THE AUTOMOBILE SALESMAN'S KNOWLEDGE
OF THE PRODUCT AS A OETERMINANT 0F SUCCESS
IN THE SELLING OF AUTOMOBILES

Thesis for the Degree of Ph. D.
MICHIGAN STATE UNIVERSITY
DAVID I. VAN BLOIS

1968 '

331E545

 

This is to certify that the
thesis entitled

THE AUTOMOBILE SALESMAN ' S KNOWLEDGE
OF THE PRODUCT AS A DETERI‘VIINANT
OF SUCCESS IN THE. SELLING OF AUTOI‘VIOBILL‘JS

presented by

David I. Van Blois

has been accepted towards fulfillment
of the requirements for

Ph

.D. degree in Bu51ness

Administration

/ 1 fl
/////;P:;U§ZZK::::;C/

Ry.” proigssor I

 

Dme April 18, 1968

 

0-169

 

 

 

ABSTRACT

THE AUTOMOBILE SALESMAN'S KNOWLEDGE OF
THE PRODUCT AS A DETERMINANT OF
SUCCESS IN THE SELLING OF AUTOMOBILES

by David I. Van Blois

The final integrative link in the network which
unites the products of the manufacturer with the consumer
is the salesman. Even though personal selling is a crucial
factor in today's economic prosperity, only a limited amount
of research has been conducted on this aspect of marketing.
Therefore, in order to increase our understanding of per-
sonal selling, this study was undertaken to determine by
empirical analysis the importance of an automobile sales-
man's knowledge of the product, in conjunction with two
other factors, personality and sales techniques, in selling
effectiveness. The relative contribution of each of these
three variables to sales performance was investigated.

The data required for testing the hypotheses were
obtained by questionnaires and rating schedules adminis-
tered by company field personnel during breakfast meetings
to 150 randomly selected automobile dealerships supervised
by the Lincoln-Mercury Division of the Ford Motor Company.
The participating dealerships were stratified into clas-

sifications according to type and size.

David I. Van Blois

The questionnaire was employed as a method of meas—
uring new car salesmen's level of product knowledge, includ-
ing their ability to translate product features into consumer
benefits. The rating schedule, on the other hand, consisted
of three paired comparisons and was used to determine the
sales managers' opinions concerning their salesmen's rela-
tive strengths in the area of personality, product knowledge,
and sales techniques. The information obtained from these
two data collecting instruments was related to two measures
of sales performance; total number of automobiles sold and
profitability of sales.

Analyses of the selected variables were conducted
on an inter-dealership and an intra—dealership level using
both parametric and nonparametric statistical methods. The
conclusions derived from the two methods of data analysis
were then compared.

The hypothesis that product knowledge (in conjunc-
tion with two other assumed determinants, personality and
sales techniques) is one of the primary determinants of
sales effectiveness received very weak support. Although
the relationships were statistically significant, it was
determined that the combined influence of product knowledge,
personality, and sales techniques accounted for approxi—
mately 14 percent of the unit variance in the salesmen's
level of sales performance. It was shown that the rela-
tionship between a salesman's level of product knowledge

and his sales effectiveness was independent of the type of

. I V. f
.0. add r. a» r” f. by
no r .. A. A.
: 9v «3 I. O» s-
. o . o .r“ a . O» .u r”
a. C “a .3 .2 ”a
"v 8 3t .9“ :r. "c
. . 3

 

--.:, c...”
-ODH‘U'-’ .
C

David I. Van Blois

market served. The relationship between the salesman's
level of product knowledge and sales effectiveness within
the large metropolitan markets was dependent upon the size
of the retail outlet. In contrast, in the less densely
populated markets the product knowledge--sales performance
relationship is independent of the size of the retail outlet.
In this study three factors considered to be con—
tributory to a salesman's level of product knowledge were
amount of formal education, sales experience within the
industry, and attendance at a manufacturer-sponsored train-
ing program. It was found that a salesman's product know-

ledge was not influenced by any of these factors.

 

 

 

THE AUTOMOBILE SALESMAN'S KNOWLEDGE
OF THE PRODUCT AS A DETERMINANT

OF SUCCESS IN THE SELLING OF AUTOMOBILES

By

David I. Van Blois

A THESIS

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

DOCTOR OF PHILOSOPHY

Department of Marketing and Transportation

1968

/
c c I
_/ . A r h"

‘ _. J
c/ t7 o

w~Copyright by

DAVID IRWIN VAN BLOIS

1968

e::c>:
5 cans;

Q
‘7'"
an
“.9
4‘ o
.F‘
-.x-

 

:e ..IL : a «H. A y L. 5
M" . .c . . C S I e. I ”S ..c..
w. .l . .n e a r ., . F
r!. a e 7. .1“ e z c.
... C a. C 5 C. S . I
. .. I «5 ”C a» ‘Q .3
e a c -s . Cu 7. R D r
w . c .. . .. . . e an
e. .H. ... ..I .. ~w.. ~.<.. .1 r..
fin“ ATV hab‘ u“ nar. .‘use .6. u arm“

 

 

 

 

ACKNOWLEDGMENTS

The writer would like to express his sincere appre-
ciation to Dr. William J. B. Crissy, under whose direction
and encouragement this research was completed and from whom
this candidate received guidance and inspiration throughout
his graduate studies. A

My gratitude is extended to Dr. Frank R. Bacon, Jr.,
who served as a thesis committee member and contributed im—
mensely to the statistical methodology employed in this study,
and to Dr. Lawrence J. Giacoletto for serving as a member
of the thesis committee.

Acknowledgments are given to the Lincoln-Mercury
Division of the Ford Motor Company, especially to the Mar-
keting Research Department (now part of the Central Market-
ing Research Organization) and the Sales Promotion and Train-
ing Department for providing the necessary impetus and finan—
cial assistance; to the General Sales Office including the
General Field Managers and Zone Managers for their coopera—
tion in the implementation of the study; and to those members
of the retail dealership sales organization who willingly
participated.

To my wife for her encouragement during the prepara—

tion of this thesis, I express my sincere thanks.

ii

I!

$.

.b

L25 '1'

51‘ C? F

7
4

L

re
‘.
r».
.3

C.

Q“‘
‘a‘
S '
Q§
."
Ae‘n‘

.“
‘fi‘.
-n.
¢' .

 

 

LIST OF
LIST OF
LIST OF
Chapter
I.

II.

III.
IV.
V.

VI.

TABLE OF CONTENTS

TABLE S O O O O O O O O O O O O O O O O O

FIGURES O O O O O O O O O O O O O O O O O

APPENDICES O O O O O O O O O O O O O O I

INTRODUCTION: THE PROBLEM . . . . . . .

CRITERIA FOR MEASUREMENT OF SALESMEN'S
EFFECTIVENESS. . . . . . . . . . . . . .

STUDY METHODOLOGY 0 O O O O O O O O O O 0

ANALYSIS OF DATA COLLECTING INSTRUMENTS.

STUDY RESULTS 0 O O O O O O O O O O O O 0

CONCLUSIONS 0 O O O O O O O O O O O O O O

SELECTED BIBLIOGRAPHY. . . . . . . . . . . . . .

APPENDICES O O O O O O O O O O O O O O O O O O 0

iii

Page
iv
vi

vii

25
34
66
83
139
154

161

r I . f . .u A} C .I E C
. . Id . . .I‘ .v. A .
J. J. a. .5. .c a: 3» Q» Q» ”a A: ‘v e rem «M». m». ‘3. a”.
E L S L K S P K P .l. 2. i K K K 2. S I.
a.
1 o I o O I I o o o
.2 1. .C .5 pa. :4 .C . 2.
.3
m.

 

 

 

 

.:

.G

.C .3 Av CC
.3 9 S . .
r. . . r z.
E d 9 E
”A S D. D

«7

 

10.

ll.

12.

13.

LIST OF TABLES

Page
Benefits Derived from Product Knowledge:
Literature Summary . . . . . . . . . . . . . . 12
Stratification of Sample by Type and Size of
DealerShipS O O O O O O O O O O O O O O O O O O 46
Kendall Rank/Partial Correlation Analysis:
Sales Performance with Product Knowledge,
Personality and Sales Techniques . . . . . . . 54
Kendall Rank Correlation Analysis: Sales
Performance with Product Knowledge . . . . . . 57
Multiple/Partial Correlation Analysis: Sales
Performance with Product Knowledge, Personal-
ity, and Sales Techniques. . . . . . . . . . . 59

Kendall Rank Correlation Analysis: Product
Knowledge with Education, FMI, and Experience. 63

Kendall Rank Correlation Analysis: Sales
Performance with Product Knowledge . . . . . . 64

Sales Managers' Comparisons of the Relative
Importance of Personality, Product Knowledge,
and Sales Techniques . . . . . . . . . . . . . 79

Kendall Rank/Partial Correlation Analysis:

Sales Performance with Product Knowledge,
Personality, and Sales Techniques (All

Dealerships) . . . . . . . . . . . . . . . . . 84

Multiple/Partial Correlation Analysis: Sales
Performance with Product Knowledge, Personal-
ity, and Sales Techniques (All Dealerships). . 89

Subgroups of Dealership Classifications. . . . 98

Kendall Rank Correlation Analysis: Sales
Performance with Product Knowledge (Multi-
point Versus Single Point Dealerships) . . . . 99

Multiple/Partial Correlation Analysis: Sales
Performance with Product Knowledge, Personal-
ity, and Sales Techniques (Multi-point Versus
Single Point Dealerships). . . . . . . . . . . 103

iv

Table

14.

15.

16.

17.

18.

Page
Kendall Rank Correlation Analysis: Sales
Performance with Product Knowledge (High
Versus Low Volume Dealerships) . . . . . . . . 110
Multiple/Partial Correlation Analysis: Sales
Performance with Product Knowledge, Personal-
ity, and Sales Techniques (High Versus Low
Volume Dealerships). . . . . . . . . . . . . . 112
Subgroups of Salesmen Classifications. . . . . 120

Kendall Rank Correlation Analysis: Product
Knowledge with Education, FMI, and Experience. 130

Kendall Rank Correlation Analysis: Sales
Performance with Product Knowledge . . . . . . l37

LIST OF FIGURES

Model of the Relationships to be Studied . . .

Frequency Distribution

Frequency Distribution
Criteria . . . . . . .

Frequency Distribution
Multi and Single Point

Frequency Distribution

of Questionnaire Scores

of Sales Performance

of Questionnaire Scores:
Dealerships . . . . . .

of Questionnaire Scores:

Multi-Point High and Low Volume Dealerships. .

Frequency Distribution

of Questionnaire Scores:

Single Point High and Low Volume Dealerships .

vi

Page

23

7O

81

122

125

127

LIST OF APPENDICES
Appendix Page
A. Introductory Letter to District Sales Manager. 161
B. Instruction Sheet to the Zone Manager. . . . . 163
C. Questionnaire. . . . . . . . . . . . . . . . . 164
D. Rating Schedule. . . . . . . . . . . . . . . . 172

E. Method of Data Analysis: Paired Comparison
Matrix 0 O O O O O O O O O O O O I O O O O O O 177

F. Transformation of Ranks to Standard Scores . . 178
G. Tabulation of the Data Cards . . . . . . . . . 179

H. Assumptions Underlying Kendall Rank and Kendall
Partial Correlation Coefficients . . . . . . . 191

I. Assumptions Underlying Tests for Significance
for Kendall Rank Correlation Coefficients. . . 192

J. Assumptions Underlying the Normal "2" Test
for the Difference Between the Sample Means. . 193

K. Assumptions for Multiple/Partial Correlation
Coefficients and F Test. . . . . . . . . . . . 194

L. Testing the Significance of the Average
Kendall Rank Correlation Coefficient . . . . . 195

M. Testing the Significance of the Pearsonian
and Multiple/Partial Correlation Coefficients. 196

N. Frequency Distribution of Total Units Sold
and Profitability of Units Sold - Multi and
Single Point Dealerships . . . . . . . . . . . 197

0. Frequency Distribution of Total Units Sold
and Profitability of Units Sold - High and
Low Volume Dealerships . . . . . . . . . . . . 199

P. Testing the Difference Between the Means of
Two samples. 0 O O O O O O 0 O O O O O O O O O 203

Q. Method of Evaluating the Consistency of the
Paired Comparison Matrices . . . . . . . . . . 204

vii

CHAPTER I

INTRODUCTION: THE PROBLEM

General Problem Area

 

Our free enterprise system with its increasing levels
of productivity depends heavily upon the consumer and the
satisfaction of his wants and needs. Consumer desires ulti—
mately determine what goods and services are produced.
Marketing focuses upon understanding the consumer to deter-
mine his basic wants and needs and to awaken latent desires.
Unless marketing is successful in directing the flow of de-
sired goods and services to the consumer, production will
diminish and our economy fail.

Marketing, therefore, is an integrative process
that attempts to unite the wants of the consumer with the
productive capabilities of the manufacturer. Increased
technological complexities of consumer purchases and the
increased amount and variety of consumer expenditures have
compelled the manufacturer to develop an effective means
to match its products with the needs of consumers.

The automobile manufacturer relies upon its fran-
chised dealer network, which consists of many locally owned
and operated dealerships, to operate as an effective means

of distributing its productive capabilities. It is essential

 

. 3v um a»
. v. .C n1 e; a? .A e; .3
. c o. .4 r. st 4‘ w. e‘ .s. rd 3. w. an

: E r. C .c 2 t I C 9 ,p. E S
E _ V .3 u a. e. f. e v 5 .1 .... S I E
T, .3 e I ,fi. 7 C I S V. t S . . .. .3 r T . I

I... 3 C e e s . I .. x e . e C d p. i l .
I .I t . . H C .. C .. C. C. .. 7. r i; _
.3 .1 I .... .3 2. C A. .Z t i z. T. t. f3 i l u I z c. 1.... .C .c.
h. . T. r” a” nu mu 4 A 2 2.; . . . i. 4‘ 7» Ab . pa. 42..» net. he .3 n v M: .U... by ..v
I a 3 .l a L I T. .r r. C .c C T .

 

 

 

 

that the manufacturer establish a corporate strategy for
all facets of its marketing effort in order to develop a
unified endeavor with its dealer network. The dealership
in turn acts as an intermediary between the manufacturer
and the consumer and must assist the manufacturer in the
implementation of the primary objective, which is the solu—
tion of consumer problems in ways commensurate with the en-
hancement of the firm's profit position.

In order to effectively accommodate the buying
habits and wants of customers in its particular sales ter-
ritory, the dealership attempts to adapt quantitatively and
qualitatively the manufacturer's product offerings. It also
adapts the manufacturer's recommendations to its own unique
competitive situation concerning such matters as the pro—
motion of products and product benefits, methods of improv-
ing advertising formats, and more effective utilization of
dealership facilities. Dealership location, local market
characteristics, financial resources, and personnel capa-
bilities influence the dealer's local marketing tactics.

Each automobile dealership employs a new car sales
staff, usually consisting of a General Sales Manager, a New
Car Sales Manager, and a sales force of between two and ten
salesmen. It is the responsibility of the New Car Sales
Manager to establish individual dealership sales tactics
and to supervise the salesmen. Being familiar with the

objectives of the dealership, he guides his salesmen in

 

1 f. Y. I Z S C. H.
O . . e c . r e
.. I mix ,1 T 3. T. 0 me 9 e .
a .C i t . e .c .
. a a t i C
T C . _ .. l .l 5
AV 4“ rL .“ 5 RM; A)
r. 7 .t a r
S . . e C a . . O n.
r‘ no .n~ 9; ‘» ‘» .mfld 1. HM a» «Us MW»;
e a. S S C I i. .8 VS 2. C
C. I .. . .3 .3 e C .. I. a... p 3..
t z s A C 5 L . s. u». “we . ..
C , Q ‘

 

 

III:
. (I “a.

A

such a manner as to benefit the dealership as well as main-
tain a high level of salesman satisfaction.

The automobile salesman is the final link in the
chain which unites the manufacturer's products with the
consumer. The salesman confronts the prospect with the
products in an endeavor to consummate a sale. At this
point of personal contact, three important elements of
personal selling are believed to be manifested by the sales-
man.

One of these factors is personality. Personality
is broadly defined as the sum total of all the traits pos-
sessed by an individual.1 When we say personality is a
factor in determining selling success, however, we mean
certain personality traits are influential. In the present
study, those human qualities of personality considered per—
tinent to personal selling are: appearance, sincerity,
persuasiveness, courtesy, friendliness, an understanding
of people, and sociability. These personal qualities will
be considered in the investigation as operationally defin—
ing personality.

Another important element of personal selling is
sales techniques. The categories of selling techniques
include: obtaining prospective buyers, qualifying customers,

planning and delivering sales presentations and product

 

1Kenneth B. Haas, How to Develop Successful Salesmen

(New York: McGraw-Hill Book CST, Inc., ififl, p. 272.

f. . . .. IA
5 r u .C e 3

fit. .7. O F .. e

aw .S a. ”v. em .u.
& My mm H“ «u
.C C 5 .a .3 .a“

 

kZCblé

4".
“$5

 

.u.
,.
‘Hces 0

LC

“"v- :

6

§
b

r

C

AA .

It. '-,
s‘v

 

demonstrations, handling objections, making trade appraisals,
closing the sale, and completing the follow-up after the
sale. These will be considered in this study as operation-
ally defining sales techniques.

The third influential factor in personal selling
is the salesman's knowledge of his products. Product know—
ledge may encompass knowledge of the quality, features, and
consumer benefits of the product, knowledge of the policies
and services of the manufacturer, prices, and an understand—
ing of the industry as a whole including competing products.
In many industries, including the automobile industry, prod-
uct knowledge is usually thought of as including (1) product
features, and (2) consumer benefits associated with the fea-
tures. This concept of product knowledge will be the opera—
tional definition of product knowledge used in this thesis.

It is the intent of this study to determine the
relationship between an automobile salesman's knowledge
of the product, including his ability to translate engin-
eering advances into consumer benefits, and his sales ef-
fectiveness.2 Since the salesman's personality and sales
techniques are assumed to be dominant factors in sales ef-
fectiveness, the relationship of product knowledge to these

other two factors will be determined, and the relative effect

 

2Engineering advances and product features are con-
sidered synonymous in this study.

of all three factors as determinants of success in the per—

sonal selling of automobiles will be illustrated.

Significance of th§_Problem

Exchange is the central concept in marketing and
personal selling is the catalyst at the point of exchange.
Wherever exchange takes place, personal selling exists,
unless the transaction is completely impersonalized. It
has been stated that if marketing functions were ranked
in order of their importance to the American economy, per-
sonal selling would be paramount because of its pivotal
role in building economic and social progress.3

Every manufacturer, whether he produces automobiles
or paper clips, must dispose of his output; for unless goods
which are produced are sold, the manufacturer cannot con-
tinue to operate. It is only through the sale of his prod-
ucts that he recovers the cost of raw materials, machinery
and labor, marketing expenses, and earns a profit. In a
very real sense, therefore, success in business is directly
dependent on how well the selling function is performed,
and the success of the economy, as measured by the growth
of the Gross National Product, relates directly to business
success.

It is evident, due to the importance of personal

 

3Milton Alexander and Edward Mazze, Sales Management-
Theory and Practice (New York: Pitman Publishing Corp.,
1965), p. 3.

 

 

selling and the vital role which the salesman plays in our
economy, that a more inclusive understanding of this sub-
ject is necessary if marketing is to become more analytical
and comprehensive.

Wendell R. Smith, in Emerging_Concepts in_Marketing,

 

states that personal selling is in danger of becoming a
"dark corner" as marketing research men and those in re-
lated fields devote their time to advertising and other
marketing problems.4 The scientific content and develop-
ment of personal selling has lagged significantly behind
that of the newer, more impersonal ways of communicating
persuasively with the market. This is because this subject
was assumed to be less appropriate for the application of
scientific marketing research methodologies. What research
has been done in the past was largely concerned with sales-
manship and the management of salesmen.

While a great deal of writing has been devoted to
salesmen and their success, research related to the deter-
minants of selling effectiveness has been neglected. As
a result, many of the opinions expressed in the sales lit—
erature are very subjective and unsubstantiated with research
evidence.

Another result of the lack of research is that there

 

4Wendell R. Smith, "Role of Selling in Modern Mar-
keting,” in Emerging_Concepts in Marketing, ed. by William S.
Decker (Chicago: American Marketing Association, 1963), p.
174.

 

 

II ET TI TI ‘IiIII

 

is little agreement among educators or executives as to
what factors are most influential in determining a sales—
man's effectiveness. The relative importance of the deter-
minants they do assume to be influential also remains un—
resolved. The importance of product knowledge, as one facet
of personal selling, is stressed as the most vital element
of successful selling by some authors, while others rele—
gate it to a minor role in the determination of selling
effectiveness.5

A well-known textbook, Salesmanship Principles and
Methods, for example, states that most marketing men accept
the assumption that sales volume is directly related to
the amount of knowledge which the salesman possesses about

the product.6 Alfred Gross, another authority on salesmanship,

 

5Richard W. Husband asserts that the number one rule
of successful selling is thorough knowledge of the goods.
The Psychology gf Successful Selling (New York: Harper and
Brothers Publishers, 1953), p. 27. Richard L. Small says,
"Of all the rules of selling, 'Know the product' is probably
the most important." Salesmanship (New York: The Macmillan
Co., 1952), p. 27.

On the other hand, L. Mercer Francisco, "How Can
You Get Salesmen to Sell Today?" Sales Management, 79
(December 6, 1957), 62-68, emphasizes that salesmen are
given too much product information and not enough training
in sales techniques. Robert McMurry, "The Mystique of Super-
Salesmanship," Harvard Business Review, 39 (March-April, 1961),
p. 117, feels that both sales techniques and product infor-
mation have been over-emphasized. McMurry argues that sell—
ing is a human task and only personality and psychology can
explain successful salesmanship.

6Carlton Pederson and Milburn Wright, Salesmanship
Principles and Methods (4th ed.; Homewood: Richard D. Irwin,
Inc., 1966), p. 185.

declares that a qualified salesman must know and understand
all ramifications of information pertaining to the product
he sells.7 Willard Thompson emphasizes the importance of
personality attributes in his eight elements of successful
selling,8 while Ivey, Horvath, and Tonning's fundamentals
of selling stress the importance of sales techniques as
determinants of sales effectiveness.9

These diverse statements are partially due to the
lack of a universally accepted definition of product know-
ledge. Many authors employ the term without stating any
intended meaning, whereas others restrict it to factual
knowledge of product features. An even broader concept
would include an understanding of the consumer benefits
associated with the product.

Furthermore, those authors who define product know—

ledge as knowledge of product features and the associated

consumer benefits, disagree as to their relative importance.

 

7Alfred Gross, Salesmanship Principles and Practices

.9: Professional Sellinng2nd ed.; New York: The Ronald Press
COO, 1959), pp. 83-840

8Willard M. Thompson, Salesmanship: Concepts, Man—
agement, and Strategy (New York: John Wiley & Sons, Inc.,
1963), p. 33. According to Thompson, there are eight essen-
tial elements in effective selling: 1- salesmanship skills,
2- adequate knowledge of company and products, 3 through 8-
personality attributes.

9Paul Ivey, Walter Horvath, and Wayland Tonning,
Successful Salesmanship (4th ed.; Englewood Cliffs: Pren—
tice—Hall, Inc., 1961), pp. 6—7. Of their seventeen funda—
mentals of selling, personality is mentioned once and prod—
uct knowledge is not referred to at all.

 

 

 

 

 

Some writers argue that the principal role of the
salesman and the bases of his effectiveness are in demon—
strating or explaining what product features mean in terms
of consumer benefits. Charles A. Kirkpatrick, for example,
emphasizes the importance of consumer benefits when he crit-
icizes many salesmen for spending too much time studying
product features and not enough time learning about the
product's benefits. He states:

The salesman, instead of selling product features,
should sell the effects and results of buying . . .
the salesman must sell what the product will do,
not the product itself—-because buyers buy future
satisfaction, not merchandise.lo

This position is supported by other well-known sales-
manship authorities who stress that facts about product
characteristics are usually considered less important from
the consumer's standpoint than what the facts mean in terms

11

of benefits, expectations, and satisfactions.

A widely-used textbook, Textbook g: Salesmanship,

 

elaborates upon the importance of emphasizing consumer

 

10Charles A. Kirkpatrick, Salesmanship: Helping
Prospects Buy (4th ed.; Cincinnati: South—Western Publish-
ing Co., 1966), pp. 132, 143.

llPederson and Wright state that " . . . the buyer
is not interested in facts about the product other than to
the extent to which these facts have a bearing on the solu-
tion of his wants or needs." Salesmanshiprrinciples and
Methods, pp. 203-204. This same theory is noted by Ivey,
Horvath, and Tonning when they recommend that salesmen should
not dwell on facts about product features, but should em-
phasize what the product will do for the consumer since
buyers are primarily interested in their own wants and needs.
Successful Salesmanship, p. 226.

 

 

 

10

benefits by illustrating that after Magnaflux Corporation
adapted the view that, "The typical prospect is interested
in just one thing—-what the item you are selling will do
for him," their sales increased 80 percent.12

As Joseph W. Thompson points out, product knowledge
must be designed to persuade. For product knowledge to
persuade, it must be presented in such a way that the con-
sumer realizes the advantages of buying.13

In order to demonstrate that his product can satis-
fy the needs of the consumer, the effective salesman must
present both the features of his product and the correspond—
ing consumer benefits, as many buyers are unable to recog-
nize the merits of product innovations without assistance.14

A comprehensive knowledge of the product, including
a corresponding knowledge of its consumer benefits, permits
a salesman to deliver his product presentation in an effort-
less and convincing manner. It also allows the salesman

to concentrate on other aspects of the sale, such as learning

 

12Frederick Russell and Frank Beach, Textbook 2£_
Salesmanship_(6th ed.; New York: McGraw-Hill Book Co.,
Inc., 19597, p. 111.

13Joseph W. Thompson, Selling: A Behavioral Sci—
ence Approach (New York: McGraw-Hill Book Co., Inc., 1966),
p. 174.

 

14This assertion is supported by the following
authors: Pederson and Wright, Salesmgpship Principles and
Methods, p. 203. Gross, Salesmanship_PrincipIes and Prac-
tices 9: Professional Sellipg, p. 121. Kenneth B. Haas and
Enos C. Perry, Sales Horizons (Englewood Cliffs: Prentice-
Hall, Inc., 1958), p. 85. Russell and Beach, Textbook 9:.
Salesmanship, p. 98.

11

more about the customers' needs and observing them for signs
that they are ready to close.15 Other advantages of thorough
product knowledge frequently mentioned in the literature
include increased self confidence, greater enthusiasm, more
job satisfaction, and increased ease in overcoming objec-
tions. Table 1 summarizes many of the advantages of product
knowledge and indicates which authors consider them impor-
tant in personal selling.

The diversity of opinions concerning product know—
ledge may be related to the industrial perspective of the
various writers. The term product knowledge has been ap-
plied to all forms of personal selling without reference
to specific products or groupings of products. It could
be postulated that the amount of product knowledge a sales-
man possesses would vary among different product groupings.

This belief is evidenced by Pederson and Wright,

who state that:

The amount and kind of information a salesman may
have to acquire will depend on the

. kind of selling he performs

. nature of the product

. the character of the customer upon whom he
calls

. and the complexity of the organization of
the company he represents.16

 

15Harold C. Cash and William J. E. Crissy, “The

Use of Appeals in Selling," The Psychologygf Sellin , Vol.
3_(Flushing: Personnel Development Associates, 1957;, p. 22.

l6Pederson and Wright, Salesmanship Principles and
Methods, p. 188.

12

TABLE 1

BENEFITS DERIVED FROM PRODUCT KNOWLEDGE:
LITERATURE SUMMARY

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

p
c 17
ow c c
-H U . o
u m c m
h . :3 m 0 Q
M n m m E
u p o a o
m mtm x c 'o E c
m >£: o m c o B
u-a-a run
o 0:: u a U a.» o
c :n:: u o H .4 m c
m c) m m m :4 -m m
m ~54 D. u -H 0 C23
m m >. x m m m m 3
$82238§2W22
Product Knowledge Benefits I 0 H¢u g m U (x 3:0 m
Adapts product to consumer X X X X X X X X
Increases self confidence X X X X X X X
Stimulates enthusiasm X X X X X X X
Promotes effective sales X X X X X
presentation
Overcomes objections X X X X
Increases job satisfaction X X X X
Develops professionalism X X
Meets competitive claims X
17

Haas and Perry, Sales Horizons, p. 85.

Gross, Salesmanshingrinciples and Practices of Professional
Selling, p. 84.

Ivey, Horvath, and Tonning, Successful Salesmanship, pp.
108, 226. *

Kirkpatrick, Salesmanship: Helping_Pro§pects Buy, pp. 129-130.
Pederson and Wright, Salesmanshingrinciples and Methods,
pp. 185-186.

Bertrand R. Canfield, Salesmanship Practices and Problems
(3rd ed.; New York: McGraw-Hill Book Co., Inc., 1958), p.
358.

Russell and Beach, Textbook of Salesmanship, pp. 94, 96—97.
George Breen, Ralph B. Thompson, and Harry West, Effective
Selling (New York: Harper and Brothers, 1950), pp. 54-56.
Steven J. Shaw and Joseph W. Thompson, Salesmanship: Modern
Viewpoints 22 Personal Communication (New York: Henry Holt
& Co., Inc., 1960), p. 376.

 

 

13

Although other writers would disagree with this
list, there seems to be general agreement that the nature
and/or complexity of the product is a significant factor
in determining the amount and kind of product knowledge
required for effective salesmanship.18

Many firms presently accept this assumption. In
a survey conducted by the National Industrial Conference
Board of business opinion and experience of 153 manufactur-
ing firms, it was found that:

Companies that place the greatest emphasis upon
product knowledge are usually those having prod- 19
ucts or product applications of a technical nature.

If this theory were true, then, salesmen selling
simple consumer staples would need comparatively little
product knowledge, whereas the effectiveness of industrial
salesmen of technical equipment, for example, would depend
principally upon their product knowledge.20

Unfortunately men working in this field must resort

to abstract generalizations concerning the relative impor-

tance of product knowledge in personal selling because of

 

18Bertrand R. Canfield, Sales Administration, Prin-
ciples and Problems (4th ed.; Englewood Cliffs: Prentice—
Hall, Inc., 1961), p. 480. Ivey, Horvath, and Tonning,
Successful Salesmanship, p. 106. Husband, The Psychology
2; Successful Selling, p. 27. '

19Alexander and Mazze, Sales Management--Theory_and
Practice, p. 269.

 

 

20Thomas F. Stroh, Salesmanship: Personal Communi-
cations and Persuasion in Marketing(Homewood: Richard D.
Irwin, Inc., 1966), p. 51.

 

 

14

the limited research available upon which they can base
their discussions. Many studies have been conducted to
determine what personality traits or tests are useful in
predicting a salesman's effectiveness, but these do not
include any reference to product knowledge.21
A study, which did include product knowledge, is
reported by Kenneth Haas in ng.£g Develop Successful Sglggf
ESE! A questionnaire was sent to several hundred sales
supervisors concerning what personal traits they felt were
important for successful salesmen. In the resulting group
of twenty traits, product knowledge was found to be one
of the fundamental knowledges required in successful sell-
ing.22
A similar study is reported by William J. Tobin.
He conducted numerous surveys of salesmen, sales executives,
and psychologists concerning what makes a man a successful
salesman. Tobin found that there were twenty essential

traits, one of which was product knowledge.23

 

21Paul F. Ross and Neil M. Dunfield, "Selecting

Salesmen for an Oil Company," Personnel Psychology, 17
(Spring, 1964), 75-84. M. D. Dunnette and W. K. Kirchner,
"Psychological Test Differences Between Industrial Sales-
men and Retail Salesmen," Journal of Applied Psychology,
44 (1960), 121- 125. Thomas W. Harrell, "The Relation of
Test Scores to Sales Criteria," Personnel Psychology, 13
(1960), 65- 69.

22K. Haas, How tg'Develop Successful Salesmen, pp.

23William J. Tobin, "What Makes a Man a Successful

Salesman?" Sales Management, 83, Part 2 (October 2, 1959),

15

Unfortunately, neither Haas nor Tobin describes his
research methodology well enough to evaluate the credibil—
ity of their conclusions. The first well-defined study of
the determinants of successful selling was conducted by
Eugene J. Benge.

Benge surveyed 564 salesmen in widely varying types
of industry. His objective was to determine what factors,
personality traits, attitudes, and work habits character-
ize successful salesmen. In an effort to determine the
value of technical knowledge, the salesmen-were rated by
their sales managers either excellent or poor on five dif-
ferent traits. Benge found that on the traits, "Well-In-
formed in his Field" and "Is Good Technically," the excel-
lent salesmen exceeded the poor salesmen by 70 and 75 per—
cent, respectively. For the total five product knowledge
traits, the excellent salesmen exceeded the poor by 44 per-
cent. Benge concluded that the big differentiations between
excellent and poor salesmen occur in self confidence, the

ability to plan, industriousness, and persuasiveness. Of

 

121-124.

It is interesting to note that 19 of the essential
traits that Tobin found were identical to those developed
in Haas' study conducted two years earlier. The only dif-
ference is Haas lists dependability as necessary, whereas
Tobin found reliability an essential trait. The other 19
factors were: integrity, knowledge of product, self—man—
agement, work organization, sincerity, initiative, indus-
triousness, acceptance of responsibility, understanding
of buying motives, sales ethics, judgment, care of health,
courtesy, determination, aggressiveness, friendliness, re-
sourcefulness, persuasiveness, and appreciation of selling.

16

less value, but still important, Benge concluded, are the
differentiations of intelligence and technical knowledge.24
The only two studies to date which attempt to deter-

mine the effect of product knowledge upon sales performance

are published in Personnel Psychology concerning life insur-

 

ance salesmen.

Donald B. Baier and Robert D. Dugan studied the
relationships among a short mental ability test (Wesman
Personnel Classification Test) and a test of life insurance
knowledge (The Information Index), with three kinds of per-
formance criteria. The performance measurements were ob—
jective production records, ratings, and job levels. These
tests were administered to 596 debit insurance agents, 126
staff managers, and 42 district managers.

The Wesman Personnel Classification Test did not
correlate significantly with any of the performance criteria
within job level samples. On the other hand, The Informa-
tion Index was related to some of the criteria in the two
management samples but not in the agents sample. Baier
and Dugan concluded that the effects of technical knowledge
of life insurance and mental ability upon sales performance
in the study are probably obscured by other variables, such

as local market conditions, "natural sales ability,"

 

24Eugene J. Benge, "What Traits and Work Habits
Characterize Successful Salesmen?" in Selling. Its Broader
Dimensions, ed. by Taylor W. Meloan and John M. Rathmell
(New York: The Macmillan Co., 1960), pp. 37- 43.

17

enthusiasm, drive, and effectiveness of planning and self
management.25

In a study of insurance salesmen conducted two years
later by P. W. Thayer, J. A. Antoinetti, and T. A. Guest,
again using The Information Index as a measurement of prod-
uct knowledge, the findings were significantly different.
They investigated the correlation between product knowledge
scores and a performance index, which measured the propor-
tion of insurance that was dropped by policyholders after
having made the initial payments. The correlations for
the two performance index periods were —.29 and -.27 for
sample sizes of 62 and 49 salesmen respectively. The re-
searchers concluded that the possession of technical know-
ledge is an asset for the life insurance agent. They state
that the study supports the position of those who favor
technical training for insurance agents.26

The survey of the literature concerning the research,
which has been conducted on the importance of product know-
ledge in relationship to sales effectiveness, indicates

that more detailed studies are required if an objective

appraisal of product knowledge is to be obtained.

 

25Donald E. Baier and Robert D. Dugan, "Tests and
Performance in a Sales Organization," Personnel Psychology,
9 (Spring, 1956), l7-26.

26Paul W. Thayer, John A. Antoinetti, and Theodore A.
Guest, "Product Knowledge and Performance--A Study of Life

Insurance Agents,“ Personnel Psychology, ll (Autumn, 1958),
411-418.

 

c O
5.8:—

e

 

.c I L . .. S -1 . i E ,
6 5 .~ . “W. New .s* s; M“ 9“ “M Q» .~.H “Rum
I . .2 .l .3 , m... u
C e v m“ a f r C we no. a.» a“
. m .s. u. r. S . a: K.» we V4
. 3 . i uh. .. .. an“ Km. 2.. 6.. at c . :u

. . . i : . D. C .:.

 

18

The relationship between product knowledge and sales
effectiveness is significant to marketing men both in indus—
try and the academic professions. The industrial sales pro-
motion and training people need an accurate basis upon which
to most effectively allocate their resources in their sales-
men training programs and in the development of promotional
literature concerning their product features. The academic
marketing men, on the other hand, need to know the relative
importance of product knowledge in order to accurately de-
fine its place in understanding the concept of salesmanship,

and, perhaps, in other aspects of marketing.

Analytical Perspective

The marketing of automobiles is a very involved
process. A salesman's knowledge of his product, including
his ability to translate product features into consumer
benefits, is not the sole determinant of his effectiveness.
.In order to compete effectively, the salesman takes advan-
tage of prices, credit arrangements, service facilities
including the availability of replacement parts, service
personnel, amount of local advertising, location and extent
of the dealership facilities, including demonstrators, dis-
play areas and office space.

Just how the individual dealership combines this
multivariate competitive mix is influenced by the territor-
ial differences in sales potential, demographic character-

istics of the local market, physical differences of market

19

areas, size of the sales staff, the competitive environment,
and financial resources available to the dealership. The

salesman has very little control over these factors. There-
fore, his performance relative to that of his fellow sales—
men in the same dealership is determined in large measure

by his ability to take advantage of the resources available
to him and the conditions over which he has control; namely,

personality, sales techniques, and product knowledge.

§glg§_experience

Even though product knowledge is believed to be a
determinant of selling effectiveness over which the sales-
man has control, there may be various factors which have
a significant influence upon his level of product knowledge.
One of the most obvious factors, which could affect a sales—
man's knowledge of product features and their corresponding
consumer benefits, is his sales experience within the auto-
mobile industry. It seems logical that after exposure to
automobile innovations over a period of time, a salesman
would develop a broad understanding of product features.
Therefore, as new engineering advances develop, the ex-
perienced salesman would be in a better position to assim-
ilate and retain new product information than one with less

experience.

Education

The level of a salesman's product knowledge could

20

also be influenced by the amount of his formal education
and/or his attendance at a manufacturer-sponsored training
course. Higher education may provide a salesman with better
study habits and methods of acquiring factual information

as well as a broader perspective from which to evaluate
proddct information. Training sessions, on the other hand,
may stress the importance of learning product knowledge

so that a salesman who has attended one of these courses

may be more motivated to acquire a thorough knowledge of

the product than those who have never attended a manufac-

turer-sponsored training course.

9£hg£_factors

It is important to note that there are other fac—
tors which could be influential in the automobile salesman's
attitude towards the acquisition and importance of product
knowledge. Some of these possible contributory elements
include: mechanical aptitude, peer group composition and
interests, sales manager's attitude, quality of the infor-
mative materials provided, company and dealership advertis—
ing programs, product displays, competitive promotional

programs, and trade publications.

Type and size g; market served

 

The concept of product knowledge in the personal
selling of automobiles is analyzed further by considering

the effect of the type of market served and the size of

21

the dealership serving those markets. It is believed that
there may be differences in the amount and kind of product
knowledge employed by salesmen in large metropolitan mar-
kets as opposed to those in less densely populated areas
and that the size of the dealership in each of these areas
may have an effect upon the salesman's level of product
knowledge.

Any study of the importance of product knowledge
in the automotive industry would have to quantitatively
relate product knowledge to a measure of sales performance
in order to determine the influence of this knowledge upon
selling effectiveness. The various criteria used in the
measurement of salesmen's effectiveness, and the specific
criteria of sales performance selected for use in this study

are discussed in detail in Chapter II.

Objectives g t_he.flgy

The primary objective of this thesis is to deter-
mine the relationship between product knowledge and sales
effectiveness in the retail selling of automobiles. Inas-
much as the salesman's personality and sales techniques are
assumed to be dominant factors in sales effectiveness, the
relationship of product knowledge to these two factors will
be determined. Thus, their relative effect as determinants
of success in the personal selling of automobiles will be
illustrated. Also studied in the first objective will be

the effect of type of market and retail outlet size upon

22

the relationship between the salesman's product knowledge
and his sales performance.

The second objective of this study is the deter-
mination of the influence of various factors on a salesman's
level of product knowledge. These factors include: level
of education, the duration of sales experience within the
automobile industry, attendance at a manufacturer-sponsored
training course, the type of market, and the size of the
retail outlet. A model of the relationships described in
these two objectives is shown in Figure l.

The third objective is to determine if the two com-
ponents of a salesman's product knowledge; knowledge of
product features and knowledge of their corresponding con-
sumer benefits, have an equal effect upon his sales effec-
tiveness.

Hopefully, this study will contribute important
knowledge concerning what factors influence effectiveness
in the personal selling of automobiles. It may, by extrapo-
lation, shed new insights into the determinants of sales

success in general.

Hypotheses
The hypotheses were developed based on the cri-
terion that each could be tested empirically using appro-
priate experimental design and methodology.
The hypotheses are:

1. That product knowledge (in conjunction with

 

23

FIGURE 1
noun or I‘ll! umlousmrs to u swmo

 

p
/ 1%th
§) “‘0 JII: , ,, ‘
9) \ ‘ o IENO fi’III/I’QV
O . g ”z, ‘1.

   
 

 

 

. 0
Q“) «R

r

1 a, 5,:
° ‘9 v I

‘"g «a e

\ '5 ‘9 o
95‘ P999 /

24

two other assumed determinants, personality
and sales techniques) is one of the primary
determinants of sales effectiveness and that
each of these variables can be measured.

This is the central hypothesis of this thesis.
That the relationship between a salesman's
level of product knowledge and sales effective-
ness is influenced by the following:

A. type of market in which he sells

B. size of the retail outlet in which he sells.
That a salesman's level of product knowledge

is influenced by the following:

A. the type of market

B. the size of the retail outlet

C. his level of education

D. his attendance at a manufacturer-sponsored
training course

E. his sales experience within the industry.
That if a salesman's level of product knowledge
is divided into knowledge of product features
and knowledge of consumer benefits, each of
these two components of product knowledge con-

tributes equally to his sales effectiveness.

CHAPTER II

CRITERIA FOR MEASUREMENT OF SALESMEN'S EFFECTIVENESS

An accurate method for measuring the salesman's
performance must be selected if the relationship between
an automobile salesman's product knowledge and his sales
effectiveness is to be determined. Any measurement for
evaluating sales performance has limitations, since many
other variables influence sales effectiveness aside from
the performance of the individual salesman. These other
variables include advertising, sales promotional materials,
pricing, and the dealership's facilities and reputation.
The variations in market and competitive conditions among
geographic territories further complicate performance com—
parisons. The salesmen and their performance may also be
affected by other uncontrollable factors, such as terri—
torial differences in product loyalty and/or changes in
local economic conditions.

It is difficult to establish a technique of meas—
urement which will be successful in various types of dealer-
ships serving diversified markets. Ideally, sales perform-
ance should be measured on the basis of existing and antic—
ipated conditions in each particular territory and for each

dealership within that area. In addition, the choice of

25

26

a criterion of salesmen's effectiveness is further compli-
cated because it is generally the practice to have the sales-
men sell both new and used cars. This means that the measure
selected must place proper emphasis upon both new and used
car sales.

Due to the above factors, it was decided that the
measures of sales effectiveness used should include at least
one that would be applicable to an inter-dealership analysis;
i.e., the same regardless of dealership and market condi-
tions and at least one that could be employed in an intra-
dealership analysis which would be peculiar to the local
dealership and its market.

The inter-dealership method of analysis makes it
possible to compare one salesman's performance with that
of all the other study participants. This would increase
the sample size and hence the reliability of the findings.
However, this method assumes selling opportunities to be
equal or near equal among the dealerships.

The intra-dealership method of analysis overcomes
many of the obstacles usually encountered when comparing
the influence of certain factors upon sales effectiveness.
It enables other factors which contribute to sales perform-
ance, such as local advertising, dealership promotional
materials, and dealership facilities to be held constant.

The influence of these factors upon the salesmen within
each dealership can be assumed to be equal or near equal,

thus, the differences in selling opportunities should be

3» . - . .

Z. O .3 e E C .1 I I E C l
4 . u a» L»

T C S T .. . 2 S E I

L .. r e :_ . .: s i
K .3 e . . C .1 I .u C I . _

.. l .3 .7 :. C ... i .l .. I E I C S I I
—m .»u ow «v "I .5 p... 9.. In ~u. “L. Q» T... .3 AV Ls

 
 

 

 

.—

27

minimized.

Specific methods of measuring sales performance
had to be determined. There are various ways to measure
the salesman's level of achievement or effectiveness. Some
of the measurements considered were: absolute unit sales
volume, sales volume compared to set quota, personal impres-
sion of the salesman by his supervisor, number of contacts
made in relationship to the number of sales produced, the
number of new customers gained, gross margins secured, the
number of product demonstrations given, gross sales volume,
gross earnings, net profitability of sales, and gross profit
per customer.

Of the many criteria available to measure the sales
effectiveness of automobile salesmen, six of the above are
most frequently mentioned in the literature on this subject
as being reliable means to measure the salesman's level
of achievement. The disadvantages of each criterion are
briefly discussed below:

1. Number g£_Units Sold

The problem associated with this method is that
it does not differentiate the selling efforts
associated with differently priced products.

It is generally believed that it is more dif-
ficult to sell a new car than a used car, and
an expensive product as compared to a less ex-
pensive model. Arbitrary weights could be ap-

plied to each of the car lines and the used car

28

category in order to compensate for this dis-
advantage. A question is then posed, however,

as to the appropriateness of the assigned weights.

Sales Volume Compared thPredetermined Quota

This method of measurement is a refinement of
absolute sales volume because it compares ac-
tual sales with potential sales. Sales objec-
tives, however, are only determined for the
individual dealership, based upon consideration
of the market potential, past sales performance,
competition, advertising and sales promotion,
the sales force, and dealership facilities.

The management of each dealership then develops
its own method for allocating the sales objec-
tives in the form of sales quotas among the
salesmen. The method usually employed in the
allocation is based heavily upon past sales
performance and varies with the span of time
covered. Therefore, if this measure were used
there would not be communality among the dealer-
ships.

Gross Sales Volume

 

The gross sales volume does not account for
the price ranges of various models offered.
In addition, it does not reflect the sales of
optional equipment and accessories, the value

given for cars traded in, or the associated

29

salesman and dealership profitability.

Salesman's Earnings

 

Salesman's earnings over a period of time would
reflect the individual's sales efforts because
a large percentage of automobile salesmen are
compensated by means of commissions. All auto—
mobile dealerships, however, do not follow the
pure commission pattern of remuneration. The
system employed varies from dealership to dealer-
ship and even within dealerships. For example,
a recently employed salesman might be compen—
sated by a modest salary in addition to his
commissions, while within the same dealership,
some of the more experienced salesmen are re-
munerated by commissions only and others may
receive straight salaries.

Profits Contributed tg_the Dealership_

 

 

 

Profits are the primary purpose of a dealership,
thus the amount of profit a salesman is able

to return to the dealership is a justifiable
measurement of his performance. These figures,
however, are not systematically recorded by

the dealerships for each salesman so that the
choice of this criterion of measurement would
involve a large amount of clerical time and
effort on the part of the dealerships selected

to participate in this study.

3O

6. Ratings by Supervisors

The most subjective criterion of measurement

is supervisory ratings of salesmen. The super-
visor making such ratings would presumably con—
sider both the quantitative and the qualitative
aspects of each salesman's effectiveness. The
weights the rater applies to the many factors
considered in the evaluation, however, are ar-

bitrarily selected. Since there are no objec-

- g?- ‘D-“-..-n-n;uv__ -‘-o
I"
. ._,_,v. —
,

tive standards used, the accuracy of the ratings
fluctuates according to the supervisor's per-
ception of each of the salesmen.

There is no one criterion which is comprehensive,
letely reliable and free from defects in measuring an
mobile salesman's performance. Accordingly, it was
ded that a multiple criterion method should be employed
ncrease the reliability and validity of the appraisal
ales effectiveness.

An administrative restraint on the choice of per-
ance measurements was that at least one of those se-
ed for use in the study had to be supplied by the re-
dents. This provided the linkage between the informa-

generated by the salesmen and the data furnished by
gement. (See Chapter III for details.) This considera-
eliminated some measures that otherwise might have
used.

One of the performance measures selected was that

 

31

of total units sold over a six months' period, from January 1,
1967, to June 30, 1967. In the automobile industry, absolute
sales volume is the most commonly used technique of measur—
ing a salesman's performance. The emphasis placed upon
total units sold, according to car line, by the manufacturer
through national incentive programs and dealership super-
visory personnel in their attempts to increase productivity,
makes this criterion one which is easily recallable by all
automobile salesmen. Second, the salesman typically thinks
in these terms in judging his progress in reference to previ-
ous time periods. Thirdly, sales volume has a common mean—
ing among dealerships. There is minimal likelihood of mis—
understanding it. Fourth, it can be employed without per-
sonally identifying each salesman, consequently insuring
the salesman's freedom of expression on the questionnaire.
Many writers in the field of marketing have stressed
that, in line with the new marketing concept, mere volume
of sales is not a sufficient criterion for measurement of
a salesman's performance. Bertrand Canfield, for example,
states that using sales volume alone may fail to give a
complete understanding of a salesman's effectiveness because,
unless the profitability factor is considered, a false im—
pression of his performance may result.27 In the automobile

industry, salesmen may sell a high volume of used cars and

 

27Bertrand R. Canfield, Sales Administration, Prin—
ciples and Problems (4th ed.; Englewood Cliffs: Prentice—
Hall, Inc., 1961), p. 477.

 

n. a . .. .

a e e a .c e

T r .I . c c .

. - .2 .1 z. .r.

S e e 5; at r» o
i C C .1 “1
,mn “A S r; r; .3 mm
o. .fiu :- D. at .. . xv

1v"

 

~ ..
.3 S
.3 it
a» . a
u u .1
; a .3
I'll"!

.IA

.9. .3 s S E .
.2 ,2 C. T. ..... . I .._. I c.
“a i. L. 2. ~» 4: as Nb
a» r. ... r. :— ~C 2‘ Ce —«
S . t S a .. E S E C .. ..

RV

I
.2
A C

 

 

32

thus have a high volume of sales, but since used cars pro-
duce relatively low profits, their performance may not be
as effective as those who have a lower sales volume of more
profitable cars.

Peter Drucker states that profitability is the most
critical measure of net effectiveness.28 Certainly profit
is the ultimate goal of all business in a competitive econ-
omy. Therefore, profitability is the ultimate test of bus—
iness performance, and the most critical measurement of a
salesman's effectiveness is the profitability of his sales.

Accordingly, the second criterion employed in this
study to measure salesmen's performance is the profit con-
tributed by their sales volumes to the individual dealer—
ships based on national averages. This is obtained by first
translating the national average gross profitability for
each new car line and used cars into a proportional value
scale. The resulting numbers are used as multipliers for
each salesman's new and used car sales volumes. The sum
of these two products represents a measure of the salesman's
profitability contribution.

Retail gross profit margins for both new and used
cars vary among the various car lines and also among the
series and types of models available within each car line.

Gross margins are seasonally affected as well. New car

 

28Peter Drucker, The Practice g: Management (New
York: Harper and Brothers, 1954), p. 76.

 

33

profits, for example, are highest in the fall of the year
with the introduction of new models. They taper off dur-
ing the model year to a low point in July and August. The
gross margins incorporated into this study, therefore, were
calculated weighted national averages, based upon the six
months used as the sales period and taking account of the

models sold during this period.

CHAPTER III
STUDY METHODOLOGY

Research Desigp

The data needed for testing the hypotheses as set
forth in Chapter I were obtained by questionnaires and rat-
ing schedules administered by field personnel of the Lincoln-
Mercury Division of the Ford Motor Company to randomly se-
lected Lincoln-Mercury supervised automobile dealerships.

A questionnaire was employed as a method of measuring the
selected automobile salesmen's level of product knowledge.
The rating schedule, which consisted of three paired com-
parisons, was used to determine the sales managers' opin-
ions concerning their salesmen's relative strengths in the
areas of personality, product knowledge, and sales tech-
niques.

Administrative representatives of the company, gen-
erally referred to as zone managers, visit dealerships per-
iodically to inspect their operation and advise them with
respect to problems of accounting, personnel, service, and
other phases of dealership operations. Zone managers re-
sponsible for the selected dealerships were asked to admin-
ister the study by their General Field Managers.

In preparation for this, the data collection aspects

34

35

of the study were forwarded from the Lincoln-Mercury Divis-
ion General Office to each of its eighteen national districts
with an introductory letter to the General Field Managers
explaining:

. The purposes of the study

. Which dealerships in the district were selected
to participate

. Where and when the participating zone managers
should administer the data collecting instruments

. To whom the zone managers should administer the
questionnaire

. Who should complete the rating schedules and why

. How the zone managers should return the completed
evaluation forms to the Division

. That the information received from both the ques-
tionnaires and the rating schedules would be coded
so that individual salesmen and dealerships are
not identified.29

Each zone manager was furnished a sheet of instruc—

tions explaining his responsibilities in the administration
of the study.30 The zone managers were instructed to arrange
special breakfast meetings for the new car salesmen of the
selected Lincoln-Mercury dealerships. This would insure
proper meeting facilities where the salesmen could assemble

and complete the questionnaire without interruptions. It

was also felt that the salesmen participating in the study

 

29A copy of the Introductory Letter is included in
Appendix A.

30A copy of the Instruction Sheet is included in
Appendix B.

36

would be encouraged to do so by providing them with a com-
pany sponsored breakfast.

At these breakfast meetings, the zone managers ad—
ministered the questionnaire to the new car salesmen and
elicited the cooperation of the sales managers in complet-
ing the rating schedules.

Zone managers were incorporated into the study plan
in order to preclude the possibility of the salesmen using
source materials while completing the questionnaire. The
zone managers were also employed to gain increased coopera-
tion of the sales managers and the participating dealerships.
More reliable results and a greater percentage of completed

returns were expected as a result of this method.

Data Gathering Instruments
Questionnaire design
The purpose of the questionnaire was to determine
the following information about the salesmen:
. Their knowledge of the product

. Their ability to translate product engineering
features into consumer benefits

. Their educational level
. Their sales experience in the automobile industry

. Their most recent attendance at a manufacturer-
sponsored training course

. Their sales volume according to car line.
The questionnaire, shown in Appendix C, was divided

into three areas. The first two sections of the questionnaire

37

were designed to place approximately equal emphasis on the
Mercury, Mercury Intermediate, and Cougar car lines. Any
product feature, which might have a disproportionate em—
phasis placed upon it because of geographic or climatic
conditions in certain areas of the country, was eliminated.

The questionnaire was structured to measure both
the salesmen's knowledge of product features and their know-
ledge of the corresponding consumer benefits. This was done
in order to test the hypothesis that if a salesman's level
of product knowledge is divided into knowledge of product
features and knowledge of consumer benefits, each of these
two components contribute equally to his sales effectiveness.

The first section of the questionnaire was struc-
tured to determine the salesman's knowledge of the 1967
Mercury full-size, Mercury Intermediate, and Cougar prod-
uct features. Fifteen multiple-choice questions were equally
devoted to feature availability, vehicle design and construc—
tion, and mechanical features. The respondents were in—
structed to select the correct phrase out of four possible
answers, three of which were incorrect.

The second segment of the questionnaire measured
the ability of the salesman to translate product features
into consumer benefits. These questions were devoted equal—
ly to the areas of comfort and convenience, safety, and
vehicle performance and durability. The features examined

in this portion of the questionnaire corresponded to those

h'.‘-’_ -—' ‘1 ‘-.§
.
L .

‘7- 1..

 

 

. 6.. .v .3 x
lb : . .C r A: s r. by MW
. 1.: ~u .. 1: rs v. e .D “u. .«C .A.‘ wt. v. . :9
n? r” .3 a. . 1 C u . s“ . .. I. we. 2. s . 1.5. M.» a.» ..t €¢
Av e .u .nu ~.. ... «n» .u‘ .. .~¢ AU Q5 4: s .0 AU

m)

 

 

 

 

38

used in the first section so that the relationship between
the respondents' knowledge of product features with sales
performance and their knowledge of the corresponding con-
sumer benefits with sales performance could be determined.

Section II of the questionnaire consisted of seven
open-ended and eight multiple-choice questions. The open-
ended questions were designed to simulate the selling sit-
uation so that the salesman's ability to present product
benefits to prospective buyers could be evaluated. The
questions were constructed so that one consumer benefit
for each product feature tested was mentioned. The respond-
ent was asked to list other benefits he had found important
in selling the feature to customers. This not only allowed
the salesman to display some creativity, but it provided
insights into the information actually used in selling as
opposed to rote product knowledge. To conserve time, how-
ever, only half of the consumer benefit questions were open-
ended. The multiple—choice questions in Section II were
structured the same as those in Section I with only one
correct answer out of four for each question.

The third section of the questionnaire was devoted
to obtaining classification information. Each salesman
was asked to designate his own educational level, length
of experience as an automobile salesman, and length of time
which had elapsed since his attendance at a Ford Marketing

Institute training session. The last question asked each

S— A:

9-
aa e
T. . .
.. “VA
.1 Tu.
2. AV
:5 h!»

each

‘7
gt:

.& A
.3
Q»

A

7‘
Q.‘

 

tgf‘
‘~

:~

C.

‘
x
y

.I
C

 

 

 

 

39

respondent to list the total number of automobiles he sold
from January 1, 1967, through June 30, 1967, for each of
the following categories: Mercury, Mercury Intermediate,

Mercury Cougar, Lincoln—Continental, and Used Cars.31

Rating schedules design

23

The rating schedules, shown in Appendix D, were

completed by the sales managers for each of their salesmen
in each of the participating dealerships. These schedules

consisted of three paired comparison matrices which were

www— .——-..—.s-
. o

employed to determine the sales managers' opinions of their
salesmen with regard to their (1) personality, (2) product
knowledge, and (3) knowledge of sales techniques.

The paired comparison matrices were used because
they are a valid measuring device in cases when the subjects
to be compared can be judged only subjectively and when
the differences between the subjects to be compared are
relatively small.32 This methodology was more advantageous
than to have the sales managers simply rank their salesmen
in each of the above categories because the paired compari-
son system provides an internal consistency check of each

of the sales manager's ratings of his salesmen.

 

31Even though salesmen handling Lincoln-Continentals
only and used cars only were excluded from the study, those
car groups had to be included in this question because in
some dealerships the salesmen, who primarily sell the three
Mercury car lines, also sell some Lincoln-Continentals and
used cars.

32H. A. David, The Method gngaired Comparisons
(New York: Hafner Publishing Company, 1963), p. 9.

 

40

In order to minimize differences among sales man—
agers in interpretation of the three categories, each was
operationally defined for the purposes of the study as
follows:

Personality: appearance, sincerity, persuasiveness,
courtesy, friendliness, an understand-
ing of people, and sociability

Product Knowledge: knowledge of product features
and their respective consumer
benefits

Sales Techniques: obtaining prospective buyers,
qualifying customers, planning
and delivering sales presenta-
tions and product demonstrations,
handling objections, making trade
appraisals, closing the sale,
and completing the follow-up
after the sale.

{CI-"w” III-3’

The sales managers listed their salesmen in alpha-
betical order on each of the three paired comparisons, so
that the data collected from the schedules could be linked
with that compiled from the questionnaires. From the list
of salesmen provided by each sales manager, the researcher
could refer to internal company records to determine their
sales volume in terms of total units sold. These sales
volumes, which corresponded to those provided by the sales-
men on their individual questionnaires, provided the link
between the two data collecting instruments.

The instructions provided for the sales managers
on the rating schedules directed them to rate each salesman
as having more (M) or less (L) of the characteristic being

evaluated than each of the other salesmen included in the

fro:

Y

s‘ud‘
be: of

 

 

“c c. .: ... .s r 1n .

LI. 9 e r 1 .c . v. p . e 3 .C
C I e t c c.» T. v. e .0

Au .. i .C . 3. u : a .n 3. c» .1 .

u . r1. .3 .C u . a ”a ..a —v. a C .5

..u s . e a: . u e .. s r; .2 .. .

o a a .. . X .3 o a ~ ; t L» 1. . 1. . p

u' Tm .L n. pup .. o .3 Q... an ab «V ~ .

 

 

 

 

 

 

41

study from their dealership. Therefore, by adding the num-
ber of M's each salesman received in each category, the
participants could be ranked in relationship to one another.
(See Appendix E.)

The number of salesmen to be evaluated by the sales
managers varied according to the size of the dealership.

It is obvious that sales managers in the larger dealerships
would have more difficulty completing the paired comparison
matrices. Only full-time car salesmen in attendance at

the breakfast meetings were included in the rating sched-
ules. Therefore, the number of salesmen which the sales
managers had to evaluate were within an operative limit,
i.e., a maximum of ten or eleven men.

In order to provide a means of evaluating the reli-
ability of the paired comparisons as a data collecting in-
strument, the final portion of the rating schedules was
designed to have the sales managers judge the importance
of personality, knowledge of sales techniques, and knowledge
of the product in the personal selling of automobiles.

This was done by requesting that the sales managers dis-
tribute ten points among the three components to the degree
of importance they placed upon them in automobile sales.

Postpaid envelopes were provided so that the com—
pleted questionnaires and rating schedules could be for-
warded by the participating zone managers directly to the

researcher at the Lincoln—Mercury Division General Office

in Dearborn, Michigan.

42

Pretest g£_Instruments

 

A pretest of the 1967 Product Information Study
was conducted in seven Lincoln-Mercury dealerships of the
Detroit District in June, 1967. The Detroit District sam-
ple included dealerships located in both Michigan and Ohio.
Zone managers administered the questionnaires and rating 3
schedules to the new car sales staffs of the participating

dealerships during breakfast meetings.

 

The objectives of the pretest were to determine:

. The general feasibility of the study, more spe- F
cifically

. The intelligibility of instructions for its
administration

. The willingness and ability of salesmen and
sales managers to furnish the desired informa-
tion

. The reliability of the questionnaire and rating
schedules~

Examination of the data yielded from the pretest
showed that the study as conceived was feasible, that in-
structions were understood, and that both the questionnaire
and rating schedules were administerable. The questionnaire
provided a differentiation on product knowledge with scores
ranging from 9 to 27 (out of 30 maximum). The rating sched-
ules were completed accurately with very few inconsistencies.
There was no indication that the sales managers were using
the same criterion of comparison in completing the three
different rating schedules.

Of the seven participating dealerships, three

43

furnished complete information. The salesmen in the other
four dealerships did not supply sufficient information in
Section III of the questionnaire concerning their sales
performance. As a result, these dealerships were not ac—
ceptable for analysis. It was believed that the salesmen
were hesitant to divulge their performance record. There-
fore, two refinements were made before the national survey
was conducted. A clause was added to Section III of the
questionnaire informing the participants that the informa-
tion received from that section was for statistical purposes
only, and that the individual salesmen would not be iden-
tified. The problem of insufficient sales information was
reduced further by sending a sheet of instructions, included
in Appendix B, with the dealership package of questionnaires
and rating charts to the zone managers. They were asked to
check the questionnaires as they were collected to see that
all questions in Section III were completed.

An analysis of the scores on each question in the
questionnaire revealed that only one question (Section II,
Part II, question 11) was being misinterpreted. That ques-

tion was revised before the national study was instituted.

Sample Structure
The Lincoln-Mercury franchised dealer network con—
sists of approximately 2,500 dealers, of which 1,100 are
supervised by the Lincoln-Mercury Division. The other 1,400

dealers are classified as "dual dealers" merchandising both

44

Lincoln-Mercury Division and Ford Division car lines under
the supervision of the Ford Division. Even though the dual
dealerships represent a greater number of dealers, these
dealers occupy low potential market areas and contribute
a relatively small share to the total Division's volume.

The dealers supervised by the Lincoln-Mercury Di-
vision are generally located in high potential market areas,
i.e., the major metropolitan markets and large towns. It
is logical to expect that the dealerships located in highly
populated urban areas enjoy a relatively greater volume of
registrations per outlet as compared to the average industry
dealership.

Dealerships are normally categorized according to
the type of market served. Multi—point dealerships are
those found in metropolitan areas in which there is more
than one Lincoln-Mercury dealership to serve that geograph-
ical territory. These dealerships can be owned and operated
by one individual or as in most multi-point areas, they are
individually owned and operated. On the other hand, single
point dealerships are those located in all other areas in
which there is only one Lincoln-Mercury dealership.

In order to test the hypotheses (2A and 28) that
the relationship between a salesman's level of product know-
ledge and sales effectiveness is influenced by the type of
market and the size of the retail outlet in which he sells,

the dealerships were divided into appropriate homogeneous

J .D.!
F. . a «\v . R. ab
9 T . . D. : . t 3. MN Mu;
T. T .. .4 .5 1: .1 S C
a» «v a» ... .q QC ._.. x: «V
4... A. .1 .2 A» .3 3 Lu 1‘

 

 

 

 

 

 

45

groupings. The most suitable method of categorizing auto-
mobile dealerships for this analysis is as follows:

. Division of dealerships into multi-point and
single point groupings

. Further division of both the multi-point and
the single point dealerships into high and low
volume dealerships, according to their new car
sales planning volume.

The Lincoln—Mercury Division has approximately 300
supervised multi-point dealerships located in 51 metropol-
itan areas. Internal company studies have shown that these
dealerships, relative to the single point locations, have
certain characteristics. For example, they have:

. More intense competitive pressures on unit

grosses and a greater pressure on wage rates
and employee benefits

. Relatively higher total advertising budgets

. Greater difficulty retailing used cars

. More customer volatility and less chance of
building repeat business.

The single point dealerships, on the other hand,
generally serve a smaller market potential, usually one
community located some distance away from the nearest city.
It is generally agreed or thought that this type of dealer-
ship relies more upon reputation as "a good place to do
business." In order to maintain a loyal clientele for re-

peat business, for example, single point dealerships may

 

33The sales planning volume of individual dealer—

ships is determined by the automobile manufacturer annually
based upon past performance and anticipated national and
local market conditions.

46

emphasize such factors as service facilities, fair trade-in
practices, and participation in civic activities.

Since the selected dealerships were to be sub-divided
into four distinct classifications in order to test the
hypotheses concerning the effects of type and size of dealer-
ship, the sample size had to be large enough in each of the
four sub-groupings to obtain a representative cross section
of the population. Because the average number of new car
salesmen per dealership is five, a sample size of 150 dealer-
ships was selected to insure a sample of more than 100 sales—
men in each of the four dealership classifications.

The stratification of the 150 dealerships, which
were individually selected within each stratum by means of

a random number table, is shown in Table 2:

TABLE 2

STRATIFICATION OF SAMPLE BY TYPE
AND SIZE OF DEALERSHIPS

 

 

 

 

Population: 1,100 Dealerships
300 Multi—point 800 Single point
70 80
Selected Selected
High Low High Low
Volume Volume Volume Volume
35 35 53 27
Selected Selected Selected Selected

 

 

 

 

 

 

The criterion for the division of the sample of

.the 150 dealerships between multi—point and single point

47

dealerships was the percentage of new car sales volume con-
tributed to the Division total by each type of dealership.
The multi-point dealerships contribute 47 percent to Lincoln—
Mercury's new car sales volume, whereas the single point
dealerships account for 53 percent. Based upon this cri—
terion, 70 multi-point and 80 single point dealerships were
selected to participate in the study.

Both the multi-point and the single point dealer-
ships were sub-divided into high and low volume dealerships
according to their new car sales planning volume, which is
established by the manufacturer. The mean new car sales
planning volume for multi-point dealerships is 567 units.
Therefore, dealerships with a planning volume of 567 or
above were treated as high volume dealerships, and those
with less than 567 were considered low volume dealerships.

In order to obtain a representative sample of the
70 multi-point dealerships, 35 high volume and 35 low volume
dealerships were selected to participate in the study.

In the single point dealer classification, the mean
new car sales planning volume is 193 units. Ideally the
80 single point dealerships should have been equally sub-
divided into 40 high and 40 low volume dealerships. The
number of salesmen in single point dealerships which have
a planning volume of less than 193 units, however, is too
low (usually only one full-time and one or two part—time

salesmen) to effectively apply the paired comparison method

48

of analysis. Therefore, of necessity, greater emphasis
was placed upon high volume single point dealerships. Of
the 80 single point dealerships included in the study, 53

high volume and 27 low volume dealerships were selected.

Methods g; Data Analysis

 

The following outline explains how the information
received from the questionnaire, the rating schedule, and
internal company records was tabulated and recorded. The
recording of the data in preparation for analyses was com—
pleted by the researcher except for selected variables which
were recorded by the computer during the processing of the
data.

I. Questionnaire Information

(1) Salesmen's effectiveness:

A. Prgparation for Parametric Statistical
Tests
Sales volumes as received on each ques-
tionnaire were recorded along with the
computed corresponding profitability, as
coded entries in the tabulation records.

B. Preparation for Nonparametric Statistical
Tests
Sales volumes and the calculated profita—
bility were listed from the highest level
to the lowest and given a corresponding
rank; 1, 2, 3, etc. The ranks for each
of the salesmen were recorded by the com-
puter during the processing of the data.

 

(2) Salesmen's knowledge of the product, includ—
ing their ability to translate product en-
gineering advances into consumer benefits:

A. Preparation for Parametric Statistical
Tests
The scores for each salesman for each
question included in the questionnaire
were recorded.

(3)

(4)

(5)

49

B. Prgparation for Nonparametric Statistical
Tests
Salesmen's scores were listed from the
highest to the lowest, and given a corres-
ponding rank. The rank received by each
salesman was recorded by the computer
during the processing of the data.

 

Salesmen's knowledge of product features:

A. Preparation for Parametric Statistical
Tests
The scores for each salesman for each
question in Section I of the questionnaire
were recorded.

 

B. Preparation for Nonparametric Statistical
Tests
Salesmen's scores for Section I of the
questionnaire were listed from the high-
est to the lowest, and given a correspond-
ing rank. The rank received by each sales-
man was recorded by the computer during
the processing of the data.

Salesmen's ability to translate product en-
gineering advances into consumer benefits:

A. Preparation for Parametric Statistical
Tests
The scores for each salesman for each
question in Section II of the question-
naire were recorded.

B. Preparation for Nonparametric Statistical
Tests
Salesmen's scores for Section II of the
questionnaire were listed from the high-
est to the lowest, and given a correspond-
ing rank. The rank received by each sales-
man was recorded by the computer during
the processing of the data.

Salesmen's level of education, experience
as an automobile salesman, and attendance
at an FMI training session:

A. Preparation for Nonparametric Statistical
Tests
Each of these measurements for each sales-
man was recorded as outlined below:

‘u-

 

 

 

 

50

Level 23 Education Code

 

 

College graduate

Some college training
High school graduate

Some high school training
Grade school

HwaUl

Experience g§.gg
Automobile Salesman Code

Over ten years

Five to ten years
Three to five years
One to three years
Under one year

i—‘NUJJI‘U'!

Attendance gt g_Ford
Motor Institute Session Code

 

Less than three months
Three to twelve months
One to five years
Greater than five years
Never attended

D—‘NUJbU'I

II. Rating Schedule Information

(1) Preparation for Nonparametric Statistical Tests
Salesmen's knowledge of the product, salesmen's
personality, and salesmen's knowledge of sales
techniques.

Each schedule was analyzed to determine the rank-
ing of each salesman in relationship to other
rated salesmen in the dealership. These rankings
for each of the above variables were recorded

for each salesman.

(2) Preparation for Parametric Statistical Tests

The salesmen's rankings for each of the vari-
ables; product knowledge, personality, and sales
techniques, were transformed to standard scores
by means of a conversion table, shown in Appendix
F. The values found in the table were recorded

 

34An example of this method of data analysis is shown
in Appendix E.

51

for each salesman.35
III. Dealership Classification Information
The type and size of dealership which each salesman
represented, according to internal company records,
was recorded in the tabulation records.

IV. Components of Personal Selling Weighted Information
Each sales manager's assigned, numerical weights for
the three components of personal selling were recorded
in the tabulation records.

After the data from both the questionnaire and the
paired comparison matrices were tabulated, they were punched
onto data cards in preparation for analysis on the Michigan
State University 3600 Control Data Computer. The informa—
tion obtained from each individual respondent was recorded
on a separate data card. Sixty-six of the available eighty

columns were used to record the information for each par-

ticipant.3

 

35In order to conduct an inter-dealership analysis,
the salesmen's rankings for each of the variables; product
knowledge, personality, and sales techniques, had to be
combined. This was accomplished by assuming that the under-
lying variable represented by each of the ranks is normally
distributed. Then by means of the transformation table,
shown in Appendix F, each set of ranks was transformed into
a set of scores from a normal distribution in which 6': 30
and 14: 10. The resulting scores represent a continuous
scale that is normally distributed.

Professor Edwin E. Ghiselli has a discussion of
this method of transmuting scores of individual groups to
standard scores in Chapter Four of The Theory g£_Psyghologr
ical Measurement (New York: McGraw-Hill Book Co., 19647:
pp. 69-101. '

36A tabulation of the data cards is shown in Ap-
pendix G along with a decoding key.

 

52

Analysis g: £29.Qg£g

Data analysis was necessary in order to test the
central hypothesis of this thesis, which is that product
knowledge, in conjunction with two other assumed determin-
ants, personality and sales techniques, is a primary deter-
minant of sales effectiveness in the retail selling of
automobiles. Both parametric and nonparametric statistical
analyses were conducted on the data.37 The results of each
method of analysis were compared to determine if a signif—
icant difference existed.

The primary statistical technique employed in this
study is correlation analysis. It should be pointed out
that correlation analysis does not prove causation. It can
show that a theory is tenable, but it cannot prove it is

true. Correlation analysis can reveal significant relation-

ships between factors. These relationships, in turn, help

 

37Data assumptions constituting the various para-

metric statistical tests used are:
. The observations are independent.
. The observations are drawn from normally distrib-
uted populations.
. These populations have the same variance.

There are no specific conditions which must be met
concerning the parameters of the population from which the
sample was drawn in a nonparametric statistical model. It
is assumed, however, that the observations are independent
and that the variable under study has underlying continuity.

The above is from Sidney Siegal's Nonparametric
Statistics for the Behavioral Sciences (New York: McGraw-
Hill Book Co., Inc., 1956), p. 19 and p. 31.

 

53

the researcher in his search for the reasons or causes which

create the relationships.38

To test the central hypothesis
TIntra-dealership analysis)

The nonparametric analysis involved the computation
of both Kendall rank and Kendall partial correlation coef-
ficients of selected dependent and independent variables on
an intra-dealership basis.39 The correlation matrix for
this analysis is shown in Table 3.

The two dependent variables were measures of a sales-
man's sales performance; total units sold and profitability
of units sold. The independent variables were the two meas-
ures of a salesman's product knowledge obtained through
the questionnaire and paired comparison matrix, his person-
ality, and sales techniques, which were also obtained by
means of the paired comparison matrices completed by the
sales managers.

To test the central hypothesis using this method
of analysis, the dependent variables of sales performance
were correlated with the two measures of a salesman's prod-

uct knowledge, his personality, and sales techniques. Then

 

38Further discussion of the meaning of correlation
analysis is found in Richard C. Clelland et al., Basic
Statistics with Business _pplications (New York: John
Wiley & Sons, Inc., 1966), p. 464.

39The assumptions underlying the Kendall rank and
Kendall partial correlation coefficients are outlined in
Appendix H.

54

TABLE 3

KENDALL RANK/PARTIAL CORRELATION ANALYSIS:
SALES PERFORMANCE WITH PRODUCT KNOWLEDGE,
PERSONALITY, AND SALES TECHNIQUES

Independent Variables

Dependent Variables

 

Sales
Performance
(Total Units

Sold)

Sales
Performance:
(Profitability
of Units Sold)

 

Product Knowledge

 

Product Knowledge

 

Product Knowledge with
Personality Partialled
Out

 

Product Knowledge with
S. Tech. Partialled Out

 

Product Knowledge with
Personality Partialled
Out

 

Product Knowledge with
5. Tech. Partialled Out

 

Personality
(Paired Comparison)

 

Sales Techniques
(Paired Comparison)

 

 

 

a Product Knowledge as Measured by the Questionnaire

b Product Knowledge as Measured by the Paired Com—

parison

c Product Knowledge as Measured by the Questionnaire,
Personality and Sales Techniques as Measured by

the Paired Comparisons

d Product Knowledge,

Personality,

and Sales Tech-

niques as Measured by the Paired Comparisons

 

 

 

 

 

4" .

fif‘
by

ri~~
tnuof.

 

55

by means of the Kendall partial correlation analysis, the
degree of independence of the independent variable, product
knowledge, from the other independent variables was studied.
The Kendall rank and Kendall partial correlation
coefficients resulting from the intra-dealership analyses
for each relationship were totaled. Then the geometric
means of the coefficients were computed by dividing the
sum of the coefficients for each relationship by the total
number of dealerships included in the analysis.
Tests for significance were performed on the average
Kendall rank correlation coefficients to determine the rela—
tive importance of the three independent variables with the

two dependent variables.40

To test hypgthesis 25
TIntra-dealership analysis)

 

Hypothesis 2A states that the relationship between
a salesman's product knowledge and his sales performance
is independent of the type of market served by the dealer—
ship. To test this, the relationship between the dependent
variables, the two measures of sales effectiveness, and the
independent variables, product knowledge, as measured by

both the questionnaire and the paired comparison matrix,

 

40The assumptions underlying the Test of Signifi-
cance of Kendall Rank Correlation Coefficients are outlined
in Appendix I.

There is no test for significance for Kendall par-
tial rank correlation coefficients.

56

were summed according to multi-point--single point dealer-
ship classifications. These summed correlation coefficients
were divided by the total number of dealerships in each
classification. Tests for significance were calculated to
determine if there was a significant relationship between
the dependent and independent variables in multi-point deal-
erships and in single point dealerships. The correlation
matrix for this analysis is shown in Table 4.

Normal "z" tests were ‘ conducted to determine if
there is a statistically significant difference in the cor—

relation levels for these two types of dealerships.41

To test hypothesis 2B
TIntra—dealership afiglysis)

To investigate further the relationship between a
salesman's product knowledge and his sales effectiveness,
the effect of the size of the retail outlet within the
multi-point--single point dealership classification was
studied. This was accomplished by determining whether the
relationship between a salesman's product knowledge and his
sales effectiveness is influenced by the size of the re-

tail outlet in both multi-point and single point dealerships.

 

41The assumptions underlying the normal "2“ test
are outlined in Appendix J. Since one of the assumptions
for the normal "z" test is that the variables being eval-
uated must be approximately normally distributed, the Kol-
mogorov-Smirnov tests were conducted on the variable, prod-
uct knowledge as determined by the various dealership group-
ings. Another assumption of the normal "2" test is that
the populations have the same variance. Therefore, the F
test was performed to determine if the samples have approx-
imately the same variance.

 

 

 

 

   

57

TABLE 4

KENDALL RANK CORRELATION ANALYSIS: SALES
PERFORMANCE WITH PRODUCT KNOWLEDGE

Dependent Variables

 

Sales Performance:
Sales Performance: (Profitability of
Independent Variables (Total Units Sold) Units Sold)

 

Product Knowledge as
Measured by the
Questionnaire A B C D E F A B C D E F

 

 

Product Knowledge as
Measured by the

 

 

 

 

 

 

 

 

 

 

 

 

 

Paired Comparisons A B C D E F A B C D E F
Key:
A Multi-point Dealerships
B Multi-point High Volume Dealerships
C Multi-point Low Volume Dealerships
D Single Point Dealerships
E Single Point High Volume Dealerships
F Single Point Low Volume Dealerships

 

The Kendall rank correlation coefficients between sales
performance and product knowledge were summed according
to dealership size within the single point and multi-point
groupings. The summed correlation coefficients were divided
by the total number of dealerships in each classification
to obtain the average correlation coefficients for each
sample subgroup. Tests for significance were computed to
ascertain if there was a significant relationship between
the dependent variables, the two measures of sales perform—
ance, and the independent variables, product knowledge, as
measured by both data collecting instruments, in the high
volume dealerships and in the low volume dealerships for
both multi-point and single point groupings. The correla-
tion matrix for this analysis is shown in Table 4. Normal
"2" tests were conducted to determine if there is a sta-
tistically significant difference in the correlation levels
for these two sizes of dealerships within both the multi-
point and the single point dealership classifications.
To test the central hypothesis
TIHIEF:dealership analysis)

The parametric method of testing the central hypoth~
esis was conducted by means of multiple and partial corre—

42

lation coefficient analyses. The correlation matrix for

this analysis is shown in Table 5.

 

42The assumptions underlying the multiple and par~
tial correlation coefficients are outlined in Appendix K.

59

TABLE 5

MULTIPLE/PARTIAL CORRELATION ANALYSIS:
SALES PERFORMANCE WITH PRODUCT KNOWLEDGE,
PERSONALITY, AND SALES TECHNIQUES

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Dependent Variables
Sales - 'Sales
Performance: Performance:
(Total Units (Profitability
Sold) of Units Sold)
Independent Variables A B C D E F G A B C D E F G
Product Knowledge &
Sales Techniques &
Personality
Product Knowledge with
Sales Techniques & Per-
sonality Partialled Out
Personality with Product
Knowledge & Sales Tech-
niques Partialled Out
Sales Techniques with
Product Knowledge &
Personality Partialled
\Out
Product Knowledge & y
Sales Techniques &
Personality
Product Knowledge with
Sales Techniques & Per-
sonality Partialled Out
Personality with Product
Knowledge & Sales Tech-
niques Partialled Out
Sales Techniques with
Product Knowledge &
Personality Partialled
{Out
Key:

a Product Knowledge A All Dealerships
as Measured by the B Multi-point Dealerships
Questionnaire C Multi-point High Volume Dealer-

ships ‘

b Product Knowledge D Multi-point Low Volume Dealerships
as Measured by the E Single Point Dealerships
Paired Comparison F Single Point High Volume Dealer-

ships
G Single Point Low Volume Dealerships

CT.

.3 . p . . I I q l I . . ‘.d I: ‘N h.
2. T a. I C. e x .l I .l .D V m... C .l a. R. to W.

D. m i m

 

 

60

Multiple correlation coefficients were computed
on an inter-dealership basis between the dependent variables,
sales performance as measured by both total units sold and
profitability of units sold, and the independent variables,
product knowledge as measured by the questionnaire and the
paired comparison matrix, personality, and sales techniques.
Then partial correlation coefficients were calculated be-
tween the dependent variables, the two measures of sales
effectiveness, and one of the independent variables with
the other two independent variables partialled out as fol—
lows:

. Product knowledge, as measured by both the ques-

tionnaire and the paired comparison matrix, with

personality and sales techniques partialled out

. Personality with product knowledge and sales tech-
niques partialled out

. Sales techniques with personality and product
knowledge partialled out

The F ratio and the level of significance were com-
puted for each of the multiple and partial correlation co-

efficients to determine the degree of significance.43

 

43The assumptions underlying the F test are outlined
in Appendix J. Since one of the assumptions for the F test
requires that the variables under investigation are approx-
imately normally distributed, the variables required must
be analyzed to determine their degree of normality. The
variables of sales performance and product knowledge as
measured by the questionnaire were analyzed by means
of the Kolmogorov—Smirnov test LU determine if they approx-
imate normal distributions. The Kolmogorov-Smirnov tests
are based upon Technical Report 41.02 entitled, "Nonpara-
metric Measures of Randomness and Goodness of Fit Kolmogorov-
Smirnov and Runs Tests" prepared by the Michigan State Uni-
versity Computer Institute for Social Science Research. Since

 

ables ial

It also
in relat
as measu

sold.

r—3
0
r+
(D
(n
n

r—o
9"
(t
(D
'1

l
()._

61

This method of partialling out the independent vari—

ables indicates the degree of interdependence of the vari-

ables; product knowledge, personality, and sales techniques.

It also indicates the relative importance of these variables

in relation to the dependent variables, sales effectiveness

as measured by total units sold and profitability of units

sold. a

To test hypotheses 25 Egg gs

Inter-dealership analysis)

To test the hypotheses concerning the effect of type a

(2A) and size (2B) of dealership classifications upon the

relationship between a salesman's product knowledge and his

sales performance, the above parametric method of analysis

was repeated by grouping the dealerships into the categories

of:

Multi-point

Multi-point high volume

Multi-point low volume

Single point

Single point high volume
Single point low volume.

The correlation matrix for this analysis is shown in Table 5.

gg_test hypothesis 35
In order to investigate the hypothesis that a sales-

man's level of product knowledge is influenced by the type

the variables product knowledge, as measured by the paired
comparison matrices, personality, and sales techniques,
were transformed to standard scores based on normal distri-
butions, it was not necessary to determine their degree

of normality.

‘5

 

‘Ii ' 'n

 

 

C)

*1
C)

62

of market served by the dealership (3A), the mean and the
standard deviation of the scores received in the multi-point
dealerships were compared to those received in the single

point dealerships by means of the normal "z" test.

Eg.test hypothesis 33

The dealerships were subdivided into high and low a
volume multi-point dealerships and high and low volume single 1
point dealerships in order to analyze the hypothesis that
a salesman's level of product knowledge is influenced by a

the size of the retail outlet for both of these dealership
classifications (3B). The mean and the standard deviation
of the scores received on the questionnaire were computed
for each of the dealership groupings. Then normal "z" tests
were calculated to determine if there was a significant
difference in the level of product knowledge in high volume
versus low volume multi-point dealerships, and high volume

versus low volume single point dealerships.

22 test hypotheses 39, 32, Egg 33

A salesman's level of product knowledge is influ-
enced by: his level of education, attendance at a manufac-
turer-sponsored training course, and his sales experience
within the industry. These hypotheses were tested by
Kendall rank correlation coefficients between the depend—
ent variables; product knowledge, as measured by the ques-
tionnaire and the paired comparison matrix, and the inde-

Pendent variables; level of education, attendance at a

 

 

 

 

63

manufacturer-sponsored training course (FMI), and sales
experience within the automobile industry. These relation-
ships are shown in Table 6.

The correlation coefficients were first computed
on an intra-dealership basis, then summed and divided by
the average correlation coefficient values. Tests for sig-
nificance were calculated among each of the relationships
of the dependent variable; product knowledge, with each of
the independent variables; education, experience, and at-
tendance at FMI. Therefore, the relative importance of

these independent variables was determined.

TABLE 6
KENDALL RANK CORRELATION ANALYSIS:

PRODUCT KNOWLEDGE WITH EDUCATION,
FMI, AND EXPERIENCE

Dependent Variables

 

Product Knowledge Product Knowledge
Independent Variables (Questionnaire) (Paired Comparison)

 

Level of Education

 

Attendance at
FMI Course

 

 

Experience in the
Automobile Industry

 

 

 

 

w : —. MAI—- -L-“h-JIW
' d
’-

64

1 test hypothesis 4.

 

Another hypothesis to be tested by the nonpara-

ztric method of analysis is that a salesman's knowledge

= product features and his knowledge of the consumer bene-
_ts resulting from the features contribute equally to his
11es effectiveness (4). The salesman's knowledge of prod-
:t features as measured in Section I of the questionnaire
1d his knowledge of the corresponding consumer benefits,

5 measured in Section II, were both correlated with the

vo measures of sales effectiveness; total units sold, and
:ofitability of units sold. The correlation matrix for

113 analysis is shown in Table 7.

TABLE 7

KENDALL RANK CORRELATION ANALYSIS:
SALES PERFORMANCE WITH PRODUCT KNOWLEDGE

Dependent Variables

 

Sales Performance:
Sales Performance: (Profitability of
idependent Variables (Total Units Sold) Units Sold)

I;

 

‘oduct Knowledge
Questionnaire
Section I

~

 

‘oduct Knowledge
Iuestionnaire
Section II

 

 

 

 

This analysis was conducted on an individual dealer-
.ip basis and then the resulting averages for the total

imple were calculated. Tests for significance were conducted

 

65

to determine if a significant relationship exists between
the salesman's product knowledge as measured in Section I
of the questionnaire, his sales performance and his know-
ledge of consumer benefits as measured in Section II of
the questionnaire, and his sales effectiveness. Normal
"2" tests were conducted to determine if there is a
statistically significant difference in the correlation
levels in these two divisions of product knowledge in re-

lationship with sales performance.

“—_ —‘_.._._. —_——E

Treatment 2:.Egg_gg£g

The Kendall rank and Kendall partial rank correla-
tion coefficients, as outlined in Tables 3, 4, 6, and 7,
were calculated on an intra—dealership basis. The calcu—
lations were based upon a computer program described in
Technical Report 47 entitled "Rank Correlation Coefficients,"
prepared by the Michigan State University Computer Institute
for Social Science Research.

Computation of the multiple and partial correlation
coefficients between the dependent variable and the inde-
pendent variables (Table 5) was based upon a computer pro—
gram described in Statistical Description No. 7, "Calcula—
tion of Least Squares (Regression) Problems on the LS Rou-
tine," prepared by Michigan State University Agricultural

Experiment Station.

CHAPTER IV
ANALYSIS OF DATA COLLECTING INSTRUMENTS

The questionnaires and paired comparison matrices,
including introductory letters to the General Field Managers,
were mailed July 27, 1967. The first returns were received
the second week of August. All but four districts were
completed by August 25. On that date, follow-up telegrams
were sent to the remaining four districts. By September 11,
1967, 142 of the original 150 dealerships selected had par-
ticipated in the study.

Eight dealerships did not respond and three more
returns were not usable due to insufficient information.

An additional ten dealerships were removed from the analyses
because it was impossible to accurately link their question—
naires to the corresponding paired comparisons. The linkage
could not be performed because of discrepancies between

the salesmen's sales performance figures shown on the ques-
tionnaires and those recorded in company records, which
contain monthly sales records of the salesmen employed in
all Lincoln-Mercury supervised dealerships. The resulting
sample size totaled 129 dealerships, or an 86 percent usable
return.

The number of salesmen who participated in the study

66

67

was 708. Some questionnaires, however, were not incorporated
into the study results for the following reasons:

. The salesman's length of selling experience at
the dealership was less than six months.

. The question concerning the salesman's sales per~
formance in Section III of the questionnaire was
incomplete.

. The salesman submitting the questionnaire did not
meet the criterion established for a Mercury new
car salesman.

. The salesman completing the questionnaire was
not evaluated in the sales manager's paired com-
parisons.

. The salesman's sales performance was not recorded
on internal company records.

The results of this study were based on the 524
remaining salesmen. Careful examination of the question-
naires received from these salesmen revealed that they were
completed fully and there was no indication of any use of
reference materials or collaboration among the participants.

One explanation of the salesmen's cooperation is the
effectiveness of the methodology of employing breakfast
meetings whereby the study was conducted by company repre—
sentatives. Surroundings were appropriate for completion
of the questionnaire and void of any interruptions. It ap-
pears that the use of this methodology to collect the data
enhanced the salesmen's willingness to participate, thereby
increasing the trustworthiness of their responses. Thus,

a more accurate measure of the salesmen's product knowledge

was obtained.

68

Questionnaire Section 33.- Open-Ended Regponses

The open-ended portion of the questionnaire concern-
ig the consumer benefits resulting from selected product
aatures assisted in the determination of the salesmen's
:tual rather than rote product knowledge by permitting
iem to express in their own words the consumer benefits
iey felt appropriate to certain features available in 1967
incoln-Mercury products. Unlike the multiple choice ques-
ions, where alternative answers were provided, the open-
ided section required recall of product knowledge rather
nan recognition by the salesmen.

Respondents' answers to the open-ended section were
idged incorrect if they were blank, too vague, contained
ncorrect information, or if they were a reiteration of
he illustrative benefit provided in the question.

Examination of the open-ended responses revealed
wide variety of benefits associated with each of the
elected features. In fact many salesmen were able to
ist several consumer advantages for some product features.
a a national scale, however, there were no product features
nich appeared to be either extremely difficult or ex+remely
asy for the salesmen to discuss.

Analysis of this section of the questionnaire re-
ealed that differences in the salesmen's level of product
nowledge could be accurately determined by grading the
pen-ended questions as correct or incorrect without giv-

ng additional credit to those salesmen who listed more

 

69

44
than one correct response.

Product Knowledge - Measured by the Questionnaire

 

In order to determine if the variable product know-
ledge, as measured by the questionnaire, meets the conditions
necessary for the parametric analysis (multiple/partial
correlation analysis and F tests) employed in this study,
the frequency distribution of the 524 questionnaire scores
were plotted as shown in Figure 2. The mean score received

on the questionnaire was 21.2 (71%) out of a possible 30

“II—‘7‘“ “‘k ’2‘“

points, with a standard deviation of 4.5.

The parametric method of analysis used in this study
requires that the observations must be drawn from normally
distributed populations. The Kolmogorov-Smirnov test was
conducted on the distribution shown in Figure 2. It was
found that the variable, product knowledge, as measured by
the questionnaire, does not meet the requirement in that
it does not approximate a normal distribution curve at a
level of significance of .05.45

Since product knowledge as measured by the ques—

tionnaire was not shown to be normally distributed, it was

 

. 44An interesting facet of the open-ended portion
of the questionnaire is that some respondents elected to
elaborate concerning the advantages and/or disadvantages
they had encountered in selling one of the features. These
detailed unsolicited responses indicate the personal in-
volvement experienced by some of the salesmen participating
in the study.

45D = .11; D less than .06 is necessary to show
that the variable is normally distributed at the .05 «30
level of significance and with an n of 524.

70

FIGURE 2

FREQUENCY DISTRIBUTION OF
QUESTIONNAIRE SCORES

Number of scores
within each
increment

6
0. 1

42 T

36

3O

24 m

12

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

,ri flnnflfl“
’5 10 15 20 25

Questionnaire Scores

0

tn: 524
Mean Score: 21.2
Standard Deviation: 4.53

 

 

 

 

 

 

71

necessary to transform the abscissa value (questionnaire
score) so that an appropriate normal distribution is ob-
tained. The transformation, logarithm to the base 10 of
the original scores, was used.46 Such a transformation
does not alter the order of the individuals in terms of
their questionnaire scores, but it merely changes the mag-
nitude of the differences among them.

The transformation of original scores to other
values does not insure that the new distribution will be
normal. Therefore, it was necessary to conduct the Kolmo-
gorov-Smirnov test to determine if normality had been
achieved. The test indicated that the transformed dis-
tribution of product knowledge, as measured by the ques-
tionnaire, does approximate a normal distribution at a .05

level of significance.47

Two Measurements 23 Product Knowledge Compared

 

The results of the sales managers' paired compari-
sons, when related to the salesmen's questionnaire scores,
produced an average Kendall rank correlation coefficient
of .26. A test for significance indicated that there was
not a statistically significant relationship between the

questionnaire scores and the sales managers' paired

 

46The transformations were based upon a computer
program described in Statistical Description No. 19, "Data
Transformations," prepared by the Michigan State University
Agricultural Experiment Station.

47D = .05; D less than .06 is necessary to show
that the variable with an n of 524 is normally distributed
at the .05 (00 level of significance.

72

comparisons of product knowledge.48

In order to determine the level of significance
in the nonparametric statistical analyses, a relatively
small n value was unavoidable. This is because in non—
parametric statistics there is no method of grouping rank
data into a continuous scale. Inasmuch as there were 524

participants representing 129 dealerships, the average

number of salesmen participating per dealership was four.

Working with an n of four, therefore, the average Kendall

w_._._.____~_.1__ .

rank correlation coefficient between the questionnaire
scores and the sales managers' paired comparisons would
have had to be .85 or larger in order to show a statistic-
ally significant relationship between the two measures of
product knowledge.

By means of parametric statistics, whereby the
dealerships were grouped by use of standardized scores,
a Pearsonian correlation coefficient between the question~
naire scores and the sales managers' paired comparisons

of .33 was calculated.49 The F statistic was computed to

 

48To determine if the average Kendall rank corre-
lation coefficient is significant, it is necessary to deter—
mine the significance of an assumed representative sample
which has the same correlation coefficient and an n value
equal to the average n for all dealerships. Then by deter-
mining if the average Kendall correlation coefficient in
the representative sample is or is not significant, it can
be inferred that the average Kendall correlation coefficient
of the total sample is or is not significant. p = 44; .05
K”) or less is necessary to show that the two variables
are significantly related. (See Appendix L for an illus—
tration of this test.)

49Since the standardized scores are derived from

73

determine if there was a statistically significant relation-
ship between these two variables.50 A statistically sig-
nificant relationship between product knowledge, as meas-
ured by the questionnaire, and product knowledge, as deter-
mined by the sales managers' paired comparisons, was found.

Even though the parametric statistical analysis
revealed that there was a statistically significant rela-
tionship between the two product knowledge measurements,
the correlation was low.

There are several factors which explain the rela-
tively low correlation between the two instruments used to
measure the salesmen's product knowledge. The paired com—
parisons were analyzed and recorded by counting the number
of times one salesman was rated as having more product know-
ledge than another. Therefore, in some dealerships where
in the sales manager's opinion the levels of the salesmen's
product knowledge were very similar, one or even two sales—
men could receive the same overall rating as another. The
paired comparison matrix provided no way to differentiate
the product knowledge between or among these men.

A similar situation occurred when the sales manager

 

an assumed normal distribution, the condition of normality
of the variable, product knowledge as measured by the paired
comparisons, is achieved.

50F = 63.8; 3.86 (F) or above is necessary to show
a significant relationship with l and 522 degrees of free~
dom and a 5% level of significance. (See Appendix M for
an illustration of this test.) .

74

felt the level of product knowledge held by two salesmen
was the same, or so close that he could not differentiate
which was higher. In both cases, the resulting ties had
to be broken.51 The questionnaire, on the other hand, pro-
vided individual scores which formed the basis for definite
comparisons of the salesmen's levels of product knowledge
as there were very few identical questionnaire scores within
dealerships.

It is important to remember that the questionnaire
was designed to measure product knowledge of the 1967
Lincoln—Mercury product line, whereas the sales managers
made the paired comparisons using the salesmen's total prod-
uct knowledge. Even though the product information tested
was applicable to the 1967 models, it was not so restrictive
as to limit its application only to that year. A manager
making the paired comparisons would have a tendency to rate
his men with predominant consideration for their recent
demonstration of product knowledge as opposed to that of
a former time.

Influence g£_experience upon product
knowledge measurements

 

One might postulate that some sales managers when

making the comparisons of their salesmen according to

 

51This was done by assigning each of them the aver-
age of the ranks which would have been assigned had no ties
occurred. For example, if two salesmen received the rank
of two, then by this method of breaking ties, each would
receive a rank value of 2.5.

75

product knowledge, might equate experience as an automobile
salesman with product knowledge. When the relationship
between the sales managers' rankings of product knowledge
and the salesmen's experience as automobile salesmen was
examined, however, the average Kendall rank correlation
coefficient was only .07. This would indicate that there
was no significant relationship between experience and the
sales managers' product knowledge paired comparisons.52
It was also found that there was no correlation between

the salesmen's experience and their product knowledge, as
measured by the questionnaire.53 Therefore, it was not

the influence of the salesmen's experience upon the sales
managers' rankings that caused the lack of agreement between
the two measures of product knowledge.

Influence g£_sales technigues gpon
product knowledge measurements

 

It might also be conjectured that the sales managers,
when making the product knowledge comparisons. were unable
to completely differentiate their salesmen's knowledge of
product features and their corresponding consumer benefits
from the sales techniques wherein this product knowledge
is employed. In order to test this possibility, a corre-

lation was computed between the sales managers' product

 

52p = .52; .05 «30 or less is necessary to show
that the two variables are significantly related.

53r = .00; p = .625; .05 (09 or less is necessary
to show that the two variables are significantly related.

 

 

 

76

knowledge paired comparisons and their rankings of sales
techniques. The average Kendall rank correlation coeffici—
ent of .47 indicates that there is no statistically signif—
icant relationship between the sales managers' ratings of
their salesmen's product knowledge and their rankings of
sales techniques.54

Using the parametric method, a Pearsonian correla-
tion of .59 was found between the sales managers' product
knowledge paired comparisons and their rankings of sales
techniques.55 To determine if there was a statistically
significant relationship between these two variables, the
F statistic was computed. It found there was a significant
relationship between the sales managers' paired comparisons
of their salesmen's product knowledge and those of sales
techniques.56 The coefficient of determination (r2) was
found to be 35 percent, indicating that 35 percent of the
unit variance of the sales managers' product knowledge paired
comparisons is explained by the variable sales techniques.

The lack of agreement between the two methods of

data analysis is caused by the low average Kendall rank

 

54p = .25; .05 (09 or less is necessary to show
that the two variables are significantly related.
55

The variable sales techniques for the parametric
analysis was obtained by transforming the rank scores into
a set of scores from a normal distribution, thereby achiev—
ing the condition of normality.

56F = 278; 3.86 (F) or above is necessary to show
a significant relationship with l and 522 degrees of free-
dom and a 5% level of significance.

 

 

 

   

77

correlation coefficient and the unavoidable small value
of n.

If the results of the parametric analyses are ac-
cepted, we must conclude that the sales managers were not
able to keep the two factors independent of one another.
However, there is also a statistically significant corre-
lation between product knowledge as measured by the ques-
tionnaire and the sales managers' paired comparison ratings
of sales techniques (r = .24).57

The observation that the sales managers were not
able to keep product knowledge and sales techniques as
mutually exclusive factors is explicable when the inter-
action between knowledge of the product and the employment
of this knowledge through the usage of sales techniques
is examined. In some instances the two are very closely
related. For example, product knowledge forms the basis
by which salesmen are able to handle objections raised by
prospective buyers during product presentations, but the
actual method by which salesmen handle these objections
are considered sales techniques.

A second situation in which product knowledge and
sales techniques are related concerns customer appraisal
and the tailoring of the sales presentation to effectively

meet consumer needs. The ability of salesmen to adequately

 

57F = 31.81; 3.86 or above is necessary to show a
significant relationship with l and 522 degrees of freedom
and a 5% level of significance.

 

 

 

 

|I

78

qualify potential buyers and their desires is defined as
a component of sales techniques. However, the employment
of product knowledge is where the salesmen point out the
features and benefits of their product which will be of
most interest to individual customers.

In fact, the automobile salesman's whole product
presentation is based upon the interaction of product know—
ledge and sales techniques. The salesman must understand
the features of his product in order to make successful
presentations and demonstrations, but again, the methods

he employed comprise sales techniques.

Reliability 2£_the Rating Schedules

 

 

In order to provide a means of evaluating the reli—
ability of the rating schedules as a data collecting instru-
ment, each sales manager participating in the study was
asked to judge the relative importance of personality, prod-
uct knowledge, and sales techniques, as determinants of
success in automobile selling. On a ten point scale, the
mean values of the three determinants resulting from the
sample of 129 sales managers were as follows: personality
3.7, product knowledge 2.6, and sales techniques 3.7.

Table 8 shows the distribution pattern for each of the
three determinants.

To determine if the sales managers were differen-
tiating among these three variables, a chi-square test for

independence was conducted. The calculated value for‘X?

 

 

 

79

was 79.78. Since the frequency table (Table 8) is three

by seven, there are twelve degrees of freedom. The observed
value ofixg is significant beyond the .001 level. Inasmuch
as p <:.OOl is less than .05 (09, it can be concluded that
the sales managers made statistically significant differ-
entiations among personality, product knowledge, and sales

techniques.

TABLE 8
SALES MANAGER'S COMPARISONS OF THE RELATIVE
IMPORTANCE OF PERSONALITY, PRODUCT
KNOWLEDGE, AND SALES TECHNIQUES

Number 2: Sales Managers

weights 2;; a a ‘ 2 .5. .6. 1:9. fleas
Personality 2 14 46 33 24 6 4 3.7
Product Knowledge 6 S7 56 7 2 l O 2.6
Sales Techniques 3 14 36 45 26 4 l 3.7

All three variables; personality, product knowledge,
and sales techniques, were considered by the sales managers
significantly important as determinants of success in the
personal selling of automobiles. Product knowledge, how-
ever, was found to be statistically less important to sales
success, according to the sales managers' weightings, than
either personality or sales techniques. (These findings
were derived by means of X? tests for statistical differ-
ences in means.)

The weightings were relatively consistent in that

80

88 percent of the sales managers weighted product knowledge
at either two or three out of a possible ten points; while
83 percent of the participants weighted personality and
sales techniques either three, four, or five. The differ—
ence of assigned values was greatest for personality and
sales techniques which ranged from one to nine and zero

to seven, respectively. Product knowledge received a low
value of one and a high of six.

Since all the salesmen were grouped together for
the inter-dealership analysis, it was important that the
salesmen be evaluated by their sales managers on a rela—
tively comparable basis. The consistency shown in the
sales managers' weightings of the relative importance of
the three variables employed in the paired comparison

matrices revealed that this condition existed.

Sales Performance Measurements

The two measures of sales performance, total units
sold and profitability of units sold were examined to de-
termine if the various levels of performance obtained by
the salesmen approximated normal distributions. This con—
dition is necessary if the variables are to be analyzed by
the multiple/partial correlation method of data analysis.
The frequency distributions of the salesmen's performance
as determined by the two measurement criteria were plotted,
as shown in Figure 3. It was found by means of the Kolmo-

gorov-Smirnov test that both of the sales performance

 

 

81

FIGURE 3
Number of Sales-
men within each FREQUENCY DISTRIBUTION OF
Increment SALES PERFORMANCE CRITERIA:
86 A TOTAL UNITS SOLD

70

54

48

32

16

86

70

54

48

32

16

n: 524
_ Mean: 55.5
Standard Deviation: 38.21

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.. HHHflnnmmmmn .

Number of Units Sold 252

 

l PROFITABILITY OF UNITS SOLD

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

W

n: 524

Mean: 103.9

Standard Deviation: 65.96

-F“
1F"—
FF-
,fl Hnnflnnnrn H.

PROFITABILITY OF UNITS SOLD 378

82

measurements did not approximate normal distributions at
the .05 level of significance.58 In order to transmute
the sales performance levels to other values, which are
distributed in accordance with a normal frequency distri-
bution, the log to the base 10 was taken on both measures
of sales performance. The resulting distributions were
found to approximate normal distributions at the .05 level
of significance, which qualified the transformed data to

be analyzed by means of parametric statistical analysis.59

 

58Total units sold, D = .13; profitability of units
sold, D = .12; D less than .06 is necessary to show that
the variable with an n of 524 is normally distributed at
the .05 (a) level of significance.

59Tota1 units sold, D = .05; profitability of units
sold, D = .05; D less than .06 is necessary to show that
the variable with an n of 524 is normally distributed at
the .05 (00 level of significance.

CHAPTER V

STUDY RESULTS

The purpose of this chapter is to present the re-
sults of testing the hypotheses. The findings of both the
inter- and the intra-dealership methods of data analyses

are discussed for each hypothesis under investigation.

 

Product Knowled e — Sales Performance
Relationship (Intra-Dealership
Analysis)

 

The hypothesis that product knowledge in conjunc-
tion with two other assumed determinants, personality and
sales techniques, is one of the primary determinants of
sales effectiveness was tested by correlating the dependent
variables (sales performance as measured by total units
sold and profitability of units sold) with each of the in-
dependent variables (product knowledge, personality, and
sales techniques).

The average Kendall rank correlation coefficient
between the questionnaire scores on product knowledge and
sales performance, when measured by total units sold, was
.12 as shown in Table 9. When this same measure of prod-
uct knowledge was related to sales performance, as measured
by profitability of units sold, the average Kendall rank

correlation was .22. Neither of these correlation

83

84

TABLE 9

KENDALL RANK/PARTIAL CORRELATION ANALYSIS:
SALES PERFORMANCE WITH PRODUCT KNOWLEDGE,
PERSONALITY, AND SALES TECHNIQUES
(ALL DEALERSHIPS)
(Based on an Average n of 4)

 

 

 

 

 

 

 

 

 

 

Dependent Variables
Sales Sales
Performance: Performance:
(Total Units (Profitability
Independent Variables Sold) of Units Sold)
a{Product Knowledge .12 .22
b{Product Knowledge .22 .36
fProduct Knowledge with
Personality Partialled
}0ut ‘ .12 .16
d
Product Knowledge with
is. Tech. Partialled Out .08 .13
Product Knowledge with
Personality Partialled
Out .03 .11
d
Product Knowledge with
S. Tech. Partialled Out .12 .20
Personality
(Paired Comparison) .22 .22
Sales Techniques
(Paired Comparison) .36 .36

 

 

 

 

Key:

a Product Knowledge as Measured by the Questionnaire

b Product Knowledge as Measured by the Paired Com-
parison

c Product Knowledge as Measured by the Questionnaire,
Personality and Sales Techniques as Measured by
the Paired Comparisons

d Product Knowledge, Personality, and Sales Techniques
as Measured by the Paired Comparisons

85

coefficients is high enough to be statistically significant.60

The difference between these two correlation levels
can be explained, however. By measuring sales performance
in terms of total units sold, equal weight is placed upon
new and used cars. Since the questionnaire was based on
new car product knowledge, the correlation between product
knowledge and sales performance as measured by profitability
of units sold more adequately describes the true relation-
ship. To substantiate this conclusion, an average Kendall
rank correlation between the salesmen's product knowledge
and total new cars sold was computed and found to be .21.

Of the two instruments used to measure the partici-
pants' product knowledge, the second measurement, the sales
managers' paired comparisons, was related to the dependent
variables (sales performance as defined by total units sold
and profitability of units sold).

The average Kendall rank correlation coefficient
between the salesmen's product knowledge, as evaluated by
the sales managers, and their sales performance, as meas-
ured by total units sold and profitability of units sold,
were .22 and .36 respectively. The tests for significance
showed that there is no statistically significant relation-

ship between product knowledge, as measured by the paired

 

6OTotal units sold, p = .53; profitability of units

sold, p = .46; .05 «y; or less is necessary to show that
the two variables are significantly related.

86

. 61
comparisons, and either measurement of sales performance.

Inasmuch as no relationship between product know-
ledge and sales performance was established by the nonpara—
metric method of data analysis, two other assumed determin-
ants of success in selling were studied. The influence
of personality and sales techniques upon the salesmen's
performance were the next two determinants to be studied.

Personality and sales techniques, as measured by
the sales managers' paired comparisons, were compared with
the salesmen's sales performance, as determined by both
total units sold and profitability of units sold. The re-
sulting average Kendall rank correlation coefficients were
.22 for personality and .36 for sales techniques, as shown
in Table 9. Tests for significance showed that no statis-
tically significant relationship exists either between per-
sonality or sales techniques with sales performance.62

The Kendall partial rank correlation analysis be-
tween the two measures of sales performance (total units

sold and profitability of units sold) and the independent

variables (personality, product knowledge, and sales

 

61Total units sold, p = .46; profitability of units
sold, p = .36; .05 «1) or less is necessary to show that
the two variables are significantly related.

62Personality-total units sold, p = .46; profita-
bility of units sold, p = .46; .05 (09 or less is necessary
to show that the two variables are significantly related.

Sales techniques—total units sold, p = .36; profit—
ability of units sold, p = .36; .05 (00 or less is neces-
sary to show that the two variables are significantly re-
lated.

87

techniques), as presented in Table 9, also appears to show
no significant relationship between any of the independent
variables and sales performance with one of the independent
variables partialled out.63

Based upon the intra-dealership method of data analysis,
there is very little evidence to support the hypothesis that
product knowledge, in conjunction with personality and sales
techniques, is one of the primary determinants of sales
effectiveness, as measured by total units sold and profit-
ability of units sold.

ngguct Knowled e - Sales Performance
Relationship Inter-Dealership
Analysis)

The parametric method of data analysis, which con-
sisted of multiple and partial correlation coefficient an-
alyses, was employed to investigate further the central
hypothesis of this thesis.

The parametric approach to analyzing the relative
importance of product knowledge with sales performance was
undertaken by first considering the relationship between
sales effectiveness and the three independent variables;
personality, product knowledge, and sales techniques, which

reflects the maximum contribution of these variables.64

 

63Unfortunately, there is no test for significance
for the Kendall partial rank correlation coefficients.

64As noted in footnote 35, the product knowledge,
as measured by paired comparison, and sales techniques

88

Then in order to determine the magnitude of the influence
of personality and sales techniques, the relationship be-
tween product knowledge and sales performance was analyzed
with both personality and sales techniques partialled out.

The multiple correlation coefficients between the
independent variables; product knowledge, as measured by
the questionnaire, personality, and sales techniques, and
the dependent variables; sales performance as measured by
total units sold, and profitability of units sold, were
.41 and .45 respectively, as shown in Table 10. The F tests
indicate that sales performance is statistically related
to product knowledge, personality, and sales techniques.65
The combined influence of these three factors, however,
explains only about 18 percent of the unit variance of the
salesmen's sales performance levels.

When sales techniques and personality were partialled
out of the multiple regression equation, the partial corre-
lation coefficients between product knowledge, as measured
by the questionnaire, and total units sold was .34 and profit-
ability of units sold was .37. The F tests revealed that

the relationships between sales performance and product

 

variables are assumed to be normally distributed. The vari-
able personality for parametric analysis was obtained by
transforming a set of rank scores into a set of scores from
a normal distribution, thereby achieving the parametric
condition of normality.

6STotal units sold, F = 35.74; profitability of
units sold, F = 44.44; 2.62 (F) or above is necessary to
show a significant relationship with 3 and 520 degrees of
freedom and a 5% level of significance.

89

TABLE 10

MULTIPLE/PARTIAL CORRELATION ANALYSIS:
SALES PERFORMANCE WITH PRODUCT KNOWLEDGE,
PERSONALITY, AND SALES TECHNIQUES
(ALL DEALERSHIPS)

 

 

 

 

 

 

 

bf

 

 

 

 

 

(11:524)
Dependent Variables
Sales Sales
Performance: Performance:
(Total Units (Profitability
Independent Variables Sold) of Units Sold)
Product Knowledge &
Sales Techniques &
Personality .41‘ .45‘
Product Knowledge with
-Sales Techniques & Per-
sonality Partialled Out .34‘ .37’
Personality with Product
Knowledge & Sales Tech-
niques Partialled Out —.01 —.01
Sales Techniques with
Product Knowledge & Per-
sonality Partialled Out .15‘ .17‘
{Product Knowledge &
Sales Techniques &
Personality .28‘ .33‘
Product Knowledge with
Sales Techniques & Per-
sonality Partialled Out .13‘ .20‘
Personality with Product
Knowledge & Sales Tech—
niques Partialled Out .03 .01
Sales Techniques with
Product Knowledge & Per-
\sonality Partialled Out .10' .09'

 

 

a Product Knowledge as Measured by the Questionnaire

b Product Knowledge as Measured by the Paired Com-

parison

' Values Statistically Significant at CC: .05.

90

knowledge, as measured by the questionnaire, with person-
ality and sales techniques partialled out are statistically

66 The coefficient of determination indicates

significant.
that about 12 percent of the unit variance in the salesmen's
levels of sales performance is explained by product knowledge
when treated as the only independent variable.

When product knowledge was measured by the sales
managers' paired comparisons, the multiple correlation co-
efficients between the independent variables; product know—
ledge, personality, and sales techniques, and the dependent
variables; sales effectiveness as measured by total units
sold and profitability of units sold, were .28 and .33 re-
spectively. The F tests showed that personality, sales
techniques, and product knowledge, as measured by the paired
comparisons, are significantly related to sales performance.67
The influence of these three variables explains about ten
percent of the unit variance of the salesmen's levels of
sales performance.

When sales techniques and personality were statis-

tically partialled out of the relationship, the partial

correlation coefficients between product knowledge, as

 

66Total units sold, F = 66.60; profitability of
units sold, F = 83.91; 3.86 (F) or above is necessary to
show a significant relationship with l and 522 degrees of
freedom and a 5% level of significance.

67Total units sold, F = 15.03; profitability of
units sold, F = 21.95; 2.62 (F) or above is necessary to
show a significant relationship with 3 and 520 degrees of
freedom and a 5% level of significance.

91

evaluated by the sales managers, and sales performance were
.13 and .20 for total units sold and profitability of units
sold (see Table 10). The F test indicates that there is
a statistically significant relationship between product
knowledge (as measured by the paired comparisons), and both
measures of sales performance (total units sold and profit-
ability of units sold), when personality and sales tech-
niques are partialled out.68
We can conclude that the relationship between prod-
uct knowledge, as measured by both the questionnaire and
the paired comparisons, and sales performance is statistic-
ally significant with and without the influence of person-
ality and sales techniques. It may be concluded that ac-
cording to the inter-dealership method of data analysis,
product knowledge is one of the determinants of sales ef-
fectiveness. Inasmuch as the combined influence of product
knowledge, personality, and sales techniques upon sales
performance accounted for only 14 percent of the unit vari-
ance of sales performance, the central hypothesis that
product knowledge, in conjunction with personality and
sales techniques, is one of the primary determinants of

sales performance is not strongly supported.

 

68Total units sold, F = 8.49; profitability of units
sold, F = 21.55; 3.86 (F) or above is necessary to show
a significant relationship with l and 522 degrees of free-
dom and a 5% level of significance.

 
 

 

 

92

Interdependency_g£_personalityflggg
galg§_techniques

To investigate further the relationship between
the three independent variables; personality, product know-
ledge, and sales techniques, the interdependency of these
variables was examined. Personality was correlated with
sales performance with product knowledge as measured by
the questionnaire and sales techniques partialled out of
the equation. The resulting partial correlation coeffici—
ents were -.01 for each of the measures of sales perform-
ance, total units sold and profitability of units sold.
The F tests indicate that when personality is treated as
the only independent variable, sales performance is not

69 This conclusion was also

significantly related to it.
arrived at when product knowledge was measured by the paired
comparisons. It was found that when personality was cor-
related with sales performance with the other two independ-
ent variables (product knowledge as measured by the paired
comparison matrix and sales techniques) statistically held
constant, the partial correlation coefficients for the two
measures of sales performance for all dealerships combined

70

were .03 and .01. (See Table 10.) This finding is

 

69Total units sold, F = .02; profitability of units
sold, F = .10; 3.86 (F) or above is necessary to show a
significant relationship with l and 522 degrees of freedom
and a 5% level of significance.

7OTotal units sold, F = .47; profitability of units
sold, F = .10; 3.86 (F) or above is necessary to show a
significant relationship with 1 and 522 degrees of freedom
and a 5% level of significance.

93

contrary to the generally accepted assumption that person-
ality is a dominant factor in sales effectiveness.

Sales techniques, on the other hand, when treated
as the only independent variable, with product knowledge
as measured by the questionnaire and personality partialled
out, showed a correlation of .15 with total units sold and
.17 with profitability of units sold. The F tests showed
that the relationships between sales techniques and sales
performance are statistically significant with the other

71 This result

two independent variables partialled out.
was supported by performing the partial correlation calcu-
lations between sales techniques and sales performance,

with product knowledge as measured by the paired compari—

72 The partial corre—

sons and personality partialled out.
lation coefficients were .10 and .09 for total units sold
and profitability of units sold respectively.

It may be concluded that the relationship between
personality and sales performance, when personality is
treated as the only independent variable, is not statistic-

ally significant. On the other hand, when sales techniques

is treated as the only independent variable, its relationship

 

71Total units sold, F = 11.80; profitability of

units sold, F = 14.92; 3.86 (F) or above is necessary to
show a significant relationship with l and 522 degrees of
freedom and a 5% level of significance.

72Total units sold, F = 5.64; profitability of units
sold, F = 4.31; 3.86 (F) or above is necessary to show a
significant relationship with l and 522 degrees of freedom
at a 5% level of significance.

94

to sales effectiveness is statistically significant.

Conclusions from Testing the Central Hypothesis

 

Product knowledge, sales techniques, and personality
were shown to have a relationship with sales performance
on the basis of the inter-dealership analysis. However,
these three factors were not shown to relate with sales
performance on the basis of the intra—dealership analysis.

Even though the inter-dealership correlation anal-
ysis showed that product knowledge, sales techniques, and
personality correlated with sales performance, the combined
influence of these three variables explains only about 14 per—
cent of the unit variance of the salesman's sales performance
levels. The other 86 percent is unexplainable by these
three variables. Therefore, the central hypothesis that
product knowledge (in conjunction with two other assumed
determinants, personality and sales techniques) is one of
the primary determinants of sales effectiveness is not
strongly supported.

It can also be concluded that the relationship be-
tween both measures of product knowledge and sales perform—
ance are relatively independent of both personality and

sales techniques.

 

Conclusions Related £9_0ther Studies
The conclusion arrived at by the parametric method
of data analysis that product knowledge is related to sales

performance agrees with the study conducted by Eugene Benge

95

of 564 salesmen in widely different types of industry.73

Benge found as a result of his survey that on the traits,
"Well-Informed in his Field" and "Is Good Technically,"
the excellent salesmen exceeded the poor salesmen by 70
and 75 percent, respectively. The total product knowledge
traits in his survey, however, showed that the excellent
salesmen exceeded the poor men by 44 percent. Therefore,
Benge concluded that product knowledge is important in suc-
cessful salesmanship but less important than self—confidence,
the ability to plan, industriousness, and persuasiveness.
On the other hand, the results of this research
obtained from parametric statistical analyses do not con-
cur with the Baier and Dugan study concerning life insur-
ance salesmen.74 Baier and Dugan related a short mental
ability test (Wesman Personnel Classification Test) and
The Information Index, which is a test of life insurance
knowledge, with three kinds of performance criteria. The
tests were administered to 596 debit insurance agents, 126
staff managers, and 42 district managers. They found that
the Wesman Classification Test did not correlate signif-
icantly with any of the performance criteria, whereas The

Information Index was related to some of the performance

 

73Benge, "What Traits and Work Habits Characterize
Successful Salesmen?" Sales Management, 77, 1956, pp. 54-56.

74Baier and Dugan, "Tests and Performance in a Sales
Organization," Personnel Psych. 9, 1966, pp. 17-26.

 

96

criteria in the two management samples, but not in the agents
sample. Baier and Dugan, therefore, concluded that the
effects of technical knowledge of life insurance upon sales
performance must have been obscured by other variables.

The conclusion reached from the parametric statis—
tics concerning the importance of product knowledge in an
automobile salesman's performance, however, does agree with
the conclusions of a study of life insurance salesmen made
by Thayer, Antoinetti, and Guest.75 They used The Infor-
mation Index as a measurement of product knowledge and cor-
related the scores with a performance index which measured
the proportion of insurance that was dropped by policyholders
after having made the initial payments. The correlations
for the two performance index periods were -.29 and —.27
for sample sizes of 62 and 49 salesmen respectively. The
researchers concluded that product knowledge is an asset
for life insurance salesmen.

In comparing the results of this study with authors
who have not presented research findings to support their
position, wide disagreement concerning the role of product
knowledge in the determination of sales effectiveness was
found. Some writers in the field of personal selling, for
example, have asserted that product knowledge is the most

vital element of successful selling.76 On the other hand,

 

7SThayer, Antoinetti, and Guest, "Product Knowledge
and Performance," Personnel Psych. 11, 1958, pp. 411-417.

76Husband, Psychology_g£_Successful Selling, p. 27,
and Small, Salesmanship, p. 27.

 

 

 

97

Francisco and McMurry argue that product knowledge has been
over-emphasized in relationship to sales effectiveness.
Most authors in the field of personal selling, however,
support the position that product knowledge is one of the
fundamentals necessary for sales effectiveness, along with
sales techniques and personality. This study in the auto-
mobile industry would not support too strongly the thesis
that product knowledge, in conjunction with personality
and sales techniques, is a primary determinant of sales
success.

Product Knowledge - Sales Performance

Relationshipby_$ype and Size
g£.Dealership

 

 

 

 

In order to test the hypotheses that the relation-
ship between a salesman's level of product knowledge and
sales effectiveness is influenced by the:

. type of market in which he sells (hypothesis 2A)

. size of the retail outlet in which he sells
(hypothesis 28).

the 129 participating dealerships were subdivided into four
subgroupings: multi—point - low volume, multi—point - high
volume, single point - low volume, and single point - high
volume. Table 11 shows the number of dealerships in each

subgroup and the corresponding percentage of each major

 

77Francisco, "How Can You Get Salesmen to Sell
Today?" Sales Management 79, 1957, pp. 62-68, and McMurry,
"The Mystique of Super Salesmanship," Harvard Business 52:
view, 39, 1961, pp. 113-122.

 

 

98

classification to the total. Then the relationship between
a salesman's product knowledge and his sales effectiveness
was studied within each of these classifications by means

of both parametric and nonparametric analysis of the data.

TABLE 11
SUBGROUPS OF DEALERSHIP CLASSIFICATIONS
Classification of Dealerships

Multi:point Single Point Total

High Volume 36- 49 85 (66%)
Size of
L 3 Volume 24 20 44 (34%)
Dealerships
Total 60 (47%) 69 (53%) 129 (100%)

Product knowledger—sales performance relationship
by type 2; dealership (intra-dealership analysis)

 

The data were analyzed according to the type of
retail market, using the nonparametric method of data an-
alysis. The resulting average Kendall rank correlation
coefficients between the salesmen's product knowledge, as
measured by the questionnaire, and their sales performance,
when measured by total units sold and profitability of units
sold, were .22 and .27 for the multi-point dealerships,
whereas the single point dealerships were .02 and .14 re—
spectively. (See Table 12.) Tests for significance indi-
cated that none of these correlation coefficients was high

enough to show a statistically significant relationship

9

TABLE

9

12

KENDALL RANK CORRELATION ANALYSIS:
SALES PERFORMANCE WITH PRODUCT KNOWLEDGE
(MULTI-POINT VERSUS SINGLE POINT DEALERSHIPS)

Dependent Variables

 

 

 

 

Sales Sales
Performance Performance:
(Total Units (Profitability
Independent Variables Sold) of Units Sold)
A B A B
Product Knowledge as
Measured by the Question-
naire .22 .02 .27 .14
Product Knowledge as
Measured by the Paired
Comparison Matrix .28 .17 .43 .27

 

 

 

 

 

Key:

A Multi-point Dealerships (n=5)

B Single point Dealerships (n=4)

 

100

between product knowledge and sales performance.78

When product knowledge (as measured by the paired
comparisons) was related with sales performance (as measured
by total units sold and profitability of units sold) in the
multi-point dealerships, the average Kendall rank correla-
tion coefficients were .28 and .43 whereas those in the
single point dealerships were .17 and .27 respectively.

(See Table 12.) None of the relationships were found to
be statistically significant.79

Inasmuch as the relationships between product know-
ledge and sales performance are not significant, it is im-
possible to determine by means of the intra-dealership an-
alysis if the relationship between a salesman's level of
product knowledge and his sales effectiveness is dependent
upon the type of market served by the dealership.

Product knowledge - sales performance relationship
by type 2: dealership—TIHter-dealership analysiET

Parametric statistics, which were based upon an

 

78Multi-point dealerships--total units sold; p =

.46; profitability of units sold; p = .42, .05 (00 or less
is necessary to show that the two variables are signifi-
cantly related.

Single point dealerships-~total units sold; p =
.61; profitability of units sold; p = .52, .05 «1) or less
is necessary to show that the two variables are signifi—
cantly related.

79Multi-point dealerships-m-total units sold, p =
.41; profitability of units sold, p = .32; .05 (00 or less
is necessary to show that the two variables are signifi-
cantly related.

Single point dealerships m-total units sold, p =
.50; profitability of units sold, p = .42; .05 fly) or less
is necessary to show that the two variables are signifi-
cantly related.

101

inter-dealership analysis, were employed to investigate

whether the type of market served influenced the relation-
ship between product knowledge and sales effectiveness.80
Studying the multi—point dealerships, it was found that
the multiple correlation coefficients between the sales-

men's product knowledge (as measured by the questionnaire),

sales techniques, and personality with sales performance

 

80Since multiple/partial correlations and F tests
require that the variables under study be normally distrib-
uted, Kolmogorov-Smirnov tests for normality were performed
on selected variables for the different dealership group-
ings. The frequency distribution curves for the variables,
total units sold and profitability of units sold are shown
in Appendix N.

The tests for normality indicated that these two
variables for the two dealership groupings, multi-point
and single point, were not normally distributed at either
the .05 or the .01 level of significance. Therefore, the
two measures of sales performance, total units sold and
profitability of units sold, were transformed to more nor-
mal distributions by means of taking the log to the base
10. The resulting distributions were found to approximate
normal distributions at the .05 level of significance.

The distribution curves and the tests for normal-
ity for the variable, product knowledge as measured by the
questionnaire, are shown later in this chapter (pp. 121—122).
The results of these tests indicate that product knowledge
as measured by the questionnaire is normally distributed
at the .01 level of significance for both of these dealer-
ship groupings. In an attempt to increase the normality
of these distributions, they were transformed by means of
taking the log to the base 10. Because of the shape of
the distributions, however, this transformation did not
favorably influence their normality. Therefore, the re-
liability of the findings based upon these distributions
are not as reliable as they would be if the distributions
were normally distributed at the .05 level of significance.

The variables product knowledge as measured by the
paired comparisons, sales techniques, and personality were
obtained by transforming the rank scores into a set of
scores from a normal distribution, thereby achieving the
condition of normality.

102

were .19 for both total units sold and profitability of
units sold. (See Table 13.) The F tests indicated that
there is a significant relationship between the three in-
dependent variables and sales performance.81

When personality and sales techniques were par-
tialled out of the equation, product knowledge, as meas—
ured by the questionnaire, was statistically related to
both measures of sales performance.82 The partial corre-
lation coefficients between product knowledge, as measured
by the questionnaire, when personality and sales techniques
were partialled out, with total units sold was .14 and .16
with profitability of units sold. Product knowledge is
shown to explain about three percent of the unit variance
in the salesmen's sales performance.

The independent variable product knowledge, as
measured by the sales managers' paired comparisons, when
related to sales performance along with the two other in-
dependent variables, personality and sales techniques,
resulted in multiple correlation coefficients of .18 and
.23 respectively. (See Table 13.) Tests for significance
showed both correlation coefficients to be statistically

41

81Total units sold, F = 3.35; profitability of
units sold, F = 3.36; 2.64 (F) or above is necessary to
show a significant relationship with 3 and 276 degrees
of freedom and a 5% level of significance.

82Total units sold, F = 5.20; profitability of
units sold, F = 6.83; 3.88 (F) or above is necessary to
show a significant relationship with l and 278 degrees of
freedom and a 5% level of significance.

103

TABLE 13

MULTIPLE/PARTIAL CORRELATION ANALYSIS:
SALES PERFORMANCE WITH PRODUCT KNOWLEDGE,
PERSONALITY, AND SALES TECHNIQUES
(MULTI—POINT VERSUS SINGLE POINT DEALERSHIPS)

 

Dependent Variables
Sales Sales
Performance: Performance:
(Total Units (Profitability
Sold) of Units Sold)
Independent Variables A B A B
Product Knowledge &
Sales Techniques &
Personality .19' .50‘ .19‘ .55‘
Product Knowledge with
Sales Techniques & Per-
a sonality Partialled Out .14‘ ..21’ .16‘ .24'
Personality with Product
Knowledge & Sales Tech-
niques Partialled Out .10 —.02 .06 .01
Sales Techniques with
Product Knowledge & Per-
sonality Partialled Out -.02 .36‘ .02 .38‘
Product Knowledge &
Sales Techniques &
Personality .18‘ .47‘ .23‘ .52‘
Product Knowledge with
Sales Techniques & Per-
b sonality Partialled Out .12 .10 .21‘ .16‘
Personality with Product
Knowledge & Sales Tech-
niques Partialled Out .09 -.02 .03 .01
Sales Techniques with
Product Knowledge & Per—
sonality Partialled Out -.08 .32‘ —.08 .31‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Key:

a Product Knowledge
as Measured by the
Questionnaire

b Product Knowledge
as Measured by the
Paired Comparison

‘ Values Statistically Significant at 04:

A Multi-point Dealerships
(n=280)

B Single Point Dealerships
(n=244)

.05

 

104

significant.83 When product knowledge, as measured by the
paired comparisons, was correlated with sales performance
with personality and sales techniques partialled out, the
multiple correlation coefficients were .12 and .21, for total
units sold and profitability of units sold. Tests for sig—
nificance showed that the relationships between product
knowledge and both measures of sales performance are sta-
tistically significant.84

Therefore, it may be concluded that in the multi-
point dealership classification there is a significant re-
lationship between both measures of product knowledge and
sales performance, as measured by both total units sold
and profitability of units sold.

In the single point dealerships, it was found that
the relationship between the three independent variables;
product knowledge, as measured by the questionnaire, per-
sonality, and sales techniques, and the dependent variables;
sales performance as measured by total units sold and
profitability of units sold were .50 and .55 respectively.
(See Table 13.) The F tests show that both R's are

83Total units sold, F = 3.05; profitability of
units sold, F = 5.18; 2.64 (F) or above is necessary to
show a significant relationship with 3 and 276 degrees
of freedom and a 5% level of significance.

84Total units sold, F = 4.29; profitability of
units sold, F = 12.25; 3.88 (F) or above is necessary to
show a significant relationship with 1 and 278 degrees
of freedom and a 5% level of significance.

105

statistically significant.85 When personality and sales

techniques were partialled out of the multiple regression
equation, the resulting partial correlation coefficients
between product knowledge, as measured by the questionnaire,
and sales performance were .21 for total units sold and
.24 for profitability of units sold. The F tests showed
that these correlation coefficients are statistically sig-
nificant.86

Product knowledge (as measured by the paired com-
parison matrices), sales techniques, and personality were
related to sales performance (as measured by total units
sold and profitability of units sold) with resulting cor-
relation coefficients of .48 and .52 which are significant
at the .05 level of significance.87 Again when sales tech-
niques and personality were partialled out of the equation
by means of partial correlation analysis, the resulting

partial coefficients were .10 and .16 between product know-

ledge, as measured by the paired comparisons, and total

 

85Total units sold, F = 26.19; profitability of
units sold, F = 34.04; 2.64 (F) or above is necessary to
show a significant relationship with 3 and 240 degrees of
freedom and a 5% level of significance.

86Total units sold, F = 11.07; profitability of
units sold, F = 15.11; 3.88 (F) or above is necessary to
show a significant relationship with 1 and 242 degrees of
freedom and a 5% level of significance.

87Total units sold, F = 22.93; profitability of
units sold, F = 30.15; 2.64 (F) or above is necessary to
show a significant relationship with 3 and 240 degrees of
freedom and a 5% level of significance.

106

units sold and profitability of units sold respectively.
Product knowledge as measured by the paired comparisons
is not statistically related to total units sold, but the
relationship between this variable and profitability of
units sold is statistically significant.88

The results of the single point dealership anal-
ysis indicates that both measures of product knowledge sig—
nificantly correlate with sales performance as measured
by profitability of units sold. It was found that product
knowledge, as measured by the questionnaire, does signif-
icantly correlate with sales performance as measured by
total units sold, but when product knowledge was measured
by the paired comparisons, it does not significantly cor-
relate with the total units sold criterion of sales per-
formance.

As has been discussed previously, the intra-dealer-
ship analysis has shown that the type of dealership is not
related to the relationship between a salesman's level of
product knowledge and his sales performance. This finding
does not support hypothesis 2A.

On the other hand, the inter-dealership analysis

showed when product knowledge was treated as the only

 

88Total units sold, F = 2.21; profitability of units
sold, F = 6.42; 3.88 (F) for above is necessary to show a
significant relationship with l and 242 degrees of freedom
and a 5% level of significance.

Based upon 1 and 242 degrees of freedom, the par-
tial correlation coefficient has to be .12 or greater to
be statistically significant.

107

independent variable in both multi-point and single point
dealerships, it could explain about four percent of the
unit variance in the salesmen's sales performance levels.
It can be concluded that the type of market served does
not significantly influence the relationship between a
salesman's level of product knowledge and his sales ef—
fectiveness.
Product knowledge - gglgg performance
relationship by_§igg'g£_dealership

The hypothesis that the relationship between a
salesman's level of product knowledge and sales effective-
ness is influenced by the size of the retail outlet in which
he sells, within the multi-point—-single point dealership
classifications was tested. In order to do this, the data
were organized into four classifications of dealership size,
based upon new car planning volume as follows:

. multi—point high volume versus multi—point low
volume

. single point high volume versus single point low
volume.

The relationship between a salesman's product knowledge
and his sales effectiveness was studied within each of
these classifications by means of both parametric and non-

parametric analysis of the data.89

 

89Since multiple/partial correlation analysis and
F tests require that the variables under study be normally
distributed, Kolmogorov-Smirnov tests for normality were
performed on selected variables for the different dealer—
ship groupings.

The frequency distribution curves for the variables,

108

Multi:point dealerships
(Intra-dealership analysis)

 

According to nonparametric statistics, the average

Kendall rank correlation coefficients between the salesmen's

 

total units sold and profitability of units sold are shown
in Appendix 0. The tests for normality indicated that these
two variables were normally distributed at the .05 level

of significance for the single point low volume dealership
classification.

Since these two variables were shown to be normally
distributed at only the .01 level of significance in the
multi-point high volume classification, the data were trans-
formed by means of taking the log to the base 10 for both
total units sold and profitability of units sold. The re-
sults of testing for normality of the transformed distribu-
tions indicated that the transformation did not favorably
influence their normality. Therefore, the findings presented
are based upon the non-transformed data and are not as re-
liable as they would be if the distributions were normally
distributed at the .05 level of significance.

On the other hand, the tests for normality indicated
that total units sold and profitability of units sold in
both single point high volume dealerships and multi-point
low volume dealerships were normally distributed at the .01
level of significance. In an attempt to increase the nor-
mality of these distributions, they were transformed by
means of taking the log to the base 10. The resulting dis-
tributions were found to be normally distributed at the .05
level of significance.

The distribution curves and the tests for normality
for the variable product knowledge, as measured by the ques-
tionnaire, are shown later in this chapter (pp. 124 through
128). The results of these tests indicate that product
as measured by the questionnaire was normally distributed
at the .05 level of significance in both the single point
low volume and the multi—point low volume dealership clas—
sifications.

Whereas in the single point high volume and the
multi-point high volume dealerships, product knowledge as
measured by the questionnaire was normally distributed at
the .01 level of significance. In an attempt to increase
the normality of the distributions, they were transformed
by means of taking the log to the base 10. Because of the

109

product knowledge, as measured by the questionnaire, and
their sales performance, as measured both by total units
sold and profitability of units sold, were .18 and .21 for
the multi-point high volume dealerships and .24 and .36
for the multi-point low volume dealerships. Tests for sig-
nificance revealed there is not a statistically significant
relationship between product knowledge and sales perform—
ance in either multi-point high volume or multi-point low
volume dealerships.9o
When product knowledge, as measured by the paired
comparisons, was related to both measures of sales effec-

tiveness in the multi—point high and low volume dealerships,

the average Kendall rank correlation coefficients were not

 

shape of the distributions, however, this transformation
did not favorably influence their normality. Therefore,
the findings presented are based upon the non-transformed
data and are not as reliable as they would be if the dis-
tributions were normally distributed at the .05 level of
significance.

The variables product knowledge as measured by the
paired comparisons, sales techniques, and personality, were
obtained by transforming the rank scores into a set of
scores from a normal distribution, thereby achieving the
condition of normality.

90Multi-point high volume--total units sold; p =
.43; profitability of units sold; p = .40; .05 (09 or less
is necessary to show that the two variables are signifi-
cantly related.

Multi-point low volume--total units sold; p = .45;
profitability of units sold, p = .36; .05 (00 or less is
necessary to show that the two variables are significantly
related.

110

statistically significant.91 (See Table 14.)

TABLE 14
KENDALL RANK CORRELATION ANALYSIS:

SALES PERFORMANCE WITH PRODUCT KNOWLEDGE
(HIGH VERSUS LOW VOLUME DEALERSHIPS)

Dependent Variables

 

Sales Sales
Performance: Performance:
(Total Units (Profitability
Sold) of Units Sold)
Independent
Variables A B C D A B C D

 

Product Knowledge:
Ques. Sect. I & II .18 .24 .06 -.09 .21 .36 .21 -.06

 

 

Product Knowledge:
Paired Comparison
Matrix .30 .27 .17 .12 .38 .50 .30 .16

 

 

 

 

 

 

 

 

 

Key:
A Multi-point High Volume Dealerships (n=5)
B Multi-point Low Volume Dealerships (n=4)
C Single Point High Volume Dealerships (n=4)

D Single Point Low Volume Dealerships (n=3)

 

91Multi-point high volume--total units sold (r =
.30); p = .32; profitability of units sold (r = .38); p =
.32; .05 (00 or less is necessary to show that the two
‘variables are significantly related.

Multi—point low volume--total units sold (r = .27);
p = .42; profitability of units sold (r = .32); p = .32;
.05 (00 or less is necessary to show that the two variables
are significantly related.

 

111

Since the intra-dealership analysis showed that
there is no statistically significant relationship between
product knowledge and sales performance in either multi-
point high volume or multi—point low volume dealerships,
it can be concluded that the product knowledge-—sales
performance relationship is independent of the size of
the retail outlet in the multi-point dealership classifi-

cation.

Multi-point dealerships
(Inter-dealership analysis)

When the parametric method of data analysis was
employed in the multi-point high volume dealerships, the
multiple correlation coefficients between product knowledge,
as measured by the questionnaire, sales techniques, and
personality with sales performance, were .25 for total units
sold and .21 for profitability of units sold. (See Table
15.) The F tests indicated that these relationships are
statistically significant.92 When personality and sales

techniques were partialled out, the partial correlation

 

92Total units sold, F = 4.46; profitability of
units sold, F = 3.05; 2.66 (F) or above is necessary to
show a significant relationship with 3 and 176 degrees of
freedom and a 5% level of significance.

112

TABLE 15

MULTIPLE/PARTIAL CORRELATION ANALYSIS:
SALES PERFORMANCE WITH PRODUCT KNOWLEDGE,
PERSONALITY, AND SALES TECHNIQUES

(HIGH VERSUS LOW VOLUME DEALERSHIPS)

Dependent Variables

 

 

Independent Variables

A

Sales

Performance:
(Total Units

Sold)

B

C

A

Sales

Performance:
(Profitability
of Units Sold)

B

C

 

Product Knowledge
& Sales Techniques
& Personality

.25‘I

.25

.28‘

.44‘l

$.21;

.31'

.30’

.47‘

 

Product Knowledge
with Sales Tech-
niques & Personality
Partialled Out

.08

.24’

"002

L.02

.01

.301

.01

b"007

 

Personality with
Product Knowledge &
Sales Techniques
Partialled Out

.22‘

.03

-011

P.12

.18‘

.02

‘008

—013

 

Sales Techniques
with Product Know-
ledge & Personality
Partialled Out

.05

.00

.28‘

.44‘

.OO

.05

.29“

.47‘

 

Product Knowledge &
Sales Techniques
& Personality

.28‘

.23

.28‘

.45‘

.28‘

.301

.31‘

.47‘

 

Product Knowledge
with Sales Tech-
niques & Personality
Partialled Out

.15‘

.22‘l

.01

"011

.19.

.294

-009

“-002

 

Personality with
Product Knowledge
& Sales Techniques
Partialled Out

.17‘

.04

.01

-010

.11

.00

-009

-013

 

 

f"?

Sales Techniques
with Product Know-
ledge & Personality
Partialled Out

 

L

.10

-014

 

 

.24‘

.45.

 

 

‘006

 

-014

 

.23‘

.44'

 

 

Key:
Product Knowledge
as Measured by
the Questionnaire

Product Knowledge
as Measured by the
Paired Comparison

A

(n=180)

B

(n=100)

C

(n=ll3)

D
(

n=51)

‘ Values Statisticallv Significant at 06:

.05

Multi—point High Volume Dealerships
Multi-point Low Volume Dealerships
Single Point High Volume Dealerships

Single Point Low Volume Dealerships

 

113

coefficients between product knowledge as measured by the
questionnaire and sales performance were .08 and .01 for
total units sold and profitability of units sold. The F
tests showed that product knowledge, when treated as an
independent variable, does not correlate with sales per-
formance in multi—point high volume dealerships.93
The relationships between sales performance as
measured by total units sold and profitability of units
sold and product knowledge, as measured by the question-
naire, personality, and sales techniques, in multi—point
low volume dealerships were correlated at .25 and .31.
Even though the F tests indicated that the multiple regres-
sion correlation coefficient with total units sold is not
statistically significant, it is statistically significant
with profitability of units sold.94
The relationship between sales performance and prod-
uct knowledge as measured by the questionnaire, with the

influence of personality and sales techniques partialled

out is statistically significant.95 These partial

 

93Total units sold, F = 1.20; profitability of
units sold, F = .02; 3.90 (F) or above is necessary to
show a significant relationship with 1 and 178 degrees
of freedom and a 5% level of significance.

94Total units sold, F = 2.13; profitability of
units sold, F = 3.33; 2.70 (F) or above is necessary to
show a significant relationship with 3 and 96 degrees of
freedom and a 5% level of significance.

95Total units sold, F = 6.08; profitability of
units sold, F = 9.69; 3.94 (F) or above is necessary to
show a significant relationship with 1 and 98 degrees of
freedom and a 5% level of significance.

114

correlation coefficients were .24 for total units sold and
.30 for profitability of units sold. As shown in Table 15,
when sales techniques and personality were treated as the
only independent variables, they had no significant influ—
ence upon the multiple regression correlation coefficient.
This explains the low level of the multiple correlation
coefficients, i.e., .25 and .31 when all three independent
variables were related with sales performance.96

Based on the above inter-dealership analysis with
product knowledge as measured by the questionnaire, it can
be concluded that the size of the retail outlet within the
multi-point dealership classification does have some rela-
tion to the product knowledge--sales performance relation-
ship.

The results were somewhat different, however, when
product knowledge, as measured by the paired comparison
matrix, was used in the parametric method of data analysis.
The multiple correlation between product knowledge, person—
ality, and sales techniques with each of the two measures

of performance was .28 in the multi-point high volume

 

96Sales techniques--tota1 units sold (r = .00);
F = .00; profitability of units sold (r = .05); F = .28;
3.94 (F) or above is necessary to show a significant re—
lationship with l and 98 degrees of freedom and a 5% level
of significance.

Personality--tota1 units sold (r = .03); F = .08;
profitability of units sold (r = -.02); F = .04; 3.94 (F)
or above is necessary to show a significant relationship
with l and 98 degrees of freedom and a 5% level of sig—
nificance.

115

dealership classification. The F tests showed that the
independent variables are significantly related to the de—
pendent variables.97 When sales techniques and personality
were partialled out of the regression equation, the partial
correlation coefficients were .15 and .19 for total units
sold and profitability of units sold respectively. (See
Table 15.) These correlation coefficients are statistic-
ally significant at the .05 level.98 Therefore, it can

be concluded that in the multi-point high volume dealer-
ships, there is some relationship between product knowledge
as measured by the paired comparison matrices and sales
performance.

The findings in the multi-point low volume dealer-
ships indicate that there is no relationship between product
knowledge, as measured by the paired comparisons, and sales
performance as measured by total units sold. However, there
is a statistically significant relationship between these

variables and profitability of units sold.99 The multiple

 

97Total units sold, F = 5.67; profitability of units
sold, F = 5.59; 2.66 (F) or above is necessary to show a
significant relationship with 3 and 176 degrees of freedom
and a 5% level of significance.

98Total units sold, F = 4.62; profitability of units
sold, F = 7.30; 3.90 (F) or above is necessary to show a
significant relationship with l and 178 degrees of freedom
and a 5% level of significance.

99Total units sold, F = 1.73; profitability of units
sold, F = 3.08; 2.70 (F) or above is necessary to show a
significant relationship with 3 and 96 degrees of freedom
and a 5% level of significance.

 

 

 

116

correlations between product knowledge, personality, and
sales techniques with sales performance, as measured by
total units sold and profitability of units sold, were .23
and .30. When personality and sales techniques were par-
tialled out of the equation, the partial correlation coef—
ficients between product knowledge and sales performance
were .22 and .29 for total units sold and profitability

of units sold respectively. The F tests revealed that the
relationships between product knowledge, as measured by
the paired comparisons, and both measures of sales perform-
ance are statistically significant.100

Based on the above inter-dealership analysis with
product knowledge as measured by the paired comparisons,
it can be concluded that the size of the retail outlet
within the multi-point dealership classification does not
have any influence upon the product knowledge—-sales per—
formance relationship.

This conclusion contradicts the previous finding
derived from the parametric analyses using the questionnaire
as the measure of product knowledge. As discussed in Chap-
ter IV, there is a statistically significant relationship
between the sales managers' paired comparisons of product

knowledge and those of sales techniques. It seems

 

lOOTotal units sold, F = 4.88; profitability of
units sold, F = 8.96; 3.94 (F) or above is necessary to
show a significant relationship with l and 98 degrees of
freedom and a 5% level of significance.

117

reasonable to conclude that this inter-relationship may
have caused the aforementioned contradiction. If we accept
this explanation, the findings obtained by use of the ques-
tionnaire as the measure of product knowledge are more
valid. Using the questionnaire as the product knowledge
measurement, it can be concluded that the relationship
between the salesman's level of product knowledge and his
sales effectiveness is related to the size of the retail
outlet within the multi—point dealership classification.
With the intra-dealership analysis, however, the indication
was that dealership size has no influence upon a salesman's
product knowledge--sales performance relationship.
Single point dealerships (Intra— and
inter-dealership analysis)

To determine if hypothesis 28 could be supported
in the single point classification of dealerships, the
foregoing nonparametric and parametric analyses were re—
peated. The purpose was to ascertain if the relationship
between product knowledge and sales performance was influ-
enced by the size of the retail outlet measured by the
dealerships new car planning volume. As previously dis—
cussed, according to nonparametric statistics there was
no statistically significant relationship between product
knowledge, measured both by the questionnaire and the paired
comparison matrix, in the single point dealership classifi-
cation. It was found that when the single point dealerships

were grouped into high and low volume classifications, there

118

was no statistically significant correlation between product
knowledge, as measured by both data collecting instruments,
and sales effectiveness (shown in Table 14).101 As a re-
sult, it can be concluded that the size of the retail out-
let within the single point dealership classification does
not have any influence upon the relationship between product
knowledge and sales performance.

The same conclusion was achieved when the data were
analyzed using parametric methods. The findings indicated
that there was no statistically significant relationship
between product knowledge, as measured by the questionnaire,

and the sales managers' paired comparisons, and sales per-

formance with personality and sales techniques partialled

 

101single point high volume--product knowledge as

measured by the questionnaire: total units sold (r = .06);
p = .58; profitability of units sold (r = .12); p = .54;
.05 (00 or less is necessary to show that the two variables
are significantly related.

Single point high volume—-product knowledge as meas-
ured by the paired comparisons: total units sold (r = .17);
p = .50; profitability of units sold (r = .30); p = .40;
.05 (00 or less is necessary to show that the two variables
are significantly related.

Single point low volume--product knowledge as meas—
ured by the questionnaire: total units sold (r = —.09);
p = .59; profitability of units sold (r = -.06); p = .60;
.05 (00 or less is necessary to show that the two variables
are significantly related.

Single point low volume--product knowledge as meas-
ured by the paired comparisons: total units sold (r = .12);
p = .58; profitability of units sold (r = .16); p = .56;
.05 «3) or less is necessary to show that the two variables
are significantly related.

 

 

 

 

 

119

102

out in the single point dealership classification. (See

Table 15.)

To summarize, the parametric method of data anal-
ysis has shown that within the multi-point classification
of dealerships, the size of the retail outlet does have some
influence upon the relationship between a salesman's level
of product knowledge and his sales effectiveness. Within
the single point dealership category, however, no signifi-
cant relationships were found.

The nonparametric statistical analyses showed that
there was no significant relationship between a salesman's
level of product knowledge and his sales performance in

either high or low volume dealerships for both single point

g

102single point high volume--product knowledge as

measured by the questionnaire: total units sold (r = .02);
F = .05; profitability of units sold (r = .01); F = .02;
3.95 (F) or above is necessary to show a significant rela-
tionship with l and 191 degrees of freedom and a 5% level
of significance.

Single point high volume—-product knowledge as meas-
ured by the paired comparisons: total units sold (r = .01);
F = .04; profitability of units sold (r = .09); F = 1.38;
3.95 (F) or above is necessary to show a significant rela-
tionship with l and 191 degrees of freedom and a 5% level
of significance.

Single point low volume--product knowledge as meas-
ured by the questionnaire: total units sold (r = -.02);
F = .02; profitability of units sold (r = -.07); F = .26;
4.03 (F) or above is necessary to show a significant rela-
tionship with l and 49 degrees of freedom and a 5% level
of significance.

Single point low volume-—product knowledge as meas-
Lutnd by the paired comparisons: total units sold (r = -.1l);
F = .57; profitability of units sold (r = —.02); F = .02;
4-053 (F) or above is necessary to show a significant rela—
tion ship with l and 49 degrees of freedom at a 5% level
0f Significance. '

120

and multi-point dealership classifications. As a result,
this stratification of the dealerships did not relate to
the product knowledge--sales effectiveness relationship.

If dealership size was measured, for example, by
the number of salesmen, a significant relationship between
product knowledge and sales effectiveness might have been
shown by the nonparametric method of data analysis.

Product Knowledge g§_g;Function g£_
Dealership EYES.222.§$E§.

Due to the diversity of the automobile market with
respect to demographic factors, it was postulated that the
salesmen's product knowledge would vary according to the
classification of the dealership. Therefore, the effect
of the size and type of dealership upon the salesmen's prod-
uct knowledge was determined.

In order to statistically evaluate these relation-
ships, the 524salesmen participating in this study were

subdivided into the four subgroupings as shown in Table 16.

TABLE 16
SUBGROUPS OF SALESMEN CLASSIFICATIONS
Type of Dealerships

Multi—point Single point

High Volume 180 193
Size of
Dealerships Low Volume 100 51

Total 280 244

121

Product knowledge — function g:
dealershipytype

 

In order to support the hypothesis that a salesman's
product knowledge is influenced by the type of market in
which he sells, the mean score on the product knowledge
questionnaire of salesmen in multi-point areas was compared
to the questionnaire scores received from salesmen who sell
in single point dealerships. The mean score received on
the questionnaire in the multi-point dealerships was 21.1
(out of a possible 30), whereas the salesmen in the single
point dealerships received a mean score of 21.4. The dis-
tribution of the questionnaire scores for both types of
dealerships is shown in Figure4.

To determine if the mean scores are significantly
different, it was necessary to conduct a normal "2" test.
The normal "z" test, however, assumes the distributions
are normal and the variance equal. In order to determine
if the data met the normality assumptions, a Kolmogorov-
Smirnov test was conducted and the results indicated that
the distributions were normal at a significance level of

103

.01. The F test for homogeneity of variance indicated

that the variances of the two samples were approximately

 

103Multi-point dealerships, D = .08; D less than
.10 is necessary to show that the variable is normally
distributed at the .01 «x; level of significance with an
n of 280.

Single point dealerships, D = .09; D less than .10
is necessary to show that the variable is normally dis-
tributed at the .01 (08 level of significance with an n
of 244.

122

FIGURE 4

FREQUENCY DISTRIBUTION OF

Number of Scores QUESTIONNAIRE scones

Within each Multi—point Dealerships

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Increment
30)
24 T
n: 280 —
Mean Score: 21.1
18 Standard Deviation: 4.31
12
6
,JT I—IF1 F111F1[] [11]
0 5 10 15 20
Questionnaire Scores
Number of Scores
Within each
Increment Single Point Dealerships
30 A
24 _
n: 244
Mean Score: 21.4
18 Standard Deviation: 4.78
12
6
{P [1 [li] {1_ F1[]
0 10 15 20 25 29

Questionnaire Scores

123

equal.104 Therefore, the normal "2" test was performed

and the results indicated that no significant difference
exists between the salesmen's level of product knowledge

in the two types of markets.105

This does not support the
hypothesis (3A) that a salesman's level of product knowledge
is influenced by the type of market in which he sells.

One explanation of this could be that all salesmen
receive the same literature, such as sales brochures, data
books, and feature booklets from the manufacturer. There-
fore, all salesmen theoretically are given equal opportuni-
ties to study and learn product knowledge. Since the size
of the sales force usually varies in proportion to the dealer-
ships volume of business, salesmen in both types of dealer-
ships may have an opportunity to present their product to
approximately the same number of people. Therefore, sales-
men in both single point and multi-point dealerships could
be exposed to the same quantityof consumer knowledge inquir—

ies with equal or near equal opportunities to respond to

these questions.

 

104F-O-12—4'782-122'F(OL- os>~124 with

243 and 279 degrees of freedom. Since the sample F (1.22)
is smaller than F (Cl: .05), the sample variances are ap-
proximately the same.

1052 = .750; 1.645 (z) or above is necessary to
show a significant difference at the 5% level of signifi-
cance. (See Appendix P for an illustration of this test.)

124

Product knowledge — function g: dealership size

 

 

 

It may be theorized that there would be a differ-
ence between the salesman's product knowledge in relation-
ship to the size of the dealership in which he sells. This
hypothesis was analyzed by first comparing the product ques-
tionnaire scores of salesmen in multi-point high volume
dealerships with those in multi-point low volume dealer-
ships. Then the single point dealerships were also divided
into high volume and low volume subgroupings and the scores
received in each classification of dealership size were com-
pared.

The distributions of scores received in the multi-
point high and low volume dealerships are shown in Figure
5. The mean score received by multi-point high volume sales—
men was 21.1 (out of a possible 30), whereas the multi-point
low volume salesmen received a mean score of 21.0. Since
the distributions were shown to approximate normal distri-
butions by means of the Kolmogorov-Smirnov tests and an F
test revealed that the sample variances were approximately

equal, a normal "2" test was performed.106 It indicated

 

106Multi-point high volume dealerships, D = .11;
D less than .12 is necessary to show that the variable is
normally distributed at the .01 level of significance and
with an n of 180.

Multi-point low volume dealerships, D = .10; D less
than .14 is necessary to show that the variable is normally
distributed at the .05 «1) level of significance and with
an n of 100.

0'12 4 442
F = = . = 1.21' F (06: .05) = 1.34 With
522 4 042 ’

bhimber of Scores

Within each

 

125

FIGURE 5

FREQUENCY DISTRIBUTION OF
QUESTIONNAIRE SCORES

Multi-point High Volume Dealerships

 

 

 

 

 

 

 

 

 

 

 

 

 

Increment
30.
24 7
n: 180
Mean Score: 21.1 _
18 Standard Deviation: 4.44
1. F
6
:JW £1f7 F11] F]f1[1 r] 77 [1:
0 5 10 15 20 25 29
Questionnaire Scores
Number of Scores
Within each
Increment
3O Multi—point Low Volume Dealerships
24
n: 100
Mean Score: 21.0
18 Standard Deviation: 4.04
12
6
O 4! F1 [1 F111 r] r
5 10 15 20 25 29

Questionnaire Scores

126

that there is no statistically significant difference in

the levels of product knowledge in these two classifications
of dealerships, multi-point high versus multi-point low vol-
ume dealerships.107

In order to determine if the salesman's level of
product knowledge is dependent upon the size of the retail
outlet in the single point dealership classification, the
distributions of the questionnaire scores were compared.
These distributions are shown in Figure 6. The mean score
received by single point high volume salesmen was 21.8,
whereas the single point low volume salesmen received a
mean score of 21.7.

In order to determine if the mean scores of two
samples are statistically equal, the conditions of normal-
ity and equal variances must be met. The F test to deter-
mine the homogeneity of variances indicated that the vari-
ance in the questionnaire scores between high and low volume
dealerships were approximately equal. Thus they qualify

for the normal "z" test.108

 

179 and 99 degrees of freedom. Since the sample F (1.21)
is smaller than F (Cl: .05), the sample variances are ap—
proximately the same.

1072 = .192; 1.645 (2) or above is necessary to
show a significant difference at the 5% level of signif-
icance.

108The single point dealerships have P =

622 3.822

degrees of freedom. Since the sample F (1.10) is smaller
than F (CL: .05), the sample variances are approximately
the same.

127

FIGURE 6

FREQUENCY DISTRIBUTION OF
QUESTIONNAIRE SCORES
ITumber of Scores
Within each

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Increment
3O 1 Single Point High Volume Dealerships
24
n: 193
Mean Score: 21.8 _ r
18 Standard Deviation: 3.98 __ 7 {
12
6
.,, nflflflflnfl . n
10 15

20 25 29
Questionnaire Scores
Number of Scores

Within each
Increment

30 1 Single Point Low Volume Dealerships
24

n: 51
18 Mean Score: 21.7

Standard Deviation: 3.82

12

 

- N H Hm.
20

5 10 15 29
Questionnaire Scores

 

 

 

 

 

128

The Kolmogorov-Smirnov tests indicated that the
distributions of the questionnaire scores were approximately
normal in both high and low single point dealership clas-
sifications, thus achieving the other condition necessary
in order to qualify the variables for the normal "z" test.109

The normal "2" test indicated that there was no sta—
tistically significant difference in the levels of product
knowledge in these two classifications of dealerships, single
point high versus single point low volume dealerships.

In conclusion, the hypothesis that a salesman's
level of product knowledge is dependent upon the size of
the retail outlet in which he sells was not supported by
the findings. In both the multi-point and single point
dealership classifications, it was found that the salesmen's
level of product knowledge in dealerships with high planning
volumes is not significantly different from the level of
product knowledge of salesmen in dealerships with relatively
low planning volumes.

Influence g£_Education, FMI, and Experience
upon Product Knowledge

 

In order to test the hypotheses that a salesman's

 

109Single point high volume, D = .11; D less than
.12 is necessary to show that the variable is normally dis—
tributed at the .01 «X) level of significance with an n
of 193.

Single point low volume, D = .12; D less than .19
is necessary to show that the variable is normally dis-
tributed at the .05 (09 level of significance with an n
of 51.

129

level of product knowledge is influenced by his:
. level of education

. attendance at a manufacturer-sponsored training
course

. sales experience within the industry,

nonparametric analyses of the data were conducted.

Education

The analysis showed average Kendall rank correla-
tion coefficients of .21 and .06 between the salesmen's
level of education and their product knowledge, as measured
by the questionnaire and the sales managers' paired compari—
sons respectively. (See Table 17.) Tests for significance
indicate that neither of these correlation coefficients is
statistically significant, therefore, indicating that there
is no relationship between the salesman's level of educa-
tion and his level of product knowledge.110 It can be con-
cluded that the complexity of the automobile does not ap-
pear to be a deterrent to the amount of information acquired
by salesmen with various levels of education. Since the
analysis revealed that no significant relationship exists
between the salesmen's educational levels and their prod-

uct knowledge, product information provided by the manufac-

turer for the salesmen may appeal equally to salesmen with

 

llOProduct knowledge (questionnaire) p = .47; prod-
uct knowledge (paired comparisons) p = .58; .05 «NJ or less
is necessary to show that the two variables are signifi-
cantly related.

130

TABLE 17
KENDALL RANK CORRELATION ANALYSIS:
PRODUCT KNOWLEDGE WITH EDUCATION,
FMI, AND EXPERIENCE

(Based upon an average n of 4)

Dependent Variables

 

Product Knowledge Product Knowledge
Independent Variables (Questionnaire) (Paired Comparison)

 

Level of Education .21 .06

 

Attendance at FMI
Course .02 .09

 

 

Experience in the
Automobile Industry -.03 .07

 

 

 

 

131

both high and low levels of education.

It can be theorized that instead of educational
level, other factors such as: personal motivation, age,
marital status, ambition, mechanical aptitude, and reading
skills, may influence a salesman's acquisition and retention

of product information.

§g£g_MarketingInstitute

As a corollary to studying the effect of formal
education upon the amount of an automobile salesman's prod-
uct knowledge, the relationship between the manufacturer's
sales training program and the salesmen's product knowledge
was analyzed. To measure this relationship, a respondent
was asked to record the length of time that had elapsed
since he last attended a Ford Motor Company Marketing In-
stitute (FMI) training course. When this measurement of
FMI was compared both with the Salesmen's questionnaire
results and the sales managers' product knowledge paired
comparisons, it was found that there is no statistically
significant relationship between the independent variable
(FMI) and the dependent variables (the two measures of prod-

111 (See Table 17.) The average Kendall

uct knowledge).
rank correlation coefficients between the independent vari-

able, attendance at FMI, and the dependent variables, the

 

lllProduct knowledge (questionnaire) p = 1.61;
product knowledge (paired comparisons) p = .56; .05 “1)
or less is necessary to show that the two variables are
significantly related.

132

two measures of product knowledge, were .02 and .09.

' A strong relationship between attendance at FMI
and the salesmen's level of product knowledge was not ex-
pected because the major emphasis of the FMI training course
is placed on the development of sales techniques. Since a
portion of the retail salesmen's course is devoted to prod-
uct knowledge and its presentation, some correspondence
between the salesmen's attendance at FMI and product know-
ledge, as measured by the questionnaire, was anticipated.

One factor which may explain the lack of a statis-
tically significant correlation between the salesmen's at—
tendance at FMI and their level of product knowledge is
that often the men who attend the training course are new
salesmen or those experienced salesmen whose performance
is marginal. Therefore, the respondents with the most re—
cent attendance at an FMI training session may be those
with relatively low levels of product knowledge.

Another factor which may have contributed to the
results of this analysis is that the data showed that many
dealerships had two or more of their salesmen attend FMI
training sessions at the same time. Since the analysis of
the relationship between attendance at FMI and the sales-
men's level of product knowledge was conducted on an intra—
dealership basis, the resulting low correlations for dealer-
ships which had two or more of their salesmen attend FMI

programs at the same time reduced the total average

133

correlation for all dealerships.112

Experience

Another factor considered important to investigate
was the influence of the automobile salesman's experience
upon his level of product knowledge. The average Kendall
rank correlation coefficients between the salesmen's experi—
ence within the industry and their product knowledge, as
measured by the questionnaire and the paired comparisons,
were -.03 and .07. (See Table 17.) Neither value was found
to be statistically significant.113

One important explanation for this result is that
salesmen with relatively little sales experience in the
automobile industry express a greater desire to learn in-
formation about product features and their corresponding

consumer benefits than do more experienced men.114 As a

 

112When dealerships having more than two salesmen
attend FMI during the same time period were removed from
the analysis, the resulting average Kendall rank correla-
tions between product knowledge, as measured by the ques-
tionnaire and the paired comparisons, and attendance at
FMI were .18 and .12. These correlation levels were shown
to be not significant. p = .50 and p = .53; .05 “1) or
less is necessary to show that the two variables are sig-
nificantly related. (Based on an average n of 4.)

113Product knowledge (questionnaire) p = .60; prod-
uct knowledge (paired comparisons) p = .57; .05 (00 or less
is necessary to show that the two variables are significant-
ly related.

114A study conducted by Rogers National Research
for the Ford Motor Institute showed that salesmen with less
than one year's experience demonstrated a much greater de-
sire to learn product knowledge than those with more than
a year's experience. Unpubl. study, Prepared for Ford Mar-
keting Institute by Rogers National Research, 1967.

134

result of this interest and the tendency of sales managers
to encourage new salesmen to familiarize themselves with
the product by use of data books, promotional literature,
and sales brochures, relatively new salesmen can acquire
equal or more knowledge of the product than the men with
greater experience.

Experience may not be a factor in the determination
of a salesman's level of product knowledge because the sales-
man with the longer sales experience may feel he can decide
which product knowledge is most valuable, in terms of in-
formation desired by consumers, and thus, he disregards
what he considers nonessential information concerning prod-
uct features and their corresponding consumer benefits.

The more inexperienced salesman, however, has no basis for
selecting only certain aspects of the provided product in-
formation so he studies all facets equally. When these
two types of salesmen were compared on the basis of the
questionnaire, which was not structured on the basis of
frequency of information requested by potential buyers,
the inexperienced salesman could score as high or higher
than the experienced man.

Another possible explanation for the inexperienced
salesmen scoring higher on the questionnaire than the ex—
perienced men is that frequently the salesmen who have sold
automobiles for a number of years may emphasize other fac-
tors, such as sales techniques and/or they may assume that

the product knowledge learned in the past is sufficient

135

to adequately meet the present product knowledge require-
ments. Therefore, these salesmen would probably score lower
on the 1967 Product Information Questionnaire than the less
experienced salesmen.

A final possible explanation for the low correla-
tion between experience and product knowledge is that all
of the salesmen in each of the dealerships measured have
had the same exposure to 1967 product information regard-
less of the length of their past experience of selling in
the automobile industry. Past experience may help the sales-
men evaluate and understand new product features and their
corresponding benefits, but the salesmen being compared
on the basis of their scores on the questionnaire, which
was composed of 1967 product information, all had the same
familiarization period for that specific product knowledge.

In summary, the results of this study do not sup-
port the hypotheses (3C, 3D, and 3E) that an automobile
salesman's level of education, his attendance at a manu-
facturer-sponsored training course, and his sales experi-
ence within the industry influence his level of product
knowledge.

Knowledge gf_Product Features and Consumer
Benefits - Sales Effectiveness

 

In an attempt to more fully understand the concept
of product knowledge, it was divided into knowledge of
product features and knowledge of consumer benefits. Meas-

ures of both of these factors were obtained by means of

136

the questionnaire. The scores received were related to
sales effectiveness in order to test the hypothesis (4)
that if a salesman's level of product knowledge is
divided into knowledge of product features and knowledge
of consumer benefits, each of these two components of prod—
uct knowledge contribute equally to his sales effectiveness.
By means of nonparametric analysis the salesman's
knowledge of product features, as measured by Section I
of the questionnaire, was correlated with the two measures
of sales effectiveness. The resulting average Kendall rank
correlation coefficients were .13 and .20 for total units
sold and profitability of units sold respectively. (See
Table 18.) Tests for significance revealed that both of
these correlation coefficients are not statistically signif-
icant, which indicates that there is no statistical rela-
tionship between knowledge of product features and a sales-
man's sales effectiveness.115
When the questionnaire scores on Section II were
related to sales performance, the average Kendall rank cor—
relation coefficients for total units sold and profitabil-
ity of units sold, were .09 and .12 respectively. These
coefficients were also tested and found not statistically

116

significant at the .05 level of significance. Therefore,

 

llsTotal units sold, p = .53; profitability of units
sold, p = .48; .05 (00 or less is necessary to show that
the two variables are significantly related.

116Total units sold, p = .56; profitability of units
sold, p = .53; .05 (00 or less is necessary to show that
the two variables are significantly related.

137

TABLE 18

KENDALL RANK CORRELATION ANALYSIS:

(Based on an average n of 4)

SALES PERFORMANCE WITH
PRODUCT KNOWLEDGE

Dependent Variables

 

Sales
Performance:
(Total Units

Independent Variables Sold)

Sales
Performance:
(Profitability
of Units Sold)

 

Product Knowledge
Questionnaire

 

 

Section I .13 .20
Product Knowledge

Questionnaire

Section II .09 .12

 

 

 

 

138

it is suggested that there is no relationship between a
salesman's knowledge of consumer benefits and his level
of sales performance.

These findings indicate that when product knowledge
is divided into knowledge of product features and know-
ledge of consumer benefits, each of these variables does
not significantly correlate with a salesman's sales effec-
tiveness. Therefore, the hypothesis that each of these
product knowledge components contributes equally to sales
performance is rejected.

The conclusion that neither knowledge of product
features or knowledge of consumer benefits is significantly
related to the salesmen's sales effectiveness is understand-
able in reference to the previously discussed relationship
between product knowledge and sales performance as deter-
mined by the intra-dealership method of data analysis.

The nonparametric analysis showed that product knowledge

and sales performance are not significantly related. There—
fore, the subdivisions of product knowledge also would not
be significantly related to sales effectiveness unless the
interaction between the two variables, knowledge of product
features and knowledge of consumer benefits, cancelled out

one another when the two variables are combined.

CHAPTER VI
CONCLUSIONS

One purpose of this chapter is to present in summary
form the conclusions derived from this study. Where appro-
priate, selected implications of the findings are presented.
The results from both the inter- and intra-dealership analyses
are noted along with a brief discussion of the differences
in the findings.

Another objective of this chapter is to present methods
by which this study might be improved and to suggest comple-
mentary research projects which could investigate further
the factors that contribute to'a salesman's sales performance

and his level of product knowledge.

Summary g£_Findings

1. Product knowledge, sales techniques, and personal-
ity were shown to have a relationship with sales performance
on the basis of the inter-dealership analysis. However, on
the basis of the intra—dealership analysis, these three fac-
tors were shown to have little relationship to sales performance.

In the intra-dealership analysis, no statistically
significant relationship was found among product knowledge,
sales techniques, and personality with sales effectiveness.

Even though the inter-dealership correlation analysis
showed that product knowledge, sales techniques, and personal—

ity correlated with sales performance, the combined influence

139

140

of these three variables explains approximately 14 percent

of the unit variance of the salesmen's sales performance.
Consequently, since the other 86 percent is unaccounted for

by these three variables, they are not shown to be very im-
portant to sales performance, at least in this investigation.
However, it should be noted that of these three factors, prod-
uct knowledge accounts for about eight percent of the unit
variance of sales performance while sales techniques explains
only two percent and personality was shown to have practically‘
no influence upon sales effectiveness when treated as the

only independent variable. The hypothesis that product know-
ledge (in conjunction with two other assumed determinants,
personality and sales techniques) is one of the primary deter-
minants of sales effectiveness is not strongly supported.

2. Personality was not shown to be related to sales
performance in either the multi-point or the single point
dealership classifications. Sales techniques, on the other
hand, was shown to have a greater influence upon sales per-
formance in the single point dealerships than in the multi—
point dealership classification. It should be pointed out
that sales techniques, when treated as the only independent
variable in the single point dealerships, accounted for
about nine percent of the unit variance in sales perform—
ance. These findings were based upon the inter-dealership
analysis.

One implication of these findings is that the sales—

men's sales techniques attribute varies according to the

141

type of market in which they are employed. A related im-
plication is that this variability might be established

in the hiring of new salesmen in the two types of markets;
i.e., different emphasis might be placed upon the importance
of the applicants' sales techniques attribute.

3. The relationship between a salesman's level
of product knowledge and his sales effectiveness is inde-
pendent of the type of market served. This conclusion was
reached by testing hypothesis 2A by means of both the intra—
and inter-dealership analysis.

Based upon the intra-dealership analysis, it was
found that there was not a statistically significant rela-
tionship between product knowledge and sales performance
in either type of market. On the other hand, the inter—
dealership analysis showed that there was a statistically
significant relationship between product knowledge and sales
performance in both the multi-point and the single point
dealership classifications. These relationships, however,
were not high enough to have practical significance. Prod-
uct knowledge, when treated as the only independent vari-
able, accounted for only four percent of the unit variance
in the salesmen's sales performance in either type of
dealership.

This conclusion, therefore, that the relationship
between a salesman's level of product knowledge and his
sales effectiveness is independent of the type of market

served suggests that it is not necessary to adapt either

142

product knowledge promotional literature or the product
knowledge segment of sales training programs according to
the type of market served.

4. The relationship between a salesman's level
of product knowledge and his sales effectiveness is some-
what dependent upon the size of the retail outlet within
the multi-point dealership classification, according to
the inter-dealership method of data analysis. However,
the intra—dealership analysis indicated that the salesman's
product knowledge—-sales performance relationship is inde-
pendent of dealership size.

Even though the interedealership analysis showed
that the relationship between product knowledge and sales
performance is somewhat dependent upon dealership size,
the findings were contradictory in that the means by which
product knowledge was measured influenced the results.
The analysis showed, for example, that the relationship
between product knowledge, as measured by the questionnaire,
and both measures of sales effectiveness is greater in
multi-point low volume dealerships than in multi-point
high volume dealerships. When product knowledge was meas—
ured by the paired comparisons, however, there was no prac—
tical difference in the product knowledge-n-sales perform-
ance relationship in either the multi-point high volume
or the multi-point low volume classifications.

These findings would indicate that it may be desir-

able to adapt product literature and/or the product knowledge

143

segment of sales training courses to the size of the retail
outlet within the multi-point dealership classification.

5. The relationship between a salesman's level
of product knowledge and sales effectiveness is independent
of the size of the retail outlet within the single point
dealership classification according to both the intra- and
the inter-dealership methods of data analyses. This find-
ing was determined as a result of testing hypothesis 28.
The implication is that it is not necessary to adapt prod-
uct literature and/or sales training programs to the size
of the retail outlet within the single point dealership
classification.

6. The salesman's level of product knowledge is
independent of the type of market in which he sells. In
other words, the level of product knowledge of salesmen
in large metropolitan markets is not significantly differ—
ent from the level of product knowledge of salesmen in less
densely populated areas.

7. The salesman's level of product knowledge is
independent of the size of the retail outlet in which he
sells in both the multi-point and the single point types
of markets. In other words, the level of product knowledge
of salesmen in dealerships with high planning volumes is
not significantly different from the level of product know—
ledge of salesmen in dealerships with relatively low plan-
ning volumes.

This finding was determined by testing hypothesis

144

3B in both the multi-point and the single point dealership
classifications.

This conclusion implies that either there is no
difference in emphasis placed upon the acquisition and im-
portance of product knowledge or there is no difference
in the need for product knowledge in the two size subgroup—
ings of both multi-point and single point dealerships.

8. Amount of formal education, sales experience
within the industry, and attendance at a manufacturer-
sponsored training program have no statistically signif-
icant influence upon a salesman's leVel of product know—
ledge, at least on the basis of the methods used to measure
these variables. These findings were determined from the
results of testing hypotheses 3C, 3D, and 3E.

The findings imply that those in charge of super—
vising the sales force should emphasize the importance and
acquisition of product knowledge with equal importance to
the inexperienced salesman, as well as those with prior
selling experience, and to lower educated salesmen as well
as those with higher levels of education.

9. When product knowledge is divided into know-
ledge of product features and knowledge of consumer bene-
fits, neither of these components of product knowledge has
a statistically significant relationship with sales effec-

tiveness, based upon the intra-dealership analysis.

145

Recommendations for Improving this Study_

 

The means of selecting the sample, controlling the
testing environment through the use of breakfast meetings,
employing company zone managers as study administrators,
and the mechanics of administration (such as the cover
letter, sheet of instructions, and the use of self-addressed
postpaid return envelopes) contributed immeasurably to the
control of extraneous factors, and to the high percentage
of usable returns (86 percent). Since the returned ques-
tionnaires generally were completely filled out, it would
appear that the salesmen were able to understand the ques-
tionnaire instructions and questions and had sufficient
time to respond. From an examination of the paired com-
parison matrices, it appears that the sales managers under-
stood the instructions and were willing to fill out the
rating schedule.

The measures of the variables product knowledge,
sales techniques, personality, and sales performance, how-
ever, may not have been completely reliable.117 There was

not perfect agreement between the two measures of product

knowledge, the questionnaire and the paired comparison

 

117It should be pointed out that the instruments
used to measure product knowledge, personality, and sales
techniques were the best available under the circumstances;
i.e., the time and cooperation required of the salesmen,
sales managers, and zone managers; the financial resources
available; the geographic dispersion of the sales force;
and the size of the research staff.

146

matrix. This implies that the respective reliabilities
of one or both of the measures of product knowledge is less

than unity.

Reliability and validityg£_the Questionnaire

 

In general, reliability of a test is a function
of its length. Administrative restraints dictated, limit-
ing the questionnaire to the number of questions included.
The Spearman-Brown formula could be applied to find out
how long it would need to be for a reliability of a given
value.

As to validity, this might be determined by first
administering the questionnaire to a group of salesmen and
then conducting in-depth interviews with the same men.
Then a comparison of their product knowledge as obtained
in the interviews to that yielded by scores on the ques-

tionnaire could be made.

Reliability g: the paired comparison matrices

A method of analyzing the reliability of product
knowledge, personality, and sales techniques, as measured
by the paired comparison matrices, is by checking the con-
sistency of the sales managers' preference judgments.

The consistency of the managers' judgments may be
evaluated by a method developed by M. G. Kendall and
Babington Smith defined as the coefficient of consistence

(C) whereby circular triads are established to analyze the

147

consistency of each series of preferences.118

A simple explanation for inconsistencies in paired
comparisons is that the judge is at least partially guess-
ing when declaring his preferences. This situation may be
caused by the evaluator's incompetence or because the fac-
tors under consideration are in fact very similar.

Another explanation is that there may be no valid
ordering of the factors even when they differ markedly.
Their merit may depend on more than one characteristic.
This forces the evaluator to mentally construct some func-
tion of the relevant characteristic and use this as a basis
of comparison. Therefore, for each additional individual
to be evaluated, there is a chance that the evaluator's
scale of relevant characteristics may be altered, thus,

a new basis for comparison is created.

There is no method of measuring sales performance
which is comprehensive, completely reliable and free from
defects. Therefore, there is no means of evaluating the
validity of the two measures of sales performance employed
in this study; total units sold and profitability of units
sold. They must be accepted as given. The accuracy of
this study concerning specified influences which affect

sales success could be increased if better sales performance

 

118For a further elaboration of this method of
analysis, consult H. A. David, Th§_Method gf Paired Com-
parisons (New York: Hafner Publishing Co., 1963), pp.
21-24. See Appendix Q for the formulas required to con—

duct this test.

 

148

measurements are perfected. Accurate measurement of the
effectiveness of the sales force is very difficult because
of the inexact nature of the salesman's job and the many
uncontrollable variables which affect performance. But
the measurement of salesmen's performance is one area in

which research is greatly needed.

Recommendations for Replicating this Study

After elaborating upon the present study by deter-
mining the reliability of the data collecting instruments
and making any necessary adjustments in them, the study,
could be replicated.

In order to determine if the conclusions reached
by this study of selected automobile salesmen are applicable
to the automobile industry as a whole, it might be desirable
to replicate the study in the other division of the same
company (Ford Division) or in competing automobile companies.

An even more critical area of research is to deter-
mine whether the relationship between product knowledge and
sales effectiveness is dependent upon the type of product
sold. It should be ascertained whether the relationship
between a salesman's knowledge of product features and their
corresponding consumer benefits and sales performance is
influenced by the nature and/or complexity of the product
being sold. This information could be obtained by repli—
cating this study in a number of industries which produce

both durable and nondurable goods, such as household

149

appliances and dairy products. The results of these studies
could then be compared with those derived from this study
of automobile salesmen.

Another area of investigation which would broaden
our understanding of the importance of product knOwledge
in its relationship with sales performance concerns the
influence of time upon this relationship. It would be in-
teresting to know if product knowledge is becoming more
or less significantly related with sales effectiveness.
One means by which this could be accomplished is by repeat-

ing this study at predetermined time intervals.119

Recommendations for Complementary
Research Projpcts

Since there are innumerable aspects of personal
selling which require research, only those specifically
related to this study will be mentioned.
Methods g£_measuripgthe product knowledgg -
sales performance relationship_

It may be advantageous to investigate the relation-
ship between a salesman's product knowledge and his sales
effectiveness by using different methods of measuring these

two variables.

1191f the product knowledge-~8ales Performance re-

lationship is found to be dependent upon time, a corollary
investigation to attempt to explain this phenomenon could
be undertaken to determine if the emphasis on product fea-
tures and/or the corresponding consumer benefits by automo-
bile advertising is also influenced by time.

150

One method by which a salesman's product knowledge
can be measured is by determining the amount of time a
salesman devotes to obtaining product knowledge.120 For
example, the amount of time a salesman spends reading prod-
uct information literature, reviewing the new models, and
discussing product features with other salesmen could be
measured either by (1) observation of the salesman's use
of his time or (2) self-timing of these factors by the
participants. Either of these measures of the salesman's
time devoted to the acquisition of product knowledge could
be correlated with a measurement of sales performance to
determine the product knowledge —-sa1es effectiveness re—
lationship.

Other techniques for measuring salesmen's product
knowledge include: the recording on tape of their sales
presentations to various prospective buyers, to obtain the
customers' evaluations of the salesmen's product knowledge,
and to employ trained researchers to conduct depth inter—
views with the selected salesmen.

On the other hand, sales performance may be meas-
ured by a number of other methods, such as sales volume

as compared to a predetermined quota, salesman's income,

 

120This time study method of analysis could be em-

ployed to study not only the acquisition of product know-
ledge but all facets of a salesman's activities. These
would include, for example, time spent delivering sales
presentations to consumers, prospecting new customers, per-
forming clerical work, and developing salesmanship skills.

151

ratings by supervisors, or actual sales as compared to
potential sales.121
An even more involved method of studying this re-
lationship would be to reassign salesmen of one product
to a completely different product line, while keeping as
many other variables constant as possible. If this method
of research was properly conducted, product knowledge would
be the only variable essentially affected by the reassign-
ments. Thus, measurement of the change in the salesmen's
performance as a result of the change in their knowledge
of the product would be permitted.

Methods 23 measuring_the personality -
sales performance relationship

 

Personality could be measured by in-depth inter-
views of the salesmen by psychologists, observations by
trained researchers of salesmen in selling situations, the
use of personality tests, customers' opinions of the sales-
men's personalities, and/or an enlargement of the paired
comparisons method. The sales managers could be requested
to complete paired comparison matrices of their salesmen
for many, separate personality traits.

Any of these other methods of personality measure-
ment could then be correlated to a measure of sales per-

formance to determine the relationship between a salesman's

 

121The disadvantages of these and other measure-
ments of sales performance are discussed in Chapter II.

152
personality and his sales effectiveness.

Methods 2: measuring sales techniques -
sales performance relationship

 

Sales techniques could be measured by the develop-
ment of a sales techniques questionnaire, in—depth inter-
views of salesmen by trained researchers, tape recordings
and/or observation of selected salesmen in selling situa-
tions, measurement of the salesmen's prospects files (for
example, the number of prospects recorded), and measurement
of the salesmen's ability to appraise selected used cars.

Any of these methods of measuring a salesman's sales
techniques could be correlated with a measure of sales per-
formance to investigate the salesman's sales techniques--
sales performance relationship.

The personality, product knowledgg, sales
techniques--sales performance relationship

 

Another area of research which should be recommended
is an investigation of personality, product knowledge, and
sales techniques as one contributing factor to a salesman's
sales performance. It is possible that by attempting to
isolate and study each of these factors as independent vari-
ables, even their combined influence upon sales effective-
ness is distorted. Therefore, a study should be undertaken
to investigate the effect of these three elements acting
as one influence rather than as three combined factors upon

salesmen's sales performance.

153

Other factors influencing_sales performance

 

Inasmuch as the findings of this study suggest that
product knowledge, personality, and sales techniques explain
only 14 percent of the unit variance of the salesmen's sales
effectiveness, it is important to investigate other factors,
such as work organization, good health, resourcefulness,
personal motivation, and ambition which may influence sales-

men's sales performance.

Other factors influencing product knowledge

In this study it was found that amount of formal
education, attendance at a manufacturer-sponsored training
program, and sales experience within the industry do not
influence a salesman's level of product knowledge. It is
important, therefore, to study the influence of other fac—
tors upon a salesman's knowledge of the product; such as
reading skills, mechanical aptitude, age, the number of
other salesmen with whom a salesman associates, and the
attitude of the sales manager towards the importance of

product knowledge.

SELECTED BIBLIOGRAPHY

BOOKS

Alexander, Milton, and Mazze, Edward M. Sales Management
Theory_and Practice. New York: Pitman Publishing
Co., 1965.

 

Aspley, John Cameron, ed. The Sales Manager's Handbook.
8th ed. Chicago: The Dartnell Corp., 1959.

Baker, Jr., Richard M., and Phifer, Gregg. Salesmanship:
Communication, Persuasion, Perception. Boston:
Allyn and Bacon, Inc., 1966.

Bender, James F. How £9 Sell Well: The Art and Science
2: Professional Salesmanship, New York: McGraw—
Hill Book Co., Inc., 1961.

 

Breen, George; Thompson, Ralph Burnham; and West, Harry.
Effective Selling. New York: Harper & Brothers,
1950. ‘

Buskirk, Richard. Principles g£_Marketing: The Management
View. New York: Holt, Rinehart, & Winston, Inc.,
1961.

 

Canfield, Bertrand R. Sales Administration, Principles
and Problems. 4th ed. Englewood Cliffs: Prentice-
Hall, Inc., 1961.

 

Canfield, Bertrand R. Salesmanship Practices and Problems.
3rd ed. New York: McGraw-Hill Book Co., Inc.,
1958.

Cash, Harold C., and Crissy, William J. E. 222.222 2:
Appeals i3 Selling, Vol. 2 of Th2 Psychology g:
Selling Series. Flushing: Personnel Development
Associates, 1957.

Clewett, Richard M. Marketing Channels for Manufactured
Products. Homewood: Richard D. Irwin, Inc., 1954.

Duddy, Edward A., and Revzan, David A. Marketing: An

Institutional Approach. New York: McGraw-HIIl
Book Co., Inc., 1953.

154

155

Francisco, L. Mercer. The More You Show the More You Sell.
Englewood Cliffs: Prentice-Hall, Inc., 1960.

 

Gross, Alfred. Salesmanship Principles and Practices of
Professional Selling. New York: The Ronald Press
Co., 1952.

Haas, Harold M. A Short Course Ln Salesmanship. New York:
Prentice-Hall, Inc., 1939.

 

 

Haas, Kenneth B. How to Develgp Successful Salesmen. New
York: McGraw-Hill Book Co., Inc., 1957.

 

Haas, Kenneth B., and Perry, Enos C. Sales Horizons.
Englewood Cliffs: Prentice-Hall, Inc., 1958.

 

Hauk, James G. "Research in Personal Selling." Science
Ln Marketing. Edited by George Schwartz. Vol. 3
of the Wiley Marketing Series, ed. by William Lazer.
New York: John Wiley & Sons, Inc., 1965.

 

 

Husband, Richard W. The Psychology_g£ Successful Selling.
New York: Harper and Brothers Publishers, 1953.

 

Ivey, Paul; Horvath, Walter; and Tonning, Wayland. Success-
ful Salesmanship. 4th ed. Englewood Cliffs:
Prentice—Hall, Inc., 1961.

 

Kirkpatrick, Charles Atkinson. Salesmanship: Helpipg_Pros—
pects Buy. 4th ed. Cincinnati: South-Western
Publishing Co., 1966.

 

 

Lapp, Charles L. Personal Supervision of Outside Salesmen
(Making Salesmen More Productiwey. Columbus: The
Bureau of Business Research, Ohio State University,
1951.

 

Lapp, Charles L. Training and Supervising Salesmen. Engle-
wood Cliffs: Prentice-Hall, Inc., 1960.

 

 

Lapp, Charles. Successful Selli_g Strategies: How to Climb
the Ladder to Sales Success. Vol. in McGraw-Hill
Series Ln Advertising and Selling, ed. by Steuart H.
Britt. New York: McGraw-Hill Book Co., Inc., 1957.

 

Lazo, Hector, and Corbin, Arnold. Management Ln Marketing.
New York: McGraw-Hill Book Co., Inc., 1961.

 

Meloan, Taylor W., and Rathmell, John M., ed. Selling:
Its Broader Dimensions. Vol. in Macmillan Market-
ing_Books, ed. by Schuyler F. Otteson. New York:
The Macmillan Co., 1960.

 

156

Pashigian, Bedros Peter. The Distribution of Automobiles,
An Economic Analysis of the Franchise System. Engle-
wood Cliffs: Prentice- Hall, Inc., 1961.

 

 

 

Pederson, Carlton, and Wright, Milburn. Salesmanship Prin-
ciples and Methods. 4th ed. Homewood: Richard D.
Irwin, Inc., 1966.

 

 

Russell, Frederick A., and Beach, Frank H. Textbook of
Salesmanship. 6th ed. New York: McGraw-Hill Book
Co., Inc., 1959.

 

Russell, Frederick; Beach, Frank; and Buskirk, Richard.
Textbook of Salesmanship. 7th ed. New York:
McGraw-Hill Book Co., Inc., 1959.

 

Shaw, Steven J., and Thompson, Joseph W. Salesmanship:
Modern Viewpoints gg Personal Communication. New
York: Henry Holt & Co., Inc., 1960.

 

 

Small, Richard Loring. Salesmanship. New York: The Mac—
millan Co., 1952.

 

Smith, Wendell R. "Role of Selling in Modern Marketing."
Emerging Concepts ig_Marketing. Edited by William S.
Decker. Chicago: American Marketing Association,
1963.

 

Stroh, Thomas F. Salesmanship: Personal Communication
and Persuasion Ln Marketing. Homewood: Richard D.
Irwin, Inc., 1966.

 

 

 

Thompson, Joseph W. Selling. A Behavioral Science Approach.
New York: McGraw—Hill Book Co., 1966.

Thompson, Willard M. Salesmanship: Concepts, Managgment,
and Strategy. New York: John Wiley & Sons, Inc.,
1963.

 

Tonning, Wayland A. How Eg Measure and Evaluate Salesmen's
Performance. Englewood Cliffs: Prentice-Hall,
Inc., 1954.

 

 

 

Tosdal, Harry R. Selling Ln Our Economy: An Economic and
Social Analy_is of Selling and Advertising. Home-
wood: Richard D. —Irwin, Inc., 1957.

 

 

157

REPORTS - PUBLISHED

Ehlers, Carrol W. The Use of Psychological Tests in Select-
ing Salesmen in the South, Public Research— Paper,
No. 18. Georgia State College of Business Adminis-
tration, July, 1960.

 

 

Umemura, G. M. Measuring Salesmen's Performance. Studies
in Business Policy #79. New York: National Indus-
trial Conference Board, 1956.

Vizza, Robert F. Measuring the Value of the Field Sales
Force. A Research Report. The Sales Executives
Club of New York, 1963.

 

PERIODICALS

Baier, Donald E., and Dugan, Robert D. "Tests and Perform-
ance in a Sales Organization," Personnel Psychology,
9 (Spring, 1956), 17-26.

 

 

Baier, Donald E., and Dugan, Robert D. "Factors in Sales
Success," Journal 2£_Applied Psychology, XLI (1957),
3 7-40 0

Belasco, James A. "The Salesman's Role Revisited," Journal

of Marketing, XXX (April, 1966), 6-ll.

 

Bell, Gerald D. "Self-Confidence and Persuasion in Car
Buying," Journal 2£_Marketing, IV (February, 1967),
46-52.

Benge, Eugene J. "What Traits and Work Habits Characterize
Successful Salesmen," Sales Management (July 15,
1956), 54-56.

Conant, Earle R. "The Customer's Side of the Street,"
Sales Marketing Today, IX (December, 1964), 15—16.

 

Crissy, William J. E., and Lapp, Charles L. "Sound Selec-
tion: First Step in Building an Effective Sales
Force," Advanced Management, 25 (March, 1960), 6-10.

Ditz, Gerhard W. "Status Problems of the Salesman," Bus—
iness Topics, 15 (Winter, 1967), 68— 80.

Drucker, Peter F. "How to Double Your Sales," Nation's
Business, 55 (March, 1967), 80—86.

 

Easton, Allan. "A Forward Step in Performance Evaluation."
Journal of Marketing, XXX (July, 1966), 26-32.

158

Francisco, L. Mercer. "How Can You Get Salesmen to Sell
Today?" Sales Management, 79 (December 6, 1957),
62-68.

French, Cecil L. "Correlates of Success in Retail Selling,"
American Journal g£_Sociology, LXVI (July, 1960),

 

Frey, John M. "The Missing Ingredients in Sales Training,"
Harvard Business Review, XXXIII (November-December,
1955), 126-132.

 

Harrell, Thomas W. "The Relation of Test Scores to Sales
Criteria," Personnel Psychology, 13 (1960), 65-69.

Kennedy, J. E. "General Device versus Specific Devices
for Selecting Car Salesmen," Journal g£_Applied
Psychology, XLII (June, 1958), 206-209.

 

Kirchner, Wayne K. "Predicting Ratings of Sales Success
with Objective Performance Information," Journal
2; Applied Psychology, XLIV (December, 1960), 398-
403.

Kirchner, Wayne K., and Dunnette, Marvin D. "Successful
Salesman as He Sees Himself," Personnel, 35
(November, 1958), 67-70.

Kirchner, Wayne K.; McElwain, Carolyn 8.; and Dunnette,
M. D. "A Note of the Relationship Between Age and
Sales Effectiveness," Journal g£_App1ied Psycholggy,
XLIV (April, 1960), 92-95.

Kline, Charles H. "The Strategy of Product Policy," Har-
varg Business Review, XXXIII (July-August, 1955),
91-1000

 

Mason, John L. "The Low Prestige of Personal Selling,"
Journal 9£_Marketing, XXIX (October, 1965), 7-10.

Mayer, David, and Greenberg, Herbert M. "What Makes a
Good Salesman," Harvard Business Review, 42 (July-
August, 1964), 119-124.

McMurry, Robert N. "How to Win or Lose Sales at the Point
of Purchase," Journal 2; Marketigg, XXIV (July,
1959), 41-49.

McMurry, Robert. “The Mystique of a Super Salesmanship,"
Harvard Business Review, XXXIX (March-April, 1961),

159

Miner, John B. "Personality and Ability Factors in Sales
Performance," Journal g£_Applied Psycholggy, XLVI
(February, 1962), 6-13.

Miracle, Gordon E. "Product Characteristics and Marketing
Strategy," Journal g£_Marketing, XXIX (January,
1965), 18-24.

 

 

Montant, Jr., Louis T. "21 Yardsticks to Measure Perform-
ance of Salesmen," Printer's Ink (July 22, 1949),
24-26 0

Russell, Jack. "A System of Sales Analysis Using Internal
Company Records," Journal g£_Marketigg, XIV (April,
1950), 675-690.

Spencer, John A. "What Makes a Buoyant, Productive Sales
Force?" Sales Management, LXXXII (February 6, 1959),

 

 

106-112.

Stewart, John B. "Functional Features in Product Strategy,"
Haryard Business Review, XXXVII (March-April, 1959),
65-78 0

Thayer, Paul W.; Antoinetti, John A.; and Guest, Theodore A.
"Product Knowledge and Performance—-A Study of Life

Insurance Agents," Personnel Psychology, 11 (Autumn,
1958), 411-418.

 

Tobin, William J. "What Makes a Man a Successful Salesman?"
Sales Management, 83 (October 2, 1959), 121-124.

Tobolski, Francis 8., and Kerr, William A. "Predictive
Value of the Empathy Test in Automobile Salesman-
ship," Journal g£_Applied Psychology, 36 (1952),
310-311.

 

"Ways to Get Salesmen to Study Product Knowledge," Indus-
trial Marketing, 42 (September, 1957), 192-194.

 

"What Buyers Expect from Salesmen," Sales Management, 82
(April 17, 1959), 96-98.

Wolff, Janet. "New Directions in Marketing: Answering
Woman's Hunger for Information," Sales Manggement,
83 (November 10, 1959), 23-26.

 

160

UNPUBLISHED MATERIALS

Goldway, Elliott M. "A Survey of Current Company Practices
in the Use of Psychological Tests for the Selection
of Salesmen." Unpublished Ph.D. dissertation, New
York University, 1955.

Kennedy, James Edward. "Validities of a Personal History
Form for Automobile Salesmen in General Compared
with Subvarities of Automobile Salesmen." Unpub-
lished Ph.D. dissertation, Michigan State Univer-
sity, 1956.

Appendix A

INTRODUCTORY LETTER TO THE GdNERAL FIELD MANAGER
(EWW LINCOLN-MERCURY DIVISION

lntra-Company Communication GENERAL OFFICE

 

JUly 2?} 1967

SUbject: 1257 Product Information Study

The Iincoln-Mercury Division is conducting a study of selected
LincolnéMercury dealers to determine the Division's effectiveness in
presenting product information to the dealer organization and to learn
more about the role of product knowledge in the selling effort.

As you know, each year a large percentage of the Division's financial
and personnel resources are devoted to the development of informational
and promotional literature concerning its products. We would like to
determine, by means of this study, the extent to which product knowledge
is employed by Lincoln-Mercury salesmen in order to reinforce or improve
our present system.

You are requested to arrange a series of breakfast meetings with new
car sales managers and full-time new car salesmen of the following
dealerships who have been selected to participate in this study:

161

Appendix A (cont.)

H
O“.
R)

- 2 — JUly 27: 1967

The salesmen should be asked to complete the enclosed salesmen's questionnaire
which will take about twenty-five minutes. Since the questions are based
upon the Mercury, Mercury Intermediate, and Cougar car lines, exclusive
Lincoln Continental salesmen should not be asked to participate. The sales-
men should complete the questionnaire without the use of any reference
materials or assistance from other salesmen.

While the salesmen are completing the questionnaires, the sales manager
should be asked to fill out the enclosed rating charts for the salesmen
participating in the study. The information received from the completed
charts will help us understand more fully the factors which determine
success in automobile selling.

The information received from both the questionnaires and rating charts will
be coded so that individual salesmen and dealerships are not identified.

Upon the completion of the questionnaires and rating charts, they should be
enclosed in the apprOpriate post-paid envelopes and returned directly to
the Division.

It is requested that you schedule meetings so as to permit return of all
questionnaires and rating charts as quickly as possible. Expenses associated
with conducting breakfast meetings should be limited to $2.75 per participant
including tax and tip and should be charged to General Office Account Number
25A-2228-226.

We believe the results of this study will guide us in our goal of better
assisting our dealer organization,and appreciate your assistance in this
regard.

  

,7 57 {1?an K‘ngr)
Concurgz " .

2 Ai/d-‘(lé/gj7fl . R: J. @gert \ \
All ' Marketing Research Manager
D. L. Galloway

Enclosure

T0:

1.

2.

Appendix B

INSTRUCTION SHEET TO THE ZONE MANAGER

INSTRUCTIONS

ZONE.MANAGERS PARTICIPATING IN THE 1967 PRODUCT INFORMATION STUDY

Arrange a breakfast meeting(s) with the new car sales er(s)
and full-time new car salesmen fer the selected dealershipis) in
your territory. Emclusive Idncoln Continental salesmen should
not be asked to participate.

Following breakfast, ask the full-time new car salesmen to
complete the enclosed questionnaires.

Please remind the salesmen participating to use no reference
material or assistance from other salesmen while completing
the questionnaires and to leave no questions blank.

While the salesmen are completing the questionnaires, ask the
Sales Manager to fill out the enclosed rating charts for the
salesmen participating in the study.

The suggested time for the completion of the questionnaires
and rating charts is approximately twenty-five minutes.

As you collect the questionnaires, please check to see that
all questions on the last page of the questionnaire are
answered.

When the questionnaires and rating charts are completed, they
should be enclosed in the appr0priate post—paid envelOpe and
returned directly to the Division.

ExPenses associated with conducting the breakfast meeting(s)
should'be limited to $2.75 per participant including tax and
tip and should be charged to General Office Account Number
25A-2228-226.

Thank you for your c00peration. We believe the results of
this study will guide us in our goal of better assisting
our dealer organization and appreciate your assistance in
this regard.

163

Appendix C

QUESTIONI‘JAIRB

EVALUATION OF 1967 PRODUCT INFORMATION

SECTION I MULTIPLE CHOICE QUESTIONS

DIRECTIONS: Select the CORRECT answer after each statement by

1.

3.

5.

The

A.
B.
C.
D.
The
A.

B.

C.
D.

circling the letter A. B. C. or D. Only one answer
is correct for each statement.

Remote Control Mirror is:

Optional on the Mercury Cougar

Standard on the Mercury Monterey and Commuter
Standard on all Mercury Intermediate series
Standard on the Mercury Montclair and Marquis

Select-Shift Merc-O-Matic is:

Standard on the Mercury Caliente

Standard on the Mercury Cougar XR?

Standard on the Mercury Park Lane. Brougham and Marquis series
Standard on the Mercury Cougar

Power Disc Brakes are standard on the:

A.
B.
C.
D.
The
A.

B.

C.
D.

Mercury Montclair and Monterey
Mercury Park Lane and Brougham
Mercury Montclair and Brougham
Mercury Commuter and Colony Park

Dual-Action Tailgate is optional on the:

Mercury Commuter and Villager
Mercury Villager and Colony Park
Mercury Colony Park and Commuter
Mercury Comet Voyager

Comfort-Stream Ventilation and Heating System is:
Standard on Mercury Park Lane and Marquis 2 door hardtops
Optional on Mercury Park Lane and Brougham 4 door hardtops

Optional on Mercury Colony Park and Commuter
Standard on Mercury Park Lane and Brougham 4 door sedans

164

7.

10.

11.

The

The

A.
B.
C.
D.

The

165
Appendix C (cont.)

Automatic Parking Brake developed for the full-size Mercury:

Releases the brake as soon as the ignition switch is turned on.
Is engaged automatically when the transmission selector is
in Neutral or Park

Is designed to eliminate accidental use of the parking brake
Can be engaged while the transmission selector is in Drive
so the transmission fluid level can be checked

Mercury Breezeway Ventilation System:

Sends inside air out through door outlets

Can be used in bad weather

Has a lighted indicator to show when the system is in use
Permits the rear window to be opened six inches

Cougar Tilt Away Steering Wheel:

Is adjustable to five different positions

Is adjustable to the right or left as desired

Swings into position automatically when the door is closed
Is adjustable by pushing the turn signal lever forward to
release lock

suspended Accelerator Pedal is:

Designed to adjust to the driver's foot size

Dasigned to move to adjust to the driver's foot angle
Suspended from the instrument panel

Located one inch above the floor

Mercury and Mercury Intermediate header mounted Day/Night

Inside Mirror is:

Made of shatter-proof glass

Designed to remain stationary upon impact

Designed to prevent the mirror from shattering

Installed on hardtops and convertibles with double-pivot arms

The Lane Changing Signal feature of the turn indicator:

DOPHP

Automatically turns on as the driver begins to change lanes
Cancels when pressure is removed from the signal lever

Is designed to substitute for the emergency flasher

Is turned on by pulling out the turn indicator lever

1125

13.

1h.

15.

The
A.

C.
D.

The

A.
B.

D.
The
A.
B.
C.
D.
The
A.

C.
D.

166
Appendix C (cont.)

Mercury Automatic Speed Control System:

Is operated by a button at the end Of the transmission
selector

Can override the car's braking system

Automatically adjusts the speed of the vehicle for
winding roads

Automatically adjusts the speed of the vehicle for
hilly roads

Whisper-Airs Conditioner:

Includes door pressure relief valves on the full-size Mercury
Replaces manually Operated fresh air vents

Is not equipped with shut-off valves

Uses a 2-speed blower to circulate cool air

full-size Mercury and Cougar AM Radio/Stereo-Sonic Tape Player:

Is completely transistorized

Has the tape player unit mounted under the instrument panel
Can only be factory installed

Is installed with two speakers in the rear

Safety Door Latches do not have:

A self-centering interlocking design

The lock striker pin mounted in rubber to cushion closing
A method of lubricating themselves

Double yokes to clamp over the striker pin

167
Appendix C (cont.)

SECTION II PART 1

DIRECTIONS: The following statements list one consumer benefit

for each feature. Please list in the space provided
other benefits you have found helpful in selling
these features.

The Dual-Action Tailgate for station wagons gives the consumer
the benefit of:

Easier entrance to the third seat
What other benefits do you find helpful in selling this feature?

 

 

The Cougar Tilt Away Steering Wheel gives the consumer the
benefit of:

The steering column adjusts to provide maximum entrance and
exit room

What other benefits do you find helpful in selling this feature?

 

The Stereo-Sonic Tape Player gives the consumer the benefit of:
Uninterrupted listening
What other benefits do you find helpful in selling this feature?

 

 

The Remote Control Mirror gives the consumer the benefit Of:
Adjustable from inside the car when the driver is in his seat

What other benefits do you find helpful in selling this feature?

 

 

168
Appendix C (cont.)

The Emergency Flasher System gives the consumer the benefit of:
Located conveniently on the steering wheel column

‘Jhat other benefits do you find helpful in selling this feature?

 

 

The Select-Shift Merc-O-Matic gives the consumer the benefit of:

.Allows the driver to up-shift and down-shift manually or go
fully automatic

What other benefits do you find helpful in selling this feature?

 

The new Upholstery Fabrics give the consumer the benefit of:

Improved soil resistance

What other benefits do you find helpful in selling this feature?

 

 

169
Appendix C (cont.)

SECTION II PART II MULTIPLE CHOICE QUESTIONS

DIRECTIONS: Select the CORRECT answer after each statement by

9.

ll.

12.

circling the letter A. B. C. or D. Only Qge answer
is correct for each statement.

The Mercury Breezeway Ventilation System permits:

A.
B.

C.
D.

The rear window to remain clear at all times due to the
overhanging roof panel

Outside air to be brought into the cowl air intake and
after circulation to exit through the backlite opening

The owner to direct the air outlets in the desired direction
Removal of smoke with all the windows closed

Whisper-Airs Conditioning is beneficial because it:

The

A.
B.
C.
D.

The
A.
B.
C.
D.
The
A.

C.
D.

Removes irritating dust. pollen and other airborn
irritants from the air

Removes exhaust fumes from the air as it enters the car
Automatically shuts off at the desired temperature

Allows the consumer to choose a combination of outside and
recirculated air

new Synchronized Windshield Washer provides:

Better cleaning due tO the larger area covered

Automatic release of the fluid when the wipers are turned On
Cleaning on all 3 wiper operating speeds

Use Of less fluid due to the smaller area covered

196? Door Latches are designed to:

Prevent the doors from freezing and/or sticking

Make it easier to tell when the door is not properly closed
Hold the door securely even when it is partially closed
Prevent accidental injury while the door is being closed

header mounted Day/Night Inside Mirror is:

Composed Of shatter-proof glass to reduce possible injury
Designed to adjust automatically to day/night driving
Designed with a flexible backing to permit easy adjustment
Permitted to break free of its mounting upon impact to
reduce possible injury

 

1J3.

1“.

15.

170
Appendix C (cont.)

Power Disc Brakes are beneficial to the customer because they:

A. Are not susceptible to effects from water

B. Provide an extra margin of stopping ability for high
speed or repeated stops

C. Improve car stability for low speed braking

D. Require lower pedal effort

The suspended Accelerator Pedal:

A. Permits the driver to more adequately adjust his speed

B. Adjusts to the driver's foot size for added comfort

C. Prevents womens' heels from catching under the pedal

D. Provides better circulation of air in the driver's foot area

Cougar's increased wheelbase provides the buyer with:

A. A larger rear compartment
B. Increased fuel capacity
C. A quieter ride

D. A better cradled ride

S ECTION III

DIRECTIONS :

1.

INFORMATIONAL FACTORS

171

Appendix C (cont.)

 

The information on this page is used for statisticali

individual salesman will be identified.

[purposes only. All information will be coded and no

A

Please check the box that applies to you and

answer the last question in the spaces provided.

What was the last grade you completed in school?

1.
2.
3.
h.
5.
How
1.
2.

3.
u.

,5.

Grade school

Some high school

High school graduate
Some college training

College graduate

DDDDD

many years have you been an automobile salesman?

Under one year

One to three years
Three to five years
Five to ten years

Over ten years

DD DUE]

How long has it been since you last attended a Ford Motor
Institute training session?

1.
2.
3.
u.
5.

Less than three months
Three to twelve months
One to five years
Greater than five years

Never attended

[:1
C]
E]

Cl
C]

What is the total number Of automobiles you sold from

January 1. 1967. through June 30. 196?. for each of the

categories listed below?

Mercury (full-size)

Mercury Intermediate-

Lincoln Continental

Used cars

Mercury Cougar

|

 

 

Appendix D
RATING SCHEDULE

INSTRUCTIONS FOR RATING CHARTS

These steps should be followed in completing the attached rating charts:

Step 1

Step 2

- Step 3

Step A

On each of the three attached charts, write the names
of your new car salesmen in alphabetical order both
vertically down the left side and horizontally across
the tap so that salesman A on the t0p is the same as
salesman A along the side, etc.

Starting with the Personality Rating Chart I and
beginning at the bottom left-hand corner, rate that
salesman as having more desirable or less desirable
personality characteristics than salesman Aflby
placing a M.(more) or I.(less) in the square provided

Similarly compare that salesman with salesmen B, C, etc.
until you have rated that salesman with every other
salesman. Cover your ratings with a card or sheet.

New move up to the next row and compare this salesman
with every other salesman. Cover your ratings.
Repeat the process until each of the salesmen has
been compared with each other salesman.

Follow the above procedure for Charts 2 and 3.

M“: gm

,ExAMPLE I W” by?
AI 8 (3 I)

 

464.44) Tm A M L-

 

may: BL M

 

S’ziiflfipficm L

 

I)

 

 

 

 

 

 

1172

173
Appendix D (cont.)

PER SONALITY CHART I

PERSONALITY: Some important personality characteristics
are-- appearance, sincerity, persuasiveness,
courtesy, friendliness, an understanding of
people, and sociability

 

.174
Appendix D (cont.)

PRODUCT KNOWLEDGE CHART 2

PRODUCT KNOWLEDGE: Product knowledge includes--
knowledge of product features and
their respective consumer benefits

 

 

 

 

 

 

 

f" C3 C5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Appendix D (cont.)

SALES TECHNIQUES

SALES TECHNIQUES: Important sales techniques include--
obtaining prospective buyers, qualifying
customers, planning and delivering sales
presentations and product demonstrations,
handling objections, making trade appraisals,
closing the sale, and completing the follow-
up nftcr the saJe

CHART 3

 

.176
Appendix D (cont.)

DIRECTIONS: In the boxes provided, please
distribute 10 points among personality,
product knowledge, and sales techniques.
The number of points you give each of
these items represents the degree of

importance you place upon it in the
personal selling of automObiles.

WMMW... I

PRODUCT KNOWLEDGE . . . .

SALES TECHNIQUES . . . . .

TOTAL 10

Appendix E

METHOD OF DATA ANALYSIS:
PAIRED COMPARISON MATRIX

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

c c c c c c
m m m m m m
E E E E E E
U) U) U) U) U) U)
m m w w m m
H H H H H H
m m m m m m
m m m m m m
A B C D E F G
Salesman A 7)? 77) 77’) 777 m
Salesman B L 777 777 Z 777
Salesman C Z A A Z Z
Salesman D Z Z 777 A Z
Salesman E Z 777 777 777 )7?
Salesman F A A 777 iv? [_
G
H
V
Number of
Salesmen M's Rank
Salesman A 5 1
Salesman B 3 3
Salesman C O 6
Salesman D l 5
Salesman E 4 2
Salesman F 2 4

177

Rank

 

10

11

12

TRANSFORMATION OF RANKS TO STANDARD SCORES

Number of Persons Ranked

Appendix F

 

In

37
23

39
3O

21

4

5

41
33
27

19

3O

10

43
35
30
25

17

g
44
37
32
28
23

16

.7.
4s
39
34
30
26
21

15

178

s
46
39
35
32
28
25
21

14

3
47
4o
36
33
30
27
24
20

13

41
37
34
31
29
26
23
19

13

41
38
35
33
30
27
25
22
19

12

42
39
36
34
32
28
26
24
21
18

12

Card
Columns

1,2

3
4,5

38

39
40
41,42
43,44
45,46
47,48
49,50,51
52,53,54
55,56
57,58
59,60
61,62
63,64
65,66

Appendix G

TABULATION OF THE DATA CARDS
(Decoding Key)

Name 2; Variable

District (Numbered 1,2,...18)

Dealer (Numbered 1,2,...9)

Salesman (Numbered 1,2,...10)

Question 1 (Questionnaire) Correct-l, Incorrect-0

Q. 2

Q. 3

Q. 4

Q. S

Q. 6
7
8
9

 

Q.

Q.

Q.

Q. 10
Q. 11
Q. 12
Q. 13
Q. 14
Q. 15
Q. 16

O O. O O
RHVFJHFH
o—aoxoooq

9999999900000
NNNNNNNN
tomxlmmbwm

. 30

Ed cation (Coded 1,2,...5) See p. 50 for meaning
of the code.

Sales experience (Coded 1,2,...5) See p. 50 for
meaning of the code.

FMI attendance (Coded 1,2,...5) See p. 50 for
meaning of the code.

Type of dealership (Single pt.-O, Multi-pt.-1)

Size (planning volume) (Low volume-O, High volume-l)

Mercury Intermediate Sales (Recorded 1,2,...80)

Cougar Sales (Recorded 1,2,...80)

Mercury Sales (Recorded 1,2,...80)

Lincoln Continental Sales (Recorded 1,2,...80)

Used Car Sales (Recorded 1,2,...150)

Profitability of Units Sold

Personality Rank (Ranked 1,2,...10)

Product Knowledge Rank (Ranked 1,2,...10)

Sales Techniques Rank (Ranked 1,2,...10)

National Personality (Standard Scores 13-48)

National Product Knowledge (Standard Scores 13-48)

National Sales Techniques (Standard Scores 13-48)

D

179

180
Appendix G

 

011011111110000001101011111111101014511020181300035
uOLJDZI109190100000001101101000101013551002050000003
012011111000100100001011110101001013341013171907041
._Q£?02141191910i100191111111104110011331013110904059
012031111010011101111111111101111013411009060103020
“9.139.921.1119 9,9..0 0.9920299219 11.1.1 1.1.9.1. .1591. 0.0.4223 10.0.3. 05. 030,0. 031.
012051111111001101101010011101111013311002060000024v
_“0950111110111101111911011000111101155119051§122§oaés
013021111010111110001101111101100113531015151607044
-”0120319111111D0100011110111110111115531002161103049
014010111010110111111011111011111114541101090829030
-__QlflQ2111191111119111111111111111i1192§1infizinSOInfib.
014031111011111101111111111110111114341105130901043
u_013091111110111i01001111111oooinlnlaaniioaosoanoasa
015011111001000100001010111000011111331106070802017
i-915021111110110010011111111000111113411104100402025
015031111010011110101111111101100014541106130400017
-__Q1§fl&1110111111111111111111101111015331103100302021
015051111110100101100000000000111114531109030404005
_-9450§9111010100110101111111010100111511103120262011
015071111011111000001111101000101014541106ou0301020
,-HoléolellOOIO10100001101101101100113500108051501045.
016021111011000000011101111000110115530107071701023
_.ifliéflfiflillfllfllflfliflfllfll11l1luinlulalflfilnl09060901022.
017011111011001100111011111010110113430109091103032
V91702111114111110111141111Lo00110111590113051002055
017031110011010100101011111001100114540110080603030
.917591;}$041001150110111111010111013290111070301039
017051111111101111101110110111111003140106080402026
01801111iciinsilnquiiiiiliiioninioiJSiooaauaalnl063
016021111000101101001011111000111005240001010100019
.-9199111111010111011910001000llOlOlllSkOQllflZflSflOflSl
019020111011011000011111111100111013520004030100023
QZlOilllii11191111111111111101100fl125§1128322116019.
021021111111100111101111111000110013551118312212012.
0210511111111qgliliggilLaiignuiiaii§5511125116Q§on6
022011111111010101111111011100111113141014313008039.
H_98QQZQiliiigilL1914190000009001100159110i0262810626.
022031111010100101010111111000111013431005110503017
H.Qۤ93119391909410999044110190009091aiaiononlooooaoa
02301011101011110111111111110111101uuuioo9471613ou6.

"m 02502111001010110100011111190111;9113010113211n235g,

 

 

v-wm

023031111010110101111111100111101014441001000000004
”wgggflllllllllifllllllllé11111110111015511033332900096i
024021101011100101100111101110111014421011231900007
-m2929911111911110101191111111011100019431003150601005
025021111010101101011111111000110005111001120506018

.n_fl%§D31111n1Q1Q11fl11Q1111a11lllflllflllkllflflfilfinbfllnlfi
0 5041110011101001011111111000001010351000030204025

026011111011111111101111011000111113211103100105fl321

.--—-———---—--—-—-——-nan-c.-.—n---—-o4._--o... —~.—m--‘—- -—o——--~_——ou .-

(cont.)

~0~wh~m _, ~o _.

!010101373737
1020202232323
.010101434303
[030302303035
.050204173525
.050203253530“
020402352535
030102213930
010201393039
020303302121
000104194119
1020302332233"
030203273327
(010401Q11901-
'050405263026
.090404303030-
'010101054545
030203343930—
050605212621
050507262615”
020302393439
010301392139~
030201213039
020101303939_.
040202253535
.010101434303“
030303303030
050409172525w
020505351717
020201232331,
010102373723
010101373237”
020202232323
010101393939”
010102393930
020203303021“
020101334101
030203273327”
000303192727»
4010502911933"
030102213930
010303392121—
020201303039
204010219H133"
~030402271933
310202313333—
3020301332741
010101313137”

181

Appendix G (cont.)

'026021111010011100001011101001111002531103160304012
027010111000010190000100000000000012441008181105030
'027020101110101111101111111100111113431008170804044
028011110011111101111111111001100014551006140601025
’“028021111110111101011110101001110114411002110601031
0290111110011101111111111110111;1112419111211402025
“”032010I11I11111111011111111001111112331111171002008
052021110000000000001111011001001005431106121494004
'032031111011111111111111111101110113541107141002002
032041110000000100101111111010001013211104220301004
"033010111011111101110111110011111114331034131206033
03502011101010010011110110100111001g511012110401024
‘“03303III10001101011f0111111000100012111012070402024
0350a1111011010100101011111009111012411902910000062
"033051110011000111001011100100101013111007060200014
05501111111111111111111110411111111253019511060201§
'7035021111111111111111110111111111013110104060701009
035031111111111111111111111111119904240100050102012
035041111111100111111111111111011014520103050400009
03601111111{41%{91931111090991111014510106191311001
."‘“036021111111011111111111111110111014510108070508032
0370111.11949919..1..Q.1.91..111411.19 1.1.1-199154.209.07100001021
'"037020101010011101001111111001010113110001070100016
036011111111111111011111110100111114420113071902056
038021111111111f11011111110011001014520108031503064
05303111111141111£99901119999991111§§§Q10§01059901§
'“”039011111011100111111001011111101111510005060700049
03902111111011111111111111.1.1..11.1.1111224109090400090.1.1.
'04‘0IIIIITII111111101111111000000115311106321804015
041021111001111111101111111111111115431100040332006

 

 

 

 

041031110000011101011111100001110010311107331200029 -

041041111100111101111111111100111010311102290102032
"‘0420II111011110110101111000011111013251118341511007
0420211110111101101111111001011110144§112225220300§
'"”U4ZOSIIIIIUUGIIUIIIUI111111101111013431109171100009
0420411110011001111011111011011111142411181§0302001
‘“‘0¢205IIII0IUTI0110111111111100111013111105151101000
044011111010111011001011111100111913511019;1§§9§g§§,
"7044021110111110111101111000111101013511005110705018
0440311010001101011011111110011001122310060805000256

"”04404011001I01000000111110IT110010122310020503000157

045010110000100000001011101101101004411108431307006.
1' 1 I 11 1111 II 111 0 I 0 O 0140 003,
045030010000000100000110111100001005531109151992QQQ;
‘"”04504111010UUIIUUOIOIIII1000000060145411050407120011
045051111011110111000111110111111014111111210200003:
”"04506IIIOUIUUIIIUIIIIO00'010010110041511001603020023
046010110011111111011111111101010013511103100906015I
4‘_.-.. 0 o 1 011' 041 100 010!
' 096011.919} 0.1.212.133.15229.9121192191111231103200800019;

 

 

020202232323
020101233737-“
010202372323
010102373724-“
02020123233
910101373Z3Z__.
010101414141
0904051919211“
020202333333
010101414141”-_
010101434343
930203303530...
010302433035
020902352535",_
040504251725
.020201333341u_-
010303412727
030205193312__.
’030102274133
:010101513737“-_
'020202232323
.010101313137"__

~3020202232323

310101323232__.
1020202303030
10303032121211”-
2010202372323
{9 .20 1.0 .12 3.3. 7.3.7- _ -.
030302272733
020203333327
.040403191927
2910101515151_n
'010101434343
:929205§Q§§2§ui-

'~040302253035

.050405172517
'020503351730
9191914151511"
040303192727
920202§§§§§11i
030404271919~
-020202353535
010101444444

.9 9.9 1Q §2§§3§3--_.
-060403162832
1955.9 9.0 9.3221926---
.050504232328
.02010gga4133
010203413327
9.319291221215111;

182

Appendix G (cont.)

 

 

 

040040111000001101101111001100111013311102080602024 030404271919
047011111011100101111111101001111113520113211703134. 010101373737
047021110011100111111111111100111113420116270600080 020202232323
0“801¥£!}110000101101011110000111011410112121502049. 010202393030
’043021111011101101101111111100111103420113161300043 020101303939
;-0“fiQ飣££919001100001010111111101113510113150707030 030303212121
049010111111001101111111111011110111330122050801002 010101373737
049021141014110100001111111000101013510115040504011 020202232323
051011111111111111001111110011010114541117351716026 .020101334141
00?}021111111111111111111111001111015441113412407011. 4010102414133
051031111111111101111111101101111014531116121501015 030203273327
-1921931119219191111100101111000111013241105260700012 -030304272719.
052011111011111111101111111011111014541107131523016 .010101444444
052021111141110111101111111010111004241103120325014 .040202202337
052031110010011100101110101000001011111115262000007 '060505162323
052099110110001100011110111000101013431113151200004 020304373228
052051111001101100101111111000110005411110141000002 050603231632
.022091111010010100111011011011101013431102110100000 2030406322816
054011101110000111101110011001110013430139311500453 060601161644
054021110011110000100000010100111013410160202713028 ,020202373737
054031111011100101101111111111111015410129190901151 .030303323232
054041111101111111111111000111111014510127200500017 ~050506232316
054051111011100101100111111001110103430120180700023 040404282828
0,054061111111011101001111101001111010030121100201015 1010105444423
055010111011111111101111111101111113550116211202015 .030403271923
055021111111111101111111111101101013230107151503023 020102334133.
055031111001111110111111111101111114430105101103019 010201413341
055041111011111101111110111101101014520113120603011. 040304192719
057010111101111111111110011101011104340104121216017 010101414141
057021111001111111111111111100111014210104120600004. .03020227333331
057031111001111010101111111011111013510101040100001 040404191919
,19§1999111019100100101110111010111114140100020100000- 0203033327273
061010101011110101101111101101101002510109161806019 020202353535
e 9619§9414941991101111110001000111015540103001513000. 030303303030,
061031110001101101001110111001111114310104160801004. 010101434343
0610411111111}01011010111111101111155301030112020012 030505301111.
001051110011011101101110000011100112430102060704006. 030000302727

-0620414199110901111000101110001011135301051216000001 A0101013939391

 

 

062020111110000000001111101000100004530104160302006 020202303030
nOégfl§9119001111111101111101001011014110101030100014 030303212121”
063010111011101111111111111101010102410115201104024 010101343434
. 05“011111010001111191111111000111014540102000903014. .02010123313111
064021110011010001100001000001111013420101010000028 010202372323
“974911191011110111101111111110110115341107101007001- 0504042320200
071021111011111111111111111001110013521100071103010 030202323737
”974039141041111101111110111101111014531106120606005 020202373737
071040011001000100000000000010010014331107140304009 040403202032
.10710511111111111111;;114111Q0900001§gg110§150002001. 0101014444043-
071061111001111111101111111111110014531100030101001 010301443244

,.07§9114101};111111111111111010111113210107171105033 030302272733

183

Appendix G (cont.)

 

”1172020111010001111101111111111110113440107060814024
072031111111010111101111101101110114240102111005020

——..- ._...4..._.._—-.—

‘072001111111001111100111111111111112510106141303007
074011111011101111011111111100111013530103020010002

"074021111001011110000000010000111103510100000200002
075011110000001001001001001011110003530111061205002

'TTTSUZIfffififfifffffffii011100100111013230108080705009
-19759§4111911411111411194110011111914510100020201000
075041111010110111101111101000111114110104100202002
”97591414111911411110111110110111911§§§Qloa920900092
076021111010111111101111111001111114230105210302002
076031111011111110101111111000101013110107150500002
076041110000101010101111111001111012520104100600002

073011111010110101}01111111011111112510001080200002

 

 

”070020111011111111101111101100011104510001020706015.

07303111101111110;l11;;1111911001015510004000300011

-.___, —.

081011111001111111100111111111110014531110141419012
081020111011111111111111111010110004531110001002011
081031111011101101101111111001010113431101020100025
Q820111}1911119191111111911190110112351111222900003

‘082021111010101101101111011001001113241114251300009
0320311111!9409109110404101100011000211113101700011

 

082041110010000110101111111100111014541109171300002-

062050100100000000101010101101010003411102020100001
082061011110100010001111111001011003531105060400005
n03391}11091QQQQ001$QQA19001990101110200401020322013
083021111011110101000110111001001103540109261503037
0330311QO191119991Q991990999911919020501191;1101021
'083041111011110100000010111001111014430110131102041
003051110011010101111111111000101103310117130403042
083060110011110011101110101011111013110105050500013

‘ 0.3 397.1 4.0 $94.1_!9.9.1..1__1_1.L1.!14.13.119011.0012390101090000 0.1. 3:
"084011111010111111101111101001110014210017111302011
... .0 8‘19?- 1. .1} 19.4.1.9!!! 1.9 9}. 9.1. 0.1.1.1919 9.9 .1}. 9.1. 4.4.21.9 9.0. 909.130.5101. 0
084030111010100101101111111100110113110004020000009
084041111011111101100010110100110114210003000000001_
085011111111111111101111111000111113240124261005029
-_ 03 5.03.1 1.1.19.0 9.1.1.1.1.1199}_1.1_1._1_1_1.9.1.1210}. .2}. .10 1.13.1106 9.2. 0.310.
086011111010010111101111101010110113530104150711006
086020110110110001111011100111111013420106020501001
“ ‘ 013701000100 01 11 010011 11 ‘11 1110111 01‘ 3‘3 10“ 1090405 '00 0‘63
087021111010101101101111101000110114310114060500030
0 0 01 00000013340105040400044
087040010111101101101111111101110015530101030113909
” ‘ 00705111101 01' 11 101011111 01' 11011110151 10104040201 02 5'
087061110000110100000111111100000013110103020200015
‘"091011111011111111111111011001111013331016152205133'
091021111011101100111111111111111013241011121501070

 

- o . .—-----’-----—---—o--------_--------¢—-‘—--—-——----—-——--—-——-——-——-----’w-—-

 

010202413333
020101334141....
040403191927
010101313737 _____
020202232323
030000211919__”
010101414141
02020.2333333___1
040303192727
000302192233 _____
020401331941
030202213333___
010101414141
.010101393939 _____
-020303302121
.03020221303OHEH
010101393939
020202303030___
4030303212121
.010105444023u_u
050103234432
030102324437 ......
040204283728

060304163228
010505452626 .....
020101394545 1
000200303930 _____
050506262621
anfln3153ofig
060302213439'
000001352115 _____
020201333341
910102910133 _____
04030g192719
03000.211221...
020202232323
010101373737 .....
010101373737
02020§2§2§23 .....
010303443232
050202233737
040506282316
0.2.9 10.137.499.111
030404322828
0§0§0§1§1§2§ _____
030504322328
010101444444°
060405162823
020202373737

184

Appendix G (cont.)

 

091051111010001101011111111100111003431002020101042
.299192111191919111990!0;11119091119100!1902090209010
092010111010010101111110111001111013111018181701061
-_9929?1$11919199991199011001901111012151012101002000
092031011011111111101111111111011013441012091200050
093011111011101101011111111101110113531001060520012
093021111110010100101011111101100014241008091001029
.2995911141111111191991111111119111110000100000400031
094020110111100111111111111110111114540106050700020
2993921141911119111111111111119110010500100020400020
095011111011101101101111111111111113210102060502016
095021110001101101101101101100111113510104000501005
096011110101011110111111111111110114510106101520089
__922021111190111100111111111000111110240119302204020
096031111111110110011111111110111115140106231301023
_-999009111110011110190000000000010004440110070403015
096051101000110100101111001010011002110103040400038
097011111111110111101111101100110000200102121001011
097020111011000101111011011000101013410100010001023
-2091001111111000101111111101001111114340102010000010
098011111011001110011011111010111014530016091504065
~0290§1110111001111011111111011111015530004081401055
098031111011101110111111111100111013510006050300064
099011111111 111 111 1 1
099021111111001101111110011101111013240100010201035
.-0??0§1111111111111111111111101111012510104010001021
101011111111101111101111111111111114511021342300024
_-19£9§£111131111111191114111110111015401019302500013
101031111111101100001111100011110113351018312200018
101041111110101111101110101011111013331012

 

311000010
102011111011011111011011101111111114531103132033005,

-2102921111911111191111111111111111010901115251709001
102031111001001100101111111000000114311103180903002
102041110111101111111101111010110114131110130703002

‘“102030111000000101000111111011110013141105210000002
103011111111100111011101100101111113250105322005003

”1103020111111101101111111111011111114210100311102005

.11°§914111111119119£1411{A1991111911§§40109€§119509;
104011111110111101101111111011111013330006111101031

1-1950§1111011111910101110011100100003110002070600051
104031110011101111001110100011110113440002050304024
104041111101110100101110011011111014110002040201023
105011101111111111101111000001001013500120211412010

2.195021111011000119111111111190191115230105120610010

 

_~-—-- -—-—-—-— .-

105031111100001111111110111111110113240107150402009
"1199011111111100101191111111100111113430004061000034
106021111111111110101110111011111013330002050700014
106031110111011101011111111101111114240092039191030
'106041111111111110101111111011111012230003030600016

2-193911111919099111911!11001011191113330139221010104

 

050605231623
040303283232”.
020101303939
030303212121“,"
010202393030
010101313732
020202232323
010101393939-“
030203213021
02030230213011-
010101373737
020202232323.“
050101174343
04020225353Smh
010403432530
030300303025”1
020503351730
020103303921.”
010301392139
030202213030”,
030202213030
010101393939”-
020303302121
020102303930_-
030303212121
010201393039. .
010102414133
030003271927
020201333341
000300192119_w
010201433543
020102354335._
040505251717
.93949930252§__
020303353030
'020302302130
010103393921
.030201213039
030101274141
010402411933”.
040204193319
0203033321211.
010201393039
010101393939
020202303030
0202013030397
020203303021
010302392130..
010103393921
030101213939

 

185

Appendix G (cont.)

 

 

:107021110110001111101111101111110104530112100802065
107031111910199199111101111101010014410101040100024
111011111010111111111010111111111015131148272025001
111021110919491999991111111009110003211119210802040
111320111010001111111111101010011013341102070324002
111 1111011100191911411110999111913231199111302090
111051110010111101101100010011111013521106130601010
111060111951419111411411101110110002511105100700007
111071110010101001110010001000101013111104060701001
111081111099099909091101001100199003211105050600095
111091110101110010011111111001110113531100010001002
1-112911}11111011111101111111101011015541110242217007
112021111010001110110111111101111113431109222004008
.4129?!!11141911111111111111910110014241109201300030
112041111101000110001111111101110014531104141507006
112921111111091191011111111011111004151103071000003
112061111110001100011100111110110004541105080300004
113011111110100101110111111010101115421109211002031

 

 

 

 

113021111101110100100011111000001111511107100806013
_111§9§11£111919110911111!011000111110501101000305029

113041111110100100011111111100110012311102000000020
_1149101190111111101111111019001}1110401105090116009
114021111100001100001111111000111014511103131004009
p-114031111111101110111111111001110014311104090401035
114041110110000101001111111100111014531104110700021
6-11“0§91999999999999OQU00119991999192531111030401005
115010110100100100000101000100111013531007350903007
115021111011111111001111111199111010221909170002008

~...1—.. ~....~.—.—~-—

115030110101100000000100100011001112511002200401032
11504111111011111100111110101011101351100324050§002
115051111111111110001011101100111013531002090102027
116011111.1.;1.110.111;01.1111110190111152.101022718113054

'“116021111111111111011101111001111014510115271201027
117011111111111100111111111101100004210112282605010

-2. fl_. n-_.__-- -~_»—..—_-—-

"117021111110100000001000000100001013330108040901108
117030111110111101101101111110111114330112242104010
117041111110101100111111111101111113510118132404006
1119291}9019011119491419111000911012400103020401124
118011111111101111101111111001111012340109061003038
1 118021111111011011111101110001101113510106061100045
“119011111111011111001111111100111014340111130800069
119021111111111101111111111111111013530107081202032
“ 0011111100111012240110100100054
119041111010101100101111111001110112240106060601019
"119051111011101101101001110110010012340104060400025
121011111011101101111010001100111013531108233433006
"121021110011011100101001000101000113511167121812002
121031111111001111101010110101010114431106141603012
90402063

 

»"--‘———---‘—- -———--—-——-—~~-.-—o~.-~-~ 0-_“>-——---—~ 4....-2 ‘Ca..-“_.. _——.-o.--—---.--.~-—————¢.———o.q

 

020201303039
010101393939_3
060404273333
030402363340..
050204283932
01050624302211”
040302333640
040303333636,2
010104474733
07060524213oum
020101404747
050400232828__.
030102324437
020201373744”_1
060205163723
01030344323222-
040403282832
030201213341_1_
020102334133
030303212727“_.
010304412719
010101434343”._
030302303035
0205033511301”
050404172525
0402032535301“-
040301253043
040500251725“,
020301353043
030103304330_.
010402432535
0101013737371“
020202232323
020402352535
02020335353Q_m
010101434343
030509301725“-
010102373723
020201232337"-
020201353543
040102254335
030303303030
950995172517
010504431725
019201513341“
030302272733
040103194127
020404331919
919191372127

 

 

186

Appendix G (cont.)

 

122021111111000100011111111101010113521005171200015
123011111000001001001000000011000103510105293000007
123021111111111111101111111101011113530105282807007
123031111110111101011111111111011015430107332103025
123041111010110111111111111011011114440103171502010
”H1450511110101101011 1111110101011014450102195394004
124011111111001000101111111010011114430101071300000
124020100001101100011111111010001114110104070000004
125010111111101111010110111100011114520100050700019
125020110110101111101111111101011012510103070600015
125 031110111111111101111111100010012420105060400019
_125940111110191111100111111010000112330102030500001
127011111111101111111111111101011010000115393700000
1&7021110101110100001011101111011013210107341500002
127031111111110111101111111110011115510100050100003
151011111011111100011111111001111014431113553006010
131021111000001101101110001010000013431111422502003
..191031111011111101011111111100100114311101100804000
132011111111111111111100111111111112511109312602060
132021111010100110111111111101001113531112161105034
132031111001101101101]10111111111113241102121106044
153011111011100110001111111100111114211007121500035
133021111001010101101111111001111013231005120602029
.1135031110000110000101111001000111013511002040000011

134011111111100111111110111111001113151105280900001
134021110109000000111110111100110113231102251001005
134031111000011001101011100000011113521102041403001
154091111111111111010110111110111013531100100204006
134051111111001111111111111100110114431101080500000
._1§§011110001111111011111111111101015511001252312046.
135021110011011111001111111001110015541009212208027
‘ 195031110111101111111111111111111014341007232010020
135041111111101111101111111011110114251005191610021
l3§0§1111§99191101011111111000111014211004212104021
136011110001000101101111111010110002520104031201079
136021111111111111191111111110111103250102051100035

 

 

_.1_-”-.~——~—~—o—--o——_

137011110101111101110111011001001012310006131201070
__192021100001001019091011000101101000000002020201025
138011110111000100101011111111111114530109241906095
11§4020111001100100101111111000001012520104140504034
141011111111101111111100100100101013441119352701056
. _141o21111111101111101110000101110013441106362701056
w"1410311110010011010010'11110111111114441107303003031

.—.—~—-——-—*-.-~-—----——”-0~—o.—A-nu---——o-----a -...--..-—.....-—...u— n..- ..1........_--_. __.»--_--

141051111000101101001111111011110015141107331201026
142011111111111111101111111111110114421102592206009

142020110010100110101111111100110111531105250925007
1420311110100111011010111111011111114331104311503004

 

142040111010100110101111101100111014311100020228001
.2152951111119091111191111111101110014431103221402013.

.——~.—— -.———.-’mm-—-'W-"~

020202232323
020402352535.
010201433543
030109304325
040305253017
050503171130.--
020101233737
010202372323”-
030302272733
,010101414141 _
020203333327
040204193319...
020101303939
030303212121 2
010202393030
010301392139m
020102303930
030203213021...
020101303939
020303302121~.-
010202393030
020203303021”-
030102213930
010301392139..
010101434343
030203303530.
050505171717
0203023530351-
040404252525
040202253535...
030105304317
020303353030“-
020404352525
010501431143--
020202232323
010101323131...
010101373737
020202232323.-_
010101373737
020202232323_-
0102 4335
0301..3043....
0403 2530
0204113525..-- .. _ _..
0505 1717
050304283532-__
020202393939
030003353235...
080307143521
040404323232.--

w—..

187

Appendix G (cont.)

 

142061111011100110111111111011110014211100210804000

- 142071111110111111011111111000111015541103080203002

-. _,_._._..__-_~.-___-_..______,__...-_..__.,,_.--______.____,_...--,_._____..._-,, ~- -..—

1142080110010100110101111111000111013431101030200002

19.3 9.1.1}.10.919.11.119_1__1_0.11.1.1.1. .1..1_ 4.1.1.11.0.1..1_49_4_1_1_09.05.011020.8.61
143021111010111110011111111101110114531100030100001
143031111010100101111111111111110014231100040000000

 

--

143041111111100111111111101111111113431100020100000
1“%9}911919999919149!999091911011101510002001000051

”144021111110110101111111000111110003530001030100030

144031111011111101001111101011111003230002000200041

“--fi-o— "n—w---——_—-~~—.-----_ .-...1....-’ -_.. .._----.~._...-..-_ _—.__._ ——

“"14501111111110011111111111110111111341011206160“097

145021111111011101101111111000110114210106050303085
145030111011111111111111111000011013240104080400044

146011111011111111011111111100110013110110081301005’

146021111111101101101111111000011112310108101803025
.199921111141191119111110111901110113110103111501034
146041111011101110011011011110000013310107151200003
146051101001101101111111111110111014510194131000001

147011111111110111111111111111111012430002070400049

-.1“7921111149011111111111111101111011240002060900035

5—4

‘147031111110111111111111111101111111110000010000008

.199911119911111111191911111109111003530001040101050
148021111011111111111111100011111103140000090502028
148030110011001101001010101011111013240001020203038
148041111000111100111111111101101014440001080200003

14805111001010010011110101010111101441000005Q300000,

148061111001111111101110101100110014110000020000000
.151911111111111111111111111119119115521119143623094
151020110110110101001111011000010014541100050740016
151031111011111111111100011010111113411106382202009
151041111011011101011111011011110114531108302302005
15195}$}9931199191919314191319111092911192391309002
151051111010111100101110101111000113411101281101002

152011111110111111101111111111111114531008182900019-

--~ —-_ ——-———--——.-»--_u—~--...~.o..—.——_-————_~—————~———-_—————_~——--

152021111010110101101111110000100013421009281100009
152031111001110111111111111111110004141004281100009

0 0 1 1100011000011000111012521003010200082
1520510100010110011111011010011010135110Q203p400050

~-—--—o——__——-u~——.-——-_-—-u...--—- —.---_.~_-——_——.—~————H~—-.-——— ——..—“ ——-——_ .....

‘153011111110000101111110101101110013230105130502081

153021111110}. 1901944111910901019.11195401031111401.0511.

““153631111111111110111110111001111114430106100300055

153041101110000110011010111011101014530102050300059

__153051111001001111101000010101111012110102030100040

153061111111.91.1111.0.1.111111.911.011.1123101090302000.1.2.

““154011111111010111101111111010101114240101060802039

154021111101011100111110111101101113310101020500047

'"15‘USIIUIIIIIIfiffffiffifi110111111113440102020200013

154041110100011111111111101001111012150100020300012
1 4 0010010300106

-12?2§l££§£91£99£9£1919£119111111411§§£QQQ€Q§QQQQLA2

 

060106254625
010601462546..
070505212828
010403411927_*
040102194133
030304212119__
020201333341
020201303039_-
010102393930
030303212121-"

, 010101393939

030302212130__
020202303030
030303303030”-
050202173535
020102354335”-
040404252525
010501031143.-
020201303039
030101213939.-
010302392130
030101324444.-
030403322832
01106062816 16.

U 040505282323
'020302313231.-

010204443728
040105254317."
050202173535
0103015330&3__

020503351130_-
030404302525
910101434343“-
030403302530
050303253011__
010502431735
020204353525".
010501442344
900503151532".
050305233223
040106284416
030404322828
9.20.2.0 12.37.317.37-
010202413333
920101334141“.
030303272727
040404191919
030202213030
910391322339.

188

Appendix G (cont.)

155031111111110111101110111110111013210006130100011
159911110111101101101011111100101111430005061502038
156020110001100010000000101001001011310002101102031
159094114119901111010011111101111113219002071109044
158011111011100101101111010101011113230013101704049
161011111010111£9£1QL111111111911019241905200§0§043
'”161021111011110101001111111111111115241005230100037
161034119019109100140111011100111004211007160501031
161041110111110101000001000011010015251001080600047
191054141019190110111111111111111114211004110200030
151061111010101001101111111101111012231004120400017
125919111211111101101111111%;0111010001901151500040
162021111110001100011111111010110104331007150800035
162031411010100101111011101100111013311000070000025
152041111011111111101111111110110013111004040300013
162051111040101101101110011010110114211000090100016
163011111011010101101111111010111115530111182917018
123021111011010101011111111100101113520101212110015
163031101011101111001111111010111013530100140227008
V165041111011111111110111111100011014430108150605005
163051111001111100111111111100101014330104170300002
194011111110111101100000111100111114540121191102022
164021111111111111101111111001110013430105120501045
1164031111010011101101111111011101004230103120502029
164041111111110111111111111101111014240104130501027
164051111011110111101111111101101004240102140401021
154061111110111111111111111101111114110104070200022
172911114111111111111110011000111113431109311107047
172021101011110101101111111100111013511101401107029

. 172051110010101100000111010100110012521101231304021

 

 

172041111101001101011111100001110012531104260702018
.1JYQQ§1111119191101111111111010110115531102170802025
172061110010111100011111111110101013541104200202032
172011111911101111111111111101111113541105332012009
173021111011111111111111111101111115141106380912005
.}73031111000000100101111111000100104551110231215002

173041101010101101100000111101110002541103361408000
.1179921111011111111111111111010111013531101050101000

175011110011011111111110111111111014531119492615002
5172931111011110111101111111110111115531106411110006
175030110010101101011111111011100015531102410712007
.175041111110001100111111111111110113441102351105002

175051110111001111011101111001100012421106230708003
17599011001919Q191011111111900010104341101040102000

 

_-—.——-—--- u... _ .—

176010111010111101111111101111110013541104320906003
7179921111911191411111141114001110004231103350402001

176031111011111111111111111010111113331102040301001
177010111011101111111111111101111114530103230804016
177021111011011101101111111011110014440104111005003

.177030110001111110001131111110111115§;Q;DLQ§051§035

.—-—-—--—--~..--—-—~-—4 ”~.5-—-..—‘ --1 .- -.--.—._~—.—-_.1,.

020103303921
010102393930
030301212139
020203303021
010101373737
010201413341.
010202413333
020303332727
000000000000
000000000000H
030101274141
010101434343__
020202353535
020503351730”.
020404352525
030305303017
020301353043
.030102304335u1
010101434343
040203253530
050404172525
010502442337
030101324444
060604161628”.
050405232823
040306283216
020203373732
010201443744.”
030203323732

'020202373731_.

060104164428
040404282828
050305233223
010101434343_-
030303303030
020202353535..
040404252525
050505171717-”
010101444444
030303323232-_
020202373737
0&0304283228..
050505232323
060406162816”.
030201213039
0202023030304
010103393921
010202393030..
030303212121
020101303939

189

Appendix G (cont.)

 

 

 

 

 

 

 

 

 

"“178011111000000000101101110100001003530102150508004 040304253025
178021111014111101101111111001110014410104110604012 010101434343
178031110011111101101110110110110014310102121102002 020202353535
178041111010101111100101111101000113540105090505003 030303303030

‘ 178051111011111111101111111011110014340103060309001 050405172517

__!8}911119119191111111110111000111113231109201504038 040102334740-
181021111111111111101111111011111115311106160904025 070204244033

q-13103011111111;101100111111100110014211105131000032 020301403647
181041111100111110011111111001011114141104200602015 070907241324

- 131050111111101111101110000000111014311103160702021 030805362030
181050111010101001101011111101110113331103160700022 010503473036
131071111111110111111114111;11111114211103210400011‘ 080606202122..
181081111010110101111110111010111114521104180300015

A-131091111011100111011111111101111113211102070101039 060708272420

"181101111101110111111110101000101114431101100101009 050405303330

-1182011111011110100111111111011101114541107270710007 020404331919+
182021111010101101101010000011111113531110170605010 030303272727
182031111011110110111111111111101114211104190303006 03020127334111
182041111010110100001111111100101014521102090105004 010102414133
.133014114919111111111;11111101111014441008251209030. 010101414141-

"183020111010111100001110111110001015231002090404021 020304332719
13303111191910110111111111101110;115341001130605002 030202273333

'“183041110010011110111111111110101014241000010001008 040403191927
184010111110111101001111111000111015531103240504022 0302032133211“
184021111011111101101111111000010014421106120703036 020301332741
-1829§1£¥1949409490400110ll1004190012531106150904009 030304272719
184041111011110101101010000011111113441104091101021 030102274133
-1599§1111911110111111111101000110114111101100301013 010405412719,
184061111000000101101111111100111013531100020105002
185011111111101101101011011001111112241002331502012 030101214141 _
.185021111110111100111111111000110014521001220604010 020404331919

-_15§Q§1411119111191101119011090111013211005100104015 010202413333.
165041111000000101011111001111100002241001140503006 ° 040303192727
185011111141119901111111111111119113530102071001015 010101444444..

"186020110111111101101000000111110013520102060801015 010101444444
186031111010111101111111111111111111430101040503005 0202021132;1_1
186041111011111101111111101000110114240102030301013 020302373237
185051110011111109401110104019110115150102020401011 0105014423441

”‘185061111110111111100110111101111015410102040200006. 020304373220
18701111193991499919}!14141010091902210101000200052 0303022727331.

“”187021111100101100101111111010110104240103100402014 010201413341
187031111010101100001111111000001112230102050307001 02010432411414
187041100000000000011111011101101002510101030100036 040303192727
’183010111111111101011111111011111119510994QSQ€9§0§5 0101013737370

““191010111100000100111111111110011113331108221600037 020202393939
191021110010001000000000000001100010521100030822003 040401303045

"“191031111111110101011111110111111112421104201201019 060104214530
191041111111110101111111111111111114311107141101020 910196454521-“

F‘I9ID5IIIIrUrffrrfifirfififffiIffiffiffiI1012431104081401020 050303263434

£2§99111££91911£99191119911111111119531;01120592039 070405153026

 

 

190

Appendix G (cont.)

191071111011111100111111101000111005151101170402016
192011111100100100111111111001111114531103110706008
192021111011101100011111111100111015111100100100002
193011111011100111101111111001011014211006250000039
'”195021111101111111111111111101111013231003250300011
1930311110011101011011111111010001: 5531002230500006
*195041111001001101111111111001110014431003070! 00024
_193051111111110111101111111111111012151001130500005
193061111101110111101101111011101014411001050400011
1194011111111101101111111111111111013541004111106013
194021111011110100101110111011111112331006150700015
194031101111101101101111111011110003511004090502012
194041111111101101101111111110111012511005110000004
195011111111111110101111101100110114330104050801050
195020110010101110101111111000111015330102040502039
.195031100011111001011111111001100114110102030301031
196011111010111100101011111011010115210002050300063
”195021110000111100111111111100111014520002000000021
197010111111111101011111101101111013210002080600034
198011111111110111111111111100111113430103191807029
190021111110110101101111111101100114340103130502067
_196030111011111101100110011111111013540101020519011
196041111111111100111111111010101003210106151005008
198051111110101101011111111000111013230103060200023

' 030505342626

010102373723
020201232337
040304203225
010101444444
020402372837
050203233732
050606231616
030505322323
010201413341
020302332733
030402271933
040101194141
020201303039
030103213921
010302392130
020101233737
010202372323
010101373737
040101254343
050203173530
020202353535
010303433030
030404302525

Appendix H

ASSUMPTIONS UNDERLYING KENDALL RANK AND KENDALL
PARTIAL CORRELATION COEFFICIENTS

The assumptions of the data required for a Kendall
rank correlation coefficient are that it is at least an
ordinal measurement of both the X and Y variables so that
every subject can be assigned a rank on both X and Y. An
ordinal scale is one in which subjects being compared for
some quality can be ranked in ascending or descending order
of the extent to which the quality in question is possessed
by the subject. The result is a complete rank ordering of
all subjects for the quality under consideration.

The assumptions for the Kendall partial rank corre—
lation coefficient are the same as those for the Kendall
rank correlation coefficient, except another variable Z
is added to the analysis. The variable 2 must have the

same qualities as X and Y.

Sidney Siegal, Nonparametric Statistics for the
Behavioral Sciences (New York: McGraw-Hill Book Co., Inc.,
1956), pp. 24, 213, 224.

191

Appendix I

ASSUMPTIONS UNDERLYING TESTS FOR SIGNIFICANCE
FOR KENDALL RANK CORRELATION COEFFICIENTS

If a selection of random samples is drawn from some
population in which X and Y are unrelated, and the samples
are ranked on X and Y, then for any given rank of the X
quality, all possible ranks of the Y quality are equally
likely. For each of the n possible rankings of Y, there
will be associated a value of the Kendall rank correlation
coefficient r. These possible values of r will range from
1 to -l, and they can be tabulated in a frequency distribu—
tion, consisting of a distribution of probabilities.

For the value of n less than eight, the distribution
of the probabilities associated with a particular correla—
tion coefficient is shown in a table entitled, "Table of
Probabilities Associated with Values as Large as Observed
Values of S in the Kendall Rank Correlation Coefficients."
(The sampling distributions of S and r are identical in a
probability sense.)

For each value of n, the significance of an observed
relation between two samples of ranks may be determined by
simply finding the value of S and then referring to the above
mentioned table to determine the probability (one-tailed)
associated with that value.

Since the value of n greater than eight was not re-
quired for the analysis performed in this study, the assump-
tions underlying its Kendall rank correlation coefficient
tests for significance need not be discussed.

Sidney Siegal, Nonparametric Statistics for the
Behavioral Sciences (New York: McGraw—Hill Book Co., Inc.,
19567, pp. 220—221.

192

Appendix J

ASSUMPTIONS UNDERLYING THE NORMAL "2" TEST
FOR THE DIFFERENCE BETWEEN THE SAMPLE MEANS
The assumptions underlying the normal "z" test for
the difference between the sample means are as follows:
1. The observations are independent.
2. The population (N) is very large.
3. The sample size (n) is equal to 30 or more.
4. The sampling ratio n/N is less than .05.

5. The populations have the same variance.

Sidney J. Armore, Introduction £9_Statistical Anal-
ysis and Inference for Ps cholo and Education (New York:
John Wiley & Sons, Inc., 19665, pp. 383- 384. ‘(Assumptions
1‘4).

 

George W. Snedecor, Statistical Methods Applied to
Experiments in A riculture and Biolo (5th ed.; Ames: Iowa
State College Press, 1556),” .§7. (Assumption 5).

193

Appendix K

ASSUMPTIONS FOR MULTIPLE/PARTIAL
CORRELATION COEFFICIENTS
The assumptions for multiple/partial correlation
coefficients are as follows:
1. The observations are independent.

2. The observations are drawn from normally dis-
tributed populations.

3. The populations have the same variance.

ASSUMPTIONS FOR THE F TEST

The assumptions required for the F test are the
first two above plus the variables involved must have been

measured in at least an interval scale.

Sidney Siegal, Non nparametric Statistics for the
Behavioral Sciences (New York: McGraw-Hill Book Co., Inc.,
1956), p. 19.

194

Appendix L

TESTING THE SIGNIFICANCE OF THE AVERAGE
KENDALL RANK CORRELATION COEFFICIENTS

1)]

S = r [1/2 n (n

where n - is the average number of salesmen in
each dealership
r - average Kendall rank correlation

coefficient

3 - a function of r and n which is required
to enter the table

For example, let n = 4 and r = .26.
S = o26[l/2°4.(4-1)] = 1056

The value of p is obtained from a "Table of Prob-
abilities Associated with Values as Large as Observed
Values of S in the Kendall Rank Correlation Coefficient."
Since p (.44) is greater than Cl(.05), the two variables
being analyzed are not significantly related, where

Ho: r=0, and H1: r>0.

'This method of testing for significance is found
in Chapter 9 of Nonparametric Statistics for the Behavioral

Sciences by Sidney Siegal (New York: McGraw-Hill Book Co.,

 

1956).

195

Appendix M

TESTING THE SIGNIFICANCE OF THE
PEARSONIAN AND MULTIPLE/PARTIAL
CORRELATION COEFFICIENTS

F = R2 . n - k-l
1 - R2 k

where R - the correlation coefficient
n - size of the sample
k - number of predictors
n1- degrees of freedom of the denominator
n2- degrees of freedom of the numerator
Degrees of freedom n1 = k and n2 = n-k-l

 

For example, let R = .33
n = 524
k = 1
Then:
.332 524 - 1 - 1
F = 1-332 1 = 63.8

The value of F (0‘: .OS) critical is obtained from
a Table of "95th Percentile Values of the F Distribution."
Since there are l and 522 degrees of freedom, F, (01: .05) =
3.86. Therefore, since the achieved value of F (63.8) is
greater than 3.86, the difference is too great to have oc-
curred by chance, and thus the associations among the vari-

ables analyzed is statistically significant, where H r = O,

0:

and H1: r>0.

This method of testing for significance is found
in Helen M. Walker and Joseph Lev, Statistical Inference

(New York: Holt, Rinehart and Winston, 1953), pp. 278, 324.

196

197
Appendix N

FREQUENCY DISTRIBUTION OF
SALES PERFORMANCE CRITERIA:

Number of Sales- TOTAL UNITS SOLD

men Within each

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Increment Multi—point Dealerships
4O
3O _
I n: 280
Mean: 56.1
_ _" Standard Deviation: 30.42
20
10
, HUN 2
0 135
Number of Units Sold
Single Point Dealerships
401
30
i
1 n: 244
Mean: 54.7
20 Standard Deviation: 45.36
10
fl Hflnflnm m1

 

 

 

 

 

 

 

 

 

 

 

 

Number of Units Sold

198
Appendix N (cont.)

FREQUENCY DISTRIBUTION OF
SALES PERFORMANCE CRITERIA:
PROFITABILITY OF UNITS SOLD

Number of Sales— Multi—point Dealerships
men Within each
Increment
4O 1 I
30
n: 280

Mean: 113.1
Standard Deviation: 64.82

20

10

 

 

 

 

 

 

 

 

 

 

10 WW 3

 

 

 

 

 

 

 

 

 

 

322
Profitability of Units Sold
Single Point Dealerships

4O 1
3O _

n: 244

Mean: 93.8
20 C Standard Deviation: 65.84
10
OHH HflI—Iflmflm [1:

378

Profitability of Units Sold

Number of

199

Appendix 0

FREQUENCY DISTRIBUTION OF
SALES PERFORMANCE CRITERIA:
TOTAL UNITS SOLD

Sale smen With—
i11 each Increment

4C)0

30

20

10

4O

3O

20

10

Mean:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

n: 180

Multi-point High Volume Dealerships

53.2

Standard Deviation: 28.51

 

,H ‘ flflflnnm

Number of Units Sold

Multi—point Low Volume Dealerships

n: 100
Mean:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

441 [][1 r1f7 F1

61.6

Standard Deviation: 33.16

 

135
Number of Units Sold

1!

200
Appendix 0 (cont.)

FREQUENCY DISTRIBUTION OF
SALES PERFORMANCE CRITERIA:
TOTAL UNITS SOLD
Number of '

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Salesmen Single Point High Volume Dealerships
Within each
Increment

40 H

30

_ n: 193
Mean: 56.6
20 7 Standard Deviation: 47.96
10 _
0H . Hnnflnnn 21.

 

 

O . 252
Number of Units Sold
Single Point Low Volume Dealerships
30
n: 51
20 Mean: 50.9
Standard Deviation: 30.26
10

0 117
Number of Units Sold

201
Appendix 0 (cont.)

FREQUENCY DISTRIBUTION OF

SALES PERFORMANCE CRITERIA:

PROFITABILITY OF UNITS SOLD
Number of

Salesmen Multi-point High Volume Dealerships
Within each
Increment
4C)11
30
n: 180

— Mean: 114.4 -
Standard Deviation: 66.02

20

10 —' —

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

, H Hflnnn

O I 308
Profitability of Units Sold

 

‘f

Multi—point Low Volume Dealerships

 

 

 

 

 

 

 

40 4
30 n: 100
Mean: 110.7
Standard Deviation: 62.76
20
10
flfl Hflflflflflflflnnm n

 

’-

0 322
Profitability of Units Sold

202
Appendix 0 (cont.)

FREQUENCY DISTRIBUTION OF

SALES PERFORMANCE CRITERIA:

PROFITABILITY OF UNITS SOLD
Number of

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Salesmen Single Point High Volume Dealerships
Within each
Increment
40 A
30
n: 193
7 Mean: 99.9
Standard Deviation: 68.60
20 _ __
'_ r—‘
10 I - —
1H Hflfl 1'] n r] r1 r1 r1 :1 i
o ’ I

profitability of Units Sold 378

Single Point Low Volume Dealerships

 

 

 

 

4O 4
30
n: 51
20 Mean: 74.2
Standard Deviation: 43.11
10
{7
0 Hill—IJ][][] {1(1f11fi rir1 _>

182
Profitability of Units Sold

Appendix P

TESTING THE DIFFERENCE BETWEEN
THE MEANS OF TWO SAMPLES

 

 

 

 

Z = X1 - X2
5 2 + s
l 2
n1 n2
where Z1 - mean of sample one In

X2 — mean of sample two

51 — standard deviation of sample one (unbiased
estimate of population

52 - standard deviation of sample two

n1 - size of sample one

n2 - size of sample two

 

For example, let X1 = 21.4
X2 : 21.1
51 = 4078
$2 = 4.31
n1 = 244
n2 = 280

z = 21.4 - 21.1

 

“4.782 4.312

244' * 260 = '750
The reject area for the H : X. - X' = O is found

0 l 2
in the Table of "Percentile Values of the Unit Normal Curve.“

If CL: .05, the reject area is z«<-l.645 or z )>l.645. Since
2 (.750) is within the accept zone, H0 is accepted, where
This method of testing the difference between the

means of two samples is found in Sidney J. Armore, Introduc-

 

tion £g_Statistical Analysis and Inference (New York: John
Wiley and Sons, 1966), pp. 384-385.
203

Appendix Q

METHOD OF EVALUATING THE CONSISTENCY
OF THE PAIRED COMPARISON MATRICES

The coefficient of consistency (C) is determined
as follows:

Step 1: Calculate T

where T = 2(ai -'§)

ai = the number of times the individual i
receives a rating More (M)

'6' = 1/2 (t -- 1), t = number of individuals
being evaluated

Step 2: Calculate c (number of circular triads)

wherec=__t(t2-l)-l/2T
24

Step 3: Calculate

Iftisodd, C=l-24c
t(t2- 15

If t is even,q= l - 24c
t(t2-45

If and only if§= 1, there are no inconsistencies.
As decreases to zero, the inconsistencies as measured by
the number of circular triads increase.

To test for significance, the T value is used. If
T is greater than the critical value of Chi Square (T, when
suitably standardized, is distributed on H0 approximately
as aIKZ with t - 1 degrees of freedom) for t - 1 degrees of
freedom and C1: .05, then the null hypothesis Ho of random-
ness is rejected and H1 (consistency) is accepted.

Source: H. A. David, The Method of Paired Com arison (New
York: Hafner Publishing_ Co., 1963), pp. 21— 24.

204

 

v
. i, .
gyfifi Ed... mi

 

tun-

 

 

       

   

MW

175

mm

 
 

ME

 

I

zlni‘figimifqigrlufium

  

03