table 4.4 are contained the observed and expected

 

 

frequency of collision involvement and record of con—
victions for moving violations. Analysis of Table 4.4

indicates:

   

52

TABLE 4.4-—Conviction—collision experience by behavioral
categories, combined male and female populations.

 

Behavioral C E F

 

Categories B H I Total
3 “19 55 71 7 l9 2
cat' 1 b 365.4 52.6 67.8 26.6 33.1 27.7 573
a 135 14 23 4 3 2
cat° 2 b 115.4 16.6 21.4 8 4 10.4 8.7 181
a 80 8 15 10 ll 4 {—
cat' 3 b 81 6 11.8 15 1 5.9 7.4 6.2 128 3
a 9 4 9 7 9 1’4
cat' 4 b 33.2 4.8 6.2 2.4 3.0 2.5 52
a 24 10 3 11 6 11
cat° 5 b 41.4 5 9 7 7 3.0 3.7 3.1 65
a 7 6 4 10 13 18
cat 6 b 37 5.3 6 9 2.7 3.3 2.8 58
Total 674 97 125 49 61 51 1057
Cell Contingencies
Cat. 1 7.97 .11 .15 14.44 6.01 23.84
Cat. 2 3.33 .41 .12 2.31 5.27 5.16
Cat. 3 .03 1.22 .00 2.85 1.75 .78
Cat. 4 17.64 .13 1.26 8.82 12.00 52.90
Cat. 5 7.31 2.85 2.87 21.33 1.43 20.13
Cat. 6 24.32 .09 1.22 19.74 28.51 82.51
Total 60.49 4.81 5.62 69.49 54.97 185.32 = 380.70

 

30 df, chi square 380.70, sig. level p > .001

a = observed frequency; b = expected frequency;
B = (0) convictions, (0) collisions; C = (0) convictions,
(l) or more collisions; E = (1) conviction, (0) collisions;
F = (l) conviction, (1) or more collisions--total collision
involvement; H = (2) or more convictions, (0) collisions-—
total number of convictions; I = (2) or more convictions,
(l) or more collisions--total collision involvement.

 

 

 

53

Behavioral categories 2 and 3 accumulated a
driving record of convictions and/or collisions
(Column B) during the test period that is not
significantly different than expected by
chance.

The number of persons with a record of zero
convictions, but one or more collisions (Column
0), and the number of persons with a record of
one conviction, but zero collisions (Column E),
is not significantly different than expected by
chance for any one of the six behavioral cate-
gories.

The number of persons in category 1 with a
record free of convictions and/or collisions
(Column B), is significantly greater than ex-
pected by chance at the 5 per cent level of
significance. In contrast the number of per-
sons with a clear driving record in categories
4, 5, and 6 is significantly less than expected
by chance.

The number of persons in categories 4 and 6
with a record of one or more convictions, plus
collision involvement (Column F), or two or
more convictions without collision involvement
(Column H), is significantly greater than ex-

pected by chance.

 

 

54

5. a. The number of persons in behavioral cate-
gory 5 with a record of one conviction and
one or more collisions (Column F), was
significantly greater than expected by
chance.

b. To a lesser degree, but still in signifi—
cantly greater numbers than expected by
chance, persons in category 5 accumulated
a record of two or more convictions plus
collisions (Column 1).

c. The number of persons in category 5 with
two or more convictions, but without
collision involvement (Column H), is not
significantly different than expected.

With chi square of 380.70 (p > .001) the null hypothesis
of no significant difference in driving record as related
to observed behavior as defined in this investigation must
be rejected. The relatively low experience of multiple
convictions or collision involvement precludes chi square
analysis for female and male pOpulations on an independent
basis.

Frequency of Conviction for
Moving3Violations

 

In Table 4.5 are contained the observed and expected
frequencies of convictions for moving violations for the

combined female and male populations which are further

 

TABLE 4.5.--Frequency of conviction for moving violations
by behavioral category, combined female and male populations.

 

 

BenaViOral B 0 DG Total

Categories
a 419 55 99

cat' 1 b 365.4 52.6 155.0 573
a 135 14 32

cat' 2 b 115.4 16.6 49.0 181
a 80 8 40

Cat° 3 b 81.6 11.8 34.6 128
a 9 4 39

cat' 4 b 33.2 4.8 14.1 52
a 24 10 31

cat' 5 b 41.4 5.9 17.6 65
a 7 6 45

Cat' 6 b 37.0 5.3 15.7 58

Total 674 97 286 1057

Cell Contingencies

Cat. 1 7.97 .11 20.23 28.30

Cat. 2 3.33 .41 5.90 9.64

Cat. 3 .03 1.22 .84 2.09

Cat. 4 17.64 .13 43.97 61.64

Cat. 5 7.31 2.85 10.20 20.36

Cat. 6 24.32 .09 54.68 79.09

Total 60.49 4.81 135.82 201.12

 

10 df, chi square 201.12, sig. level p >

a = observed frequency; b = expected frequency; B =
(0) collisions; C = (0) convictions, (l)
or more collisions; DG

(0) convictions,

violations.

(1) or more convictions for moving

 

 

56

separated by behavioral categories. In this computation

the influence of collision experience has been removed

and computation is restricted to the analysis of con-

victions for moving violations.

The null hypothesis of no significant difference

in the incidence of convictions for moving violations as

related to observed behavior must be rejected. A chi

square value of 201.12 exceeds the .001 level of confi-

dence.

The data observed in Columns B and C in Table
4.5 are the same as the data observed in Columns
B and C in Table 4.4.

Observation of Column DG demonstrates that be-
havioral categories 1, 4, and 6 contribute most
highly to the total value of chi square in that
the value of each of the cells would by itself
exceed the .001 level of significance.

The cell value for behavioral category 5 in
Column DC is significant at the .01 level.

The persons in behavioral category 1 demon—
strate a record of convictions for moving
violations significantly lower than expected

by chance.

Those persons in behavioral categories 4, 5,
and 6, have accumulated a record of con-
victions substantially higher than expected by

chance alone.

 

 

 

 

 

 

Tables 4.6 and 4.7 contain information relative to

the conviction experience for moving violations by be-
havioral category, but with the sample population fur-
ther divided according to sex.

Analysis of Table 4.6, Conviction Experience for
Moving Violations by Behavioral Category--Female Sub—
jects, reveals:

1. Females in category 5 have accumulated a

record of (0) convictions but (1) or more
collisions (Column C) in greater numbers than
expected by chance.

2. Females in category 5 have accumulated a

record of (l) or more convictions for moving

violations (Column DG) in greater numbers

than expected by chance.
In no other classification is a significant difference
indicated for the female population. With chi square
value of 28.46 (p > .001) the null hypothesis of no
significant difference in conviction experience for moving
violations by behavioral category must be rejected for the
female population.

The findings presented in Table 4.7, Conviction
Experience for Moving Violations by Behavioral Categories
-—Ma1e Subjects, demonstrate a record essentially the

same as that for the female population.

 

 

 

 

 

 

 

58

 

TABLE 4.6.--Conviction experience for moving violations by
behavioral category--female subjects.

 

 

22:21:21.2: B c m
Cat- 1 3 377.5 27.9 47.6 353
Cat- 2 S 78.2 7.7 13.1 97
Cat- 3 S 33.0 3.3 5.7 “2
Cat' 5 S 21 2 3.1 13.6 27
Total 408 41 70 519

--u_-’----—n——_-—u--—---—fl“‘----—---------------—-——-“--———

Cell Contingencies

Cat. 1 .26 .03 1.21 1.50
Cat. 2 .19 .37 .34 .90
Cat. 3 .12 .50 1.91 2.53
Cat. 5 4.91 1.24 11.38 23.53
Total 5.48 8.14 14.84 28.46

 

6 df, chi square 28.46, sig. level p > .001.

a = observed frequency; b = expected frequency;
B = (0) convictions, (0) collisions; C = (0) convictions,
(l) or more collisions; DG = (1) or more convictions for
moving violations.

 

 

 

 

59

 

TABLE 4.7.--Conviction experience for moving violations

by behavioral categories—-male subjects.

 

Behavioral

 

 

 

Categories B C DG Total
a 133 28 59

cat' 1 b 11.0 21.9 87.1 220
a 55 8 21

cat' 2 b 42.0 8.3 33.2 8”
a 49 6 31

Cat: 3 b 43.4 8.6 34.0 85
a 7 3 35

cat' 4 b 22.7 4.5 17.8 “5
a 13 4 21

cat' 5 b 19.1 3.8 ' 15.0 38
a 7 3 40

Cat' 6 b 25.2 5.0 19.8 50

Total 264 52 207 523

Cell Contingencies

Cat. 1 4.36 1.70 9.07 15.13

Cat. 2 4.02 .01 4.48 8.51

Cat. 3 .72 .78 .26 1.76

Cat. 4 10.86 .50 16.62 27.98

Cat. 5 1.94 .01 2.40 4.35

Cat. 6 13.14 .80 20.61 24.55

Total 35.04 3.80 53.44 92.28

10 df, chi square 92.28, sig. level p >
a = observed frequency; b = expected frequency;

B = (0) convictions, (0) collisions; C =
(l) or more collisions; DG = (l) or more convictions for

moving violations.

(0) convictions,

 

 

None of the male behavioral categories has

accrued a driving record of (0) collisions or
(0) convictions (Column B) significantly better
than expected by chance.

Those persons making up categories 4 and 6

have fewer persons with driving records free

of convictions or collisions than would be
expected by chance.

In contrast to the female population, the con-
viction experience is not significantly differ-
ent for those males identified with category 5
than those males identified with categories

2 and 3.

Significantly fewer than expected of those
males identified with category 1 had accured

a record of convictions (Column DG).

The males in both categories 4 and 6 had
accumulated a record of convictions in signifi-

cantly greater numbers than expected by chance.

The null hypothesis of no difference in "conviction experi-
ence for moving violations by behavioral categories, male

subjects," must be rejected. Chi square of 92.28 (p > .001).

mm” ‘ '

 

 

 

 

61

Comparison of Collision Involvemgpt by Behavioral

Categories, Combined Female and Male

 

 

Populations

Table 4.8 contains a comparison of the total number

of persons in each behavioral category with a record of

(0) collisions to those persons with a record of (l) or

more collisions.

l.

a.

The number of persons in behavioral cate-
gory l with a record of (l) or more col—
lisions is significantly less than expected
by chance (Column K).

The number of persons in behavioral cate-
gory 1 with a record free of any collision
involvement is significantly greater than
expected by chance.

The number of persons in behavioral category
2 who have no record of collisions is not
significantly different than expected by
chance.

The number of persons in behavioral category
2 who have experienced collisions is signifi—

cantly less than expected by chance.

The number of persons in behavioral category 3

with or without a record of collisions or con-

victions is not significantly different than

expected by chance.

 

 

TABLE 4.8.—-Collision involvement by behavioral categories
--combined female and male populations.

 

Behavioral BEH K

 

 

 

Categories Total
a 509 64

Cat- 1 b 466.2 106.8 573
a 161 20

Cat. 2 b 147.3 33.7 181
a 106 22

Cat. 3 b 104.1 23.9 128
a 27 25

Cat. 4 b “2.3 9.7 52
a 33 32

cat' 5 b 52.9 12.1 65
a 24 34

cat' 6 b 47.2 10.8 58

Total 860 197 1057

Cell Contingencies

Cat. 1 3.93 17.15 21.08

Cat. 2 1.27 5.57 6.84

Cat. 3 .03 .01 .04

Cat. 4 5.53 24.13 29.66

Cat. 5 7.49 32.73 40.22

Cat. 6 11.40 49.84 61.24

Total 29.65 129.43 159.08

 

5 df, chi square 159.08, sig. level p > .001

a = observed frequency; b = expected frequency;
BEH = (0), (1) or (2+) convictions, (0) collisions; K =
total number of persons involved in collisions.

 

 

 

63

4. The involvement rate in collisions for persons
in categories 4, 5, and 6 is significantly
greater than expected by chance, and con-
versely, the number of persons in categories
4, 5, and 6 without a record of collision in-
volvement is significantly less than expected
by chance.

The null hypothesis of no difference in record of
collision involvement as related to behavioral category
must be rejected. Chi square 159.08 (p > .001).

Table 4.9 contains the observed and expected
collision involvement by behavioral category for the fe—
male population. The collision experience for the female
population by behavioral category is comparable to the
experience for convictions for moving violations.

1. The collision experience for females in be-
havioral categories 1, 2, and 3 is not signifi-
cantly different than expected by chance.

2. The number of persons in category 5 who had
been involved in collisions is significantly
higher than expected by chance at the .001
level.

The null hypothesis of no significant difference in

collision involvement by behavioral category for female

subjects must be rejected. Chi square 50.031 (p > .001).

 

 

 

444!
1"
I

 

 

64

 

TABLE 4.9.--Collision involvement by behavioral categories
--female subjects.

 

 

8282:6312; 354 K Total
cat° 1 S 315 38.1 353
Cat. 2 g 32.5 18.5 97
cat° 3 8 37.5 4.5 ”2
cat° 5 8 24.1 13 9 27
Total 463 55 519

Cell Contingencies

Cat. 1 .26 2.17 2.43
Cat 2 .02 .21 .23
Cat. 3 .00 .06 .06
Cat. 5 5.11 42.48 47.59
Total 5.39 44.92 50.31

 

3 df, chi square 50.31, sig. level p > .001

a = observed frequency; b = expected frequency;
B = (0) convictions, (0) collisions; E = (1) conviction,
(0) collisions; H = (2) convictions, (0) collisions; K =
total number of persons involved in collisions.

 

 

 

 

884'
Ill

 

 

65

Table 4.10, Collision Involvement by Behavioral

Categories-—Male Subjects, indicates:

 

1. The collision involvement for the male popu-
lation in categories 1 and 2 is significantly
less than expected by chance (Column K).

2. The collision involvement of the male popu-
lation in category 3 is not significantly
different than expected by chance.

3. The male population in categories 4, 5, and 6
were involved in collisions with such frequency
as to be significantly greater than expected
by chance.

The null hypothesis of no significant difference in
collision involvement by behavioral category for the male
population must be rejected. Chi square 65.48 (p > .001).

Chargeable and Non-chargeable Collision
Experience by Behavioral Category

In Table 4.11 are contained the observed and expected
frequencies of collision involvement by behavioral cate-
gories. Specifically, comparisons are made of chargeable
only, non-chargeable only, or combinations of chargeable
and non-chargeable collisions experienced by behavioral
category.

1. The collision involvement for persons in cate-

gories 2 and 3 is not significantly different

than expected by chance.

 

 

 

 

TABLE 4.10.--Collision involvement by behavioral categories
-—male subjects.

 

 

 

 

Behavioral BEH K Total

Categories
a 185 35

Cat. 1 b 164.5 55.5 220
a 73 ll

cat° 2 b 62.8 21 2 8“ .
a 68 18

cat' 3 b 64.3 21 7 86
a 23 22 7—

cat' 4 b 33.6 11.4 “5
a 20 18

cat° 5 o 28.4 9 6 38
a 22 28

cat 6 b 37.4 12.6 50

Total 391 132 523

Cell Contingencies

Cat. 1 2.31 7.57 9.88

Cat. 2 1.66 4.91 6.57

Cat. 3 .21 .63 .84

Cat. 4 3.34 9.86 13.20

Cat. 5 2.48 7.35 9.83

Cat. 6 6.34 18.82 25.16

Total 16.34 49.14 65.48

 

5 df, chi square 65.48, sig. level p > .001

a = observed frequency; b = expected frequency;
B = (0) convictions, (0) collisions; E = (1) conviction,
(0) collisions; H = (2) convictions, (0) collisions; K =
total number of persons involved in collisions.

 

 

TABLE 4.1l.--Chargeable and non-chargeable collision in—
volvement by behavioral category—-female and male popu-
lations combined.

 

 

BehaV1°ral BEH N 0 P Total
Categories

a 509 25 38 3
cat' 1 b 466.2 42.8 51.5 12.4 573

a 161 8 10 2
025° 2 b 147.3 13.5 16.3 3.9 181

a 106 10 11 1
025° 3 b 104.1 9.6 11.5 2.8 128

a 27 10 10 5
cat' 4 b 42.3 3.9 4.7 1.1 52

a 33 ll 15 6
cat' 5 b 52 9 4.9 5.8 1 4 55

a 24 15 ll 8
cat° 6 o 47 2 4.3 5.2 1.3 58
Total 860 79 95 23 1057

Cell Contingencies

Cat. 1 3.93 7.40 3.54 10.48 25.35
Cat. 2 1.27 2.24 2.43 .92 6.86
Cat. 3 .03 .01 .02 1.16 1.22
Cat. 4 5.53 9.54 5.98 13.83 34.88
Cat. 5 7.49 7.59 14.59 15.11 44.78
Cat. 6 11.40 26.63 6.47 34.53 79.03
Total 29.65 53.41 33.03 76.03 192.12

 

15 df, chi square 192.12, sig. level p > .001+

a = observed frequency; b = expected frequency;
BEH = (0), (1) or (2+) convictions, (0) collisions; N =
chargeable collisions only; 0 = non-chargeable collisions
only; P = both chargeable and non-chargeable collisions.

 

 

 

 

68

Persons in behavioral category 1 experienced

an involvement in either chargeable or non—

chargeable collisions or a combination of the

two significantly less frequently than ex—

pected by chance.

Those persons in behavioral category 4 were:

a.

Involved in a significantly greater number
of chargeable accidents than expected by
chance.

Involved in a significantly greater number
of both chargeable and non—chargeable

accidents than expected by chance.

The number of persons in behavioral category 5

were:

a.

Involved in non—chargeable collisions only
in significantly greater numbers than ex-
pected by chance.

Involved in both chargeable and non-charge-
able collisions in significantly greater

numbers than expected by chance.

The number of persons in behavioral category 6

were:

a.

Involved in chargeable only collisions in
greater numbers than expected by chance.
Involved in both chargeable and non-charge-
able collisions in numbers greater than

expected by chance.

.4
‘I 8|.lll

 

 

 

69

The null hypothesis of no significant difference
in chargeable and non-chargeable collision involvement
by behavioral category with female and male populations
combined must be rejected. Chi square 192.12 (p > .001).
Due to the relatively low number of individuals
in the total population involved in collisions during
the period of time encompassed in the research, it is
impossible to analyze the incidence of chargeable, non-
chargeable, and chargeable plus non-chargeable collisions
by sex and behavioral category. It is possible, however,
to determine the degree of relative frequency through

observation of Tables 4.12 and 4.13.

Summary
Statistical analysis of the data reveals:

1. When students are grouped on the basis of ob-
served behavior, response to items in the Mann
Inventory are significantly different for each
behavioral category. Sixty of the 63 items
are significant above the 10 per cent level of
confidence for the sample population.

The response to items with the sample
population divided by sex reveals 37 items
that differentiate significantly for females
and 47 items for males at or above the 10
per cent level of confidence. Response to
20 to 39 specific items are identified with

any one of the six behavioral categories.

 

8‘8 (81“:

 

 

 

70

TABLE 4.12.——Chargeable and non—chargeable collision in-
volvement by behavioral category—-female population.

 

 

2:221:12: N . .
025' l 8 315.0 13.2 23.8 2.1 353
055° 2 8 82.5 3.4 8.5 1.5 97
025° 3 8 37.5 1.5 2.8 O 2 “2
cat° 5 8 23.1 5.9 3.8 1.2 27
Total 463 18 35 3 519

 

a = observed frequency; b = expected frequency;
BEH = (0), (l) or (2+) convictions, (0) collisions; N =
chargeable collisions only; 0 = non-chargeable collisions
only; P = both chargeable and non-chargeable collisions.

-w

I 9. .
‘ ‘ 39”.? ".V *‘W C WW II
’ ‘ ‘ ' ' o .. - 12 o h . .

 

.33? ‘-_

“1”

 

'l—nn_.__ -
. . 1.....

."!"'."Z'

 

 

 

 

71

TABLE 4.13.--Chargeable and non-chargeable collision in-
volvement by behavioral category--male population.

 

 

2:228:12: N O P
Cat. 1 : 323,5 23.2 24.1 7.6 219.4
Cat- 2 3 63.8 8.9 9.2 2.9 83.8
Cat. 3 2 23,3 3,1 g 4 3.0 85.8
Cat. 4 g 33.6 1,7 13.9 3.6 44.8
Cat. 5 S 28.4 2.0 4.2 3.31 37.9
Cat- 5 S 37.4 15.3 15.5 3.7 33-9
Total 391 55 59 18 523

 

a = observed frequency; b = expected frequency;
BEH = (0), (l) or (2+) convictions, (0) collisions; N =
chargeable collisions only; 0 = non—chargeable collisions
only; P = both chargeable and non-chargeable collisions.

 

 

 

 

   

,7 / A)
.fl' '

I

“1‘1 .

.—‘

 

72

 

The reliability of the M. I. was determined

by application of product moment coefficient
of correlation to two separate administrations
of the instrument. Correlation values of

.697 — .986 were derived. With 63 degress of
freedom, these values indicate a high to very
high correlation with marked or dependable
relationships.

The relationship between the behavioral cate-
gories and driving records was of the same
direction and magnitude as was the relation-
ship between the M. I. and observed behavior.
As the individuals behavior deviated from
category 1 toward category 6 his driving record

of convictions and/collisions increased.

 

 

 

 

 

CHAPTER V

SUMMARY AND RECOMMENDATIONS

The primary purpose of this study was to investi—
gate the validity and reliability of the Mann Inventory.
An effort was made to ascertain whether or not persons
classified into behavioral categories on the basis of
observed behavior would respond differently to items in
the M. I.

The secondary purpose of the study was to ascertain
whether or not the observed behavior which identified an
individual within a behavioral category would be re-
flected in subsequent driving record.

Sexton High School, Lansing, Michigan, was selected
for this investigation on the basis of:

l. A student body representing a broad cross-

section of the population.

2. A Driver Education program reaching several

hundred students each year.

3. Availability of comprehensive and professional

police traffic record files.

73

 

 

 

7M

All students enrolled in the Driver Education pro-
gram completed the Mann Inventory during the last week
of the Driver Education course. On the basis of rater
observation, throughout the semester, each rater placed
the individual Inventory response sheets of his students
into one of the six appropriate behavioral categories.

The hypotheses were tested by means of chi square,
the cell square contingency method and product—moment

correlation.

Conclusions
The following are the conclusions based upon the

findings of this investigation.

Conclusion on the Relationship
of Observed Behavior and Re-
sponse to Items in the Inven-

to y

I

l. A significant relationship exists between ob-

 

served behavior and an individual's response

 

to items in the Inventory. Findings of signifi-

 

cant differences from .10 to .001 occurred with
60 of the 63 items.

2. There appears to be a difference in response

 

to items in the Inventory on the basis of sex.
Thirty—seven items differentiated for females

and U7 items for males.

 

 

 

 

75

The reliability of the instrument for all be—

 

havioral categories was extremely high. Corre—

 

lation values of .697 to .986 were obtained
when separate test administrations were com—
pared.

Observed behavior in Driver Education can be

 

a predictor of future driver behavior. As <
individual behavior deviated from category 1
toward category 6 the individual's record of

convictions and/or collisions increased.

Behavioral Category 1. Behavior characterized by

 

well adjusted interaction with persons and consis-
tent with the norms of the society in which the
individual lives.

Behavioral Category 6. Behavior characterized by

 

a pendulum effect, vacillating between extremes
of aggression and withdrawal.

Conclusions on the Relationship

 

 

of Behavioral Categories and
Driving Record

A highly significant relationship exists between

observed behavior in Driver Education and driving record

as attested to by the absence or presence of convictions

 

for traffic offenses and/or collisions.

1.

Persons in Behavioral Category 1 were involved

in significantly fewer collisions than expected
by chance.

Persons in Behavioral Category 1 were involved
in significantly fewer times than expected by

chance for moving violations.

 

 

 

76

3. The observed frequency of involvement in
either collisions or convictions for be-
havioral categories 2 and 3 did not deviate
from the theoretical frequency.

Behavioral Category 2. Behavior generally charac-

terized by satisfactory interaction with persons

and society, but with periodic withdrawal from con-
tact with people.

 

Behavioral Category 3. Behavior generally charac-
terized by satisfactory interaction with persons
and society, but with periodic efforts toward
assertive action.

h. The conviction and collision involvement of
persons in Behavioral Categories H, 5, and 6,
is significantly greater than expected by
chance.

Behavioral Category 4. Behavior characterized by

forceful, outgoing action or vigorous efforts to
assert oneself over others.

 

Behavioravaategory 5. Behavior characterized by
withdrawal from contact with other persons.

 

5. Persons in Behavioral Category u were involved
in a disproportionately high number of charge-
able collisions.

6. Persons in Behavioral Category 5 were involved
in a disproportionately high number of non-
chargeable collisions.

7. The persons in Behavioral Catetory 6 were
involved in a disportionately high number of
chargeable or both chargeable and non-charge-

able collisions.

 

 

 

 

77

Discussion

 

Since the findings of this research have impli-
cations for pencil-paper tests, observer—rating behavior

and driver behavior further discussion is warranted.

Mann Inventory and Raters

 

On the basis of the research it appears that indi-
viduals will express their feelings and response to items
in a pencil-paper test in such a manner as to reflect
their observed behavior.

Raters with comparable professional experience,
classifying individuals independently, on the basis of
established criterion can arrive at essentially the same
categorization. For persons doing similar research,
certain possible biases should be recognized. This pro-
blem was illustrated in the present research when both
raters exhibited an identical bias toward females in
Behavioral Categories A and 6. Females exhibiting certain
characteristics, (aggressive behavior), were failed dis-
proportionately, excluding them from the sample population.
Had a female rater been included in the study, this bias
might have been less apparent.

Observed Behavior and Impli—
cations for Driver Education

Programs

When the driving records of Behavioral Categories

 

l, 2, and 3 are compared to categories H, 5, and 6, we

 

 

 

 

78

observe that 83.5 per cent of the sample population,

882 persons, experienced 273 convictions and 1u2 colli—
sions during the test period, while 16.5 per cent of the
sample population, 175 persons, experienced 253 con-
victions and 1A3 collisions. A question might well be
raised as to the effectiveness of Driver Education as

we know it today, among the three behavioral categories
A, 5, and 6, exhibiting deviant behavior.

Since we have no valid criteria against which to
measure the experience of either the 882 relatively well
adjusted students, or the 175 students who fell into the
extreme behavioral categories, we can only surmise that
Driver Education as it was taught at Sexton High School
during the period of September, 1958 through August, 1960
was apparently successful among the relatively well ad—
Justed student body. In fact, behavioral categories 2
and 3 might be considered a norm group for rarely were
they involved in traffic offenses or collisions at a
rate different than expected by chance. If this is true,
however, then the program failed to meet the needs of
students falling into behavioral categories A, 5, and 6.

The extremely high number of female students in
behavioral categories A and 6 who received failing grades
in the course tends to indicate that the raters recog-
nized the inadequacies of the program in preparing these

students.

 

 

 

79

If we accept the previously stated conclusion, it
would appear that an additional program should be pro-
vided for both female and male students in categories
4, 5, and 6.

One approach might well be the one presently em-
ployed in the typical Driver Education program if stu—
dents are screened so that those students with identi-
fiable problems of adjustment are channeled into a
second type of program.

The second type of program might well be patterned
along the lines of small group therapy sessions. One
such approach is that suggested by Myrick,l and Mann,2

respectively in, A Manual For Instructors Conducting

 

Discussions on Driver Attitudes, and Let's Talk It Over.
An approach similar to that suggested by these authors
was employed for a period of three years in the Ingham
County Driver Safety School in Lansing, Michigan. This
program was conducted for persons (primarily males), con-
victed of two or more moving violations within a period
of 12 months, and/or those persons whose license had
previously been suspended or revoked, and who were again

called in for a hearing. This program was structured in

 

lRichard Myrick, Lawrence Schlesinger and Barbara
Marx, A Manual for Instructors Conducting Discussions
on Driver Attitudes, The George Washington University,
November, 1962.

 

 

2William Mann, "Let's Talk It Over," Analogy,
Charter Issue, 1966, pp. 4—9.

 

 

 

80

such a manner that class size was limited to 12 students,
emphasis was placed upon group interaction in an attempt
to gain greater understanding of one's own basic behavior.
A follow-up study, conducted at the end of the third year,
indicated that persons classified in behavioral categories
A and 5 appeared to show substantial improvement in their
driving record as compared to their record prior to enter-
ing the school. Persons classified as behavioral category
6 appeared to improve to a lesser degree.

In conclusion, if persons with a greater potential
for involvement in collisions or for conviction of moving
violations can in fact be identified, in advance, then
it would appear that persons in the field of traffic
accident prevention must reassess not only the goals,
content, and type of driver education, but also driver
licensing, driver improvement and other of the allied

fields.

Recommendations

1. A further follow—up on the 1057 drivers in—
cluded in this study to ascertain whether or
not the relationship between observed be-
havior and driving behavior continued beyond
the test period.

2. Determine whether persons who have been called
in for driver license hearings because of un-

satisfactory driving records are identifiable

 

 

 

 

 

81

with any of the six behavioral categories em-
ployed in this study.

Determine whether individuals identified as
accident repeaters respond to items in the
Mann Inventory in such a manner as to allow
their classification according to the be-
havioral categories employed in this study.
Determine the value of the Mann Inventory in
areas outside the field of driver behavior.
Replicate the present study, but attempt to
work with those persons identified as exhibit—
ing behavior comparable to behavioral cate-
gories A, 5, and 6, to determine whether or
not, with special attention, the driving be—

havior of these groups might be improved.

 

 

 

BIBLIOGRAPHY

 

 

 

 

 

 

 

 

 

BIBLIOGRAPHY

Primary,Sources

 

Accident Facts, 1965 edition, National Safety Council,
Statistics Division, 59.

 

Adelstein, A. M. "Accident Proneness: A Criticism of
the Concept Based Upon an Analysis of Shunters'
Accidents," Journal of the Royal Statistical
Society, cxv (Series A, 19527, 354—400.

Alexander, F. Psychosomatic Medicine as quoted by L.
LeShan, "Dynamics in Accident-Prone Behavior,"
Psychiatry, 1952, 73.

 

 

Birnbach, B. S. "A Comparative Study of Accident Re—
peater and Accident-Free Pupils." Unpublished
Doctoral dissertation, New York University, 1947.

Brazell, Robert E. A Comparison of Various Behind-the-
Wheel Training Methods. Safety and Traffic Divi-
sion, Automobile Club of Michigan, 1962.

 

 

Brody, Leon. "Personal Characteristics of Chronic Vio-
lators and Accident Repeaters," Bulletin No. 152,
National Academy of Sciences, National Research
Council.

Burg, Albert. "An Investigation of Some Relationships
Between Dynamic Visual Acuity, Static Visual Acuity
and Driving Record," University of California,
Department of Engineering, April, 1964.

Cobb, P. W. Automobile Driver Tests Administered to
3,663 Persons in Connecticut, 1936437 and the
Relation of the Test Scores to the Accident
Sustained. Unpublished Report to Highway
Research Board, Washington, D. C., July, 1939.

 

 

 

 

Dunbar, F. "Medical Aspects of Accidents and Mistakes
in the Industrial Army and in the Armed Forces,"
War Medical, IV (1943), 167

 

83

 

 

 

84

Dunbar, F. Synopsis of Psychosomatic Diagnosis and
Treatment as quoted by L. LeShan, "Dynamics in
Accident—Prone Behavior," Psychiatry, 1952, p. 73.

 

Edwards, Allen L. Statistical Methods for the Be—
havioral Sciences. New York: Rinehart and Company,
Inc., 1954.

 

Eno Foundation for Highway Traffic Control. "Personal
Characteristics of Traffic-Accident Repeaters."
Saugatuck, Connecticut, 1948.

Finesilver, S. G. "National Symposium on the Deaf;
Driving and Employability." U. S. Department of
Health, Education, and Welfare, Vocational Re-
habilitation Administration. Washington: U. S.
Government Printing Office, 1964.

Greenwood, Major, Yule, G. "An Inquiry Into the Nature
of Frequency Distributions Representative of
Multiple Happenings with Particular Reference to
the Occurrence of Multiple Attacks of Disease or
of Repeated Accidents," Journal of the Royal
Statistical Society, LXXXIII (19207, 255-279.

Greenwood, Major and Woods, Hilda. "The Incidence of
Industrial Accidents with Special Reference to
Multiple Accidents," Medical Research Committee,
Industrial Fatigue Research Board, Report No. 4,

1919.

Guilford, J. P. Fundamental Statistics in Psychology
and Education. New York: McGraw-Hill Book Company,

Inc., 1950.

 

Haddon, William, Suchman, Edward and Klein, David.
Accident Research Methods and Approaches. New York:
Harper and Row Publishers, 1964.

Hanner, Charles P. "Use of Personal Data in Underwriting
Setting," Traffic Safety Research Review, National
Safety Council, VII, 3 (September, 1963).

Hanner, Charles P. "Use of Psychological Inventory in
Writing Insurance for Youthful Male Drivers,"
Traffic Safety Research Review, National Safety

Council, VII, 1 (March, 1963).

 

 

 

85

Heath, E. D. "The Relationships Between Driving Records,
Selected Personality Characteristics, and Bio—
graphical Data of Traffic Offenders and Non-
offenders." Highway Research Board, Bulletin 212,
1959.

Jacobs, Herbert H. and others. Behavioral Approaches
to Accident Research. New York: Association for
the Aid of Crippled Children, 1961.

 

Kremple, R. J. "A Study of Application Blank Items to
Discover Those Differentiating Accident-Free from

Accident—Prone Truck Drivers." Unpublished
Master's dissertation, Western Reserve University,
1951.

Lauer, A. R. "Age and Sex in Relation to Accidents."
Bulletin 60, Highway Research Board, 1952.

Lauer, A. R. et al. "Aptitude Tests for Army Motor
Vehicle Operators." Report 981, Personnel Re-
search Section, TAGO, Department of Army, Washington,
D. C., 1952.

Mann, William. "Let's Talk It Over," Analo ., Charter
Issue, Allstate Insurance Company, 1966, 4—9.

Marsh, B. W. "Aging and Driving," Traffic Engineering
Reprint, November, 1960.

McFarland, Ross A. The Major Strategies in the Prevention
of Highway Injuries and Fatalities. Harvard School
of Public Health.

 

 

Miller, Gene. "Accident Repeaters May Not Be Accident-
Prone," National Safety News, LXVII (March, 1953),
3-6.

Myrick, Richard, Lawrence Schlesinger and Barbara Marx.
A Manual For Instructors Conducting Discussions
on Driver Attitudes. The George Washington University,
November, 1962.

 

 

Schaff, John G. "Personal Attitude Survey." Unpublished
Master's dissertation, Michigan State University,
1957.

Schulzinger, M. The Accident Syndrome. Springfield,
Illinois: Charles C. Thomas Publishers, 1956.

 

 

 

86

Schuster, B. H., J. P. Guilford, An Analysis of Accident
Repeater and Chronic Violator Drivers. A paper
presented at the 47th National Safety Congress
and Exposition. Chicago, Illinois, October, 1959.

 

Shaw, Lynnete. "The Practical Use of Projective Per—
sonality Tests as Accident Predictors," Traffic
Safety Research Review, IX, 2 (June, 1965), 34.

 

Siegel, Sidney. Non-Parametric Statistics. New York:
McGraw-Hill Book Company, Inc., 1956.

Tillman, W. A., G. E. Hobbs. "The Accident Prone Auto—
mobile Driver," A Study of Psychiatric and Social
Background, American Journal of Psychiatry, CVI, 5
(1949), 321-331.

 

West, Roscoe L. "The Philosophy of the Accident Problem,"
The Journal of Educational Sociology, October, 1946.

 

Wong, W. A., and G. E. Hobbs. "Personal Factors in
Industrial Accidents," Industrial Medicine and
Surger , XVIII (July, 1949), 291—294.

 

Secondary Sources

 

Adler, A. ”The Psychology of Repeated Accidents in In-
dustry," American Journal of Psychiatry, XCVIII,
1 (1941), 99—101.

 

Allgaier, Earl. ”Human Behavior and Traffic Accidents,"
International Road Safety and Traffic Review, VI,
1 (Winter, 1958).

 

Baker, James S. "Driver Characteristics and Freeway
Accidents," Traffic Digest and Review, VII, 11
(November, 1959).

 

Bannister, H. ”Another Approach to the Problem of Acci—
dent Causation," British Journal of Psychology,
XXVIII, 3 (1938), 304—314.

 

Billion, C. E. ”Community Study of the Characteristics
of Drivers and Driver Behavior Related to Accident
Experience.” Bulletin 172, Highway Research Board,
1958.

Bletzacker, Richard W., and Thomas G. Britterham.
"An Analysis of One-Car Accidents." Bulletin
208, Highway Research Board, 1959.

 

 

 

 

 

 

87

Brady, R. O. "A Preliminary Study Into the Relationship
Between Accident Rates and Personality of Automobile
Drivers." Unpublished Master's dissertation,
Catholic University, Washington, D. C., 1948.

Brody, Leon. "Personal Factors in Safe Operation of
Motor Vehicles." New York: Center for Safety
Education, New York University, 1941.

Brown, Paul L. Driver Behavior: A Study of Personality
Characteristics. Minneapolis: University of
Minnesota Highway Department, January, 1959.

 

 

Campbell, B. J. "The Effects of Driver Improvement
Actions on Driving Behavior," Traffic Safety
Research Review, IV, 3 (September, 1959).

 

 

Case, Harry W., and Roger G. Stewart. "Development of
a Driving Attitude Scale." Bulletin 172, Highway
Research Board, 1958.

Case, Harry W., and Others. "The Habitual Traffic
Violator." Bulletin 120, Highway Research Board,
July, 1956.

Conger, John J. et al. "Personal and Interpersonal
Factors in Motor Vehicle Accidents," The American
Journal of Psychiatry, June, 1957.

 

 

Conger, John J. et a1. "Psychological and Psychophysio-
logical Factors in Motor Vehicle Accidents," Egg
Journal of the American Medical Association,
CLXIX, 14 (April 4, 1959).

 

DeSilva, H. R. et a1. "Some Psychological Factors in
Accident Repeater Drivers," Journal of Abnormal
and Social Psychology, XXXIV (January, 1939).

 

 

 

Forbes, T. W. "Psychological Factors in Traffic Accidents
on Freeways," Traffic Safety Research Review, II,
4 (December, 1958).

 

Freeman, F. et al. The Role of Human Factors in Accident
Prevention. Accident Prevention Program.
Washington: U. S. Department of Health, Education,
and Welfare, August, 1960.

 

 

Goldstein, L. G., and J. N. Mosel. "A Factor Study of
Driver's Attitudes, with Further Study on Driver
Aggression." Bulletin 172, Highway Research Board,
1958.

 

88

Greenshields, Bruce D. "Attitudes, Emotions, Accidents,"
Traffic QuarterIy, XIII, 2 (April, 1959), 221-235.

 

Harris, F. J. "Can Personality Tests Identify Accident-
Prone Employees?" Personnel Psychology, III,
(1946), 489-532.

Hersey, Rexford B. "Emotional Factors in Accidents,"
Personnel Journal, XV (1936), 59—65.

Johnson, H. M. "The Detection and Treatment of Accident-
Prone Drivers," Psychological Bulletin, XXXXIII,
6 (1946), 489-532.

 

Krech, D., and R. S. Crutchfield. Theory and Problems
of Social Psychology. New York: McGraw—Hill
Book Company, 1948.

Kummer, Jerome. "Emotion Behind the Wheel Causes Traffic
Accidents," Science Digest, May, 1963, 10-20.

LeShan, L. L. "Dynamics in Accident-Prone Behavior,"
Psychiatry, XV, 1 (1950), 73-80.

McClean, Alan. "Accidents and the Human Factor,"
Personnel Journal, XXXIV (February, 1956), 342-345.

McFarland, Ross A. et a1. Youth and the Automobile.
New York: Association for the Aid of Crippled
Children, 1960.

 

McGuire, Frederick L. "Psychological Comparison of
Automobile Drivers," U. S. Armed Forces Medical
Journal, VII, 12 (December, 1956).

 

McGuire, Frederick L. "Rosenzweig Picture Frustration
Study for Selecting Safe Drivers," U. S. Armed
Forces Medical Journal, VII, 2 (February, 1956).

 

Pond, J. H. "An Investigation of Differences Between a
High Accident Frequency Group and a Normal Accident
Frequency Group." Unpublished Master's dissertation,
Pennsylvania State College, 1950.

Rommel, R. C. S. "Personality Characteristics and
Attitudes of Youthful Accident-Repeating Drivers,"
Traffic Safety Research Review, III, 1 (March, 1959).

 

Ronis, L. A. "A Study of Characteristics Differentiating
Bus Drivers Who Have Accidents from Those Who Do
Not Have Accidents." Unpublished Master's disser—
tation, Western Reserve University, 1950.

 

89

Schlesinger, Lawrence. Grogp Dynamic Study of Driver
Attitudes and Driving Behavior. Washington, D. C.:
George Washington University, Driver Behavior
Research Project, May, 1963.

 

Selling, L. S. "The Psychiatric Findings in the Cases
of 500 Traffic Offenders and Accident-Prone
Drivers," American Journal of Psychiatry, XCVII,
l (1940), 68—79.

 

Suchman, E., and A. Scherzer. Current Research in
Childhood Accidents. New York: Association for
the Aid of Crippled Children, 1960.

Thorndike, Robert L. The Human Factor in Accidents.
Washington, D. C.: U. S. Department of Health,
Education, and Welfare, 1962.

Venables, Peter. "Car Driving Consistency and Measures
of Personality," Journal of Applied Psychology,
XXXX, 1 (February, 1956).

Werthom, F. "Mass Media and Emotional Development,"
State of Mind, I (1957).

| I I 1"“!

 

 

 

we

 

5N3!

 

 

APPENDICES

9O

 

 

7- E-J I1..E'._.—-_I--- In]... a- —__.

APPENDIX A

THE MANN INVENTORY AND RESPONSE SHEET

91

 

l‘u“!
1
.I'l‘

 

 

 

 

92

THE MANN INVENTORY

Name Age Sex Months Driving
Experience

 

 

The following statements reflect your attitude and feelings
about yourself and your relations to others. There are no
right or wrong answers. Fill in on the answer sheet the
answer that reflects your feelings the best.

Danot mark on the test booklet

 

A. always B. usually C. sometimes D. rarely E. never
1. I (like) (liked) to take part in organized extra-

curricular activities in school.
2. Young people are much better drivers than middle—aged people.
3. Policemen are sincere in enforcing the laws.

4. My parents (are) (were) reasonable in their relations
with me.

My community is a happy place to live.

I put off until tomorrow things I should do today.

I like to daydream.

. I feel full of pep when I get behind the wheel.

\OCDNChU‘I

I (live) (lived) in a home that (is) (was) happy.
10. If I see a police officer, I am more careful.

ll. Over-careful drivers cause more accidents than the
so-called reckless ones.

12. I enjoy being out late at night and sleeping mornings.
13. I get a feeling of real power when driving a car.

14. Courses in school (any grade level) are set up to meet
the needs and interests of the student.

15. I am concerned about the way my clothes look.
16. Slow drivers should be kept off the highways.

17. All new drivers should be required to take a course
in driver education.

 

 

 

18.

19.
20.

21.

22.

23.

24.

25.

26.

27.
28.

29.
30.

31.

32.

33-
34.

35.

36.

37.

38.

f A “l"’\\///
. ' \,
, _
K
.

» ——:——u—-—-—-- - ,1

93

 

Unsafe drivers should be deprived of the right to
drive.

Accidents don't just happen; they are caused.

I like to get everything out of a car that it has
in it.

The chief work of most policemen should be traffic
control.

My parents (exert) (exerted) too much control over me.

The people in my community want all traffic laws
enforced.

I have been tempted to cheat on a test.
I get impatient in heavy traffic.

There are times when it seems like everyone is against
me.

Old, defective cars should be kept off the road.

Drivers should be given more freedom in obeying
traffic signs.

People should drive when they are angry.
Passing on hills and curves is exceedingly dangerous.

It is necessary to stop at "stop" signs if no other
cars are in sight.

I like to put extras on my car to attract attention.
I am good at talking myself out of trouble.
Strong discipline in practice makes a better team.

I (am) (was) popular with most of the students in
my class.

Police officers are rougher on teen-agers than on
adults.

Teachers want to help students with their problems.

My (father) (principal driver in family) gets traffic
tickets for moving violations.

 

39.

40.

41.

42.
43.
44.
45.
46.

47.

48.
49.
50.
51.
52.
53-

54.

55.
56.
57.

58.

59.

 

94

I have as good table manners at home as when I eat
out.

I have been wrong in an argument but wouldn't admit
it to my opponent.

Society should have the right to question the way
I drive.

I like to razz a team when it is losing.

I am proud of my reputation in the community.
I am considered a friendly person.

I like most of my work.

Our family (spends) (spent) a great deal of time
together.

Attitudes toward driving are more important than
ability to handle a car.

I like to take chances when I'm driving.
Traffic laws are set up to promote safety.
Courtesy toward other drivers is important.
I like a great deal of freedom.

I don't mind being told what to do.

My grades in school (are) (were) a good indication
of my ability.

I become concerned about what other people think
of me.

I find that older people tend to be too bossy.
I feel somewhat nervous when I drive a car.

I think courtesy towards others is a good reflection
of a person's character.

I get more fun out of driving a car than in any
other activity.

The police are only trying to do the job for which
they were hired.

 

61.
62.
63.

My folks (insist) (insisted) that I spend most week—
day evenings at home.

I am considered a reliable
I like to help a person who is in trouble.

I am more courteou' than the average driver.

 

 

 

.I‘Ill‘l‘ll‘lll
ll"
3'

Name

10.

11.
12.
13.
14.
15.
16.
l7.
l8.
19.
20.

21.

96

THE MANN INVENTORY

Age

 

23. i ) ( ) ( ) ( ) ( )
24. ( ) ( ) ( ) ( ) C )
25. ( ) ( ) ( ) ( ) ( )
26. ( ) ( ) ( ) ( ) ( )
27. ( ) ( ) ( ) ( ) ( )
28. ( ) ( ) ( ) ( ) ( )
29. ( ) ( ) ( ) ( ) ( )
30. ( ) ( ) ( ) ( ) ( )
31. ( ) ( ) ( ) ( ) ( )

32.()()()()()
33.()()<)()()
34.()()()()()
35.()()()()()
36.()(
37-()( )
38.()()()
39.()()()
40.()()()
41.()()()
42.()()()()()

M3.
44.
1+5.
46.
147.
48.
49.
50.
51.
52.

53-
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.

 

 

 

 

 

 

APPENDIX B

INVENTORY ITEMS FOUND TO BE
SIGNIFICANTLY RELATED TO

BEHAVIORAL CATEGORIES

97

 

99

Cate or 6. Items 1, 6, 12, 14, 16, 20, 22, 27,
32, 35, 42, 43, 44, 45, 47, 51, 54, 56, 57 and 62.

 

~13. . iglw

 

1—1
L
i
,
)
I
I
I
r
V
5

 

713717

\llil‘lll‘lllll‘llli

II”
“
Mill
“-

0
3
0
3
9
2
1

3

 

"WWW