A STUDY OF THE CONSUMER PREFERENCE OF -
SELECTED CUT FLOWERS. ,

Thais fat the Degree of Ph. D.
MCHIGAN sums, uumksm _
Yun-I'eh Han A

19613

1115515

This is to certify that the

thesis entitled

A STUDY OF THE CONSUMER PREFERENCE 0F

SELECTED CUT FLOWERS

presented by

Yun-Teh Han

has been accepted towards fulfillment
of the requirements for

4% degree mm ture

.fi/ 49m

Major professor
[MUM

LIBRARY

Michigan Statue
University

 

ABSTRACT

A STUDY OF THE CONSUMER PREFERENCE OF
SELECTED CUT FLOWERS

by Yun-Teh Han

The florist industry is now in the budding stage of
a marketing renaissance. The latest surge of interest has
been aroused by sluggish market conditions typed as ”over-
production" and "under-consumption.”

One approach toward the alleviating of such a
situation would be an increase in the consumption of the
products for ”not occasional events" market. This accounts
for the current emphasis on consumer preferences.

This study sought information germane to present day
consumer's flower preference. As in other preference
studies, it sought consumer response to several physical
attributes of five major cut flowers: carnations, pom—pon
Chrysanthemums, standard Chrysanthemums, gladiolus, and
roses. The stimuli selected for this study were color, grade,

number, and price.

Yun-Teh Han

To index the above mentioned consumer flower prefer-
ences, the Michigan State University's Consumer Preference
Panel was utilized from 1957 until 1963.

Color preference study revealed that preference for
individual colors was not pronounced in carnations, pom-pon
Chrysanthemums, and gladiolus. Red roses and yellow standard
Chrysanthemums were the favored in these two types of flowers.

In general consumers preferred large numbers of
flowers. However, when the flowers were priced the preference—
pattern might change. This in turn depended upon whether
the price was considered high or low by the consumer.

The overall results of grade preference studies com-
pared favorably with the established practices.

A retail sales survey was conducted to test the
relationship of consumers stated preferences (in color, grade,
number, and price) of the five concerned flowers to their
marketing behavior revealed by the reporting florists in the
State of Michigan.

This study of consumer flower preferences provides
the basis for further studies of consumer preferences for

flowers. This in turn should assist the floral industry to

find new customers, and to better serve the American consumer

with a product which plays a role in nearly every society

in the world.

A STUDY OF THE CONSUMER PREFERENCE OF

SELECTED CUT FLOWERS

BY

Yun—Teh Han

A THESIS

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

DOCTOR OF PHILOSOPHY

Department of Horticulture

1963

ACKNOWLEDGEMENT S

Appreciation is acknowledged to Prof. Paul R. Krone
for the suggestions that he has made and the role that he
has played as Chairman of the guidance committee.

Gratitude is expressed to other members of the
guidance committee, Doctors H. E. Larzelers, A. L. Kenworthy,
W. J. Haney, William Lazer, and Mr. L. J. Tolle, Jr. for
their valuable advice and guidance in the presentation of
this study. Additional thanks are due to Dr. Larzelere and
Mr. Tolle who helped formulate the program of study.

Special gratitude is due to the writer's parents,
Dr. and Mrs. L. W. Han, for their unfailing encouragements

which made the empirical research posSible.

ii

TABLE OF CONTENTS

Chapter Page
I O I NT RODUfl ION O O O O O O O O O O O O O O O O O 1
Purpose of Study 1

Economic Significance of Study 2
Objectives 9

II. REVIEW OF LITERATURE . . . . . . . . . . . . . 11

The Function of the Consumer in a Free

Choice Economy 11
Consumer Preference Researdh 13
Consumer Panel as a Marketing Tool 18

III. METHODOLOGY OF CONSUMER PANEL STUDY . . . . . . 22

Materials 22
Methods 23
Scope of the Study 31

IV. RESULTS OF THE CONSUMER PANEL STUDY . . . . . . 33

Carnations 34
Pom-pon Chrysanthemums 44
Standard Chrysanthemums 53
Gladiolus 62
Roses 71

V. RETAIL SALES SURVEY . . . . . . . . . . . . . . 84

Survey Procedures 86
Characteristic of Reporting Florists 87
Results of the Survey 89

iii

Chapter

VI 0 DISCUSSION 0 O O O O O O O O O O O O O O O 0

Preference Study
The "Economic Strength" of Preference

A System of Color Preference

Relationship of'ConsumersStated Preferences
to their Marketing Behavior

Evaluation on the Present Study

Motivation of Consumer Buying and Consumer
Education

v. SWY O O O O O O O O O O O O O O O O O O O

BIBLIOBRAPHY O O O O O O O O O C O O O O 0

APPENDIX

iv

Page
93
93
96
99

112
107

109
114
121

134

Table

10.

LIST OF TABLES

Consumers' color preferences in carnations:
estimated rankings . . . . . . . . . . . .

Consumers' color preferences in carnations:
per cent of panel selecting sample as 1st-

choice and the distribution of frequency of

choices within a sample . . . . . . . . .

Consumers' number preferences of carnations:
estimated rankings

Consumers' number preferences of carnations:

per cent of panel selecting samples as lst—

choice and the distribution of frequency
of choices within a sample . . . . . . . .

Consumers' color preferences in pom-pon
Chrysanthemums: estimated ranking . . . .

Consumers' color preferences in pom-pon

Chrysanthemums: per cent of panel selecting

samples as lst—choice and the distribution
of frequency of choices within a sample .

Consumers' number preferences of pom-pon
Chrysanthemums: estimated rankings and

distribution of preferences . . . . . . .

Consumers' grades preferences of pom-pon
Chrysanthemums: estimated rankings . . .

Consumers' grades preferences of pom-pon

Chrysanthemums: per cent of panel selecting

samples as lst-choice and the distribution
of frequency of choices within a sample .

Consumers' color preferences in standard

Chrysanthemums: estimated rankings . . .

Page

36

38

42

43

46

47

50

51

52

55

Table Page

11. Consumers' color preferences in standard
Chrysanthemums: per cent of panel
selecting sample as lst—choice and the
distribution of choices within a sample . . . 56

12. Consumers' number preferences of standard
Chrysanthemums: estimated rankings . . . . . 6O

13. Consumers' number preferences of standard
Chrysanthemums: per cent of panel
selecting samples as lst-choice and the
distribution of frequency of choices with-
in a sample . . . . . . . . . . . . . . . . . 61

14. Consumers' color preferences in gladiolus:
estimated rankings . . . . . . . . . . . . . . 63

15. Consumers' color preferences in gladiolus:
per cent of panel selecting sample as lst-
choice and the distribution of frequency
of choices within a sample . . . . . . . . . . 65

16. Consumers' number preferences of gladiolus:
estimated rankings . . . . . . . . . . . . . . 68

17. Consumers' number preferences of gladiolus:
per cent of panel selecting samples as lst-
choice and the distribution of frequency
of choices within a sample . . . . . . . . . . 7O

18. Consumers' color preferences in roses:
estimated rankings . . . . . . . . . . . . . . 72

19. Consumers' color preferences in roses: per
cent of panel selecting sample as lst-
choice and the distribution of frequency
of choices within a sample . . . . . . . . . . 74

20. Consumers' number preferences of roses:
estimated rankings . . . . . . . . . . . . . . 77

21. Consumers' number preferences of roses: per
cent of panel selecting samples as lst-
choice and the distribution of frequency
of choices within a sample . . . . . . . . . . 80

vi

Table Page

22. Consumers' grades preference of roses:
estimated rankings . . . . . . . . . . . . . . 81

23. Consumers' grades preference of roses: per
cent of panel selecting samples as 1st-
choice and the distribution of frequency
of choices within a sample . . . . . . . . . . 83

24. Percentage distribution of retail florist sales
by size of city . . . . . . . . . . . . . . . 88

25. Percentage distribution of consumer flower-
buying by types of flowers . . . . . . . . . . 9O

26. Average percentage distribution of consumer
color acceptance by types of flowers . . . . . 91

27. Average percentage distribution of unit of
flower sales: by types of flowers . . . . . . 92

28. The economic strength of consumer number
preference in the three grades of roses . . . 97

vii

LIST OF GRAPHS

Graphs Page

1. The comparison of the average percentage of
the lst—choice for each color sample with
the expected average percentage: carnations
and pom-pon Chrysanthemums . . . . . . . . . . 39

2. The comparison of the average percentage of the
lst-choice for each color sample with the
expected average percentage: standard
Chrysanthemums . . . . . . . . . . . . . . . . 57

3. The comparison of the average percentage of
-‘ the lst-choice for each color sample with
the expected average percentage: gladiolus

and roses . . . . . . . . . . . . . . . . . . 66

viii

LIST OF APPENDICES

Appendix
1. A mail questionnaire- for recruiting consumer
panel members
2. A mail questionnaire of retail sales survey

ix

CHAPTER I
INTRODUCTION

Purpose of Study

The market success or failure of a given product
depends upon consumer preferences, particularly if there
are also good substitute commodities available.

In spite of this much marketing research in flori-
culture has been producer-oriented rather than consumer—
oriented. Flowers have been marketed without benefit of
investigation into the consumers' desires regarding the
product. To throw some light on the problem, this author
undertook a consumer-oriented cut flower preference study
in the period between January 1957 and February 1963. The
over-all objective was to explore consumer preferences for
cut flowers, in the Greater Detroit area. Simultaneously,
in order to supplement the information from consumers, a
retail florists survey was conducted in the entire state of
Michigan.

The researdh was designed to be intensive covering

five major cut flowers and using selected panel participants

representative of the particular area involved. It was
also designed to provide an insight into the dynamics of
consumer desire for the four kinds of cut flowers in-
volved: carnations, Chrysanthemums, gladiolus, and roses.
With adequate consumer preference information in
the hands of producers and distributors, consumers may be
better able to obtain what they want. If the information
makes posSible any reduction in waste and marketing costs,
the savings may be passed on to the consumer. Hence,
this can make a significant contribution to the improvement
of floricultural marketing and perhaps may increase the

demand for flowers.

Economic Significance of the Study

Economic Importance of Florist Crops

Florist crops are classified by the U. S. Bureau
of the Census in a rather heterogeneous classification
identified as "horticulture Specialties". The 1950 Census
of Agriculture and the Special Census of Horticultural
Specialties for the same year reveal the economic importance
of these crops. Only recently has the full significance of

these crops been generally recognized.

Possum (105), presented a report concerning the "trade
in Horticultural Specialties". There he pointed out the
position of horticultural specialties in agriculture:

During the first half of this century, horticultural
specialty crops became increasingly important in the
agriculture of every State and region of the United
States. By 1950, these crOps made up nearly 2 per cent
of the value of the Nation's sales of farm products.

. . . From 1890 to 1950, greenhouse crops were the
major kind of horticultural specialty production.
Nursery crops usually amount to about half the value
of greenhouse production. Until recently, the growing
of bulbs and flower and vegetable seeds have accounted
for the remaining segment of the total for horticultural
specialties. Since 1930, outdoor production of flowers
has become a significant factor in the total production
of horticultural-specialty farms in the United States.

Furthermore, census figures for 1949 reveal that the sales
of horticultural specialties by farmers, over 60 per cent of
which were florist crops, while 57 per cent fell in this
category in 1959.1

Florist products may be divided into three general

types as follows: (1) cut flowers, (2) potted plants,

(3) bulbs, plants, rooted cuttings and flower seed. Whereas

 

1c.f. U.S. Bureau of the Census, United States

Census of Agriculture: 1950, Special Reports, Horticultural
Specialties, Vol. 5, Part I (WaShington: Government Printing
Office). pp. 33, 429—433, 733, 755; Census of Agriculture:
1959, Special Reports, Horticultural Specialties, Vol. 5,
Part I, pp. 26—27, 45—59, 109-111, 158, 168, 175.

cut flowers accounted for about two thirds of the value of
florists'crop in 1949, they accounted for about one half in 1959.2
Cut flowers represent the most important single group of

crops classified as horticultural specialties. They accounted
for over 40 per cent of total sales in 1949, and 28 per cent

in 1959.

This study covered only the five major cut flowers
grown in this country, namely: carnation, "standard" chry-
santhemums, "pompon" Chrysanthemums, gladiolus, and rose.

And they are rated the first five ranks of cut flowers (by
value of wholesale prices) in both 1949 and 1959 Census
Reports of Horticultural Specialties.

Similar trends in the importance of florist production
are also true in the State of Michigan. Michigan's pro-
duction of Horticultural Specialties has expanded more rapidly
relative to Michigan agriculture, than that of the United
States. Von Oppenfeld (84 ), pointed out that this favorable
rate of growth has been clearly evident since 1940 and the
continuing growth of the horticultural specialty field in

Michigan offers continued opportunity for producers of flowers,

plants and bulbs.

 

2U.S. Bureau of Census, Special Reports, Horticultural
Specialties ,op. cit., 1950, pp. 33-37, 377 and 755; U.S.
Bureau of Census, Special Reports,Horticultural Specialties,
op. cit., 1959, pp. 45-49.

Commercial establishments growing horticultural
specialties in Michigan produced crops valued at $11.2
million in 1949, $19.4 million in 1959. Cut flowers and
related plants made up two-thirds of the total in 1949, one-
half in 1959. Roses continue to be the largest single money
crop, with a value at the production level totaling over
$1,178,000 in wholesale value in 1959. Pompon Chrysanthemums
were the second.most valuable crop, followed closely by
carnations and then by standard Chrysanthemums, Gladiolus
sales ranked 4th in wholesale value in the state. (U.S.
Department of Agriculture report on cut flowers, Production
and Sales, 1958-59, Intentions for 1960 in Ten Selected
States).

The greatest increase in dollar value for the period
from 1958 to 1959 occurred in the value of the gladiolus crop.
The value of gladiolus sold in the state in 1958 was $54,000,
while in 1959, the value had increased to $186,000. Never-
theless, the total increase for all five cut flowers was an

extremely moderate one. The value in 1958 being $2,249,000,

and in 1959, $2,429,000.

The Hidden Problemsiip Florigulture Marketing

The market for flowers has increased in total on a
per capita basis. Voigt (127 ), quoted Fossum's estimate of
consumer expenditure for goods and services of floriculture

as increasing 54 per cent from 1949 to 1959. Non—florist

outlets' sales rose 78 per cent, FTD orders increased 84 per
cent and FTD values of orders increased 11.6 per cent. An
increase of about 50 per cent occurred in U. S. florists'
production, per capita disposable income, all retail trade
and retail flower sales.

Despite what appeared to be a rather favorable trend,
recent wholesale price and supply conditions indicated a need
for further expansion of consumption. There were frequent
periods when production greatly exceeded demand or sales
[Liesveld (15 )].

Problems involved hinge on certain factors which
motivate the demand for flowers on the trade practices of re—
tail florists. The existing markets for floricultural pro—
ducts may be grouped into four categories [Dewey 01M))].

Three of the four categories that he grouped are'bccasional"

events.

1. Personal events 3. Holidays
a. Funerals a. Easter
b. Illnesses b. Christmas
c. Births c. Mbthers' Day
d. Weddings d. Decoration Day
e. Home entertainment e. Valentine Day

f. Thanksgiving

2. Public events 4. Other demands
a. Church services a. Impulse sales
b. Store openings b. Home decoration

c. Organization meetings
d. Dances

When the demand is largely limited to flowers for
special occasions, retail florists find themselves faced
with a relatively inelastic demand. This demand has a
"necessity" characteristic. The use of flowers on these oc-
casions is dictated by long established social customs and
is relatively independent of price. Much of the effective
demand in the form of a "designed" product is for occasions
rather than home use. Consequently, retail florists attempt
to maximize profits by selling services along with flowers
and by maintaining rather constant selling prices throughout
the year.

Sales of flowers for everyday use, at present, proba-
bly account for less than 5 per cent of all flowers sold by
retail florists [Trotter (118)]. The customer buying flowers
for everyday use is not likely to be willing to pay for the
services of designing, delivery, credit, etc., usually as-
sociated with retail florist sales.

There has been very little effort on the part of
retail florists to satisfy the demand for lowepriced flowers.
It is also true that there is no established custom in this
Country calling for the use of flowers in the home. The
development of this important potential market could material—
ly increase sales of flowers and plants. It is important

however that these sales be made at prices that will cover

present costs of production and marketing. An important
phase of the present problem then has to do with adjusting
marketing policies and orienting them to a new market which
supposedly will yield expanded sales.

The effort to develop this new market needs a fresh
approach to floral merchandising. Only in the light of the
real product classification can an interested marketer make
assumptions and learn lessons from the status quo.

Research has shown that flowers for the home convey
different images in consumers' minds as compared to their
image involved in flowers for special events. This conse-
quently will influence their behavior. It is probable that
as flower sales for home decorations increase, more of the
flowers that are purchased will be seen by the purchaser.
Thus one would expect a new set of preferences to develop.
The selection and preference of the retailer will become less
of a factor and consumer preference will become more important.

There have been quite a number of marketing research
problems studied by floricultural marketing researdhers but
only a few are oriented to the study of consumer preferences.
Furthermore, there has been no concerted effort to synthesize

these materials into a new meaningful condensed form.

 

3L. J. Tolle, Strategy Considerationsiin Changing the
Retail Outlet for Floral Products, in mimeo.

Objectives

It is recognized that the likes and dislikes of the
consumer, his opinions and prejudices, and the amount of money
he has available are all important items in determining the
merchandising methods to be used and the product to be
offered.

In buying any product the consumer is confronted with
the need for making decisions. These decisions are affected
by color, style, degree and kind of packaging, prevailing
prices and the extent to which other products may be substitu-
ted. The criteria for selecting the merchandise purchased
by consumers might not always be controlled by the marketer,
but they cannot be ignored for they function in the consumer's
decision-making.

This study is primarily designed to investigate the
physical attributes that stimulate consumer actions as they
affect their preferences for the five major cut flowers:
carnations, pom-pon Chrysanthemums, standard Chrysanthemums,
gladiolus, and roses. The stimuli under investigation were
color, size, price, quantity, and grade.

More specifically, this study was undertaken to
determine:

(1) Consumers' color preferences,

(2) Consumers' unit-of—purdhase preferences,

10

(3) The influence of prices per unit on the consumers'
preferences,

(4) Consumers' preferences for grades and a corre—
lation of these preferences to established grades,
and

(5) The influence of price on the grade preferences,

(6) Relationship of consumensstated preferences to
their marketing behavior (revealed from florists'

survey).

CHAPTER II
REVIEW OF LITERATURE

The Function of the Consumer in a
Free Choice Economy

Katona ( 10), made an analysis of the
consumer latitude in this country. He reasoned that:

. . . Economies are conceivable in which consumer
expenditures do depend soley on consumer income. In—
deed, as of fifty or one hundred years ago in the
United States, and even today in many countries, this
may constitute a fair approximation of the actual situ—
ation. One may think of a poor economy in which most
people devote all their income to subsistence. Whatever
people earn they spend on food, shelter, and clothing
in order barely to survive. In such a situation con-
sumer discretion is absent. In feudal societies it is
the landowning gentry, and in early industrial societies
the few great entrepreneurs, who direct the economy.

But in the United States today the role of the consumer
is fundamentally different. Several major recent devel-
opments have substantially increased the power of
American consumers by allowing them freedom to advance
or delay their purchases and to spend whether above or
below their current incomes.

Similarly Gordon (E3 ), pointed out: "It (free
choice) is the power to choose those goods and services they
(consumersfi‘want in an economy characterized by security and
governed by a price system. . . . The counterpart of free

choice is the concept of free conscious rejection."

11

12

In essence, Lazer, et. a1. (13 ), commented that the
business enterprise is governed largely by consumer sover-
eignty rather than by authoritarianism of either corporate
management or the government. Acceptance of the viewpoint
that, in fact, the consumer is king, implies an understanding
of the freedom of consumer choice, and the voluntary action
that underlines achievement of market goals.

This orientation is implied with Martineau's claim
(63 ), that business is simply a means for the society to
achieve its goals, not an end in itself. In a highly compet-
itive free economy one must meet the consumer's taste for
color, attractive design, convenience.

Gordon (8 ), suggested that the function of consu-
mers in a free choice economy is to use their freedom of
choice positively so as to generate the production of an
abundance of wealth. Thus consumer welfare would be enhanced.

One may easily detect a much wanted implication for
marketing strategy which derives from this system design.
Since consumer satisfaction is so important to the marketing
philosophy, the scientific study of the consumer is central
to the development of marketing strategy. The fact—founded
method of solving marketing problems is a cornerstone of the
marketing management concept. Information about numerous

marketing problems--up-to-date, information about consumers,

13

and consumer behavior, is pivotal to the effective mobili—

zation of marketing resources.
Consumer Preference Research

Consumer Preference Research in Agriculture

Jasper (108) commented on the role of marketing
research in the area of agriculture. For many years,
agricultural researdh was concentrated primarily on improving
and increasing the efficiency of agricultural production.
Progress made in this field, in some instances, created
surpluses which increased the need for improvement in
marketing methods. This, in turn, has created a need for
more information about consumer preferences, practices, and
buying habits in order that marketing improvements might be
developed on the basis of a more complete understanding of
the problem.

Seeing the need for consumer preference studies on
agricultural products, the Department of Agriculture has
undertaken a great deal of research in this since the end of
WOrld War II. In 1950 Bayton (33 ) reported on consumer
preference researCh in the Department of Agriculture. One
of the major features of the present program of analysis of
consumer preferences is its direction toward classes of pro-
ducts. Fundamental to this approach is the aim to obtain

data that will be beneficial not only to one specific group,

14

such as retailers, but that can also be used by producers,
processors, wholesalers, retailers, and consumers.

Another feature is the cooperative basis on which each
project is developed. Involved in the planning of each study
are specialists in the product area, as well as agricultural
economists, home economists, marketing specialists, and psy-
chologists. This permits each study to be designed in terms of
the major dimensions that are important to it.

The above statement is a full account of the func-
tion of consumer preference investigations in the realm of
agriculture marketing research. However, limited attention
has been given to methodology.

Howell (55) outlined the methods already tried for ob-
taining information regarding consumers' preferences: (1) Con-
sumers' preference surveys, designed to learn from consumers
their preferences with respect to select products and factors
affecting these preferences; (2) Consumer purchase survey's de-
signed to assemble data relating to volume of purchases and ex-
penditures for specific articles by individuals and by families
to be used as a basis for indicating the response of volume of
purchases and expenditures to individual incomes and to other
factors; and (3) Retail sales surveys, designed to assemble data
as to volume of sales, prices, and other factors for specific
products Which would show the response of volume of sales to

prices and other factors.

In closing, an over-all evaluation of this research

15

on consumer preferences should be brought into light. Bayton
(33) generalized his report on "Consumer Preference Research
in the Department of Agriculture" thus:

This research on consumer preferences indicates that
from the viewpoint of action-research the results of
sudh studies are of value--to Government administrators
and the private interests involved (consumers, retailers,
wholesalers, and growers).

. . . The integration of the findings into systematic
economics of demand-analysis appears to rest, however, on
somewhat tenuous grounds. Whereas these studies contrib—
ute to our knowledge of consumers with respect to
product-classes (citrus fruits, potatoes, men's clothing
etc.) We still need to develop a systematic conception
of the principles involved in the behavior of consumers,
which includes the psychological and sociological
variables that are most certainly operative in influ-

encing the behavior of consumers, irrespective of
product-class.

Consumer Preferences for Flowers

There have been quite a number of marketing research
problems studied by floricultural marketing researdhers, yet
only a few are oriented to the study of consumer preferences.
Since they are so few, an effort will be made here to present
a general outline of each contributing work.

Sherman, et. al. (115) conducted a study of consumer
preferences for roses, carnations, and Chrysanthemums. The
study used a consumer panel, representative of the population
of Columbus, Ohio. The purpose of that project was to find

out consumer preferences for roses as related to carnations

16

and Chrysanthemums. The results were presented under the
following headings: preferences for flower types, color
preference, and frequency of purchase. A similar study was
undertaken by LeClerg (112) Mississippi Agricultural Exper-
iment Station. These two investigations were done with cut
flowers (Sherman's study only included cut flowers; LeClerg's
study included both cut flowers and pot plants). Therefore,
they proved to be the most useful reference for this study.
The details of these findings are distributed throughout this
paper under appropriate headings.

Early (102) of Pennsylvania State University has also
done important work in this area. A personal interview type
of survey was conducted to study consumer flower—buying
practices. The city of York, Pennsylvania was selected be-
cause of its diversity of occupations, incomes, nationalities,
and religions. The sample yielded 1,053 completed and ac—
ceptable schedules which were collected from October 1954 to
February 1955. Since consumer concepts of a product will
influence its use, a part of that paper was devoted to the
description of the image of floral products and the image of
their plan of sales. Four areas were related to help compose
the consumers' image of floral products. The four areas were:
(1) values and attitudes toward floral gifts; (2) preference

for floral products; (3) influence of selected retail

17

florists' sales techniques; and (4) price concept of floral
products.

Concerning the second area of the presentation, the
author pointed out several factors which effect the preference
pattern over the floral products for different uses. These
were variables within the sampled consumers themselves, such
as: sex, marriage status, age etc.

In light of the above findings certain implications
and recommendations were made for future merchandising
efforts.

Sorensen (117) conducted a study designed to obtain
information on consumer buying habits; the use and preference
of plants and flowers for the home and the possibility of
changing and developing consumer buying habits. It was as-
sumed that the demand for floral products may increase if
these habits could be changed or developed. The hypothesis
was that the large undeveloped market potential for plants
and flowers in the home could be developed further. To test
this hypothesis, a consumer panel was conducted in College
Station and Bryan. A three-part objective of that study was
to find out where, when and how often the panel members
bought flowers or plants for home use. The revealed consumers'
preference indicated that the hypothesis was a sound one.

Certain preferences should be considered by the retail florist

in developing the market for flowers in the home. They were:

18

The flowers should not be arranged (according to
60 per cent of the panel),

There should be a large variety,

Care instructions should be included,

Merchandise should be priced plainly,

Medium—sized potted plants were preferred.

The second highlight of Sorensen's finding concerned
the factors affecting purchases for the home. Among the
items deserving most attention are: price, inadequate know;
ledge about certain plants and flowers, availability through—

out the season, keeping quality, care required and appearance.

Consumer Panel as a Marketing Tool

A panel is defined as a group of consumers organized
to serve with some continuity in an advisory, judiciary or
fact-finding capacity. Panels vary in accordance with
research objectives and thus may be quite different in indi-
vidual conception and operation. Data are obtained by mail,
personal interviews or consumer group meetings. There are
essentially two types of consumer panels: the product opinion
and attitude panel, and the consumer purchase panel. In some
cases features of both are combined in a single panel.

Shaffer (124 ) made an extensive study on the oper-
ation of consumer panels. In that paper, he gave a good

description of the two types of consumer panels mentioned

19

above. Here only the 'product opinion and attitude panel'
is considered.

The general purpose of the product opinion and atti—
tude panel is to determine consumers' reactions, attitudes,
preferences, and opinions concerning specific products. The
types of studies made with each class of panel may be divided
into two general categories; (1) those dealing with new and
undeveloped (in a market sense) products, and (2) those
dealing with products and services already on the market.
The former is often referred to as a product-testing panel.
The second type of study places more emphasis on what the
consumer does. The information obtained is used for programs
promoting the sale of the product. By determining what the
consumer likes and dislikes about certain products the pro—
ducer is guided in his plans to provide the market with the
things the consumer wants most. The preferences can be
identified with the different groups within the market, so
as to determine the direction of the promotional programs.
At the same time, the types of retail outlets through which
the products should move can be determined to improve the
distribution system for the particular product.

From the earlier review of literature on consumer
preferences of flowers, it was rather interesting to note

that the consumer panel is a prominent marketing tool in that

20

field of study. This is true even in other fields of study
where consumer preference research is important. One

writer commented on the role of consumer panels played in
marketing research in the November 1945 issue of Printers'

Ink Magazine: "The consumer panel is a unique market research
tool. Provided the sponsor knows how to operate successfully
and interpret results accurately, it points the way for a
decision before he makes heavy commitments in production,
distribution and advertising. It is the best instrument de-
vised to date whereby he can consistently penetrate the mind of
Mrs. Consumer and discover her definite views on a host of sub-
jects affecting product acceptance or merdhandising."

Even though the researcher may be aware of the sig-
nificance of the consumer panel in product research, yet he
may not be aware of some of the pitfalls of its use. No
wonder Carrolll called to the attention of workers on consumer
testing studies two important aspects of the problem which
appeared to have been overlooked. They were the need for
greater flexibility in planning consumer tests and the Oppor-
tunity for more efficient use of consumer test data. Indeed,
those two aspects can never be overestimated in this type of

research. Owing to the variations in the sample size demand--

 

lMavis B. Carroll, "Consumer Product Testing Statis—
tics," in Arthur D. Little, Inc., Flavor Research and Food
Acceptance, (New York: Reinhold Publishing Corp., 1958),
pp. 162—174.

 

21

the number of products given to evaluate, and the extent of
information required from the individual, flexibility is a
prime requisite in planning a test. Only by securing a
more efficient use of consumer test data, can more adequate
and dependable information be obtained.

In concluding, one may safely claim that regardless
of the recognition accorded panels as a unique market
research tool, it can be stated definitely that they provide
the key that opens the door to the solution of many problems
confronting manufacturer (or producer), distributor, and

media.

CHAPTER III
METHODOLOGY --- CONSUMER PREFERENCE STUDY

Materials

Five major cut flowers were selected for this study,
namely: carnations, pom-pon Chrysanthemums, standard
Chrysanthemums, gladiolus, and roses.

In nearly all cases, flowers were obtained from the
Lansing Florist Exchange, a wholesale florist in Lansing,
Michigan. The flowers were delivered to the laboratory of
the Horticulture Department at Michigan State University, on
the morning of the appointed testing day.

Flowers were then inspected by the researcher. Since
the investigations were held in the city of Detroit the
flowers were packed as carefully as possible to prevent
damage. They were then transported to Detroit by private
car. Although they were not refrigerated en route, no keeping
quality problems were encountered since the panels were held
during the comparative cool seasons of the year.

The flower exhibitions were prepared in the labo-
ratory of the Home Economics Department, of Wayne State

University. The flowers were placed in white cardboard

22

23

ice-cream containers (% gal. size). A simple but uniform
arrangement was then designed characteristic of the
exhibitions throughout the experiments. A strong effort was
made to minimize undesirable bias—inducing elements and at
the same time to provide for all factors other than those
being tested to be equal. Likewise, care was taken to insure
that samples from each of the treatments would be carefully
selected to avoid blemishes, variance of shape or state of
openness which might unduly bias the panel members.

A detailed description of the flower exhibitions for
each experiment is provided under the subject of 'scope of

the study'.

Methods
The panel and statistical methods will be described

here as they were common to all tests.

The Panel

 

The Michigan Consumer Preference Panel was initiated
in 1956 to establish consumer preferences relative to grades,
varieties, sizes, color, and processing techniques for
agricultural products. It is under the direction of
Dr. H. E. Larzelere, department of Agricultural Economics,
Michigan State University. The panel met three to five
times a year. The testing materials have been provided by

several cooperating departments. Among them are Poultry,

24

Science, Animal Husbandry, Dairy, Farm Crops, Food Science, and
Horticulture--all of Michigan State University and by the
Home Economics Department of Wayne State University. Oc-
casionally a commercial firm has also participated. The
horticulture department has participated in this project,
since 1957.

For the sake of convenience in discussion, the fol-
lowing presentation is subdivided into three parts, namely:
(1) Selection of the panel members, (2) justification of the
sampling method, and (3) panel procedure.

(1) Selection of the panel members: Mail question—
naires were sent to people selected at random obtained from the
latest Detroit telephone directory. About 5 per cent of
these were returned by the post office for nondelivery.
About 20 per cent of the remainder were filled out and re-
turned either from the first or from a second questionnaire.

The basic questions called for information regarding
age, education, income, and willingness to come to a display
room in Detroit to rate samples of the different products
displayed (Appendix 1). A majority of those selected for
the panel were in the middle income group($4,000-10,000).
They had obtained 12 - 13 years formal education, and were
in the 31 to 45 year age bracket.

Those who returned questionnaires (they were in the

above age, education and income groups) and who were willing

25

to take part in the panel were contacted by telephone to
schedule their visits, and to inform them of the method of
remuneration ($4.00 per visit to each panelist to cover the
costs of parking, babysitters, meals that they might have to
eat away from home and gasoline or bus fare).

(2) Justification of Samplings: This panel was
selected as a sample of a major group of inhabitants of
Detroit, Michigan. On the basis of census information it was
estimated that the age, education, income, of the groups
selected resulted in a panel representing about half the con-
sumer purchasing power in that city. Constant effort is ex-
pended toward periodically modifying the panel to maintain
it as a reasonable index of typical consumer behavior.

A special report on the social profile of Detroit
(1955) by the Survey Research Center of the University of
Michigan (87 ) together with the census information confirmed
the model group used in this study.

Although the technique of the selection was justified,
the justification for choosing the city of Detroit against
other places in the State for this particular study, must be
considered.

Von Oppenfeld (85 ), studied the retail distribution
of floricultural crops in Midhigan. His findings revealed

70 per cent of the floriculture trade was concentrated in

26

seven metropolitan markets in which 64 per cent of Michigan's
population resided. Almost half of all retail florist sales
were made to consumers in metropolitan Detroit.

Although his report was based on research done in
1954, periodic studies of official reports (Census of Agri-
culture, special Reports, Horticultural Specialties) indicates
that this pattern has remained unchanged. On this ground,
using the city of Detroit as the sampling area was amply
justified, since it was necessary for this study to be in one
geographic area.

(3) Panel procedure: Panel meetings were conducted
from two o'clock in the afternoon until ten o'clock in the
evening. Approximately 125 members attended each panel
session with 20 going through at half hour intervals. After
a briefing on the general purpose of the project, panel
members were asked to rank the samples in each series inde-
pendently of the other series. Except where indicated other-
wise, the ranking was based on the order in which the panel
member would select the various samples if they were buying
them, regardless of price (in the afternoon sessions). Price—
tags were attached in the evening when the tests were based
on price. When an individual consumer completed his ranking
of the products, his forms were checked to make sure that he

ranked all the products within eaCh series.

27

Symbols were used to link the actual samples with
the ranking on the panel report card so that numerical or
alphabetical order would not influence selection. The
symbols were alternated so that a particular figure would
not continually indicate a certain quality. The position of
the products on display were also varied in order that the
highest quality was not placed in a standard position. The
symbols used were #, %, &, *, ( ), and +.

As pointed out, in each of the tests, the panel
members were asked to rank the samples in order of their
preferences. They could show equal preference to two or
more of the samples in each series. In some cases they were
asked to answer a short questionnaire which was designed to
give some complementary information, e.g. frequency and

purpose of purchasing flowers, et cetera.

'Statistical Analysis

In the consumer panels, each individual consumer
ranked the samples tested in order of preference, from one
to three, four, five or six, and these data were summarized.
Since it is difficult to make comparisons from the rankings
(revealed by the panelists) directly for overall preferences,

the following five statistical tools were used for each panel

test:

28

(1) a coefficient of concordance, W,

(2) Kendall's Z—test,

(3) estimated ranking,

(4) per cent first preference choice,

(5) distribution of preferences within categories

(samples) of the variable-class test.

The coefficient concordance, W, ranges from zero or
no agreement in preferences among consumers, to l or complete
agreement [Kendall (12)].

Kendall's Z-test is used to test the significance of
an observed W value. That is to say to determine whether
the differences in preferences are real or whether they could
have been due to chance alone. The hypothesis stated that
observers have no community of preference for eaCh test based
on the significance of an observed value of "W". A 1 per
cent risk was Chosen in each single conclusion for all the
tests. When the observed value is greater than those of S
at probability level of l per cent, the hypothesis hence is
rejected, i.e., there is community of preference. And an
estimated ranking is set up for the test.

Kendall ( 12), recommended ranking according to the
sums of ranks alloted to the individuals. The one having the
lowest total should be ranked first; next lowest, second,

and so on. This procedure gives a "best" estimate in a

29

certain sense associated with least squares. In fact, the
sum of squares of differences between what the totals are and
What they would be if all rankings were alike is a minimum
when the ranking is estimated by this method.

A generalization based on preference-intensity for a
certain color sample was projected from the estimated rankings.
The following terms were used for this purpose to help denote
the relative position on a preference-intensity scale; high,
high—medium, medium, medium-low, and low (in a descending
order).

The per cent of first preference choice is a method
by which to select the one most popular product. In other
words, one of several alternative categories is selected, and
the most p0pular product is then called "the preferred"
product.

The distribution of preferences within the category
of the variation-class test indicates the order of choice
within a class and the proportion of the population that may
have a distinct aversion to the category of merchandise
presented.

In addition, in the number preference study for roses
one more test is employed—-the preference index. It is ob-
tained by weighting first place votes higher than second place
votes, second place higher than third, etc. These adjusted

percentages are added to give a single preference score for

30

that category (sample). The highest score in each series
indicates the sample that is preferred by the largest number
of panel members. The amount of difference between the
scores shows the degree of preference for samples.

An arithmetic mean was used in the three following
analyses to describe the entire set of data (i.e. the combined
analysis of tests): (1) per cent in first preference choice,
(2) distribution of preferences within sample classes, (3)
preference index (roses--number preference study only).

Because not all of the samples were present in eadh
test of that variety of experiment due to technical limitations,
a proper approximation is needed in order to combine a set
of data, e.g., in the study of "percentage of let choice of
X sample" and "distribution of frequency of Y sample". The
philosophy of this readjustment is to convert the base
percentage of all tests to a same "footing-place"--as if all
the concerned samples of that variable test were presented.
Hence the figures appearing in the two mentioned studies
were not the originals, but rather the "modified figures".
An arithemetic mean, therefore, was derived from those
modified (adjusted) figures to describe the location of a

set of dats without causing a distortion to the origins.

is

31

All of the tests mentioned here denote in varying

degrees the selection of the one most popular product. This

is considered to be the theme of this presentation.

Scope of the Study

Studies of consumer preferences as previously pointed

out mainly provide a means of seeking appropriate stimuli to

consumer behavior.

throughout the study.

Type of flowers

Carnations

Pom-pon
Chrysanthemums

Variable

color
preferences

number
preferences

color
preferences

number

preferences
(10" bunch)
in diameter

grade
preferences
(size of a
bunch)

Categories

white, red,

dark pink,

pink, variegated
yellow

3 flowers

Noommb

white,

dark bronze,
bronze,

pale bronze,
yellow,

light lavender,
lavender

1 bunch
2 bunches
3 bundhes

6"/bu.
8"/bu.
10"/bu.
12"/bu.
15"/bu.

This philosophy has been underlined

A detailed outline is shown below:

Other feature

day-light vs.
fluorescent

price vs.
non-price

day-light vs.
fluorescent
light

price vs.
non-price

price vs.
non-price

Type of flowers

Standard
Chrysanthemums

Gladiolus

Roses

Variable

color
preferences

number
preferences

color
preferences

number
preferences

color
preferences

number
preferences

grade
preferences
(length of
the stem)

32

Categories

white, pink,
bronze,
yellow,
lavender,

white, red,
pink,
variegated

yellow

3 flowers
4
5
6 ll
8
2

white

dark red,

red

pink,

yellow,

pale lavender

9 ll__
3 flowers
5 ll
7 ll
9 ll
12 "
12 ll...-
same as 9"
15"--
same as 9"

9"
12"
15"
21"

Other feature

day-light vs.
fluorescent
light

price vs.
non—price

Day-light vs.
fluorescent
light

price vs.
non-price

day light vs.
fluorescence
light

price vs.
non-price

price vs.
non-price

CHAPTER IV

RESULTS OF THE CONSUMER PANEL STUDY

Rodes (56), referred to a two-fold objective in
preference studies. MuCh of the literature on preference
studies is rather confusing, he commented, until it is
realized that there has been a difference in objectives
among the studies, though the difference does not appear to
have been explicitly mentioned.

Many preference studies have as their objectives the
selection of one of the several alternative products as the
product to be merchandised to a certain customer population.
Some other preference studies are concerned with merchan-
dising two or more products simultaneously.

In this study, analysis emphasized the one most
popular product.

The detailed description of the analyzing methods
used have been discussed in Chapter III—-Methodology.

Tables presented in this chapter were purposely kept
in two separate categories: Case l--afternoon session and
Case 2-—evening session. This division was caused by either

a Change of one element in the environmental condition having

33

34

taken place, e.g., lighting system: or because a new feature
was introduced to the study, e.g., price-tags. The presenr
tations were designed only to serve as a general reference
and further comparisons are extended in Chapter VI.

Since the greater concern is on the basis of each
type of selected cut flower, the main body of this report
will be divided into five parts according to the five types
of selected flowers. The presentation of the data obtained
for each of the five flowers is submitted in the following
order: carnations, pom-pon Chrysanthemums, standard Chrysan-

themums, gladiolus, and roses.

Carnations
Two major investigations were conducted for this
flower: (1) color preference and (2) number preference (unit-
of—purchase). The former one had six replicates overlapping
a range of five years--l957 to 1962, whereas the latter one

had two replicates all performed in the year of 1962.

Color Preference

Six colors were Chosen for this investigation: white,
red, dark pink, pink, variegated (red and White) and yellow.
These are considered to be the prominant colors in carnations.
Samples consisted of units of five.

The scheme of this investigation follows:

35

Date of Test Testinggsamples (color categories)

October 1957 red, dark pink, light pink, white

November 1957 red, dark pink, light pink, white, variegated
December 1957 red, dark pink, light pink, white

May 1961 red, light pink, white, variegated

February 1962 red, dark pink, variegated

November 1962 red, dark pink, light pink, white, variegated,
yellow

Table 1 summarizes the rankings of the six tests.

Table 2 indicates the percentage of lst-choice
selections for the six color samples and the distribution of
frequency of choices for each color sample.

A graphic presentation abstracted from Table 2 is
designed to provide a closer look at the data by comparing
the average percentage of the lst choice of each color sample
with the expected average (Fig. 1). There were only four
colors with above average-preference (16.7 per cent): dark
pink, variegated, white, and yellow (tested once). There was
no material difference in the three preferred colors.

An interesting and quite notable change in first
preference occurred in red carnations between the tests of
1957 and the two tests of 1961 and one test in 1962. The
1957 tests showed high preference for red carnations. The
more recent tests showed a low preference for red. Whether
it was due to "seasonal variation" or a combination of influ—

ences is unknown.

36

 

 

msmm.* ooH 6mm scam .omumm spans
unwed tofinm> N ommo
mmom.s mm 6mm spans xcam smumm
unqu IoHHm> H ommu
aoma mm:
mmsm.* om scam when: scam 6mm
usuaq sumo m ammo
Noom.s we seem wears sum scam
unwed Mama H mmmo
nmma .omo
mama.* on muHm>oz scam wears xcam sum
usqu Mama N ommu
vmma. em H ommo
umma .>oz
mmeo.* we wear: sum scam scam
prong sumo m ammo
emmH.t am aware scam 6mm scam
unmau sumo a mumo
nmma .uuo
s mesmeuoo Hussmaoo rum gum sue sum 6cm uua
Icoo no mo Honfisc Dcofifluomxm
ucowowmmoou Hmuoa mafixcwm ooumeumm mo oumn

 

 

+.mmcexcmu ooumefiumo "MCOADmcumo CH mocwuommum HoHoo uoESmcoolt.H magma

37

COConomoum CH 30H ECHooE omMCmulGCHm umeH oCm mpHSB

mummu NmmH .HmmH CH
COCwnomon CH 30H oCm mummu bmmH CH CUCoHommHm CH ECHCoEISmHfi oomentloou

oucouomowm CH ECHooEICmHS mm3IIAoUCo omumopv 30HHoM

Lemma .>oz
mo N ommo Hmouxov ooCouommum CH £0H£ ou ECHooEISUHS Cmutl=ooummoHum>=

oUCoHomon CH £0H£ mHuCoumHmCoo mCBtIxCHm xumo
"oozogm momma oBu ox» mo mmCHCCHm one

.Co 0m oCm .oCooom .umoon uxoC “umHHm oomeu mm3
AmMCmH mo Edmv HMDOD umo3oH on“ mCH>m£ oCo .mmmH .ANH V HHmoCmM .mfiw.m oom+

 

mCHusmHH RBBIICOHmmom mCHCo>o "N wmmo Ho>oH RH pm .mHm
ugmHH mmoIICOHmmom COOCHoumm "H mmmu a
mnmN.e HOH Com xCHm ouHCB .nmumm 30HHow MCHm
DSmHA IwHHm> Mnma N ommu
memm.* mm was seam spars soaamw. V ewumm, xch
quHq uoHnm> sumo H mumo
NomH .>oz
nNHH.s mm Com xCHm . Coumm
MHMQ IOHHM> N OmMU
mmOH.¥ No Com xCHm ..om»wm
MHMQ l0flHM> H OmMU

NomH .Qmm

38

.Q oCm m CH meutumH mo
ommuCounom mCHCCommoHHoo CHmCu Hmsqo DOC HHH3 onEMm CHmquU w you

 

 

 

 

 

U CCm C CH CSOCm CUHOCotumH mo mommuCoouom may .mooCouomonmnoHu Ho omsmoom
. .AquHH

ooCoomoCosHm oxCHon ouHSB ECCSV mCHquHH N33 CDH3 COHmmom mCHCo>o ”N ommo

.uanthmC CH Conmom COOCHoumm ”H ommo
m.HH m.HH m.m b.5N w.mH m.mH «.mH H.0H m.¢H m.mH m.0N m.wH m.ON m.5H 30HHoN
m.hH 0.0H N.bH N.mH m.NH m.HN N.0N b.0H h.mH m.eH n.m N.hH b.0N m.mN CoummoHum>
m.oH ¢.bH m.NN o.mH m.mH m.NH m.m ¢.mH m.mH H.mH H.NN H.mH m.NH m.HH MCHm
n.eH o.mH 0.0H h.hH m.0N m.HN 5.0N mva H.¢H m.HH h.mH m.mH ¢.¢N o.MN xCHm Mama
m.mH o.oH m.HN m.©H m.mH o.oH m.mH m.mH m.oH o.mH o.mH m.mH m.mH N.mH Com
h.mH N.©H N.vH m.vH m.mH N.0N oJmH H.mH m.wH 5.0H m.hH m.mH H.NH ¢.HH ouHCS

qu0 Mom Ame DCmU Homflmg qu0 Com Ana uCoo HmmAKu
Cum Cum Cue ohm CCN umH oUHOCUIumH Cum Cum Cue Cum UCN CmH oUHOCUIDmH monECm
ooCoComwnm molnCOHuCQHHumHQ .Q .U oUCmemon mo COHCCQHHpmHQ .m .< HoHoo
.Ntommu, meommO.

 

 

.onEmw m CCH3 mooHOCU mo moCmdwoum mo COHDCQHHumHC on» oCm CUHOCUIDmH
mm onEmm mCHuomHmm HmCmm Ho qu0 Com

umCoHDMCumo CH CUCouomoum COHoo HoECmCOOII.N oHnme

\
. 3». s, 3.. ob

l...\. .ltvii t l

   

UN gum. v? «Q

  

 

-- - - - TNT” xvi:

  
     

 

mint m E 2 Tm: SE 2am \Zml

y. El «3.». «.qu not shes

 
 

USN v
uwsmktm \

 

 

WZOCQZECQ

N wwnu I _ menu
. . "mrwktmuguk Wmnuvapux mmcuumxm of. «.23
sign :5 Sum E 3.53 .3 at u... Russia. uwmrusw 2: us Entire 2: Eu

 

      
 

 
 

      

 

.
['5

w-

t’%) 971%?-

(7/0) ”NW-.151

40

The distribution of frequency of choices of these
six color samples did not give a clue to any new findings
other than the facts related above.

Sherman, et. al. (115) recommended a "write-in"
technique for testing flower color preferences--which was
not associated with the particular samples of flowers. Hence,
the answers would be influenced in no way. Le Clerg (112)
adopted this method for his survey in 1958. This research
method was contrary to the method employed here in this
study (appearance of actual flowers). Therefore an experiment
was conducted in February 1963 using the "write—in" method
to test Whether the two research methods would lead to dif—
ferent results. Part of the results of that panel along with

Sherman's and Le Clerg's findings are reported here:

No Pref-
Red Yellow Pink White Mixed Other erence
Sherman 46.4 .9 8.9 12.5 10.7 4.5 16.1
Le Clerg 38.4 2.1 31.9 19.1 2.8 0.4 5.3
Write-in 17.8 5.3 16.4 16.4 4.0 15.8 8.6
panel (dark) (variegated)
15.8
(light)

Apparently there was no agreement between the three "write-in"
experiments. Yet there was great similarity between the
February "write-in" panel and the regular "visual" panel of

this study—-where the products were seen (c.f., Fig. l).

I97

41

In short, the above analysis revealed that there was
no significantly "preferred" color in carnations. Panel
members however did show relatively higher preferences for
dark pink and variegated (red and white) over other colors

(Table 2).

Ngmber Preference

Six "units-of—purchase" categories were chosen for
this investigation: each sample consisted of 3, 4, 5, 6, 8,
or 12 flowers. Emphasis was stressed on the small units.
Popular comment indicates that the demand for cut flowers is
different from the present 'designed' floral products con-
sisting of large units. Dark pink carnations were used.

The scheme of this investigation was:

Date of Test Testing Samples Special Feature
(unit categories) (in case 2)

May 1959 5, 8, and 12 *

February 1963 3, 4, 5, 6, 8, and 12 30¢ each

60¢ each (in doz.)

Table 3 summarizes the results of the two tests in
rankings. The results showed no significant price—effect on
the unit-of—purchase. The units of three had low preference
in both cases, whereas the units of five and eight showed

high preference in the two cases.

42

Table 3.--Consumer number preference+ of Carnations:
estimated rankings.

 

 

Date of Estimated Ranking Total Coefficient
Experiment number of of concord-
lst 2nd 3rd 4th 5th 6th consumer ance W

 

May 1959
Case 1 48 .0013
Case 2 5 8 12 82 *.3532
February
1963
Case 1 8 5 6 12 4 3 63 *.3787
Case 2 8 5 6 4 12 3 86 *.3532

 

*Significant at 1% level

+Numbers indicate numbers of flowers in a unit

Case 1: afternoon session (not priced)

Case 2: evening session (priced)

In Table 4 is presented the analysis of the per-
centage of lst choice of the six unit samples and the
distribution of frequency of choices for each unit sample.
No attempt was made to draw a further conclusion on the lst-
choice analysis, since it was considered with only one
replicate. Furthermore, an erroneous price-tag appeared on
the units of twelve.

However, the study of the "distribution of frequency
of choices within a sample" indicated:

units of 3 and 4 were predominantly in the low ranks
preference in both cases,

ACCUHHQV COHmmom mCHCo>o "N ommo

ACCUHCQ HOCV Conmom COOCCouHm “H ommo

 

43

m.mN b.HN N.mH N.ON N.o m.m m.m h.mH h.mH H.MH N.®H v.MH m.mH o.mH CoNOC

o.m v.NH m.©H H.m o.mN ¢.nN ¢.>N ¢.m N.¢H >.vH N.0H o.hN 5.0m o.mN m

m.N H.m m.NH N.om m.vm @.HH o.HH m.o m.m ¢.mN m.mm b.0N m.mH m.¢H o

H.0H H.0H b.5H N.mH m.hH ¢.oN v.0N ¢.m m.@H ¢.mH o.mN ¢.oH m.mH H.mH m

H.mH m.mH m.mN ¢.nH m.NH m.m m.m H.v o.>¢ m.mH m.hH m.m m.v m.¢ e

m.ov ¢.Hm o.¢H m.m 0.0 m.N m.N m.Ho m.mH m.vH N.m N.m o.H ¢.H m
“Coo Com Only DCwo Mom Aha

 

Cum Cpm Cue Cum CCN DmH ooHOCU Cum Cum Cue .Uum UCN umH mUHOCU monECm
CUCoComon mo COHDCQHCDmHQ CmH oUCoHomoum mo COHDCQHHumHD umH uHCD

 

 

N ommu H ommo

 

.onEmm m CHCuHs mooHOCo mo moCmswouw Ho COHuanuumHo oCu CCm CUHOCU umH mm
onEMm mCHuooHom HoCmm mo uCoo Com "mCOHumCCmo CH CUCCHoHon HoQECC HoECmCOUII.¢ oHQmB

44

units of 5 and 6 were predominantly in the high-
medium to medium ranks. This was more
evident in Case 2,

units of 8 were predominantly in the high ranks in
both cases,

unit of a dozen showed two modal preferences, one in
high the other in low.

Therefore within the scope of this study, the unit

of eight was preferred by more panel members than the other

five units.

Pom-pon Chrysanthemums
Three major investigations were conducted for this
flower: color preference, number preference (unit-of—
purchase), and grade preference (diameter of the bunch). The

next few paragraphs are devoted to the discussion of these

three studies, respectively.

Color Preference

Seven colors were chosen for this investigation,
namely: white, dark bronze, bronze, pale bronze, yellow,
light lavender, and lavender. These are considered to be the
prominant colors in pom-pons. The samples (each bunch) were
10" in diameter.

The sCheme of this investigation was:
Date of Test Testing Samples (color categories)
October 1957 white, yellow, bronze, pale bronze, light

lavender

November 1957 white, yellow, bronze, dark bronze, lavender

December 1957 white, bronze, dark bronze, lavender
February 1962 white, light lavender

45

November 1962 white, yellow, dark bronze, pale bronze,
lavender, light lavender

Table 5 summarizes the results of the five tests in

rankings.

The findings in the two cases showed:

Light lavender ran high-medium to high in preference

Pale bronze was high—medium in preference

White ran medium-low to high in preference

Bronze ran medium to low in preference

Dark bronze ran low, except the Nov. 1957 test (case 2)

Lavender varied in the extremes

Yellow varied through the preference scale

Table 6 shows a study of the percentage of panelists
Who assigned lst choice to each of the seven color samples
and the distribution of frequency of choices for each color
sample.

A graphic presentation abstracted from Table 6 is
submitted in Fig. 1. It is designed to provide a closer look
at the data by comparing the average percentage of the first
choice of eaCh color sample with the expected average.

There were four colors with the above average prefer-
ence (14.3 per cent). In Case 1 (afternoon sessions), they
were pale bronze, dark bronze, yellow, and light lavender,
but in Case 2 (evening sessions) they were pale bronze, light
lavender, white, and yellow. However, panel members showed

a higher preference for pale bronze than for the other prefer-

red colors in both cases. Furthermore, high frequency of

AmCHDQmHH x33 mCHmCV COHmmmm mCHCw>m "N mmmu
ADCmHH mop CHV COHmwom COOCHoumm ”H ommu

Hm>mH RH um ucmonchHms

46

 

HHmN.*

HOH

HOUCO>MQ ONCOHm

muHCs

ONCOHm

HooCo>MH

3OHHON

 

 

N ommo
Mumn onm DHOHH
mooN.* mm oNConm HooCo>mH oDHCS oNCoum HoCCo>mq BoHHow H ommu
Humo onm UCmHH NomH HmQEo>oz
meno.e mm CoCCo>mq ouHCB
DSmHH N 0mmo
mnoo.* N@ H mmmu
momH mumsnamm
mmHH.e om oNCoum oNCoum muHCB HopCo>mH
Mqu N omwo
mHmH.« we onCoum oNCoum ouHCZ CooCo>mH H wmmu
Muma hmmH Coflaooon
HheH.* on BoHHow ouCoum HoCCo>mH oNCoum ouHCB
men N ommu
HNmo.s em BoHHmw CooCo>mH oNCoum ouHCS oNCoum H ommo
Mumn hmmH HwQEo>oz
NmHH.* mo oNCoum opHCB BoHHoM oNCoum uoCCm>mH
mHmm “Cqu m sumo
mnhH.* Hm oNCoum ouHCB BoHHo> oNCoum HooCo>mH H ommo
mHmm pana smmH umnopuo
3 oUCm HoECmCoo Cum Cum CDC Chm CCN umH uCoEHuomxm
ICCooCoo no mo HoQECC mo oCmn

mCHMCmm ooumEHumm

ucmHonmmoo Hmuoe

 

 

.mmCHXCwH omumEHumw umComIEom CH mUCoummon HoHoo HmECmCOUtt.m mHQMB

47

AmCHusmHH x33 mCHmCV COHmmom mCHCm>o “N ommu
ADCmHH moo CHV COHmmow COOCHmumm "H ommo

 

 

 

 

 

H.vH ¢.HN C.¢H 0.NH m.0H m.vH v.NH H.NH H.0H m.hH H.0H m.0H 0.0H h.mH 0.HH >.HH CooCo>mH
0.VH 0.N. 0.0. «.0 0.0H n.0m b.0H N.mH 0.0H 0.0- 0.v b.0H 0.0H m.HN 0.0H 0.0H CoCCo>mH
unmaa
0.¢H N.0H 0.HN ¢.¢H 0.0H 0.0H H.mH >.¢H 0.0H N.0H H.MH 0.NH m.mH b.0H 0.0H 0.mH 3oHHow
0.0H 0.N 0.mH 0.0H v.MH 0.0. 0.0N. .v.nN 0.¢H N.H 0.0H N.HH H.HH ¢.b 0.0m N.Hm oNCoum
onm
0.0H 0.0H 8.5H v.5H ¢.NH v.0H m.n 0.5 0.0H 0.0H 0.0H 0.0H 0.0H m.MH 0.0 0.0 oNCoum
0.¢H 0.vH 0.0 H.MH N.mN b.0H m.n m.> 0.0H 0.0H 0.0H 0.0H N.HN N.NH 0.0H v.0H oNConm
Hymn
H.0H >.NH m.NH 0.mH N.vH w.vH 0.0H H.0H H.¢H m.NH m.mH C.mH 0.0H H.0H N.¢H 0.MH muHCS

uCoU Hog ARV uCoo uCoo Mom Axv quo

I. ComAxv I. Com AME
Cub CDO Cum Cue ohm CCN umH CUHOCU Cum £D0 Cum Cue Chm CCN umH oUHono monEmw
wUCoHoHon mo COHDCQHCDmHQ umCHm CUCCHommCm mo COHCCQHHumHQ umnflh HOHOU

HN mmCUQ HH ommov
.onEmm

m CHCCH3 mmoHOCU mo >0Codqoum mo COHDCQHHumHo oCm oUHOQUIDmH mm onEmm mCHuooH

low mHoQEoE HoCmm mo uCou “mm "mECEosqumhuzo Comtaom CH oUCmHmmon HOHOU HoECmCOOII.0 oHQmB

48

preference for pale bronze was found in the first preference--
37.7 per cent (in case 1) and 25.5 per cent (in case 2) as
compared to the next 6 lower ranks (c.f. Table 6).

A "write-in" panel for color preference in pom—pons
was conducted in February 1963. The results of that panel
were quite different from the "visual" panels. The results
of the "write-in" panel showed the following preferences:
yellow was preferred by 26 per cent of the panel members, no
color preference by 19 per cent, white by 17 per cent, dark
bronze by 10 per cent, bronze by 10 per cent, pale bronze by
7 per cent, mixed by 4 per cent, dark lavender by 3 per cent,
pink by 3 per cent, and lavender by 2 per cent.

The inconsistency in these two methods of testing may
have arisen because of the inaccurate color concepts possessed
by the consumers expecially with the colors 'lavender' and
'bronze'. The panelists may also have been hindered by their
limited knowledge on the distinction between pom-pons and
standard Chrysanthemums.

When the preference scores of the three levels of
color saturation in bronze were combined, this figure indicated

26 per cent preferred the color of bronze. This may help
to interpret the deviation between the two findings.

Conclusion: consumers were not color conscious in

selecting pom-pons. Whether this was due to limited knowledge

49

about the product was not known. Pale bronze was more

popular than the other six colors (see Table 6).

Number Preference

Three units-of—purchase were chosen for this investi-
gation: 1 bunch, 2 bunches, and 3 bunches. The size of the
bunch was 10" in diameter and the color was White.

The scheme of this investigation was:

Date of Test Testing Samples Special Feature
(unit—of—purchase) (in case 2)
February 1963 l, 2, 3 bunches $3.79 per bunch

The results of the findings were analyzed in ranking
and preference distributions within a sample and among
samples (Table 7).

Although this study was conducted only once, the
results indicated significantly the preference for three
bunches when no price was placed on the samples. The price—
mechanism reversed the results from a high preference for the
three bunch units to the high preference for a one bunch

unit (Table 7).

Grade Preference

The grading system was determined by the size of the
bunch. Five sizes were used for this study: bunches 6", 8",
10", 12", and 15" (in 'diameter ). Samples Were in

lavender in both tests.

50

ACCUHCQ v Conmom mCHCo>o "N ommo

ACCUHHQ COCV COHmmom COOCCoumm "H ommo

 

 

 

 

 

 

 

 

m.e0 H.0H m.mH H.mH m.0 H.mm 0.4m m.mm uwgucsn m
m.mH m.m0 o.mH 0.mH m.~e H.mm m.HH m.0H mmnucsn m
m.0H e.HH m.00 n.00 0.04 ~.- H.Hm m.om Conan H
uCoo Com uCoU Com DCoo Com uCoo Com
sum cam nuH moHoCuuuuH sum 6cm uuH moHoCouuuH umHmsmm
oUCoHoHon mo ComusnmuumHn wUConomoum M0 COHDCQHHDmHD DHCD
.Ntmmmufl H ommu
oUCwuomon mo COHUCQHHumHQ
Hm>mH aH um ucmuHchmHm
k.
memm.a 0m manages m mmsocsn m Conan H m mumo
memH.« m0 umsucsn N Conan H umnocsn m H sumo
momH mumsunmm
3 CUCC HoECmCOU ohm oCN umH quEHHomxm mo oumn
toHooCoo no mo HoQECC
ucmHoHummoo Hmuoe mCchmm.empmsHuum

 

 

mCchmm nmumsHuum

.mECEmfluCMmmnnu ComIEom CH CUCCHommHm HoQECC HoECmCOOII.h oHQCB

51

The scheme of this investigation was:

Date of Test Testing Samples (size of the bunch)

February 1962 6" ($2.15), 8" ($2.75), 10" ($3.50), 12" ($4.25)
February 1963 6’' ($2.29), 8" ($2.99), 10" ($3.79),
12" ($4.49), 15” ($5.59)

Table 8 summarizes the results of the two tests.

Table 8.--Consumer grade preferences of pom—pons: estimated

 

 

rankings.
Estimated Ranking Total Coefficient
Date of number of of concord-
Experiment lst 2nd 3rd 4th 5th consumers ance W

 

February 1962

Case 1 8" 12" 10" 6" 62 *.5539

Case 2 10" 8" 12" 6" 85 *.2480
February 1963

Case 1 15" 10" 12" 6" 8" 63 *.3864

Case 2 10" 12" 8" 15" 6" 86 *.1008

 

Case 1: afternoon session (not priced)

Case 2: evening session (priced)

The finding showed:

6” was low in preference in both cases,

8” varied in preference in case 1 and was medium in
preference in case 2,

10" moved from high-medium in preference in case 1 to
high in case 2,

12" ran medium in preference in two cases

15" was high in case 1 yet low in case 2

Table 9 presents a study of the per cent of the five
grade samples scoring lst-choice and the distribution of

frequency of choices for each grade sample.

52

ACCUHCQV COHmmom mCHCo>o "N ommu

 

 

ApooHum DOCV COHmmom COOCHouHm "H ommu
5.00 0.5H ¢.>H 0.0H 0.0N N.HN N.m 0.v 0.0H m.0 v.00 b.N0 =mH
0.0H 0.0N 0.0H N.mN 0.mH 0.0H 0.0H 0.0H 0.5N 0.0m H.0H 0.5H =NH
0.HH N.v m.¢N N.mN 0.0m 0.00 N.mH n.h 0.mm b.5m 0.0H m.hH =0H
v.0H 0.0N 0.0m r.mN N.mH 0.mH N.Nm 0.0N n.0H 0.0H 0.0H «.5H :0
0.0N N.¢m 0.mH 0.0 v.0H 0.0H 0.8N ¢.mv 0.0H 0.0 «.0 0.0 CopoEMHo CH :0
“Coo Com Axv X.Axv pCoo Com ARV $.Axv
Cam ape sum cam uuH onoCo Cum Cue sum cam uuH moHoCo umeamm
ooCouomoCm mo COHCCQHHDmHQ umH COCoHomoum mo COHDCQHHumHQ umH ocmno

 

 

MN CWMU. H @mMU.

 

 

.onEmm
m CHCMH3 mooHono mo moCosvoum mo COHDCQHHumHo any CCm CUHOQUIuwH mm monEMm
mCHuooHom HoCmm Ho uCoo Com umECECCUCmmmnnu ComIEom mo CUCCHmmon mcmum HoECmCOOII.0 oHQma

53

The study to determine the "per cent of each grade
scoring first place" revealed several interesting points.
There was price-effect in the 15" bunch. When price was not
a factor the 15" bunch scored high at 62.7%, when price was
a factor it drOpped to the medium level (21.2). On the other
hand, when price was not a factor, the 10" bunch scored 17.3.
When it was a factor it increased to 34.6 per cent. Although
the preference for the 6" bunch was increased from 6.0 per
cent in case 1, to 14.3 per cent in case 2, the apparent
preference score for this small bunch was still low compared
with the others. The 8" and 12" bunches indicated no price—
effect. They remained in the preference range.

The above statement was confirmed by the third
analysis--"the distribution of frequency within a sample?.

Conclusion; when price was not a factor the consumers
indicated the 15" bunch to be their preferred size (62.7 per
cent of consumers in this group). When price became a factor

the preferred size changed to the 10" bunch (Table 9, case 2).

Standard Chrysanthemums
Two major investigations were conducted with this
flower: color preference and number preference (unit-of-
purchase). The following paragraphs are devoted to a discus-

sion of the results of these two studies.

54

Color Preference

Five colors were chosen for this investigation:
namely white, pink, bronze, yellow, and lavender. These are
the prominant colors in standard chrysanthemums. Samples
were in units of five in all tests.

The sCheme of this investigation was:

Date of Test Testing samples (color categories)
October 1957 white, yellow, bronze, lavender

November 1957 white, yellow, bronze, lavender
December 1957 white, yellow, bronze, pink

Table 10 summarizes the results of the three tests
in rankings.

The findings of the two cases (sessions) showed:

Yellow was consistently high in preference.

Pink (tested once) was medium in preference.
Bronze and lavender were medium-low in preference.
White varied through the preference scale.

Table 11 indicates the percentage of lst-choice
selections for the five color samples and the distribution
of frequency of Choices for each color sample. A figured
presentation abstracted from Table 11 is included to
graphically compare the average percentage of lst-choice
selections of each color sample with the expected average

(Fig. 2).

55

Avon: mCHuflmHH N330 COHmmwm mCHCo>m "N ommo

AuanHt>m0 CHV COHmmom COOCnoumm "H wmmo

H33 sH .t... ucmoHuHumHm

 

 

 

.1
eHmH.s om museum muHrz xch onHm» m wumo
oneH.* we muHCs xch mucoum soHHmw H mumo
HmmH nonsense
mamm.* on mucoum Coecm>ma muHss 3oHHmw N sumo
oomH.s em umecm>mq mucoum muHsz 3033 H sumo
nmmH amnem>oz
Hmem.s 00 mucoum umecm>mq muHss onHmw m sumo
~no~.s Hm umecm>ma museum mpng 3oHHow H sumo
HmmH Cupopuo
3 mUCm HoECmCoo Hue Chm UCN umH mHmEmm
IonooCoo Ho Ho ConECC HoHOO
uanonmmoo Hmuoe mCchmm eupmermm

 

 

.mmCHMCmH CowEHumo

umECEoCqumquU onmpCmum CH oUCoummon COHOU umECmCOOII.0H oHQMB

56

 

 

AComC mCHquHH x330 COHmmom mCHCo>o "N ommo
ApflmHHtmmo CH0 COHmmow COOCHoCmm "H 0000

0.0N N.0N m.>m N.hH 0.0 H.0 0.0N 0.mN ¢.Nm v.0H H.0 0.0 HCCCo>mH
0.0N H.H 0.0H 0.mN m.Nv 0.0a 0.0N b.m 0.NH h.vN 5.0m N.>m 3oHHow
0.0N h.N¢ m.nH 0.0 H.0H 0.0 0.0N 0.0m m.>H H.NH 0.0H N.0H oNCoum
0.0N 0.0H 0.hN 0.0H 0.0H 0.0H 0.0N n.HN 0.0H 0.0m h.HH >.HH MCHm
0.0N >.NH m.hH 0.0N ¢.HN 0.0N 0.0N 0.0N 0.0N n.NN 0.0H 0.0H ouHCB

uCoo Com Axv X.Axv CCoo Com ARV .X Axv
Cum Cue sum cam umH moHono Cum Cue cum 6cm, uuH «UHogo umHCeum uoHoo
moconomoum mo COHuanuumHn lumH CUCoHomonm mo COHDCQHHumHQ IDmH

 

1m mumow,

HH mumoa

 

HoCmm mo uCoo Com

.onEmm m CHCDHB
mooHOCo mo SOCosqonm mo COHDCQHHumHC o3» oCm oUHOCUtumH mm mmHmEmm mCHuooHom
umECEofiqumausu ChmoCmum CH COCoCoHon COHoo HoECmCOOII.HH oHan

57

\whcmxm.‘ 302%; 00.3th «at $.53

 

 

$9.3. Q \‘
n Numkxm

  
  

 

 

0202m152¢mxm$ 030230

 

a wenseouhvnx vmwxmxwx meumlkm 92 3.5; oISNW .53 «NW
LQK Mounxxvtkvd Us *0 ”@t8\& UTNLMSQ 2% WC tnhtm‘nxtaw CHE .N .mmuH

Q4

\2

b1

MN

om

Ma

Q}

‘2.

(%J 3310;]: -4SI

58

There was only one color yellow with an "above
average" preference (20.0 per cent) in case 1 (afternoon
sessions). Two colors--yellow and White ranked above the
expected average in case 2 (evening sessions). In both cases
yellow was highly preferred over the other four colors. In
case 1 (afternoon sessions) the model preference was for
yellow: 37 per cent preferred yellow, 19 per cent preferred
bronze, 16 per cent preferred white, 12 per cent preferred
pink, and 6 per cent preferred lavender. In case 2, (evening
sessions) yellow was more popular than white (Table 11).

These findings all agreed with the results of the
analysis of "distribution of frequency of choices" in Table
11. Only yellow had high frequencies in the first two ranks
(case l--afternoon sessions: 63 per cent; case 2--evening
sessions: 68 per cent).

The following table shows the results of three "write-
in" experiments—~Sherman's finding (115), LeClerg's finding

(112), and the Feb. experiment, 1963.

Yellow Pink White Mixed Other No Answer
Sherman 56.6 .9 12.4 6.2 6.2 17.7
LeClerg 56.4 3.9 12.6 9.8 11.2 6.0
February 41.6 4.7 12.8 4.7 20.8 (Bronze) 11.4

1963 4.0 (Lavender)

59

There was agreement in the overall results of the three
"write-in" studies. The February panel conducted as a part
of this study indicated bronze was next to yellow in popularity.
Conclusion:
All of the tests indicated unanimously that yellow

was the preferred color .

Number Preference

Six units-of-preference were chosen for this inves-
tigation: 3, 4, 5, 6, 8 and 12 standard mums. All samples
were yellow in color.

The scheme of this investigation was:

Date of Test Testing Samples Special Feature
(unit-of—purchase) (case 2 only)
February 1962 3, 4*, 5, 6 90¢ each
May 1962 3, 4, 5, 6 90¢ each
February 1963 3, 4, 5, 6, 8, 12 83¢ each

*mispricing, 34¢ instead

Table 12 summarizes the results of the three tests.

The tests showed a great price-effect on the consumer
number preference in standard chrysanthemums. In general,
when price was not a factor the large units were more prefer-
red than the small units. However, when price became a factor,
the small units became more popular. There was only one ex-

ception, the unit of three was low in preference throughout.

 

Date
Ema

Feb:

the

'U
"i

ls

in

fit

Ca:

60

Table 12.--Consumer number preference+ for standard chrysan-
themums: estimated rankings.

 

 

 

Estimated Ranking Total Coefficient

Date of number of of concord-
Experiment lst 2nd 3rd 4th 5th 6th consumer ance W
February 1962

Case 1 5 6 3 4 62 *.3340

Case 2 4 3 5 6 85 *.0704
May 1962

Case 1 5 6 4 3 40 *.3201

Case 2 4 5 3 6 78 *.0815
February 1963

Case 1 12 6 8 4 5 3 63 *.2504

Case 2 6 5 4 8 3 12 86 *.1508

 

*Significant at 1% level

+Numbers indicate the number of flowers in a unit.
Case 1: afternoon session (not priced)
Case 2: evening session (priced)

The shift in popularity when price became a factor in
the small units (of 4, 5, and 6) was dependent upon the offering-
price.

Table 13 presents an analysis of the percentage of
lst-choice and the distribution of frequency of choices With-
in a sample. In case 1 where the flowers were not priced,
units of 12 were preferred significantly over the remaining
five units (with 49 per cent of consumers in this group). In

case 2, when the flowers were priced the popularity of the

ACCUHHQV COHmmom mCHCo>o “N ommo

ACooHCm DOCV COHmmom COOCHoDmm "H ommu

.mCHoHHm DUCHHOUCHIIMDCU N00H humCCCmm mCHCCHoxm
CH.

 

61

 

 

 

5.50 H.mH 0.0H m.m 0.5 0.0H 0.0H N.m 0.0 H.HH 0.5H m.0 0.¢m 0.0a ACmNoov NH
H.0 N.hm 0.0H 0.b 0.0H 0.NH 0.NH H.HH h.NH 0.0H 0.h H.0m 0.0H m.eH 0
n.0H N.NH 0.0N v.0H m.mH N.¢N 0.mN n.0H N.NH b.5H 0.0H 0.0H 0.0N 0.0H 0
0.0H N.HH 0.0 0.00 N.0H 0.0H N.mH ¢.mH 0.>H h.mH m.HH 0.0N b.5N N.vN m
0.0H 0.0H 0.0H 0.NH 0.0N 0.mH 0.0H 0.0 0.NN m.mN 0.0N >.mH N.h 0.0 e
0.HN 0.0H m.0H H.NH 0.0 0.0N v.0H 0.0m H.0H 0.NN v.0H 0.NH 0.0 0.0 muo3on m
DCoo Com Amp X Amp DCoo Com AMV x Amy
CDO CDm CDC Cum CCN DmH ooHOCU CDO CD0 CD0 Cum CCN DmH oUHOCU monEMm DHCD
COCouomonm mo COHDCCHCDmHQ IDmH COCoHomon mo CoHDCCHCDmHQ IDmH
“N ommoe NH mmmo,

 

 

.onEMm m CHCDH3
mooHOCo mo moCoCUmum mo CoHDCCHHDmHo 0CD CCm CUHOCUIDmH mm monEmm mCHDuoHom
HoCmm mo DCoo Com umECEoCDCmm>HCU CCMCCMDm CH COCoComonm CCCECC HoECmCOUII.mH oHCmB

62

units of twelve fell to 11 per cent and the units of six be-
came more popular (Table 13). While the units of three and
four doubled thepreference score as the price factor was
introduced (units of three-—7 per cent to 16 per cent; units
of four--6 per cent to 13 per cent), no price—effect was found
in the units of eight.

The analysis of the distribution of frequency of

choices conveyed the same message as the preceding paragraph.

Gladiolus
Two major investigations were conducted using this
flower: color preference and number preference (unit-of-
purchase). The next few paragraphs are devoted to a dis-

cussion of the findings in these two studies.

Color Preference

Six colors were chosen for this investigation, namely:
white, red, pink, variegated (pink and White), yellow, and
lavender. They were the prominant colors in gladiolus.
Samples were in units of five in all tests.

The scheme of this investigation was:

Date of Test Testing Samples (color categories)

May 1962 white, yellow, red, lavender, variegated
September 1962 white, yellow, red, lavender, pink
November 1962 white, yellow, red, pink

Table 14 summarizes the results of the three tests.

63

Hooms mCHDCmHH x330 COHmmom mCHCo>o “N owmo

ADCmHthmC CH0 COHmwom COOCHoDmm "H ommu

Hm>mH NH Du DumOHchmHm

.1

 

 

 

 

0000.4 HOH pom oDHCB CCHC 3oHHoN N ommo
0550.* 00 Com oDHCB xCHm 3oHHow H omwo
NO0H HoCEo>oz
m000.s e0 oDHCB Com 3oHHow HoCCo>mH CCHC N ommo
0050.4 0v oDHCB CoCCo>mH Com CCHC 30HHow H ommu
memH nonsmummm
000N.a 05 BoHHow uoCCo>mH Com CoDmmoHum> oDHC3 N owmo
05¢N.a 00 BoHHow HopCo>mH poDmmoHnm> Com oDHCZ H ommo
«mmH mm:
3 ooCm CoECmCoo CD0 CD0 Cum CCN DmH DCoEHHomxm
louooCoo 00 mo uoCECC 00 meQ
DCoHoHHHwoo HMDOB mCHMCmm CoDmEHDmm

 

 

.mmCHCCmu ooDmEHDmo

umCHoHanw CH oUCoHomoum HoHoo HoECmCoot|.¢H oHCmB

64
The findings of the two cases (sessions) showed:

Pink ran high or high-medium in preference,

Variegated (tested once) ran high-medium to medium in
preference,

Lavender was in the medium range of preference,
Red ran medium to low in preference,

Yellow varied in the extreme in preference (high and
low),

White varied along the scale of preference.

Table 15 shows the per cent of panelists Who selected
each of the six color samples as their lst-Choice and the
distribution of frequency of choices for each color sample.

A graphic presentation abstracted from this table compares
the average percentage of the first choice of the color
samples with the expected average (Fig. 3).

Four colors were above the average-preference (16.7 per
cent) in case 1 (afternoon sessions)--white, pink, red, and
yellow. Only two out of the four remained above the average
in preference in case 2 (evening sessions)--white and pink.

The "distribution of frequency" of Choices of the six
color samples did not yield a favorable position for either
White gladiolus or pink ones. No individual "preferred"
color was clearly indicated in gladiolus. However both the
white and pink gladiolus (salmon in trade terms) were more
popular than the other four colors. The overall results com-
pared favorably with a similar study conducted by the

Mississippi Agricultural Experiment Station, (112).

65

 

 

Avon: mCHDCmHH x33v Conmom mCHCo>w “N ommo

HDCUHH >00 CH0 COHmmom COOCHoDmm "H ommu
0.0H 5.0H 0.MN 0.0H 0.0H 0.NH 5.HH 0.0H 0.NN 0.0N 0.0H m.0H N.5 0.0 HooCo>MH
0.0H 0.0N 0.0H 0.5H 0.0H 5.0H H.0H 0.0H 0.0H 0.0H 0.0H 0.0N 0.0H 0.5H onHoM
0.0H 0.0 H.0H 5.0N 0.HN H.0H m.mH 0.0H m.0H 0.0H 0.0N 0.HN N.v 0.¢ CODmmoHHm>
N.0H 0.HH 0.0H 0.0H 0.0H H.mN 0.mN N.0H 0.0H N.0H 0.HN 5.0H 0.0N N.NN MCHC
N.0H N.0H ¢.0H H.0H 0.HN 0.HH 0.0H N.0H 0.0H H.NN H.0H 0.0H N.0H H.0H Com
N.0H H.0H 0.0H 5.0H H.mH 0.0N m.¢N N.0H 0.0H 0.NH m.5H 0.0H 0.NN 5.HN oDHC3

DCoo Com Axv Hxv DCoo Com Hxv N.Hxv

CDO CD0 CD0 Gum CCN DmH o0HOCU CDO CDO CDC 0C0 CCN DmH ooHOCU memEmm
COConomon mo COHDCCHCDmHQ IDmH oonHomoum mo CoHDCCHHDmHQ IDmH HoHoo

 

AN owmuu

 

 

 

mm OHmEmm mCHDUOHmm Hocmm mo DCoo Com

.onEmm m CHCDH3 mooHOCu mo MUCosvoCm mo COHDCCHCDmHC oCD CCm CUHOCUIDmH

"msHOHCMHO CH COCoHomoum HOHOU uoECmCOOtt.mH oHCCB

66

33:3 as 0. 33E ft aux he tam via

 

metusqu 3 «50% 30k Mans xfid Vow. mks:

DIUIEOUUEEIIIEII-

 

WEHQHQSM

.03 souLQ 050.65? ngomkkm 22 $.13 wismm.
Lieu wunw L0H puma? JEN oi CS sinktmotmmx 05910.5. Nike rournkxéu 255 mfu‘

Home

 

‘4‘

a
r%2 Pawn 2251

‘41

 

67

A "write-in" panel was conducted in February 1963.
The results of that panel indicated 17 per cent preferred
yellow, 16 per cent preferred white, 16 per cent preferred
pink, 12 per cent no color preferred, 12 per cent preferred
red, 11 per cent preferred variegated, 11 per cent preferred
mixed, and 5 per cent preferred lavender.

The findings of the two research methods ("visual" vs.
"write—in") had a great resemblance in this case. The above

conclusion seems then to be confirmed.

Number Preference

Six "units-of—purchase" were chosen for this investi-
gation, consisting of 3, 4, 5, 6, 8, and 12 gladiolus.
Samples were in yellow (May, 1962) and in white for the re-
maining two tests.

The scheme of this investigation was:

Date of Test Testing Samples Special feature
(unit-of-purchase) (in case 2)
May 1962 3, 4, 5, 6 26¢ each flower
September 1962 4, 5, 6, 8 21¢ each flower
February 1963 3, 4, 5, 6, 8, 12 31¢ each flower

Table 16 summarizes the results of the three tests.
And shows the rankings for the various units. The units of
three and eight remained in the same position on the prefer-

ence scale in the two cases. The units of three were least

E'v‘

in

68

Table 16.--Consumer number preference+ of Gladiolus:
estimated rankings.

 

 

Estimated Ranking Total Coefficient
Date of number of of concord-
Experiment lst 2nd 3rd 4th 5th 6th consumer ance W

 

May 1962
Case 1 40 .0360
Case 2 5 6 4 3 78 *.1164
September 1962
Case 1 6 8 4 5 49 *.3198
Case 2 6 8 5 4 84 *.3008
February’l963
Case 1 12 8 6 4 5 3 63 *.2791
Case 2 6 8 4 5 12 3 86 *.1251

 

*Significant at 1% level

+Numbers indicate the number of flowers in an unit.

Case 1: afternoon session (not priced)

Case 2: evening session (priced)
preferred and the units of eight were high-medium in prefer—
ence. The units of four and twelve became less popular as
the price factor was introduced. This negative effect was
even more obvious in the units of twelve Which ranked t0ps
in preference when the flowers were not pricedH-but moved to
medium-low in preference when price became a factor. The
units of five and six became more favorably accepted. More—

over, units of six ranked tops in case 2.

69

Table 17 presents an analysis of the percentage of
lst—choices for the six unit samples and the distribution of
frequency of choices for each unit sample. Both units of
three and four showed low ratings in preference in two cases
(not priced and priced). The units of six and eight each
shared one fourth of the first preference ratings in after-
noon sessions (not priced) and evening sessions (priced).

The unit of twelve was the "modal preference" number(with 50
per cent of consumers in this group) in the afternoon sessions.
Only 22 per cent of the panelists preferred it in the evening

sessions.

70

ACCUHCQ DOCV COHmwom COOCHODHM "H ommu

ACOUHDQV COHmmom mCHCo>o ”N ommo

 

 

0.00 0.NH 5.¢ 0.NH 0.HH H.NN H.NN 0.5 N.m 0.0 5.NH 0.0H 0.00 0.00 ACoNoov NH
0.0H 0.5H N.0H H.5 0.0N 0.0N 0.0N 5.0H N.NH N.MH N.0H 5.0N N.0H 0.5H 0
0.0 0.0H 5.HH 0.5H 5.0H 0.0N 0.0N N.NH 5.0H 0.HH H.0H H.0H 0.0N 0.0N O
0.NH N.0H 0.0H 0.NN 0.0H 0.0H 0.0 5.5H 0.HN 0.0N N.0H 0.0H N.m 0.0 0
¢.Hm H.¢N 0.0N 5.0H 0.0H 0.HH. 0.0H v.HH 0.0N 0.0H H.0H 5.0H ¢.mH 0.HH C
¢.Hm N.0H 0.HN 0.5 0.0 0.0H 0.NH N.Hm 0.5H 0.0H 0.0H 0.0 5.0H N.m muoson m
DCoo Com HMO N Hwy DCoo Com AMC H.HMV
CDO CDO CDC cum CCN DmH ooHOCU CDO CDO CDC Cum CCN DmH oUHOCU monEmm
ooCoComon Ho COHDCCHCDmHQ IDmH COCouomon mo CoHDCCHCDmHQ IDmH DHCD

 

Cm mmmov

 

“H sumo”

 

 

monEmm mCHDooHom HoCmm mo DCmU Com

.onEmm m CHCDHB mooHOCU mo MUCoswoum mo COHDCCHHDmHo oCD CCm CUHOCUIDmH mm
umCHOHCmHO CH COCCComon HoCECC HoECmCOOIt.5H CHCCB

Bats

Sep‘

NDVI

71

Roses
Three major investigations were conducted with this
flower: color preference, number preference (unit-of-
purchase), and grade preference (stem-length). The following
paragraphs are devoted to a discussion of the findings in

these three studies.

Color Preference

Six colors were chosen for this investigation: white,
dark red, red (Better Times), pink, yellow, and pale lavender
(Sterling-Silver). The samples were in units of five with
15" stem-length.

The scheme of this investigation was:

Date of Test Testing Samples (color categories)

September 1962 dark red, red (Better Times), pink, yellow

November 1962 dark red, red (Better Times), pink, yellow
white, pale lavender (Sterling-Silver)

Table 18 summarizes the results of the two tests.
The results for the two sessions showed:

Dark red was high in preference,
Yellow ran high-medium to medium in preference
Red (Better Times) was medium in preference,

Pink ranked low in preference except in Nov. 1962
of case 2 (evening session),

White and pale lavender (Sterling—Silver)--(both
tested once) were low in preference

72

Avon: mCHDCmHH x330 COHmmom mCHCo>o "N ommu
HDCmHH mmo CHV COHmmom COOCHoDmm “H ommo

Hw>mH NH Du DCMUHchmHm
an

 

5N00.* HOH HoCCo>mq ODHCB Com 3oHHow CCHC Com N ommo
onm Mama
0000.4 50 CoCCo>mH CCHC oDHCB Com 3OHHoH. Com H ommu
onm Cumn NOmH CoCEo>oz
O0OH.« v0 CCHm Com BoHHow pom N ommo
Cumn
HNON.* me CCHm Com onHow Com H ommu

Hume memH umnsmummm

 

CUCm CoECmCoo CD0 CD0 CD0 Chm CCN DmH DCoEHHomxm
ICHOUCOU no mo HoCECC mo ODmn

 

DcwHonmmoo Hmuoe mCchmm euumeHDmm

 

 

.mmCHCCmu UCDCEHDmo "momom CH mooCoHomoum HOHOU HoECmCOOII.0H CHCCB

73

Table 19 shows the percent of panelists who placed
each of the six colors first and the distribution of frequency
of their choices for each color sample. Fig. 3 presents data
abstracted from this table designed to provide a comparison
of the number of first choices recorded for each of the color
samples with the expected average.

Two colors were found to have an "above average"
(16.7 per cent) preference (afternoon and evening sessions)
in two cases. They were dark red and pale lavender. Dark
red was more pOpular than pale lavender (Table 19).

The analysis of the distribution of frequency of
choices for dark red roses, showed the preference score for
dark red concentrated in the first two ranks . In pale
lavender the modal preference score occurred in the last
rank (Table 19). This explains the contradictory presen—
tation between the ranking analysis and the first preference
analysis for this particular color. Though pale lavender
roses did show a relatively high percentage of preference in
the first-choice category in the color sample test, however
within its own color class a greater proportion of peOple
disliked the color. Thus pale lavender appeared low in the
estimated ranking study. (The author noticed that those
panelists revealed a high preference in this color partially

because it is considered a novelty to them;)

74

Coons mCHDCmHH x330 COmeom mCHCo>w "N ommo

ADCmHH NCO CHV COHmmom COOCHoDDm.uH mmmu

 

 

O.H¢ 0.mH 0.0 0.0H 0.5 0.0H 0.0H 0.Hm 0.0H 0.0H H.0H N.5 5.HN 0.0N HoCCo>mH
onm
0.0H 0.0H 0.HH 0.0N 0.0H 0.5H 0.0H 0.NH 0.5H 5.0H 5.NH N.0N 5.0H N.0H onHoM
0.NH 0.0H 0.NN 0.0H 5.0H 0.0H N.0H 0.0H 5.0N 0.0H 0.NN H.¢H N.0 0.0 CCHC
5.0H N.5H 0.NN 0.0N 0.0H 0.HH 0.0H 5.0H 0.5H 5.5H 0.0N N.vH N.0 0-0 HmoEHB CCDDomv
Com
0.0H 0.0H 0.HH N.0 5.0N 0.0N m.0N 5.NH N.HH 0.NH 0.0H 0.0N 0.mm 0.Nm Com Cnmn
0.5H 5.0N 0.0H 0.HH 0.NH 0.0 0.0 0.0H 0.5H 0.0N 0.0H 0.0H 0.0H 0.NH oDHCB
DCoo Mom Hxv x Axv DCoo Com Hxv N.Hxv
CDO CDO CDC cum CCN DmH ooHOCU CDO CDO CDv Cum CCN DmH oUHOCU monEmm HOHoo
ooCouoHoum Do CoHDCCHHDmHQ IDmH COCouomoum mo COHDCCHHDmHQ IDmH

 

Ii

.N ommo

\

.wammoe

 

.onEmm m CHCDH3 mooHOCo mo hoCoCUon mo COHDCCHHDmHU mCD 0C0 CUHOCU
IDmH mm CHmEmw mCHDomHom HmCmm mo DCoo mom

"momom CH oUCoHomem HOHOU HmECmCOUII.0H mHCmB

75

The following table shows the results of three
"write-in" experiments--Sherman's finding (115), LeClerg's

finding (112) and the test of February 1963.

Red Yellow Pink Mixed Other No answer
Sherman 54.5 22.3 .9 5.4 10.7
LeClerg 56.0 16.6 14.4 1.9 3.9
Feb. 30.8 19.9 6.4 3.9 1.9 (Pale Lavender) 4.5
1963 26.9 (dark 5.8 (White)
red)

There was agreement between the findings of the three
"write-in" methods. It generally corresponded with the
preceding findings of the "visual" experiments.

There was a superficial deviation between the findings
of the "visual" and "write-in" panel on red and dark red.

This may have been due to the fact that the variety of red
roses used in the "visual" panel--Better Times is not a dis-
tinct red. (Because of the popularity of this variety in the
flower market, its varietal name was used to denote a certain
degree of color saturation of red as characterized in this va-
riety.) It was inferior (not distinct) only in the sense of
comparison with some other red varieties. It still remained
competitive with other samples in the tests. Yet when the
panel members were asked about their color preference in the
"write-in" experiment case, they naturally associated the color

"red with the superior (distinct) red roses and with other

76

colors. Red roses are very popular flowers in this country.
In conclusion the red roses, in a broad sense, would
be considered to be "the most popular" color in roses. Yellow

roses were second in popularity.

Number Preference

Five units—of—purchase for three different grades of
roses were chosen for this investigation: 3, 5, 7, 9, and 12
(a dozen). The variety Better Times was used for this test.
Flowers with stems 9", 12", and 15" long were used. They

were treated as three independent series in all the testing

panels.
The scheme of this investigation was:
Date of Test Testing Samples Special Feature
Number in each unit (Price at retail in case 2)
January 9": 3, 5, 7, 9, 12 24¢ each flower
1957 12": 3, 5, 7, 9, 12 30¢ each flower
15": 3, 5, 7, 9, 12 36¢ each flower
February 9“: 3, 5, 7, 9, 12 16¢ eaCh flower
1957 12": 3, 5, 7, 9, 12 20¢ each flower
15": 3, 5, 7, 9, 12 24¢ each flower
March 9": 3, 5, 7, 9, 12 20¢ each flower
1957 12": 3, 5, 7, 9, 12 24¢ each flower
15": 3, 5, 7, 9, 12 30¢ each flower
April 9": 3, 5, 7, 9, 12 20¢ each flower
1957 12": 3, 5, 7, 9, 12 24¢ each flower
15": 3, 5, 7, 9, 12 30¢ each flower
May 9": 3, 5, 7, 9, 12 16¢ each flower
1957 12": 3, 5, 7, 9, 12 20¢ each flower
15": 3, 5, 7, 9, 12 24¢ each flower

77
Table 20 summarizes the results of the five tests.

+
Table 20.--Consumer number preference in Roses: estimated

 

ranking.
Estimated Ranking Total Coefficient
Date of number of of concord—
Experiments lst 2nd 3rd 4th 5th consumer ance W

 

9" Jan. 1957

Case 1
Case 2 7 9 5 3 12 90 *.0817
Feb. 1957
Case 1 7 9 5 12 3, 48 *.1614
Case 2 7 9 12 5 3 80 *.1352
Mar. 1957
Case 1 5,7 9 12 3 53 *.0839
Case 2 7 9 5 12 3 68 *.0807
Apr. 1957
Case 1 5,7 9 3 12 51 *.1753
Case 2 5 7 12 9 3 85 *.1254
May 1957
Case 1 5 9 7 3 12 64 *.0646
Case 2 9 7 12 5 3 84 *.0974
12" Jan. 1957
Case 1
Case 2 5 7 9 3 12 90 *.0564
Feb. 1957
Case 1 9 7 5 12 3 48 *.2036
Case 2 9 7 5 12 3 80 » *.1480
Mar. 1957
Case 1 7 5 9 12 3 53 *.3675
Case 2 5 9 7 12 3 68 *.0497
Apr. 1957
Case 1 7 5,9 3 12 51 *.0708

Case 2 7 9 5 3 12 85 *.1740

78

Table 20.--Continued

 

 

 

 

Estimated Ranking Total . .
Date of number of Coeff1c1ent
Experiments lst.2nd 3rd 4th 5th consumer Of concord-
ance W
May 1957
Case 1 7 5 9 3 12 64 *.1555
Case 2 12 7 5 9 3 84 *.0732
15" Jan. 1957
Case 1
Case 2 5 7 9 3 12 90 *.0591
Feb. 1957
Case 1 7 9 5 l2 3 48 *.2325
Case 2 7 9 5 12 3 80 *.0629
Mar. 1957
Case 1 7 5 9 12 3 53 *.1456
Case 2 5‘ 7 9 3 12 68 *.0713
Apr. 1957
Case 1 5 7 3 9 12 51 *.1200
Case 2 5 7 9 12 3 85 *.1328
May 1957
Case 1 7 5 9 3 12 64 *.3352
Case 2 7 12 9 5 3 84 *.1180

 

*Significant at 1% level

+ . . . .
Numbers indicate the number of flowers in an unit.
Case 1: afternoon session (not priced)

Case 2: evening session (priced)

The data revealed in the three grades of Better Times roses
that there was no great price-effect on the consumer number
preference. The preferred "unit-of-purchase" was seven,

regardless of grade and whether or not the flowers were priced.

79

Table 21 shows the percentage of panelists who
selected each of the five units in the three grades as their
lst—choice. It shows also the distribution of frequency of
choices for each unit sample in three grades. The unit
consisting of three roses was the least preferred number in
the three grades in both case 1 and 2 (afternoon session-—not
priced, and evening session-priced).

For 9" roses, the unit of seven and twelve achieved
the same relative high preference score (Table 21, case l--
not priced). There was no price—effect on the number
preference.

In the 12" roses, the preference was quite evenly
distributed among the unit —5, -7, -9, -12. No price effect
was found in the number preference.

The same pattern of preference distribution was found
in 15" roses as in the 12". Price however had some bearing
on the acceptibility of the units of none and twelve in 15"
roses.

In conclusion, the tests showed no preference for any
specific number of Better Times roses in any of three grades.
The unit of three was the least preferred number. No profound

price—effect based on the number of roses in the unit was

indicated.

80

ACCUHCQV COHmmom mCHCo>o "N ommo

ACCUHCQ DOCV COHmmom COOCHCDDC "H «000

 

 

 

0.Nm 0.0H 0.0H 0.0H 0.0N v.0N 0.Nv N.NH 0.0 0.0H 0.0H 0.5H NH

0.5 0.Hm 0.HN 0.0N 5.0H 0.0H 0.0 0.00 N.0H 0.0N 5.0N 0.HN m

0.0 0.HH 0.00 0.0N 0.0N 0.HN N.v 0.5 @900 0.HN 0.0N 0.0N 5

5.0 0.0N 5.0H 0.0N 0.0N 0.NN 0.5 5.0N 0.5H 0.0N N.NN H.0N O

0.50 0.0H 0.0H H.0H 5.0H H.0H 0.00 0.5H 0.0H 0.0H 0.0H N.5H m =0H
0.0m 0.5H 0.NH 5.NH 0.5N v.mN N.0¢ 5.0H 0.0H 0.0 0.0N v.0H NH

0.5 0.0N 0.0N 0.HN 0.0N e.mN 0.0 0.0N 0.0H 0.0N H.0N N.NN 0

0.5 0.NH N.0N 0.0N m.HN N.0N 0.5 5.0 N.¢0 0.NN 0.0N 0.0N 5

0.5 0.0N 5.mN 0.0N 0.NN 0.0H 0.0 0.NN 0.0m 0.¢N 0.0H 0.0N 0

5.00 0.0H 0.0H N.HH 5.0H H.0H ¢.5m 0.0H 0.0H 0.0H 0.0H 0.NH m =NH
v.0N 0.0H 5.NH 0.NH 5.0N 0.0N 0.0m 0.NH 0.0H 5.0H N.0N N.0N NH

H.5 v.¢N 0.HN 0.0N 0.NN 5.0H v.5 v.0N 0.0H 0.0N 0.HN 0.0H 0

0.0 0.HH 0.Nm 5.vN 0.0N N.NN 0.0 5.0 0.00 0.0H 0.0N 0.0N 5

0.0 0.0N 0.5H 5.5N 5.0H 0.0H 0.0 H.5N 0.0H 0.0N N.0N 0.0H O

5.00 0.0H v.0H 0.0H 0.0N 0.5H 0.00 0.0H 0.HH 0.HN 5.NH 0.HH mnoonm m :0

DCoo Com Hxv N.Hxv DCoo Com Hxv N.Hxv
CDO CDC Chm CCN DmH oCHOCU CDO CDC Ohm CCN DmH oCHOCU monEmm DHCD
COCoHomon mo COHDCCHHDmHQ tDmH CUCCHCOCHC mo COHDCCHHDmHQ IDmH
1N ommow .H ommofl

 

 

.mHmEmm w CHCDHB mooHOCo Do moCosqoum mo COHDCCHHDCHC oCD CCm ooHOCo
IDmH mm monEMm mCHDumHom HoCmm mo DCoU Com «momom mo COCmHomon HCCECC HOECmCOOII.HN oHCmB

81

Grade Preference

As in commercial practice the grades used in this
study were based on the stem-length. Four lengths were
selected for this investigation: 9", 12", 15", 21". The
variety Better Times was used here. The samples were in
units of five and the indicated prices were for the whole

group of five flowers.

The sCheme of this investigation was:

Date of Test Testing Samples (stem-length)

February 1960 12" (95¢), 15" ($1.39), 21" ($1.95)
September 1962 9" ($1.15), 12" ($1.49), 15" ($1.89)
February 1963 9" ($1.39), 12" ($1.79), 15" ($2.29)

Table 22 summarizes the results of the three tests.

Table 22.--Consumer grade preference of Roses: estimated

 

 

 

rankings
Estimated Total Coefficient
Ranking number of of concordr
Grade Samples lst 2nd 3rd consumer ance W
February 1960
Case 1 12" 15" 21" 57 *.1922
Case 2 12" 15" 21" 79 *.2059
September 1962
Case 1 15" 12" 9" 49 *.2403
Case 2 12" 9" 15" 84 *.1634
February 1963
Case 1 15" 12" 9" 63 *.3612
Case 2 12" 15" 9" 86 *.1711

 

*Significant at 1% level
Case 1: afternoon (not priced)

Case 2: evening (priced)

82

The tests showed a great price-effect on consumer
grade preference. In case 1 when the flowers were not priced
the 15" grade was preferred, whereas in case 2 where the
flowers were priced the 12" grade was preferred.

Table 23 shows the per cent of panelists who selected
each of the four grade samples as their lst-choice and the
distribution of frequency within the samples. The findings
revealed both 9" and 21"roses were low in preference in cases
1 and 2. In case 1 the 15" roses were more popular than 12"
(15" was preferred by 38 per cent panelists, 12" by 26 per
cent, 21" by 12 per cent, and 9" by 10 per cent). In case 2
the 12" became more popular than the 15" (12" was preferred
by 32 per cent panelists, 15" by 29 per cent, 9" by 16 per

cent, and 21" by 11 per cent).

83

HCoUHHmV COHmmom mCHCo>o "N ommo

ACCUHCQ DOCV COHmmom COOCHoDmm ”H @000

 

 

0.0N H.5e 0.0H 0.HH H.HH 0.0N 0.N0 0.0 0.NH 0.HH =HN

0.0N N.0H m.5N O.Hm 0.0N 0.0N 5.0 N.Hm 0.00 0.00 :0H

0.0N 0.0H 0.Hm m.mm m.Nm 0.0N 5.mH 0.00 0.5N m.0N =NH

0.0N 0.N0 0.0H 0.0H 5.0H 0.0N 0.00 H.0H N.0H 0.0H =0
DCoo Com Hxv x Hxv DCoo Com Hxv N.Hxv

sue sum new DuH moHoso Cue sum esm DuH moHoCo mmHCsmm

CUCoCommum Do CoHDCCHHDmHQ IDmH COCouoDoCm Do COHDCCHHDmHQ tDmH opmno

 

.N ommow

 

sH MWMUO

 

 

.onEmm m CHCDH3 mooHOCo mo NUCCCCCCD Do COHDCCHHDmHo oCD OCC CUHOCU

IDmH mm monEmm mCHDomHmm HoCmm Do DCoU Dom "momom CH oUCouoHon ocmum HmECmCOOII.MN oHCmB

CHAPTER V
RETAIL SALES SURVEY

Howell (55) listed three methods already tried out for
obtaining information regarding consumers' preferences:

(1) consumers' preference surveys, (2) consumer purchase
surveys, and (3) retail sales surveys. Two of the three
methods were employed in this study; namely the consumers'
preference survey and the retail sales survey.

In the preceding chapters the discussion was centered
on a study of the consumers' preference survey. This phase
was designed to present the secondary phase of the research—-
a retail sales survey. \

Generally, a retail sales survey is planned to assemble
data regarding volume of sales, prices, and other factors for
a specific product, which would show the relationship of sales
volume to prices and to other factors. In this study, in
addition to the conventional point of interest—-price effects,
a few other specific factors were included.

These specific features were related to the character-
istic of the flower itself, e.g., color of the flower, or the

general appearance of the merchandise.

84

85

As indicated previously this survey was purposely
designed to supplement the consumer preference study.

Rhodes (73) stated: "The fact that product A's
average preference score is found to exceed 'significantly’
product B's average preference score does not necessarily
mean that more people prefer A than prefer B or that more of
A will be sold than of B." Hence, it is rather desirable to
compare the findings of the two types of survey, and to see
how closely the results of the consumer stated preferences
compare to their marketing behavior. As there may be a
discrepancy between what people say they desire and What they
really desire, the relationships between "saying" and "doing"
were investigated. The hypothesis tested is that consumers'
stated preferences differ from their actual preferences for
certain flowers, color and grades.

The report submitted in this chapter is mainly designed
to relate the results of the retail sales survey. The results
of the consumer trials and the retail sales survey will be com-
pared in the next Chapter under the heading of ”relationship
of consumer stated preferences to their marketing behavior."

In this chapter the presentation proceeds as follows:
(1) survey procedures, (2) Characteristics of the partici-

pating florists, and (3) results.

86

Survey Procedures

Surveyed florists were selected as follows: question-
naires were sent by mail to 657 names obtained from the
latest Michigan State Florist Associations' Membership Direc-
tory. Only the establishments classified as retail growers
or retail florists were contacted.

Von Oppenfeld (85) pointed out that in Michigan 78 per
cent of all sales were reported from "retail—grower" and
"retail florists" outlets. Hence the selection of this
sampling population is justified in this survey.

The basic questions asked concerned: (1) general
information of the surveyed florists, (2) purpose of consumers'
flower purchasing, (3) consumer color preferences (acceptance
in this case), (4) size of units of purchase ( Appendix 2).
The latter two investigations were limited in the following
flowers: carnations, pom-pons Chrysanthemums, standard
Chrysanthemums, gladiolus, and roses.

One hundred forty three out of the 657 florists re—
sponded--representing about 22 per cent of those contacted.
Only 123 of the returned questionnaires were included in the

final study, since the remaining 20 did not provide: clear and

concise information.

87

Characteristics of the Reporting Florists

Sales Volume
The distribution of retail florist shops by size of

establishment in this survey was as follows:

Size of retail establishments Distribution of establishments

 

(dollar volume) (number)
30,000 or less 52
31,000 - 60,000 38
61,000 — 90,000 17
91,000 — 120,000 10

121,000 or more 6

This profile of distribution of retail florist shops
by size of establishments characterizes the National picture
[Trotter (118) and the 1958 Census of Business on Retail
Trade--Sa1es sizes]. Therefore, it was considered a good

sample.

Retail Florist Sales by Size of City

VOn Oppenfeld (119) showed that for the United States
there is a relationship between the size of cities and the
volume of retail florist sales. This survey confirmed his

statement (Table 24).

88

 

 

0.00 0.00 0.00 OHOE no I 000.HNH

0.0H 0.0H 0.05 0.0H 000.0NH I 000.H0

0.HH 0.0 ¢.0N 0.HH O.MN 000.00 I 000.H0

0.0H 5.N H.0 0.0H 0.0 0.0N H.Om 5.N 000.00 I 000.Hm

0.0 0.0 0.N 0.0H 0.0H 0.0N 0.00 000H HO 000.00
OHOE HO 000.0wv 000.0MN 000.0HH 000.00 000.0N 000.¢H mmOH Ho HHMHHOOV

000.000 000.0CN 000.0NH 000.00 000.0m 000.0H 000.0 000.¢ OECHO> mOHmm

 

mo mmHDHo 0C onmm 00 CoHDCCHCDmHC ommDCoUHom

 

 

.NDHU mo oNHm NC monm DmHHon HHmDoD Ho COHDCCHHDmHC omeCoUmeII.VN CHCCB

89

Store Traffic and Location of Stores

Fortychree per cent of the reporting florists indi-
cated their stores were located on the roadside, 37 per cent
in neighborhood shopping districts, and 20 per cent in
central shopping districts.

Thirty-seven per cent of the florists indicated that
their traffic causing feature was a main highway, 17 per cent
drew major business from a hospital, 12 per cent from a
cemetary, 10 per cent from a shopping—center, 8 per cent
from schools, and 5 per cent from office—buildings.

The above findings were not in agreement with reports

by Knight (111), and Von Oppenfeld (119).

Results of the Survey

Consumer Baying Habits

Eighty-two per cent of the reporting florists had
60 per cent or higher of their customer's orders placed by
telephone. The purposes for consumer flower-buying in
carnations, Chrysanthemums, gladiolus, and roses are revealed
in Table 25. These figures reflected the importance of the
funeral and wedding business to the retail florist. Roses

had more diverse usage than other flowers.

90

 

0.0H O.m 0.0H H.4H 0.HH msonmuuo sense

 

m.HH 0.NH e.m 0.HH 0.0H mus Hmenmsm>m
0.0 0.0 0.0 0.0 0.0H hummuw>HCCC
m.m 5.0 H.0 5.0 H.sH NmsCDsHm
0.HH 0.0H e.mH m.e s.0 soHDmuooue HmHossssoo
0.0H N.0H m.m 0.0H 0.NH meHesus
m.OH H.0H s.» m.~H m.mH HmuHCuom
m.~m 0.04 s.mm m.sm m.m~ m>mso s Hmuwssm
DCwU Com HMH
mCHOHomHO mComIEom H.0Dmv was: mCoHDmCHCU momom mCOHmmUUO

 

muo3on mo mmmhe

 

 

wum3on no moth hC 0CH>CCIHCBOHD HCECmCoo mo COHDCCHHDme mmmDCooHomII.0N.oHCmB

Consumer Color

91

Acceptance

Table 26 summarizes consumer color acceptance in the

four major flowers revealed by the 123 reporting florists.

Table 26.--Average percentage distribution of consumer color
acceptance—-by types of flowers.

 

 

Type of flower:

Roses:

Carnations:

Mums (std.):

Pom-pons:

Gladiolus:

Red (58.0) Dark red (33.9)
No color specified (23.8) Pink (9.5) Yellow (8.8)
white (8.8) mixed color (7.5)

no color specified (34.3) White (23.3) Red (22.0)
Light Pink (21.0) mixed color (16.6)

Other [dyed] (10.8) Dark Pink (9.7)

Variegated (8.2) Yellow (7.0)

No color specified (35.2) Yellow 28.5)
White (26.2) Bronze (14.7) mixed color (14.7)
Lavender (9.2) Pink (8.2) Other (7.6)

No color specified (41.1) Yellow (24.0)

White (23.3) Mixed color (21.5)

Pale Bronze (10.0) Dark Bronze (9.5)

Lavender (7.9) Pink (6.9) Dark Lavender (6.0)

No color specified (50.3) Mixed color (23.4)
White (20.0) Pink (17.8) Red (15.0) Yellow (11.6)
Variegated (9.4) Lavender (8.0)

 

The Unit of Flower Sales

Table 27 summarizes the unit of flower sales for the

five major flowers indicated by the 123 reporting florists.

92

Table 27:——Average percentage distribution of unit of flower
sales: by types of cut flowers.

 

 

Size of Unit Purchase

 

 

Type of 3 4 5 6 7 - ll dozen

flower Flowers Flowers Flowers Flowers Flowers or more
or less

Roses 13.0 5.1 4.2 13.2 5.0 75.3

Carnations 16.8 5.1 5.6 18.5 12.1 66.2

Mums (std.) 20.8 8.2 14.4 27.4 18.3 49.3

Gladiolus 7.7 4.4 5.0 18.9 17.9 76.0
1 bunch 2 bunches 3 bunches

Pom-pons 73.5 19.3 19.5

 

CHAPTER VI
DISCUSSION

Preference Study

This paper is designed to measure the consumer pref-
erence for flowers, i.e., study consumer desire for selected
flowers. The measurement of preference was designed to
determine the particular situation or situations in which
preferences can be feasibly ascertained, to select the ones
desired, and to interpret the results accordingly.

However, confusion abounds in the literature
describing investigations of "consumer acceptanCe" and
"consumer preference". The basic problems are the same two
that cause confusion throughout the Whole field of investi-
gation research. The first is the failure to define precisely,
and to suitably restrict the subject of investigation. The
second is the failure to employ a technique which can meet
the objective.

The term "acceptance" of a particular commodity most
frequently involves either a measure of the quantities bought
or of the number of consumers who buy the commodity when the

"usual" alternatives are offered in conjunction therewith,

93

94

and When the price and income structures are known and fixed.
The time interval, the pOpulation of consumers, and the units
of the classes or alternatives are then stated or assumed to
be known.

The word "prefer" indicates a certain ranking. An
assertion that one member of a class of alternatives is
preferred to another clearly identifies at least two members
of the class. Thus it would appear that if either individual
or group consumer preference for a number of commodities were
to be measured, a ranking of these commodities by the consumers
should be sought. The class of alternatives may well consist
in part of those constituting the "Usual" environment, but in
a measurement of preference among a group of commodities every
consumer must rank every member of the group.

Preference studies frequently answer the question of
why consumers buy or do not buy a product, and are therefore
the logical precursors of many acceptance studies. They limit
the forms of a product which must be subjected to acceptance
measurement and demonstrate the existance of alternable
preferences founded upon ignorance, prejudice, or fancy
[Burrows (46)].

The preference study denoted here is still different
from the traditional marginal utility theory which is also a

study based on the revealed preferred frequencies. While the

95

distinction between preference and utility may appear to be
verbal, the distinction is apparent in the types of questions
asked in consumer research. In preference inquiry the re-
spondent is asked how much he would like to have the article
in question, not how much satisfaction he experiences from it.
Measurement of preference reflects not only degree of satis-
faction but also the dynamics of habit, social pressure,
advertising influence and any other factors which determine
consumer choice[ Benson (36)].

Churchman (5) in his article “The consumer and his
interest", 1946, throws some light on this delicate subject.
First he defined the consumer in terms of his purposes-—which
are different from the purposes of production, since interests
refer to purposes, and an action serves the interest of the
consumer if it assists him in the accomplishment of his
consumer-ends. It might be thought that since one can claim
success in trapping the consumer within the tyranny of his
words, he can almost automatically grasp the meaning of the
consumer's interest as well. That is to say shall "consumer
interest" be equivalent to "consumer desire"? If one then
wants to know whether a certain product serves the consumer
interest, it is only necessary to measure preferences, eval-
uate biological factors such as health, well-being, etc., and

in general run the entire gamut of pe0ple's desires.

96

Churchman claimed that one of the most serious fallacies of
the consumer problem is to assume that the measurement of

consumer interest depends soley upon the measurement of con-
sumer desire. For only those consumer purposes which serve

him best can be used in considering consumer interest.

The "Economic Strength” of Preferences

The "economic strengt " of preferences refers here
to the extent to which the preferences among a set of products
are a function of their price relationships. This strength
is very relevant to the problem of which and how many grades
or selling—units should be offered to a customer population
of divergent preferences.

The determination of the economic strength of prefer-
ences was particularly important in the uni-product selection
problem (choosing one of several alternatives to be merchan-
dised). This was especially true when the most p0pular
product was the most expensive product to be merchandised.
Such determination is important among the whole set of pro-
ducts in a multi-product marketing situation. The price
relations and the market channels for all of the products
need to be estimated.

An example shown here was the determination of
economic strength of preference in the uni-product selection

problem (Table 28, c.f., Table 21). The retail sales survey

97

Table 28.--The economic strength of consumer number prefer-
ence in the three grades of roses.

 

 

 

Date of Unit of purchase
Grades experiment Price 3 5 7 9 12
per cent of lst-choice
9" 1. Feb., May 1957 16¢ each 14.0 10.8 22.8 25.4 27.0
2. Mar., Apr.1957 20¢ each 15.8 24.2 19.2 15.3 25.4
3. Jan. 1957 24¢ each 29.0 14.0 27.1 12.1 17.8
12" 1. Feb., May 1957 20¢ each 12.1 17.3 19.8 23.0 28.4
2. Mar., Apr.l957 24¢ each 12.1 21.4 19.1 27.8 19.6
3. Jan. 1957 30¢ each 21.8 17.3 24.5 15.5 20.9
15" 1. Feb., May 1957 24¢ each 13.0 12.4 29.6 19.8 30.2
2. Mar., Apr 1957 30¢ each 14.6 31.6 15.2 13.4 24.9
3. Jan. 1957 36¢ each 25.2 24.3 18.0 10.8 21.6

 

revealed that 75 per cent

12 or more rather than some smaller units.

of the roses were sold in units of

No doubt this

most popular unit was the most expensive product offered.

It thus fulfilled the justification of "economic strength"

of preferences in this study.

Table 28 indicates no preferred preference unit in

any of the 9", 12",

prices.

When price was a factor larger units (six or more) were

preferred in the low price range.

or 15" roses over a range of the tested

It was clear that consumers were price-conscious.

The units of five became

popular as the price raised to the medium of price-range.

The units of three were more p0pu1ar when the price range

was higher.

None of the units in any of the three grades of

98

roses got a "strong" preference in the light of its economic

strength preference analysis.

However, the units of twelve showed relatively
"greater" strength than the other smaller units. This was
countered by the two different motivations in consumer flower
buying. In some instances, such as when a congratulatory
gift is purchased, the consumer wants to satisfy a desire
for prestige and status. The price of the product then be-
comes important (and/or the price is not so important as the
quantity), because under these circumstances, the flowers
must be elegant enough and perhaps large enough to make an
effective showing. However, when flowers are to be purchased
for the home the consumer shifts his image to one of low
prices for economy purposes.

The above finding of no strong "economic strength"
preferences of consumer buying units in roses projected a
general pattern of "economic strength" preferences in other
studied flowers. On the other hand, the panelists revealed
"modal expenditure" for the studied flowers, such as:
$3.50—$3.79 for pom-pon Chrysanthemums (Table 7 and 8), and
$3.60r$4.98 for standard Chrysanthemums (Table 12).

Table 28 also indirectly indicates that a
potential home—use market exists in all grades of roses.
Price was recognized here as an important factor

which materially affected consumbers' buying

99

habits where roses were involved. This coincided with the
results of another test (Feb. 1963 write-in test), which
indicates that 43 per cent of the panel members do not use
more cut flowers in home decoration because of the high re-
tail price of flowers. Twenty-eight per cent indicated that
they do not buy more frequently because of the "short shelf-
1ife of flowers". Twelve per cent indicated that they were
not accustomed to using flowers. Six per cent indicated
that flowers were not readily available. Five per cent indi-
cated limited knowledge about flowers. Four per cent relied
on their own garden supply, and 2 per cent claimed personal
dislike for flowers. The second main reason was also in-
directly related to the first important reason. Price played
an important role in discouraging consumers from buying
flowers for home-use.

More studies are needed to develop marketing methods
which will secure a larger share of the consumer's dollars
for flowers. The study of ”economic strengt " of preferences
is a good approach. This study will solve many problems in
the multi-product selection program. That program is a most

urgent task which a progressive marketer must face.

A System of Color Preferences
It is well known that most flowers purchased today

are not seen by the purchaser. It seems possible however

100

that with the changing market condition, consumers will see
more floral products in the future. Thus one can expect a
whole new series of preferences to arise. The choice in the
future will not be entirely up to the retailer as it is today.
Color preferences may be the first to undergo change.

The color terms used to denote particular colors in
the color chart were not invested with a special technical
meaning. Terms in ”common" usage were employed. However a
horticulture color chart (121) was consulted in the latter
part of the color preference tests.

The study showed that color preferences varied with
regard to individual flowers, although with some flowers the
"preferred color” was not clearly defined, e.g. carnation,
pom-pon Chrysanthemum and gladiolus.

In general there are few factors that would effect
the expressed color preference. These can be grouped into
three categories, namely: the subjective variables, the
objective variables, and the environmental variables.

The subjective variables here denote such things as
personal taste [Grange (50)], and gender of the person
[Guilford (53); Pasto, et. al., (68), ]. The objective vari-
ables included all the qualitative properties of the objective
stimulus [Grange (50)]. It was noted in the panel study a

slight quality difference (other than color) in the samples

101

would affect the consumers color preference. This explains
why there were some inconsistency in the findings of their
replicates on color preference test of that flower. Other
objective variables are such as: the hue, brightness and
saturation of a context [McCormuk, et. al., (65)]. This
might be.one of the subtle factor which induced a inconsis-
tency in the two methods of testing (visual vs. write-in) in
color test on pom—pon chrysanthemums.

Finally, Riesen (74) pointed out that man's response
to color may depend predominantly on specific environmental
factors. Here, only two environmental elements were extended
into further discussion.

The first element had to do with the role of light
on color preference. Sanders (75) stated: ". . . preferred
color for any object may differ from the color of the object
in day light due to psychological influences. The range of
acceptable colors depends on the object and the extent to
which its quality varies with color." Moreover the role of
light on color preference has more than a psychological
influence. Sorensen, et. al. (116) pointed out: "Some flower
colors, such as red and bronze, look 'dead' under the usual
daylight fluorescent lighting in supermarkets."

Unfortunately the lighting of the laboratory under

which this study was conducted was probably the type Sorensen

102

referred to (warm White deluxe fluorescent, i.e., WWX). The
findings reported here substantiate the observations made by
Sorensen. However the red color probably did not appear half
as 'dead' as the bronze under the WWX, since only the result
with gladiolus showed a material difference between the day
light session and the evening session (using WWX lighting)
(c.f. Fig. 3). The preference for bronze pom—pon and
standard Chrysanthemums decreased considerably in the evening
When fluorescent tubes provided the only light source (c.f.
Fig. 1 and 2). On the contrary, it seemed that the preference
for all white flowers except white roses increased when they
were exhibited under artificial light (c.f. Fig. l and 2).
The second environmental element involved was the
temperature. Bryant (44) indicated the effect of temperature
on consumer color preferences in roses. He reported:
“ . . . red is best except during extremely hot periods in
the Summer months when cooler colors such as pink and yellow
are desired.“ This seasonal-variation in color preference
was not noted in this study, perhaps due to the fact that
the panels were conducted only in the relatively cool

seasons.

Relationship of Consumer Stated Preferences
to Their Marketing Behavior

This project was designed to test the hypothesis

t1“lat consumers' stated preferenCes differ from their actual

103

preferences which were revealed by the retail florists. Much
of the theory referred to the field of social psychology as
interpreted by Myrdal (18). Generally, most people have a
desire to please. According to Myrdal, people want to be
rational and Objective in their beliefs. They are inclined
to express only those beliefs for which they have reasons.

In exchanging ideas, people prefer to give good, logical, or
popular reasons for a particular belief or action. These
reasons may not be the true reasons. It is this situation
which creates evaluation problems in a preference study.

In this research it was anticipated that the above
mentioned psychological factor would not be very important in
creating a discrepancy between what people ‘say and What they
do. The type of answer sought in this research were more or
less outside the ordinary scope of things about which people
have a desire to please. Yet the potential influence of this
factor was acknowledged.

It was felt that there are two factors which heavily
influence the discrepancy between the consumers' stated
flower preferences and their actual preferences. One of the
two influential factors is the consumer's economic behavior
and its effect on their Choice of the floral products. The
other factor was that more than half of the flowers sold at

retail are used for funerals and weddings. Under these

104

circumstances, the consumer frequently does not assert a
preference and the choice of flowers used is left up to the
retailer.

Psychologically panel members might have been subjected
to all of these same compromises and desires to conform, and
to all these same practices of "obligation buying" and
"dependence on retail florists".

Three main aspects were involved in testing the
relationship of consumer stated preferences to their marketing
behavior, using the four selected types of flowers. These
aspects were: (1) color preferences, (2) number preferences,

and (3) grade preferences.

Colgp,P;efe£ences

The hypothesis was accepted in this test--consumers'
stated preferences differ from their actual preferences which
were revealed by florists.

In the retail sales survey, "no color preferences"
were noted with the modal preference and even appeared to be
the prominant Choice for all concerned flowers except roses.
These findings did not agree with the consumer survey. Even
with pom-pon Chrysanthemums, WhiCh were rated with the highest
”no color preference" score, the ”no color preference” still

was less popular than colored samples of that flower.

105

The color that ranked second in the retail sales
survey for all selected flowers except roses was "white".
For the types of flowers studied, it rated high as a color
in the consumer survey, but it was less popular in the re-
tail sales survey.

Another deviation involved in the rating of "mixed
colors”. In the consumer survey mixed colors rated a low
preference yet it ranked high in actual consumer buying.

In short, there seemed to be no close relationship
between the consumers' stated color preferences and their
actual marketing behavior. This discrepancy was greater in
those flowers where consumers did not show a pronounced
preference for color, e.g.: carnations, pom-pon chrysanthe-

mums, and gladiolus.

Number Preference

The hypothesis was rejected in this test--consumers'
stated preferences do not differ from their actual preferences
which were revealed by florists. The overall results of the
retail sales survey for all flowers tested compared favorably
with the results of the consumer studies (c.f. Table 36 with
the per cent of lst-choice in all instances in the case 2
observations in Table 4, 7, 13, 17, and 21).

The prices indicated in various flower exhibits in

the consumer panel studies were comparable with the

106

prevailing retail price-range of that flower (revealed from re-
tail sales survey). Only the standard Chrysanthemums were
priced higher than the prevailing price range (revealed from
the florists' survey). Other flowers were priced lower than
the prevailing price range. It is possible that this explained
Why only the unit of twelve standard mums was not the most p0pu-
lar unit as revealed in the consumer preference study. This was
in contrast to the results indicated in the retail sales survey.

Concluding, a close relationship was evident between
the "unit-of—purchase" preferences stated by consumers and
their actual "buying unit". The number preference ran largely
to dozens and half-dozens except with pom—pon chrysanthemums.
In consumer studies the two popular units were very competi-
tive in the number preference Whereas in the retail sales
survey the units of twelve (one dozen) were indicated to be
one to three times more favorably received than the units of

six (half-dozen).

.ggade Preference

The hypothesis was rejected in this test--consumers'
stated preferences do not differ from their actual preferences
which were revealed by florists. The overall results of the
consumer studies on grade preference in pom-pon Chrysanthemums
(size of the bunch) and roses (stem—length) compared favorably
with the established retail practice [See case 2 (evening

session-priced) in Table 9 and 23 ].

107

Pom-pon Chrysanthemums are frequently sold in bunches
of approximately 10" in diameter and as indicated in the
consumer survey this was a desirable size compared to the
other four alternative sizes [Table 9 case 2 (priced): 10"
was preferred by 35 per cent of the panelists, 15" by 21 per
cent, 12” by 16 per cent, 8" by 15 per cent, and 6" by 14 per
cent].

Twelve inch and 15” roses are p0pular on the rose
market whereas 9" roses are in lesser demand. This was also

born out in the consumer study.
Evaluation of the Present Study

In the beginning of the paper, it was pointed out
that this study was designed as a study to ascertain the
preferences of consumers for four types of flowers. A set
of accurate and repeatable human judgements was sought. This
is a challenging field for investigation, Things that influ-
ence human judgement can be very subtle. It is the investi-
gator's task to provide a proper setting for these judgements.
Hence, it is highly desirable to have the work evaluated in
the light of learned experience in the laboratory. This will
sharpen the research tool for future studies.

Two approaChes were employed for this evaluation—-
mainly a study of the planning phase and of certain specific

problems of control.

108

Planning Phase

The planning phase is of paramount importance.
Vaguely conceived objectives provide only vague results.
Varying different objectives in a ”preference study" indi-
cate a need for differing methods of analysis.

This study was a uni—product selection project.
Thus the methods of analysis and presentation used were
appropriate according to the study objectives [Rhodes
(72)]. For a detailed discussion of the relative merits of
an "M rating" correlation technique in selecting the most
popular product see: Friedman (48). Brown (43),-and Bliss

et. a1. (40). Kendall (12).

Specific Problem of_§ont£gl

The following problems having to do with control
were involved.

(1) Control of the setting: "Control of the setting"
referrs to the physical surroundings of the judge. In the
previous discussion lighting was shown to have a bearing on
the preference for certain colors. Therefore only that
lighting which provides "neutral outdoor effect" is recom—
mended.

(2) Control of samples: The phrase "control of the
sample to be judged" referrs to the condition of the samples

themselves. Probably the most important problem under this

109

heading is the control of irrelevant characteristics of the
samples. It was a serious and practical problem to maintain
an "all other things being equal" condition among the samples,
such as: the state of openness of the exhibited flower
samples; the corresponding brightness, hue and saturation of
the color, etc. This was partially responsible for the
"technical limitation" claimed in Chapter IV. It was not
possible to get samples of all kinds and colors submitted to
the test for each experiment. This limitation might be
partially overcome by introducing photographs into the con-
sumer preference studies. There are a number of advantages
to using photographs that are readily apparent [Gaarder (49)]:
(l) The samples used in the study can be selected more
objectively, (2) With photographs the test can be repeated,
getting directly comparable results, (3) It would be possible
to provide replications in order to test a respondent's
consistancy in two rankings of the same set of samples. The
photograph approach might be desired particularly to determine
preferences for non-uniform perishable items that are judged

to a large extent by their surface characteristics.

Motivation of Consumer Buying
and Consumer Education

Motivation of Consumer Buying

Here is acruote from the editor of the May 1961 issue

of "The Economist" concerning consumers' choice. He stated

110

"Since the war the steadiness, not to say the voraciousness,
of the buying habits of Americans--and the stability of their
incomes--have done much to moderate the severity of recessions
and now once more the consumer is being looked to for a sharp
push toward higher levels of prosperity."

Katona (57) also pointed out that in this society so
rich, fluid, and skillful, the American consumer is not
satiated. He continued: "The American consumer has not be-
come disenchanted with the installment buying plan. He has
not recently elevated savings to be his principal goal in
life. By and large, the American consumer still has unfilled
needs and still is willing to spend or borrow to fill them."

Dichter (47) declared that there have been four major
shifts in consumer thinking, strongly affecting buying atti-
tudes: "(JJ Americans are throwing off the puritanical cloak
and purchasing for pleasure, without twinge of conscience;

(2) they are saying, 'why should not I have this or that?’
allowing emotional appeals to influence their purchase;

(3) they are more mature in their buying attitudes, thinking
ahead in long-range terms; (4) there is desire for individu-
ality, self-expression and recognition." In short, they buy
to satisfy psychological desires shaped by world conditions,
enlightened attidues. They are enjoying new freedoms in self-
indulgence and expanding personal wants far beyond material

needs.

111

This view derives from the Gestalt psychologists, in
holding that the individuarsswants always to be seen in.a
favorable light, and that the maintenance and enhancement of
the individual is the most fundamental of all drives. Each
product projects a certain image of its own in the consumers'
mind. Hence the consumers, in the process of making a pur-
chase decision, are matching the appropriate product-image to
their own [Alderson (1)] .

The consumer's concept of a product develops through
experience and becomes reinforced in light of the meaning
attached to the product itself, to the situation in which
the product is used, and to the assumed reaction of the
persons associated with its use.

With flowers, for instance, a decision to buy might
be influenced by an individual's concept of the use of flowers
for certain people, his ideas about the characteristics of
flowers and his feelings toward the place of purchase. These
factors help constitute the person's image of flowers and
serve as a guide in the decision—making process (Early (102)).
The merchandiser needs to know the consumer's concept of a
‘product since it is this image that helps to motivate a person

to desire or reject a product.

Yet, it is clear that a proper definition of the

relevant attitudes will require more detailed research on the

112

personality level regarding the interrelations of motives,
incentive, expectancy, cognition, learning, and their relation
to action. This requires all the ingenuity and experience

the biologist, psychologist, sociologist, and economist can
bring to bear.

In short, this is a climate of attitudes, which is
the basis of modern public relations. Those businessmen will
thrive that create good images by adjusting to the changing
public attitudes. At the same time, these businessmen need
to improve themselves, since, in trying to live up to the
images they project, they must change themselves to fit the

image.

Consumer Education

It has been suggested that the next frontier for
marketing is an inner one, the marketing of the mind and the
personal development of consumers. Seeing, in the absence of
a technology of consumption, most peOple tend to use socially
approved symbols of achievement as their guides in consuming.
Abroad, attempts to mimic royalty in dress and food were so
strong that they called forth sumptuary legislation. In
America, respect for economic power led to such widespread
worShip of its symbols and wasteful expenditures as to give
rise to Thorstein Veblen's "conspicuous consumption” disser-

tation [Kelley (11)]. Hence, one of the roles of marketing

113

in the future may be that of encouraging increasing expendi-
tures of both dollars and time to develop consumers intel—
lectually, socially, and morally. Marketing, during a period
of increasing leisure, may well become a significant cultural
force. Marketing may provide the impetus for an improvement
of consumer tasts and increase in their cognizance and ap—
preciation of aesthetic values. That is to say the function
of consumer education.

As in the floricultural marketing field, the consumer
education program should not only aim to provide the consumer
with a wider experience with flowers but also to introduce
them to flower-appreciation.

Thus this project is looked upon not only as a bene-
factor to the market deve10pment program of the floral
industry but also ultimately to bring enrichment to the

consumers' life.

CHAPTER VII

SUMMARY

The purpose of this study was to determine consumer
preferences for the following five cut flowers. The findings
were presented as average percentages of the panel members
selecting units for their lst-choice.

In the text, case I referred to the afternoon sessions
When natural day light was available and the flowers were not
priced, and case 2 referred to the evening sessions when
artificial lighting was used (warm White de luxe, G.E. Tubes)
in color preference testsand when the flowers were priced inf“
number and grade preference tests.

Consumer preferences were recorded for:

(1) Carnations--

A. Color preference: The preference for individual
colors was not pronounced. The modal preference
was for dark pink and variegated in the "visual
product" panel (In case 1: 24 per cent selected
"variegated" color, 23 per cent selected dark
pink, 18 per cent selected yellow, 15 per cent
selected red, 12 per cent chose pink, and 11 per

cent chose White; In case 2: 21 per cent chose

114

115

dark pink, 20 per cent chose variegated, 19 per
cent chose white, 18 per cent chose yellow, 15 per
cent chose red, and 10 per cent chose pink).

The color preference for carnations was even
more diverse among the colored samples in the
"write-in” panel (18 per cent of panelists for
red; 16 per cent for dark pink, variegated, white,
and light pink; 9 per cent for no color preference 7
5 per cent for yellow; and 4 per cent for mixed
colors).

Quantity (number) preference: Price (price per
group) seemed to have no effect on the preference
for units of five, six, and eight. The preference
for units of twelve was lowered as the price in-
creased. The preference for the small units
(three and four) increased as the price increased.
Units of eight were more popular than twelve in
case 1:(28 per cent of panelists for units of
eight, 19 per cent for units of twelve, 18 per cent
for units of five, 14 per cent for units of six,

4 per cent for units of four, and 1 per cent for
units of three). Units of eight was barely more
popular than units of five in case 2: (27 per cent

of panelists for units of eight, 26 per cent for

116

units of five; 12 per cent for units of six, 10
per cent for units of twelve, 9 per cent for units
of four, and 2 per cent for units of three). The
results in case 2 might have been modified since
the units of twelve were over—priced in proportion

to other units.

(2) Pom-pon Chrysanthemums--

A.

Color preference: The preference for individual
colors was not pronounced. Pale bronze was more
popular than other colored samples but the prefer-
ence for pale bronze was not pronounced. (c.f.
Table 6). In the ”write-in” test yellow and bronze
(including light, medium, and dark shade of bronze)
were more popular than "no color preference" and
five other colors (26:26:19:l7:4:3:3:2).

Number preference: A price effect was found in

the bunches that were 10” in diameter. In case 1,
the modal preference of panelists was for three
bunches (with 52 per cent of panelists). In case
2, the modal preference of panelists was for one
bunch (with 66 per cent of panelists).

Grade preference (size of the bunch): The modal
preference of panelists in case 1 was for 15" bunch

(with 63 per cent of panelists). The effect of

117

price was apparent. The 10" bunch became more
popular than the 15" bunch when the flowers were
priced (c.f. Table 9, case 2). Consumer grade
preference compared favorably with the established

practice.

(3) Standard Chrysanthemums--

A.

Color preference: The preference for yellow
standard mums did not appear as definite as in
other similar studies conducted by the Ohio
Agricultural Experiment Station (115), and
Mississippi Agricultural Station (112). However
yellow was more p0pu1ar than white, pink, bronze,
and lavender (case 1: 37 per cent of panelists
preferred yellow; case 2: 41 per cent preferred
yellow). In the "write-in" test, the modal
preference of panelists was for yellow (With 42
per cent of panelists).

Quantity (number) preference: Significant effect
of pricing was found here. In case 1, the modal
preference of panelists was for units of twelve
(with 49, per cent of panelists). However, in
case 2, units of six were more p0pular than units
of eight, twelve, and three other purchasing—units

(24:16:15:13:13:1l).

(4)

(5)

118

Gladiolus--

A.

Roses
A.

Color preference: Preference for individual color
was not pronounced. It was more or less evenly
distributed among the colors. In the "write'in"
panel the "no color preference“ and "mixed colors"
were almost equally as popular as the individual
colors (17 per cent for yellow, 16 per cent for
white, 16 per cent for pink, 12 per cent for no
color, 12 per cent for red, 11 per cent for mixed and
variegated, 5 per cent for lavender).

Quantity (number) preference: Price seemed to have
no effect on the preference for units of four,
five, and six. In case 1 the modal preference of
panelists was for units of twelve (with 50 per cent
of panelists). In case 2 the units of twelve be-
came less p0pular than units of six and eight but
still was preferred more than two other purchasing

units (24:22:13:10:9).

Color preference: The p0pularity of red roses was
borne out in the results of the "write-in" panel.
Fifty eight per cent of those surveyed singled out
the red rose (both dark red and Better Times red)
as the favorite colored rose. However in the

”visual product" study, the preference for red

119

roses appeared less definite than in the "write-
in" test. Dark red was more popular than pale
lavender (variety Sterling-Silver) and four other
colors in both case 1 and 2 (case 1: 33 per cent
of panelists preferred dark red; case 2: 28 per
cent of panelists preferred dark red). Although
pale lavender had a relatively high percentage in
first preference it also showed a strong negative
attitude exhibited by a great number of panel mem—
bers (c.f. Table 19 Distribution of frequencies of
preference). This color was the preferred color
for some people, and the least preferred for many
others. The variety Better Times was used as the
sample for red roses. It was not considered to be
a superior variety among the red rose varieties
although the quality of the sample of Better times
was competitive to other colored samples.
Quantity (number) preference: No price-effect in
the 9" (except the units of three) and 12" roses
was noted. The quantity preference in all grades
was quite evenly distributed among the testing
unit-samples. Pricing had no effect on the

preference scores in units of three, five, or

seven in 15" roses. However, it seemed to increase

120

the preference for units of twelve and conversely
lowered the preference for units of nine. The
units of twelve were more popular than seven, five,
three, and nine in case 2 (26:22:22:16:14).

Grade preference (stem-length): Better Times Roses
with 15" stem were more popular than those with

12" stems in case 1 (38:26:12:10), the 12" roses
were more p0pular than 15' in case 2 (32:29:16:1l).
Twenty—one roses ranked low in preference . The

9" roses were not so p0pular as the 12" and 15".

10.

BIBLIOGRAPHY

Books

Alderson, W. Marketing Behavior and Executive Action.
Homewood, 111,: Richard D. Irwin, Inc., 1957.

Atkinson, J.“ (ed.) The Assessment of Human Motiyes.
New YOrk: Dvan Nostrand Co., 1958.

 

Beckman, T.N. and W.R. Davidson. Marketing. 7th ed.
New York: The Ronald Press Co. 1962.

Cambell, P. The Consumerylnterest, A.Study in Consumer
Economics. New York: Harper & Brothers, Pub.,
1949.

Churchman, C. W;, R. L. Ackoff, and M. Wax. Measurement
of Consumer Interest. Phi1., Pa.: U. of Penn.
Press, 1947.

 

Clark, L. H. Consumer Behavior. Vol. 3. New York:
Harper & Brothers. 1958.

 

Clements, F. and T. Meyers. ”How They Tell What we
‘Want." .Yearbook of Agriculture. United States
Department of Agriculture, Washington, D. C., 1954,
pp. 207-211.

Gordon, L. J. The function of the consumegyin a free
choice economy. The Kazanjian Foundation Lectures
1957. westport, Conn: The Calvin K. Kazanjian
Economics Foundation, Inc., 1958.

Henell, 0. Marketing Aspects ofHogsewiyes' Knowledge
of Goods. The Institute for Marketing and manage—
ment Research, 1953.

Katona, G. The Powerfnl Consumer. ngchological
Studies of the American Econgmy. New York: McGrawe
Hill Book Co., Inc., 1960.

121

ll.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

122

Kelley, P. C. Consumer Economics. Homewood, 111.:
Richard D. Irwin, Inc. 1953.

Kendall, M. G. The Problems of M Ranking_. 2nd. ed.
New York: Hafner Publishing Co., 1955.

Lazer, W. and E. J. Kelley. Managerial Marketing:
Perspectives angyVieWpoints. Rev. Ed. Homewood,
111.: Richard D. Irwin, Inc. 1962.

Leonhard, D. Consume§_Resaarch with Projective
TeChnigues. Henandoah, Iowa: Wbrld Publishing Co.,
1955.

.Liesveld,.J. H. The Retail Florist. New York: the
Macmillan Co., 1954.

Little, Arthur D., Inc. Flavor Research and Food
Acceptance. New York: Reinhold Publishing Corp.,
1958.

Luck, D. S., H. G. Wales, and D. A. Taylor. Marketing
Research. 2nd. ed. Englewood Cliffs, New Jersey:
Prentice—Hall, Inc., 1961.

Myrdal, G., R. Sterner, and A. Rose. An American
.Qilemma, The Negro Problem and Modern Democracy.
New York: Harper and Brothers, Pub., 1944, pp. 1027-
1064. '

Post, K. Flogist Crop Production and Marketing. New
York: Harper & Brothers, 1960.

Siegel, S. Nonparametric Statistics--§gr the Behavior
Science. New YCrk: McGraw-Hill Book Co., Inc.,
1956.

Thorndike, E. L. (ed.) The Psychology of wantgy
Interest, and Attitudes. New York: D. Appleton-
Century Crafts, Inc., 1959.

‘Wallich, H. C. The Cost gngreedom. New York: Harper
& Brothers, 1960. '

WOlf, H. A., W. G. Jennings, W. N. Jones, L. W. Latimer,
B. J. Rothwell, II, R. F. Vancil. Mbtivation
ResearCh. Boston, Mass.: Motivation Research
Associates, 1955.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

123

Periodicals
Abrahamsen, M. "Determining consumer preferences for
sweet potatoes: methods and results.” Journal of

 

Farm Economy. Vol. 31, Part I, 1949, pp. 635-639.

Alderson, W. "Psychology for Marketing and Economics."
Journal of_Marketing. Vol. 17, No. 2, Oct. 1952,
pp. 119-35.

Anderson, R. L. "Use of Contingency Tables in the
Analysis of Consumer Preference Studies." Biometrics.
Dec. 1959, pp. 582-90.

 

Anon. "Consumer panels as a marketing tool." Printer's
Ink. Vol. 213, Nov. 9, 1945, pp. 25, 318, 140.

 

Anon. "Consumers' Choice." The Economig. Vol. 199,
May 13, 1961, pp. 676.

Anon. "People: What's Behind Their Choices in Buying,
in WOrking." Businessweek. Aug. 14 (Part I: 50-61),
Aug. 21 (Part II: 130-143), Aug. 28 (Part III: 119-

), 1954.

Anon. "The 'I am me' consumer." Businessweek. Dec.
Dec. 23, 1961. pp. 38-9.

 

Applebaum, W. "Studying customer behavior in retail
stores." Journal of Marketing. Vol. 16, Oct. 1951,
pp. 172—78.

Baker, D. J. and C. H. Berry. "The price elasticity of
demand for fluid skin milk." Journal of_Farm
Economics. Vol. 35, Feb. 1953, pp. 124-29.

Bayton, J. A- “Consumer preference research in the
Department of Agriculture." Agricultural Economic
Research. Vol. II, No. 4. Oct. 1950, pp. 105-112.

 

Bayton, J. A. "Motivation, cognition, learning basic
factors in consumer behavior." Journal of
Marketing. Vol. 22, Jan. 1958. pp. 282—89.

Benson, P. H. "A model for the analysis of consumer
preference and an exploratory test." Journal of
‘Applied Psychology. Vol. 39, No. 5, Oct. 1955,
pp. 375-81.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

124

Benson, P. H. "A Psychometric Approach to Predicting
Consumer Preference." Personnel Psychology. Vol. 13,
1960. pp. 71-80.

Bilkey, W. J. “A.psychological approach to consumer
behavior analysis.” Journal of Marketing. Vol. 18,
No. 1, July 1953, pp. 18-25.

Bland, F. E. "Effectiveness of education programs on

meat consumption.“ Georgia Agricultural Research.
Ag. Exp. Sta., Ga. St. U. Vol. 3, No. 4, Spring
1962, pp. 8-10.

Blanenship, A. B. ”Creativity in consumer Research."
Journal of Marketing. V01. 25, Oct. 1962, pp. 34-38.

Bliss, C. I., M. L. Greenwood, and M. H. McKenrick.
”A comparison of scoring methods of taste tests
with mealiness of potatoes." Food Technology.
Vol. 7, 1953, pp. 491-95.

Boddewyn, J. “Galbraith's Wicked Wants." Jonrnal of
Marketing. Vol. 25, Oct. 1961, pp. 14-18.

Bradley, R. A. "Some Statistical Methods in Taste
Testing and Equality Evaluation." Biometrics.
Vol. 9, Part I, 1953, pp. 22-38.

Brown, G. H. "Measuring consumer attitudes toward
products." Journal of Marketing. Vol. 14, Apr.
1950. pp. 691-98.

Bryant, W. and J. Swingen. "Flowers . . . something
new for self-service--prepackaged.” .gpe-Pack-Age.
May 1948, pp. 15, 18-21, 22.

Burrows, G. L. ”Consumer Acceptance or Consumer
Preference." Agragultugal Economic Research.
Vol. 4, Apr. 1952. pp. 52-56.

Dedolph, R. R. and H. E. Larzelere. "Tomatoes-color
development and consumer preference." Qnarterly
Bulletin. Agricultural Experiment Station, Michigan
State University. Vol. 45, No. 2, Aug. 1961,
pp. 219-222.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

125

Dichter, E. A. "What Are the Real Reasons People Buy
To-day?" Sales Management. Feb. 1 (Part I: 36-
38, 86-89), Feb. 15 (Part II: 46-55), 1955.

Friedman, M. "A comparison of alternative tests
significance for the problem of M rankings.”
Annual Mathematic Statistag. Vol. 11, 1940,
pp. 86-92.

Gaarder, R. 0. and N. V. Strand. "Use of photographs
in consumer preference studies of pork.” Journal
of Farm Economics. Vol. 39, Feb. 1957. pP- 59-66.

Grange, G. W. "An experiment study of colour prefer-
ences." Journal Genetic Psychology. Vol. 52,
1955: pp. 3-20.

Grattan, C. H. "The Complex world of the Consumer."
Consumer Reports. Vol. 20, Aug. 1955, pp. 384-5.

Greig, W. S., E. L. Bedford, and H. Larzelere.
”Consumer preferences among apple varieties in
fresh and processed forms." Quarterly Bulletin.
Agricultural Experiment Station, Michigan State
University. Vol. 44, No. 3, Feb. 1962, pp. 505-526.

Guilford, J. P. and P. Smith. "A.system of color
preferences." American Journal Psychology.
Vol. 72, Dec. 1959, pp. 487-502.

Hansen, N. J. and I. I. Holland. "A.consumer prefer-
ences study of charcoal." ‘gowaygtate Joannal_gf
Science. Vol. 34, No. 4, May, 1960, pp. 709-712.

Howell, L. D. "Some phases of consumer preference and
demand research." Journal ongarm Economy. Vol.
36, Nov. 1954, pp. 641-653.

 

John, M. E. "Classification of values that serve as
motivators to consumer purchases.” Journal a;
Farm Econpmy Vol. 38, Nov. 1956, pp. 956-63.

Katona, G. "Have consumers stopped.wanting?" Printers'
Ink. Feb. 17, 1961, pp. 20, 22, 24, 26.

Krugman, H. B. "The learning of consumer preference."
Journal of Marketing. V01. 26, No. 2, Apr. 1962,
p. 31.

 

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

126

Kuehn, A- A. and R. L. Day. "Strategy of product
quality." Harvard Business Review. Vol. 40,
Nov. 1962. PP. 100-110.

Larzelere, H. B. and R. R. DeDolph. "Consumer accept—
ance of greenhouse-grown and southern field-grown
tomatoes." Iguarterly Bailetin. Agricultural
Experiment Station, Michigan State University.
Vol. 44, No. 3, pp. 554-558.

Larzelere, H. E. and R. D. Gibb. "Consumers' Opinions
of quality in pork chops.“ .Quarterly Bulletin.
Agricultural Experiment Station, MiChigan State
University. V01. 39, No. 2, Nov. 1956, pp. 327-
333.

Lesly, P. "The new consumer who thinks for himself."
Sales Marketing. Vol. 84, Mr. 18, 1960, pp. 33,
130, 132

Martineau, P. D. ”Martineau presents fresh rebuttal
to critics' Charges of waste, obsolescence in U.S.
Economic system.” Journal of Marketing. V01. 26,
No. 1, Jan. 1962, pp. 27.

Matson, Ed. and P. Krone. "Per Capita Sales Vary in
Eight Cities." MiChigan Florist. No. 247,
Nov. 1951. pp. 9-21.

McCormick, E. J., R. E. BlanChard, and G. G. Karas.
"The effect of hue, brightness and saturation on
color preference considering various contexts."
.gioceedings Indiana Academy Science. Vol. 67,
1957, p. 297.

Mills, W. C., Jr., L. E. Dawson and H. E. Larzelere.
"Consumer preferences for turkeys with various
grade labels." Quarterly Bulletin. Agricultural
Experiment Station, MiChigan State University.
Vol. 43, N0. 2, pp. 249-260.

Newman, P. "Basic Assumptions in Preference Theory."
Economic Journal, Vol. 71, Sept. 1961, pp. 535-538.

Pasto, T. An, and P. Kivisto. ”Group differences in
color choice and rejection." Journal of Clinicai
Psychology. Vol. 12, 1956, pp. 379-81.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

127

Peryam, D. R. and N. J. Gutman. "Variation in
Preference Ratings for Food Served at Meals."
Food Technology. Vol. 12, Jan. 1958, pp. 30-33.

Politz, A. and W. E. Deming. "On the necessity to pre-

 

sent consumer preferences as predictions." The
Journal of Marketing. Vol. 18, No. 1, July, 1953,
pp. 1-5.

Releigh, T. "Consumer motivation study advises special
occasion rose marketing.” ‘giorist Exchange &
Horticulture Trade world. Je. 28, 1958, pp. 9,29.

 

Rhodes, V. J. "The Measurement of Consumer Preferences."
Journal of Farm Economics. Vol. 37, Nov. 1955,
pp. 638-51.

Rhodes, V. J. "The variability of consumer preferences."
Journal of Farm Economics. V01. 41, Aug. 1959,
pp. 519-27.

Riesen, A. H. "The role of light and color in psycho-
biology." Acta Psychologica. Vol. 11, 1955,
pp. 221-22.

Sanders, C. L. "Color Preferences for Natural Objects."
Engineering. Vol. 54, July 1959, pp. 452-56.

Shaffer, J. D. ”Some observations concerning the
relationship of consumer research to consumer
education.” Journal of Farm Economy. Vol. 34,
No. 4, Nov. 1952, pp. 551-55.

Sobol, M. G. "Panel mortality and panel bias."
Journal oi American Statistic Assogiation. Vol. 54,
No. 285, 1959, pp. 52-68.

Southworth, H. M. "Progress in developing a set of
principles in the area of consumer behavior."
Journai of Farm Economy. Vol. 36, Dec. 1954,
pp. 1071-82.

Stonborough, T. H. W. "Fixed Panels in consumer
research." Journal of Marketing. Vol. 7, Oct. 1942,
pp. 129-138

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

128

Thurstone, L. L. "A law of Comparative Judgment."
Psychology Review. Vol. 34, 1927, pp. 273-286.

Thurstone, L. L. "The indifference function.” Journal
of Social Psychology. V01. 2, May 1931, pp. 139-
167.

Thurstone, L. L. "Experiment Methods in Food Testing."

Journal of Applied Psychology. Vol. 35, No. 3,
Je. 1951. PP. 141-45.

Tobin, B. F. "Consumer preferences are key to marketing
trends." The U. S. Egg anapPoultiy. V01. 54, No. 9,
Sept. 1948, pp. 12-13, 26-28.

Von Oppenfeld, H. "Michigan Floricultural Production."
Quarterly Bulletin. Agricultural Experiment
Station, Michigan State College. V01. 37 (No. 1),
Aug. 1954, pp. 35-51.

Von Oppenfeld, H. "Retail Distribution of floricultural
crops in Michigan.“ .Qnarterlyggulletin. Agri-
cultural Experiment Station, Michigan State College,
E. Lansing, MiChigan. V01. 37 (No. 1), Aug. 1954,
pp. 52-59.

White, I. S. “The functions of advertising in our
culture.” Journal of Marketing. Vol. 24, J1. 1959,
pp. 8-14.

Bulletins

Anon. "A social profile of Detroit: 1955." Detroit
Area Study, The Universityyof Michigany Survey
ResearCh Center, Institute for Social Research,
1956, PP. l-46.

Anon. ”Citrus preferences among household consumers in
Louisville and in Nelson County, Kentucky." United
States Bureau of Agriculture Economics, washington,
D. C., Ag. Inf. Bul. No. 2, 1950.

Anon. ”Consumer preferences for frozen peas in relation
to standards for grades.” United States Department
of Agriculture, washington, D. C., Mktg. Res. Rept.
No. 280, 1958.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

129

Anon. "Potato preferences among household consumers."
United States Department of Agriculture, Washington,
D. C., Misc. Publ. No. 667, 1948.

Anon. Report of the Definition Committee of the
American Marketing Association, published by the
Association, Chicago, 1961.

Anon. "women's preferences among selected textile pro-
ducts." United States Department of Agriculture,
washington, D. C., Misc. Publ. No. 641, 1947.

Bell, H. P. "Preferences for Canned Grapefruit Juices."
United States Department of Agriculture,‘WaShington,
D. C., Mktg. Res. Rept. No. 108, 1955.

Bliss, C. I., E. 0. Anderson, and R. E. Marland. "A
technique for testing consumer preferences, with
special reference to the constituents of ice
cream.” Conn. (Storrs) Ag. Exp. Sta., Bul. 251,
1943.

Branson, R. E., M. Jacobs, and R. Hall. "Frozen grape-
fruit sections; evaluating a new product by retail
sales audit and household survey." United States
Department of Agriculture, waShington, D. C.,

Mktg. Res. Rept. No. 110, 1955.

Bunk, M. E. and R. N. Hampton. ”Problems in selling
prepackaged roses in grocery supermarkets and
variety stores.“ Cornell U., Ag. Exp. Sta. Ithaca,
N. Y., A. E. 930, Oct. 1953.

Burrell, H. "The aqua-pak—-a new package for selling
flowers.” Cornell U., Ag. Exp. Sta. Ithaca,
N. Y., A. E. 1060, Ap. 1957.

Crosby, G. and A. L. Larson. "Relationship of Okla-
homa Livestock Producers Stated Marketing Prefer-
ences to Marketing Behavior." Okla. St. U. Exp.
Sta., Processed Service p. 356, J1. 1960.

DeloaCh, D. B. "The cut flower industry: an analysis
of its growth potential." Calif. Ag. Exp. Sta.,
mime. Rept. No. 214, Ja. 1959.

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.

130

Dewey, A. "Retail demands for floricultural products.
"U. of Conn. Exp. Sta., College of Ag., Bul. 343,
Apr. 1959.

Dwoskin, P. B. and M. Jacobs. "Potato flakes, a new
form of dehydrated mashed potatoes; market position
and consumer acceptance in Binghamton, Endicott, and
Johnson City, New York." United States Department
of Agriculture, Washington, D. C., Mktg. Res. Rept.
No. 186, 1957.

Early, J. W. "The sale of flowers . . . preference
factors and merchandising methods." Penn. Ag.
Exp. Sta., A. E. & R. S. 16, Je. 1958.

Early, J. and C. E. Trotter. "Consumer preferences
for floral products." Penn. Ag. Exp. Sta., Prog.
Rept. 143, Jan. 1956.

Fossum, M. T. "Floricultural and ornamental horti-
cultural economics." Society of American Florists
and Ornamental Horticulturists, 1950.

Fossum, M. T. ”Trade in horticultural specialties
1890-1950." United States Department of Agriculture,
Ag. Mktg. Serv., waShington, D. C., Mktg. Res.
Rept. No. 33, May, 1957.

Hampton, R. N. "MerChandising flowers in retail florist
sh0ps." Cornell U. Ag. Exp. St., Ithaca, N. Y.,
A. E. 1004, Oct. 1955.

Hampton, R. N. and J. L. Kupka. "Problems in retail
pricing and packaging of flowers--for sale in
self service outlets." Cornell U., Ag;. Exp. Sta.
Ithaca, N. Y., A. E. 1003, Oct. 1955.

Jasper, A. W. "Some highlights from consumer egg
studies." United States Department of Agriculture,
washington, D. C. Inf. Bul. No. 110, 1953.

Jasper, A. W. and R. E. Cray. "Consumer Preferences,
Practices and Demands in Purchasing Eggs and Poultry
in Columbus, Ohio, 1950." Ohio Ag. Exp. Sta. Res.,

Bul. 736, Oct. 1953

131

110. Kelly, R. A. "Floricultural sales in mass market out-
lets." U. of Ill. Ag. Exp. Sta., Bul. 675, Aug.
1961.

111. Knight, W. R. "An attempt to determine the effects of
advertising and other factors on florists' sales."
Mich. St. College, Dept. of Ag. Econ., July, 1952.

112. LeClerg, R. E. "Consumer preference for cut flowers
and pot plants in Mississippi." Miss. Ag. Exp.
Sta., Bul. 565, Sept. 1958.

113. Schuyler, F. O., R. B. Miner, J. D. Long, and W. O.
Yoder. "Orientation to business research."

Indiana University, Indiana Readings in Business,
No. 12, 1956.

114. Sherman, R. W. and M. Baker. "Sale of potted plants
and cut flowers through supermarkets." Ohio Ag.
Exp. Sta., A. E. 336, 1961.

115. Sherman, R. W., D. C. Kiplinger, and H. E. Williams.
"Consumer preferences for cut roses, carnations,
and chrysanthemums." Ohio Ag. Exp. Sta., Res.
Cir. 31, Mar. 1956.

116. Sorensen, H. B., A. F. Dewerth, and E. R. Jensen.
"Production methods and new markets for Texas
florist crops." Texas Ag. Exp. Sta., Bul. 900,
Ap. 1958.

117. Sorensen, H. B., R. E. Odom and A. F. Dewerth. "Con—
sumer preference for flowers and ornamental plants.
Texas Ag. Exp. Sta., MP-489, Feb. 1961.

118. Trotter, W. K. "Problems in marketing florist crop."
Cornell University Agricultural Experiment Station,
Ithaca, N. Y., A. E. 983, Je. 1955.

119. Von Oppenfeld, H. "A study of the retail florist
business." Cornell U., Dept. of Ag. Econ., A. E.
764, Jan. 1951.

120. Von Oppenfeld, H., J. F. Schwartz, and P. R. Krone.
"Merchandising in retail flower shop." Mich. Ag.
Exp. Sta. Spec., Bul. 412, Apr. 1957.

121.

122.

123.

124.

125.

126.

127.

128.

132

Wilson, R. F. Horticultural Colour Chart. The British
Colour Council in collaboration with The Royal
Horticultural Society. 1938.

Unpublished Material

Early, J. W. "Consumer Attitude and Preferences for
Floral Products." Unpublished M. S. Thesis,
Graduate School of Pennsylvania State University,
1956.

Rhodes, V. J. "A Theoretical and Empirical Investigation
of Consumer Preferences for Beef by Grades in
Metropolitan St. Louis." Unpublished Ph.D. Thesis.
Graduate School of Harvard University, 1955.

Shaffer, J. D. "Methodological Basis for the Operation
of a Consumer Purchase Panel." Unpublished Ph.D.
Thesis, Graduate School of Michigan State College,
1952.

Tolle, L. J., Jr. "Strategy considerations in Changing
the Retail Outlet for Floral Products." Memeo,
undated.

Speedhes

Tolle, L. J., Jr. "Projections for Floriculture.” A
speech presented at the wholesalers session, Mich.
St. Florist Assn. meeting, Detroit on Feb. 21, 1961.

Voigt, A. 0. "Marketing of Flowers and Ornamentals."
A report to the Horticultural Agindustry Conference,
Penn. St. U. on May 4, 1961.

Miscellaneous

Anon. "Census of Agriculture, Special Reports,
Horticultural Specialties." V01. V, Part I,
Bureau of Census, Washington, D. C., 1950,
PP. 33-37, 429-433, 733, 755.

133

129. Anon. "Census of Agriculture, Special Reports,
Horticultural Specialties." Vol. V, Part I.
Bureau of Census, Washington, D. C., 1959, pp. 26—
27, 45-49, 109-111, 158, 168, 175.

130. Anon. "Census of Business, Retail Trade--Sales Sizes."
Bureau of Census, Washington, D. C., 1958, pp. 2—35.

131. Anon. "Cut Flowers, Production & Sales." Statistical
Reporting Service, Crop Reporting Service, Crop
Reporting Board, washington, D. C., 1958-59,
Intentions for 1960 in Ten Selected States.

JAPPEHUIEX 1

MICHIGAN STATE UNIVERSITY EAST LANSING

 

DEPARTMENT OF AGRICULTURAL ECONOMICS

Dear Homemaker:

We at Michigan State University are carrying on a research
project to help the food industry produce, process and sell the
quality of various food products that you as a consumer prefer.

We are doing this by having groups of interested consumers come to
a display room arranged in Detroit with the cooperation of Wayne
State University.

We will make a list of all those who indicate on the enclosed
questionnaire that they are willing to take part in the study. From
this list we will select the names of a number of people to visit
the display room. We are asking questions about your age, education,
and income so that we can get a better idea of family characteristics
in Detroit. This information in turn will make our research results
more applicable to a large number of families in the city of Detroit.

If you are one of those chosen, we hope that you can come to
the display room and give us your opinion of various samples of eggS,
poultry, apples, potatoes and other products. We feel'sure the
visit will be most interesting to you. Furthermore, you will be
helping to improve the general marketiquality and grades of these

products. '
Very truly yours,
Henry E Larzelere
Associate Professor in
Agricultural Economics
HBL:jds

P.S. Please fill out the questionnaire and send it back whether or

not you are able to take part.
H.H.L.

134'

First of all, how many persons are there in your household? That is, how
many eat their meals regularly in your home? 1.

What are the approximate ages of the heads of your household? Please
check the one which fits best.

2. Female head 3. Male head
( ) a. Under 30 ( ) a. Under 30
( ) b. 31-us ( ) b. al—us
( ) c. ue-so ( ) c. us-eo
( ) d. Over 60 ( ) d. Over 60

About how many years of formal education were completed?

u. By the female head 5. By the male head
( ) a. 0—8 years ( ) a. 0-8 years
( ) b. 9-11 years ( ) b. 9-11 years
( ) c. 12-13 years ( ) c. 12-13 years
( ) d. 14 or more ( ) d. 14 or more

How many members of your household are employed more than one-third of the
time outside of home? 6.

AbOut what is the total yearly income of all members of your household from
all sources after federal income taxes were deducted? Please check the cate-
gory below which fits your best estimate.

7. ( ) a. Under $2,000 ( ) d. 5,u01—7,ooo
( ) b. 2,000-u,ooo ( ) e. 7,001-10,ooo
( ) c. u,001-5,u00 ( ) f. 10,000 a over

8. Who does most of the food buying for your family? (Check one)
Male Female Both Other
head head

 

 

 

 

9. How often do you usually purchase these items? (Check one in each column)

Eggs Chicken Processed Potatoes Lamb Turkey
fryers Apples
Once a week

 

 

Once in 2 weeks
Once a month
Once a year

Never

 

 

 

 

 

 

If you are selected would you be willing to visit our display room at a time
that is convenient for both of us?

 

Name Phone No.

 

Address

 

Thank you for your help.

APPENDIX 2

monies: 3mm mammal??? East Lansing

 

College of Agriculture . Department of Horticulture

November 12 , 1962

Dear Sir:

My name is Linda Han. I am a graduate student studying
Floriculture Marketing at Michigan State University. We are
conducting a research project studying consumer preferences
in cut flowers and the resulting effect of super market sales
on retail florist businesses. By so doing, we hope to gain
more knowledge about your customers' preferences for cut flowers
in retail florist shops as opposed to mass market outlets.
When we have smmrized this information, we will send a
copy to you, if you request it. Your answer will be treated
in the strictest confidence. I am sure you will find direct
benefits to your business by the knowledge gained.

We are asking you to help us to check our "laboratory test"
data with your actual sales records. Please fill out all the
blanks in the enclosed questionnaire sheets using your best
Judgment and with the assistance of your records. We shall
compare your replies and the laboratory test results and see
if there is any deviation between them. If there is a material
difference, then we shall study in which ways these two
sources of information differ.

You have been chosen for this particular study, because you
are operating the very sort of business which is desirable for

this particular study. We sincerely thank you for your coopera-
tion and hope to hear from you within 10 days.

Very truly yours,
Sim/W // 171/
Linda Han
Ehc.

F-l
general Information

1. My business is: (check one)
Retail
Retail-grower __

2. My last-year sales volume ($) is between:

30, OOO-or-less
31,000- 60,000
61,000- 90,000
91,000-120,000
121,000-150,000
151,000-or-more
3. What is the size of your trading area (approximate population)? Check one.

h,999-or-less
5, 000- 11:, 999
15, 000- 29, 999
30,000- 59,999
60, COO-119,000
120,000-239,999
21:0, GOO-M9, 999__
SO0,000~or-more f

 

h. Location of my store is: (check one)
Roadside
Neighborhood shopping district
Shopping center
Central shopping district (down-town)

 

 

 

 

5. My store is near: (check those which apply)
Hospital
Cemetery
Shopping center
Office building:
School university)
Highway
Other traffic causing feature (Specify)

 

6. Do you operate a greenhouse? Yes No

7. Do you offer cash-and-carry specials-neither unarranged cut flower and/or
small plants? Yes__ Nor

If so, how often? (check one)

Daily
Once a week
Once a month
Holida s _
Other Specify)

 

8. Estimate the percentage of customer's orders placed by:
Telephone %
In person j

9.

10.

ll.

12.

F2

Of orders placed in rson, the items most frequently (number of salsa)
bought are: (check-one )

Arrangement of cut flowers
Potted flowering plants
Foliage plants

Giftware

 

 

 

when your customers specify a type of cut flower (loose, wrapped, or
boxed only), what types most frequently are requested? Estimate in
percentage of their dollar-value importance.

 

 

 

 

 

Carnation AA 4%
Ole 'olus fl
mum Std.) z
Orchid__ 4%
Pompon z
Rose %
Snapdragon z
Other cut flowers %

 

Approximately what is the sales volume of each of the following kind of
cut flowers(made-up and loose cut flowers) in your total business
sales volume? (by percentage of the dollar value)

Roses__ ” fl

 

 

 

 

Mums_ %
Carnations_g %
Gladiolus A_%
Pompous %
A11 ot or flowers %

 

'—.-——A .. -.-——-..A—.. ‘

What are the price ranges you offer for the following cut flowers
(0H1Y'the loose flowers-boxed or bunched)?

 

 

 

 

Roses /doz.
Mums _/doz.
Pompons__g A_JK10" bu.
Carnations /doz.
Gladiolus 4__/doz.

 

Customer Preferences

1.

Estimate the percentage of your’business (in.number of sales) coming from
the following sources for each type of cut flowers. (on annual basis)

Rose Carnation Mum(std.) Pompon Gladiolus

Funeral & grave

 

 

 

 

Hospital

 

 

Wedding

 

 

Commercial or

 

indus. decoration

 

 

Birthday

 

 

4
1

Anniversary

 

 

Every-day use

 

 

 

I I
n
. l
' I
. I
e
e
V
' l
o
.
i
.-
.‘
a A. - -r-.-‘ - ,, a l __ . ‘ .
t .
, .
x
.
o
o
”e ‘Vom-ncm ”fin-yum, \4.‘ .q,. ,. ~—, . q — ‘ ‘I’F.""’ ".;A '~’_ . . - A 4
a . . . , . H- ‘ . « v ‘1
’ D
x
I
c

  

*v‘s.- ,-~.r

)‘»:l" . . "'v
.‘ Jn's"j-~.hu]1‘

\

J;-

.3»; "or arm

. ~_
‘ . , .
0 Y e
.— v ..__' D’r :__ Q
. . \‘
. 1.. ‘ 5’
_, . A '\
,-bt’"“-’
..v
I
. . .

   

[‘45
‘wt-5

““4 «hm-fly.

"1
"it"
‘I

'u

4

'9 sonar

 

h

’ ‘1 .-~ s... .: -~- -~ ‘.~. -, .
.‘s as. mat:.f-.'{m.’:" Ll. . 491:4 ‘3 (up. 2f.v_.-:)_~:rr.p.f .7 -. 69175! 9111.232“.

I o _! ,'~ " 1‘. 4’4‘.‘
’. . ‘ ‘ I,- ~ .,' w' J", ‘1}
t "3 2.." '1... f‘flfif’ld *‘w'
,4'.’ 5,. -
' “w:
I
.
o -‘ ‘O

“ . .4 4 _-
félftz'tr'

 

e
, .
.
.
n .' -a..‘-'es'. "!~"
_-. .
,_ .,
. «. .. a 'v- up,
.
.
s
.
.
v
e
- s ..
I

 

anw‘b‘m' "‘

2. Estimate in percentage, how often do your customers specify each
following colors of roses.

No color Specified 5
Better Times

Dark red _J£
Yellow

 

 

Other (specify)

 

%
Pink J
Nflxed color %

 

3. Estimate in percentage, how often do your customers specify'each following

colors of carnations.

No color Specified 5‘

Red
Novelty

Light pink
Dark Pink
White

Yellow

Other (Specify)

 

 

 

 

 

Mixed color

Ame-L

 

h. Estimate in percentage, how often do your customers specify each following

colors of pggpons.

No color specified %
White i
Dark Lavender___
Lavender

Dark bronco

Pale bronze
Yellow

Other (specify)
Mixed color lrfi

 

 

 

interstate

 

5. Estimate in percentage, how often do your customers specify each
following colors of gladiolus.

No color specified 1,
Lavender

Red

Novelty

White

Yellow

Other (Specify)
Mixed color

 

 

 

 

 

aa‘cfi‘cfi‘ofiak‘aa

 

Fh

6. Estimate in percentage, how often do your customers specify each following
colors of Ell-m Std.).

No color preference f
Yellow

Lavender

Pink

White

Bronce

Other (specify)
Mixed color

 

 

 

3Q) m‘ofibfl‘ofi

 

Size 2f Units purchase

1. The percentage of roses sold in the following units-of-purchase is:

 

 

 

 

 

flowers-or-less Z
flowers _%
5 flowers g_%
6 flowers %
7-11 flowers %
doz.-or-more %

 

2. The percentage of mumcs (std.) sold in the following units-of-purchase is:

3 flowers-or-less
h flowers
5 flowers
6 flowers

. 7-11 flowers
doz.-or-more

 

 

 

‘cfl‘efl Lainey; 3Q

 

3. The percentage of pompgns sold in the following units-of-purchase is:

 

 

one bu. (10") 4%
two bu. (10") }
three-or-more bu. (10") %

 

h. The percentage of carnations sold:in the following units-of-purchase is:

 

 

 

 

 

3 flowers-or-less Ag!
h flowers %
5 flowers %
6 flowers %
7-11 flowers %
doz.-or-more _%

 

S. The percentage of gladiolus sold in the following units-or-purchase is:

3 flowers-or-less
h flowers

5 flowers

6 flowers

7-11 flowers

 

 

laah‘nahfioa

 

r
.0--

 

s:

R0031 U f' C

1.4!
:‘i‘ Y

 

 

 

 

R

"'illifitfijfluflilllfllfllflfmlitwfglfififS