; WM

FASHION. O'PIN‘ON LEADERSHIP AND
FASHION. ADOPTION m RELATION
To SOCIAL PARnchTIOM AND
FAVORABLENESS TOWARD NEW STYLES
m umveesm WOMEN’S momma .

Thesis for the Degree of M. A.
.MlCHiGAN STATE UNIVERSlTY
CAROL ANN. MYERS
.197 1

.
.1 Jw'o
w.—

 

   

t

.‘

 
  

LIBRARY

Michigan State
UanCI‘Slty 5mg;

 

 

 

 

 

 

 

 

MSU

LIBRARIES

V

 

 

 

 

RETURNING MATERIALS:

P1ace in book drop to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped be1ow.

 

 

 

 

ABSTRACT

FASHION OPINION LEADERSHIP AND FASHION ADOPTION
IN RELATION TO SOCIAL PARTICIPATION AND
FAVORABLENESS TOWARD NEW STYLES IN
UNIVERSITY WOMEN'S CLOTHING

BY

Carol Ann Myers

The main purpose of this study was to compare the
characteristics of three types of prestigious fashion
leaders among university women (fashion opinion leaders,
fashion innovators, and fashion dual leaders) and fashion
non-leaders on favorableness toward new styles and social
participation. A second objective was to compare each of
the four sub-categories of respondents above with the rest
of the subjects on the favorableness toward new styles
variable in order to determine whether certain sub-
categories were more emulative of group norms than others.

Rogers' generalization (Diffusion of Innovations,

 

1962) that opinion leaders must not be too different from
others in order to communicate and carry out their influ-
ential role was the basis for the proposition that fashion
opinion leaders would not differ significantly from others

on favorableness toward new styles. The other three

sub-categories were hypothesized to be more free to show

non-conformity to group norms.

Carol Ann Myers

Questionnaires were mailed to 500 university women
students who were randomly selected from the 12,765 women
who met criteria for the p0pulation. Replies from 243
women were used in the statistical analysis of the data.
The questionnaire included the following measures: (1)
Schrank's Fashion Opinion Leadership measure, (2) a
modification of Schrank's Fashion Innovativeness measure
to assess fashion adoption, (3) a Favorableness Toward New
Styles measure which was developed for the study, (4) a
modification of Chapin's Social Participation Scale, and
(5) a modification of the McGuire-White Index of Social
Status based on the income, occupation, and education of
the main wage earner. Background information concerning
the respondents' year in university and marital status was
obtained from the university enrollment records.

Data were analyzed primarily by Chi-square tests
and post-hoc Chi-square comparisons of differences between‘
groups. Pearson r correlations verified the Chi-square
test results and measured the effects of marital status
and year in university on the test variables (see 1-4
above). The .01 level of significance was the basis for
the acceptance or rejectance of the hypotheses.

The findings of this study indicated that marital
status, year in university, and socio-economic status

were not influencing any of the test variables. The data

revealed that highly significant positive relationships

existed among fashion adoption, fashion opinion leadership,

Carol Ann Myers

and favorableness toward new styles. Neither fashion
opinion leadership nor fashion adoption, however, was
significantly related to social participation as measured
by formal organizational participation in the university
area, although slight positive relationships were found
among the variables.

Respondents were divided into four mutually
exclusive sub-categories based on their scores on the
fashion adoption and fashion opinion leadership measures.
Fashion Opinion leaders, fashion innovators, fashion dual
leaders, and fashion non-leaders did not differ signifi-
cantly on social participation. Significant differences,
however, were found among the four sub-categories on
favorableness toward new styles. Further analysis of the
data revealed that both fashion dual leaders and fashion
opinion leaders were significantly different from fashion
non-leaders.

Each of the four sub—categories was compared with
the rest of the subjects on favorableness toward new
styles. Fashion opinion leaders and fashion innovators
were not significantly different, while fashion dual
leaders and fashion non-leaders were significantly differ-
ent from the rest of the subjects. The findings seemed to
indicate that fashion Opinion leaders and fashion
innovators show more emulation of group norms than fashion

dual leaders or fashion non-leaders.

FASHION OPINION LEADERSHIP AND FASHION ADOPTION
IN RELATION TO SOCIAL PARTICIPATION AND
FAVORABLENESS TOWARD NEW STYLES IN

UNIVERSITY WOMEN'S CLOTHING

BY

Carol Ann Myers

A THESIS

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

MASTER OF ARTS

Department of Human Environment and Design

1971

ACKNOWLEDGMENTS

The writer wishes to express sincere gratitude to
her advisor, Dr. Anna M. Creekmore, for her encouragement
and guidance in directing this study; to each of the
members of her committee, Dr. Robert R. Rice, Dr. Joanne B.
Eicher, and Dr. Holly L. Schrank, for their interest and
helpful suggestions; to Dr. Mary Ellen McSweeney for her
advice concerning the statistical treatment of the data;
to the graduate students who served as judges, particularly
Joyce L. Allred and Mary Ann Tucker for their enthusiasm
and willing assistance; and to her parents, Mr. and Mrs.
Harold E. Myers, whose understanding and encouragement

made this study possible.

ii

TABLE OF CONTENTS

Chapter Page

I 0 INTRODUCTION 0 O O O O O O O O I I 1

II. REVIEW OF THE LITERATURE . . . . . . . 8

The Meaning of Fashion . . . . . . . 8
Characteristics of Prestigious Fashion

Leaders . . . . . . . . . . . 21

Methodologies Used to Measure Fashion
Acceptance and Prestigious Fashion

Leaders . . . . . . . . . . . 40

III. STATEMENT OF THE PROBLEM . . . . . . . 45

Definition of Terms . . . . . . . . 45

Hypotheses . . . . . . . . . . . 50

Assumptions. . . . . . . . . . . 57

IV 0 PROCEDURE 0 O O O O O O O O O O O 59

Selection of the Sample. . . . . . . 59

Selection and Development of the Measures. 61

Pre-Testing the Initial Measures. . . . 80

Administration of the Final Measures . . 81

Statistical Analysis of the Data. . . . 82

V. RESULTS AND DISCUSSION. . . . . . . . 86

Description of the Sample . . . . . . 86

Descriptive Data Concerning Each of

the Variables . . . . . . . . . 90
Relationship Between Fashion Adoption and

Fashion Opinion Leadership . . . . . 101

Relationship Between Fashion Opinion

Leadership and Favorableness Toward

New Styles . . . . . . . . . . 103
Relationship Between Fashion Opinion

Leadership and Social Participation . . 105
Relationship Between Fashion Adoption
and Social Participation. . . . . . 106

iii

Chapter Page

Relationship Between Fashion Adoption and

Favorableness Toward New Styles . . . . 108
Significance of Differences Between
Sub-Categories on Social Participation. . 109

Significance of Differences Between

Sub- -Categories on Favorableness Toward

New Styles . . . . . . . . . . 111
Significance of the Difference Between

Fashion Opinion Leaders and the Rest of

the Subjects on Favorableness Toward New

Styles . . . . . . . . . . . 114
Significance of the Differences Between Each

of the Following Sub- -Categories and the

Rest of the Subjects on Favorableness

Toward New Styles: Fashion Innovators,

Fashion Dual Leaders, and Fashion

Non-Leaders. . . . . . . . . . . 116

VI. SUMMARY AND CONCLUSIONS . . . . . . . . 119

Summary. . . . . . . . . . . . . 119

Implications . . . . . . . . . . . 126

Recommendations . . . . . . . . . . 132

BIBLIOGRAPHY . . . . . . . . . . . . . . 138
APPENDICES

Appendix
A. Questionnaire . . . . . . . . . . . 143
B. Fashion Curves . . . . . . . . . . . 151

iv

Table

1.

10.

11.

LIST OF TABLES

Chi-Square Values and Levels of Significance
for the Test Variables and Socio-Economic
Status 0 I O I I O I O I O O O O

Summary of the Measures of Central Tendency
and Dispersion for the Test Variables. . .

Percentage of Women Rating Each of the Styles
on the Fashion-Cycle Ranking Index. . . .

Percentage Distribution and Mean Number of
Styles Bought New or Sewn in Total
Wardrobe . . . . . . . . . . . .

Frequency and Percentage of Women Discontinuing
wearing Style List Items to Total Number of
Women Owning Style List Items . . . . .

Relationship of Women's Fashion Adoption and
Fashion Opinion Leadership . . . . . .

Relationship of Women's Fashion Opinion
Leadership and Favorableness Toward New
stYleS O C O O O I O O I O I O I

Relationship of Women's Fashion Opinion
Leadership and Social Participation . . .

Relationship of Women's Fashion Adoption and
Social Participation . . . . . . . .

Relationship of Women's Fashion Adoption and
Favorableness Toward New Styles. . . . .

Mean Scores on Fashion AdOption for Each
Sub-Category and the Rest of the Subjects .

Page

89

91

94

96

98

101

103

105

107

108

110

Table

12. Significance
Categories

13. Significance
Categories
Styles .

of Differences Between Sub-
on Social Participation . . .

of Differences Between Sub-
on Favorableness Toward New

14. Confidence Intervals for Significance of
Differences Between Pairs of Sub-

Categories
Styles .

on Favorableness Toward New

15. Confidence Intervals for Significance of the

Difference

Between Fashion Opinion Leaders

and the Rest of the Subjects on Favorable-
ness Toward New Styles . . . . . . .

16. Confidence Intervals for Significance of
Differences Between Each of the Following

Sub-Categories and the Rest of the Subjects

on Favorableness Toward New Styles:
Fashion Innovators, Fashion Dual Leaders,
and Fashion Non-Leaders. . . . . . .

vi

Page

111

112

113

115

117

Figure
1.

B.1.

LIST OF FIGURES

Four Sub-Categories of Respondents .

Fashion Curves for Styles Appearing in
Vogue, 1966-70 . . . . . . .

Fashion Curves for Styles Appearing in
vogue, 1970—71 0 o o o o o 0

Fashion Curves for Style List Items,
1968-71 . . . . . . . . .

vii

Page

48

151

152

153

CHAPTER I

INTRODUCTION

Social scientists have subscribed to the idea that
the basis for fashion change as well as the controlling
determinants are two polar motivations: (l) emulation or
the desire to be like the group, and (2) differentiation
or the desire for individual distinction.l Therefore,
theories of fashion change should logically integrate
these two polar concepts.

Fashion theorists, however, have generally
emphasized either emulation or differentiation rather than
the relationship between the two. As a result, researchers

investigating fashion adoption and the diffusion of new

 

1Paul H. Nystrom, Economics of Fashion (New York:
The Ronald Press Company, 1928), pp. 59-60, 80; Edward
Sapir, "Fashion," Vol. VI, EncyclOpaedia of the Social
Sciences (New York: The Macmillan Company, 1931), pp.
139-44 (hereinafter referred to as "Fashion"); Herbert
Blumer, "Fashion: From Class Differentiation to Collective
Selection," Sociological Quarterly, X (Summer, 1969),
275-91 (hereinafter referred to as "Collective Selection");
Georg Simmel, "Fashion," The American Journal of Sociology,
LXII, No. 6 (May, 1957), 541—58 (hereinafter referred to
as "Fashion"), reprinted from the International Quarterly,
X (October, 1904), 130-55; J. C. Flugel, The Psychologyyof
Clothes (New York: International Universities Press, Inc.,
1969), pp. 138—40 (hereinafter referred to as Psychology
of Clothes).

 

styles in women's clothing generally have been concerned
with the causal, rather than reciprocal, relationship
between certain prestigious individuals, who are influ-
ential in fashion acceptance, and the process of fashion
change.

Prestigious fashion leaders are often considered
to be nationally or internationally famous, or upper-class
society women. Little research has been done to identify
and profile other types of prestigious fashion leaders,
such as those who are among the first to adopt, or person-
ally influence others to adopt, new styles in their local
social system. The local prestigious leader may be more
influential in the diffusion and adoption of new styles
than the stereotyped leader. If new styles are diffused
within our society on both levels, a complete theory of
fashion change must explain the roles and characteristics
of both types of prestigious fashion leaders. Therefore,
in this study, the diffusion of new styles in women's
clothing will be investigated to determine what possible
role emulation or differentiation may play in the acceptance
of these new styles by prestigious individuals and their
followers within a relatively confined social system, a
university campus.

Young stated in 1937, that fashion ". . . must be
sufficiently rapid to outmode previous years but it must be

sufficiently slow to prevent the leaders from outdistancfing

their followers."1 The extent to which prestigious

persons may, while maintaining their influential position,
either differentiate themselves from others by wearing new
styles, or conversely, reflect emulation by wearing
clothing similar to their followers, is not known. Jack
and Schiffer, in 1947, investigated the reciprocal roles

of fashion promoters and followers and found that the
extremes of clothing styles as shown in the mass media,
were not accepted by the woman on the street. The

question Jack and Schiffer sought to answer was: "what

are the limits within which norm creators—-in this case
fashion designers—-must remain if they are to be followed."2
This study seeks to follow the diffusion process one step
further than Jack and Schiffer to contribute to answering
the question: will prestigious fashion leaders be followed
if wearing or promoting extreme styles?

Today each new fashion season introduces greater
numbers and varieties of clothing styles to consumers, yet
many of these styles are completely rejected. Distributers
of fashion need guidelines relative to how much, when,

and what types of new styles may produce favorable

 

lAgnes Brooks Young, Recurring Cycles of Fashion
(New York: Harper and Row, 1937), p. 4. (Hereinafter
referred to as Recurring Cycles.)

 

2Nancy K. Jack and Betty Schiffer, "The Limits of
Fashion Control," American Sociological Review, XIII (1948),
730-38. (Hereinafter referred to as "Limits of Fashion.")

 

reactions from consumers for more intelligent management
Of store Operations for the betterment of both the retailer
and the consumer.

Although several empirical studies of fashion
leadership have been conducted, the majority have examined
individuals' conformity to the mode, rather than comparing
different types of prestigious fashion leaders and their
attitudes and adoption of new styles. In this study two
recently introduced garment styles are used as a means to
assess the differences among the sub-categories of
respondents on their attitudes and behaviors. The new
styles are: (l) mid-calf lengths of dresses and pants,
and (2) "hot" pants, very short shorts. Differences among
the attitudes of the four sub-categories of respondents
may reveal differences in motives of emulation and differ-
entiation, and thereby lead to a better understanding of
the mechanism of fashion change.

The popular literature in the late 60's and early
V70's has suggested that university women are independent
and "do their own thing" in dress. However, casual
observations of college campuses seem to indicate that
dress is fairly uniform for many, but some women stand
out because of their fashionable dress. This apparent
discrepancy may indicate that innovators have more
favorable attitudes toward new clothing styles than

others. Or, possibly certain groups within the population

may also have differing needs for fashionable clothing due
‘tO differences in social participation.

Rogers,1 a rural sociologist and communications
theorist, has developed a theory of the diffusion and
adoption of innovations which includes generalizations
concerning the characteristics of certain typical
prestigious leaders in relation to group norms on inno-
vativeness. Rogers' theory has been utilized in a few
studies of leadership in the adoption of new fabrics and
clothing styles.2 While these studies have supported
some of Rogers' generalizations concerning the character-
istics of Opinion leaders and innovators, many variables
including attitudes toward new styles and social par-
ticipation have been given little attention.

The main objective of this study was to determine,
using Rogers' theory of the adoption and diffusion of
innovations, if there are differences among three types

of prestigious fashion leaders (fashion opinion leaders,

 

lEverett Rogers, Diffusion of Innovations (New
York: The Free Press, 1962): pp. 303-14.

 

2George M. Beal and Everett M. Rogers, "Infor-
mational Sources in the Adoption Process of New Fabrics,"
Journal of Home Economics, XXXXIX, No. 8 (October, 1957),
630-34 (hereinafter referred to as "Information Sources");
Holly Lois Schrank, "Fashion Innovativeness and Fashion
Opinion Leadership as Related to Social Insecurity,
Attitudes Toward Conformity, Clothing Interest and Socio-
economic Level" (unpublished Ph.D. dissertation, Ohio
State University, 1970) (hereinafter referred to as
"Fashion Innovation").

fashion innovators, and fashion dual leaders) and fashion
non-leaders on the following selected variables: (1)
social participation, and (2) favorableness toward new
styles.

A second objective was to compare each of the four
sub-categories of respondents above with the rest of the
subjects on favorableness toward new styles, in order to
determine whether certain sub-categories are more
emulative of group norms than others. If fashion opinion
leaders' attitudes toward new styles are similar to the
attitudes of the rest of the subjects, then Rogers' theory,
that the more influential leaders show a greater amount of
emulation or conformity to group norms than other
prestigious leaders, is valid.

A third objective was developed during the
execution of the present study when a need became evident
for: (1) the classification and clarification of the
fashion terminology employed, and (2) the development and
refinement of measures of the variables studied for
further research.

Knowledge of the characteristics of prestigious
fashion leaders and the limits of fashion control (the
balance of the forces, emulation and differentiation),
could help home economists to: (1) cooperate with
marketers in narrowing the gap between research and

development and actual consumer use of new fibers,

fabrics, and garments, (2) spread consumer information
concerning the selection, use and care of textiles and
apparel, (3) speculate about and perhaps predict fads

and modes, and (4) assist consumers in making their style
preferences and other apparel needs known to designers,
manufacturers and retailers. An investigation of the
roles of prestigious fashion leaders would be a contri-
bution to existing knowledge concerning fashion change,

human behavior, and social control.

CHAPTER II

REVIEW OF THE LITERATURE

The review of the pertinent literature including
theory and empirical research will be presented in the
following sections: (1) the meaning of fashion, (2)
theories of fashion acceptance, (3) characteristics of
prestigious fashion leaders, and (4) methodologies used
to measure fashion acceptance and prestigious fashion

leaders.

The Meaning of Fashion

Terms such as fashion, style, fad, taste, mode,
and fashion trend have had conflicting and popularized
meanings. For example, various authors have used the
following definitions of the term "fashion": fashion is
transitory usage, a series of recurring changes, a social
custom, a form of luxury, a department of the mores, col-
lective behavior, and a form of social regulation. Young
stated that the various definitions of fashion could
Ibe conceptually categorized in three ways: (1) a product,

(2) the process of change, and (3) collective behavior.

A11 definitions, however, dealt with qualities of fashion
rather than quantities.l Precise linguistic definitions
are needed for research interpretation and replication,2
and these definitions must be operationalized in quanti-
tative terms to be of use.

Fashion may be quantitatively defined in two ways:

3 or (2) fashion as the newest

(1) fashion as the mode,
style.4 Modal fashion, or the prevailing styles at any
given time, is determined by a fashion count of that style
which is worn most frequently. Quantifying fashion, when
defined as the newest styles, involves procedures such as
determining and calculating percentages of both new styles
worn and offerings in fashion publications and clothing
stores. Fashion counts of styles worn least frequently
may not locate new styles due to intervening factors such

as income or the fact that the particular style represents

a declining mode.

 

lYoung, Recurring Cycles, p. 201.

 

2Jean D. Schlater, Frances M. Magrabi, and Joanne B.
Eicher, "Social Science Methodology," Journal of Home Eco-
nomics, LV, No. 6 (June, 1963), 424-25.

 

3Nystrom, Economics of Fashion, p. 4; Jeanette
Jarnow and Beatrice Judelle, Inside the Fashion Business
(New York: John Wiley and Sons, 1965), p. 2.

 

 

4Emory S. Bogardus, Essentials of Social Psy-
chology (Los Angeles: J. R. Miller Press, 1923), p. 121.

 

10

Since the primary focus of this study was on the
processes of the diffusion and adoption of innovations,
fashion was conceptually defined as a process or change.

New styles or innovations were considered synonyms for the
term fashion. Fashion was operationally defined in this
study as: an innovation in styles or some element of a
style which has recently been introduced, which is perceived
as new, and which a small percentage of the social system
has adopted at one point in time.

The distinction between fashions and short-lived
fashions or fads, becomes evident after the fact and is
dependent on the nature of their acceptance cycle. Wasson
generalized that modal fashions when compared to fads have:
" . . . an initially slower rise to popularity, a plateau
of continuing pOpularity and a slow rather than an abrupt
decline."l

Bogardus' twenty-five-year study of fads revealed
that they were adopted quickly and declined rapidly, the
majority surviving less than six months.2 After his two-
year study of fads Janney made the following distinction
between fads and fashion: "A fad is here defined as a

precipitate but short-lived deviation in some article or

1Chester R. Wasson, "How Predictable Are Fashion
.and Other Product Life Cycles?" Journal of Marketing,
XXXII (July, 1968), 36-43.

2Emory S. Bogardus, "Social Psychology of Fads,"
glgurnal of Applied Sociology, VIII (1923), 239-43.

11

articles of clothing . . . fashions are slower to originate,
last longer and are less striking in deviation from the
current mode."1 Blumer believed that one of the funda-
mental ways in which fashion differs from fads was that
“fads have no line of historical continuity; each springs
up independently of a forerunner and gives rise to no
successor."2

Since this study was concerned with that stage in
the process of the diffusion of new_styles when only a
small percentage of individuals have adopted the style,
the distinction between fads and fashions can only be
speculated about but cannot be predicted. New styles are
‘neither fads nor modes but may become one or the other
after a period of time. In order to understand the nature
of the relationships between the diffusion and adoption of
new styles and the characteristics of individuals who first
adopt them, theories of fashion acceptance must be studied.

These theories are discussed in the following section.

Theories of Fashion Acceptance

 

The "trickle-down" theory of fashion diffusion,

developed in the late 1800's by economists and social

 

1J. E. Janney, "Fad and Fashion Leadership Among
IJndergraduate Wbmen," Journal of Abnormal and Social Psycho-
.lo , XXXVI (1941), 275-78. (Hereinafter referred to as
Fa and Fashion Leadership.")

 

2Blumer, "Collective Selection," p. 283.

12

psychologists, had been more or less accepted by home
economists and others until recent years. Simmel, a
sociologist writing in 1904, was among the first to
delineate the "trickle-down" theory and its concommitant
view of class differentiation and the acceptance of new
clothing styles. He described the process Of fashion
change as follows:

Just as soon as the lower classes begin to copy
their [the elite's] style, thereby crossing the line
of demarcation the upper classes have drawn and
destroying the uniformity of their coherence, the
upper classes run away from this style and adopt a
new one, which in its turn differentiates them from
the masses: and thus the game goes merrily on.1

Although Simmel's concern was not how new styles were
copied by the lower classes he implied that prestigious
fashion leadership was involved in the process of fashion
acceptance.

Simmel believed that fashion existed because of the
need to reconcile two opposing forces: (1) differentiation
and (2) uniformity. He discussed two types of individuals
representative Of these two polar forces:

1. The "Dude," who relying on personal conviction,
experiments with more extreme clothing styles

than others wear. He "leads the way" yet allows

himself to be led by the group.

2. The "Imitator," who exerts little individuality

but wears clothing similar to that worn by
society.2

 

1

Simmel, "Fashion,' pp. 543-45.

21bid., pp. 543, 549.

13

Simmel stated that although the social demands of fashion
appear to influence both types of individuals, these
demands appeared most exaggerated in the Dude. Simmel's
statement that "infinite" proportions of the two polar
types of individuals exist implies that several types of
prestigious fashion leaders may, in fact, exist.l Further-
more, some of these prestigious leader types may be less
conforming in their dress than others.

The problem of reconciling the prestigious leaders'
freedom in the selection of clothing styles with the pres-
sure Of society to conform was also recognized by "trickle-
down" exponents Flugel and Sapir. In 1929, Flugel cited
several historical examples of fashions which had failed
to become generally accepted, in order to make this gener-
alization about fashion: "Fashions, if they are to be
successful must be in accordance with certain ideals cur-
rent at the time they are launched." Flugel also stated
that there is a direct relationship between the prestige
Of the launcher of fashions and the degree of difference
that he can bridge in his attempted innovation.2 Sapir,

a social scientist, stated in 1931, that features of fashion
which do not conform to the unconscious social ideals of a
given culture are relatively insecure. "The fashion de-

signer must have a sure feeling for established custom and

 

1Ibid., p. 542.

2Flugel, Psychology of Clothes, pp. 47-54, 143.

14

the degree to which he can safely depart from it." Sapir
believed fashion was custom in the guise of departure from
custom.1 Although the two authors presented their views
in slightly different ways, both seem to indicate that
group norms do have some limiting or controlling effect on
the promoters' and innovators' selection of new styles.
While Simmel, Flugel, Sapir and other "trickle-down"
theorists seem to have acknowledged the influence of group

norms on prestigious fashion leaders they were primarily

 

concerned with the converse influence, that of the
prestigious leaders on the group.

1 Some shortcomings of the "trickle-down" theory
were pointed out by Whyte when he proposed his theory of
"inconspicuous consumption" in 1956. Whyte wrote about the
pressures of the corporate organization on the attitudes
and consumption of the American suburbanite. He stated
that the philosophy of "keeping up with the Joneses" and
the vaunting of worldly goods has been replaced with the
idea of "keeping up with the group":

. . . it is the group that determines whan a luxury
becomes a necessity. Just as the group punishes its

members for buying prematurely, so it punishes them
for not buying.2

 

1Sapir, "Fashion,' p. 141.

2William H. Whyte, Jr., The Organization Man
(Garden City, N.Y.: Doubleday and Company, Inc., 1957),
pp. 346-47.

15

Whyte's theory has implications for the influence of
group norms on prestigious fashion leaders.

Changes in United States' society in recent years
indicate that the theory of the vertical flow of fashions
from a prestigious elite to the lower socio-economic
classes may require further review. In his rebuttle to
the "trickle-down" theory in 1964, King presents the
following evidence against the traditional theory:

1) income redistribution has resulted in an enlarged
middle class with increased purchasing power, 2¥~mass.
communication media accelerate the.spread of fashion
_awareness, 3) new styles are introduced nearly simul-
taneously in Paris and to the American mass consumer
~in a wide variety of price lines, 4) price lines vary
only in subtle differences in quality rather than in
obvious style features.1

King's empirical data on the adoption of women's
millinery indicated that personal fashion information

exchange moved horizontally within the same social status
level. He found that the "innovators," or the first 35 per
cent of the 303 Boston women to purchase new hats, were
not consistently "upper class," but were of higher socio-
economic levels than the late buyers. Furthermore, "in—
fluentials" were dispersed throughout all socio-economic

levels and the largest numbers of these fashion Opinion

 

1Charles W. King, "Fashion Adoption: A Rebuttal
to the 'Trickle-Down' Theory," Proceedings of the Winter
Conference of the American Marketing Association (Boston,
Mass., December 27-28, 1963), pp. 108-25. THereinafter
referred to as "A Rebuttal.")

 

16

leaders were in the late buyer groups. Although King
recognized that there was some "trickling-down" via the
ranks of the designers, manufacturers, and retailers, he
proposed a "trickle-across" theory to explain the process
of fashion adoption within a social system.1 Therefore,
according to King's :trickle-across" theory of fashion,-
emulation and differentiation areelikely to occur within
rather than between social strata as the "trickle-down"
theorists believed.
A later theory termed "collective selection" which
was proposed by Blumer in 1969, also seems to reject
certain aspects of the "trickle-down" theory. Although
Blumer acknowledged the importance of an elite group in
the acceptance of fashion, he believed that fashion
transcends the elite. Blumer stated that the prestige of
the elite affects but does not control taste or "cause"
fashion as the "trickle-down" theorists believed: .
. . . it is the suitability or potential fashionable-
ness of the design which allows the prestige of the
elite to be attached to it. The design has to corre-
spond to the direction of incipient taste of the
fashion consuming public.2

Blumer explained that the endorsement of a new style by

the elite enhanced the probability of its adoption but he

added that not all prestigious persons were innovators and

 

lIbid., pp. 115, 121-22.

2Blumer, "Collective Selection," p. 280.

17

innovators were not necessarily those persons with the
highest prestige.l He implied that prestigious fashion
leaders were able to introduce only those fashions which
were consistent with developing tastes.

Since theories of class differentiation apparently
do not adequately explain the process of fashion change
and since Blumer did not specify the details of the
mechanism of fashion change on the level of the individual,
a theoretical framework is needed with which to investigate
fashion change and prestigious fashion leadership on the
local level.

An interdisciplinary theory from the fields of
rural sociology and communications, developed by Rogers in
1962, deals with the adoption of innovations by individuals
within a social system. Rogers' theory is applicable to
studies of prestigious fashion leaders and their acceptance
of textiles and apparel. Rogers has summarized over 500
research studies on the adoption of innovations in diverse
fields such as medical sociology, anthropology, agri-
culture, education, marketing, and rural sociology.2

Rogers suggested that classification of individuals
in adopter categories should be on the basis of time of

adoption in relation to other individuals within the

 

lIbid., p. 281.

2Rogers, Diffusion of Innovations, p. 5.

"WV—4 q .-.

18

social system. Research findings show that adoption

follows a normal distribution over time and an
Therefore, Rogers based the

"s" curve

when plotted cumulatively.
fo llowing adopter categories on standard deviations of
(1) innovators, the first 2 1/2 per

the normal curve:

cent to adopt, (2) early adopters, the next 13 1/2 per

cent to adopt, (3) early majority, the next 17 1/2 per
cent, (4) late majority, the next 51 per cent to adopt,
l

and (5) laggards or the last 16 per cent to adopt.

Rogers stated that both adopters and rejectors of

innovations may be classified bytwo types of Opinion

leadership: (1) active, and (2) passive. He defined

Opinion leaders as those individuals who have a greater

or unequal share of influence upon others. Thus, Rogers'
(1) active

model contained the four following categories:

adopters, who adopt and influence; (2) active rejectors,
Who reject and influence; (3) passive adopters, who adOpt

and do not influence; and (4) passive rejectors, who

[1 e ither adOpt nor influence.
Rogers suggested that differences in innovative-

ne SS among the four types of leaders may be discovered by

a. n Individuals
1 1').
a. l 137 view change favorably and are predisposed to adOpt

annalysis of the norms of the social system.

social systems with norms oriented toward change gener-

 

\

l1pm., pp. 152, 164-65. 21bid., p. 210.

19

new ideas more rapidly than individuals in traditional

systems. Rogers implies that although all individuals in

a social system generally reflect the norms of the social

system to some extent, the opinion leader may show a

greater amount of conformity to those norms than the

- l
J. nnovator .

Both Schrank2 and Sproles3 stated that many of

the shortcomings of most current research and theories
concerning fashion change are due to the failure to dis-

tinguish between the roles of different types of pres-

tigious fashion leaders. Most investigators have been

concerned with only one leader's role. For example,
Katz and Lazarsfeld4 and SummersS investigated the
f ashion Opinion leader but not the innovator. Only those

 

11bid., pp. 197, 245-47.

2Schrank, "Fashion Innovation," p. 80.

George Bryan Sproles, "A Profile Analysis of the
D urable Press Clothing Information Communicator"

(unpub-
l i sted Master's thesis, Purdue University, 1968), pp. 124,
21...:E:;1_,

(Hereinafter referred to as "Durable Press Infor-
ma t ion Communicator . ")

4Elihu Katz and Paul Lazarsfeld, Personal Influence:
EE13§255§3_ Part Played by People in the Flow of Mass Communi-
:=EE‘-1::jfbns

M TNew York: The Free Press, 1955), pp. 247-70.
He: re1nafter referred to as Personal Influence.)

cl 5John O. Summers, "The Identity of Women's

R ching Fashion Opinion Leaders," Journal of Marketing

W, VII (May, 1970), 178-85. THereinafter referred
Q as Fashion Opinion Leaders.")

20

gs'tudies by Goodell,l King,2 Robertson and Myers,3 and
s;<:‘hrank4 included data on fashion Opinion leaders,
j_Jnnovators, and their assumed followers or non-leaders.
GTIne findings of the latter studies indicated that a
czcsmbination leader type, the influential or "active
21<iopter" suggested by Rogers, may exist. Upon close
examination, four groups similar to those described by
Iiogers, seem also to have existed in Janney's study of
:fiaddists.5 Goodell stated that Rogers' concept of passive
éadopters (fashion innovators) was consistent with emulative
.anshion theories.6 Rogers' theory, therefore, appears to
]::e suitable for research dealing with prestigious fashion
:]_eaders.
The theories of fashion acceptance which were dis-

¢::=ussed in this section of the literature were: (1) the

 

lAnne Stubenrauch Goodell, "Comparison of Two

Techniques for the Identification of Fashion Leaders"
(unpublished Master's thesis, Ohio State University, 1967).
( Hereinafter referred to as "Two Techniques.")

2King, "A Rebuttal."

3Thomas S. Robertson and James H. Myers, "Person-
iE’-¥:l—.‘:i.‘ty Correlates of Opinion Leadership and Innovative
‘Lllzyring Behavior," Journal of Marketing Research, VI (May,
1 9 6 9) , 164-68. (Hereinafter referred to as "Innovative
Buying Behavior . ")

4Schrank, "Fashion Innovation."

5Janney, "Fad and Fashion Leadership," pp. 275—78.

6Goodell, "Two Techniques," p. 10.

21

classic "trickle-down" theory including Simmel's views
(of class differentiation, and Flugel's and Sapir's ideas
about influence of the group on the styles selected by the
fashion promoters and prestigious fashion leaders; (2)
Whyte's theory of "inconspicuous consumption"; (3) Blumer's
theory of "collective-selection"; (4) the "trickle-across"
theory proposed by King; and (5) Rogers' theory of the
diffusion and adoption of innovations. Regardless Of
differences in theoretical perspectives, writers seem to
agree that fashion diffusion is a process involving both
emulation and differentiation.
Rogers' theory, however, appears to be the only
«Jne which attempts to define and operationalize the
:ijmterrelationships between these two polar concepts, as he
:1reelates individuals innovativeness to group norms on inno-
Vuréativeness. Therefore, the generalizations Rogers made
(:rcancerning the characteristics of innovators and Opinion
Il_<E:aders have been utilized as the theoretical basis for
‘t2411lis study. Selected characteristics of prestigious
fashion leaders who were derived from Rogers' model will
be discussed in the following section of the review of

t h e literature .

Characteristics of Prestigious
Fashion Leaders

Researchers investigating prestigious fashion
IJEEEaders have generally studied these influentials on one

(’15 two dimensions: (1) influentials compared to

22

non-influentials (Rogers,1 King,2 Goodell,3 and Sumers4);
sand (2) influentials compared to all other respondents in
-the sample (Sproless). A limited number of research
studies could be found presenting comparisons among the
three types of prestigious fashion influentials (fashion
innovators, fashion opinion leaders, and fashion dual
leaders) and fashion non-leaders. Therefore, since the
Objective of this study was to make comparisons among the
characteristics of the prestigious fashion leaders, reviews
of research findings were grouped according to selected
characteristics for each type of fashion influential.
The pertinent empirical findings and theoretical
:interpretations concerning selected characteristics of
prestigious fashion leaders will be discussed and summarized
.jLJn the three sub-sections which follow: (1) fashion
<:>Ipinion leaders, (2) fashion innovators, and (3) fashion
<iiumnal leaders. The characteristics discussed will be

l]_.j;mited to those which are relevant to the variables to be

1Rogers, Diffusion of Innovations.

2King, "A Rebuttle."

3Goodell, "Two Techniques."

Summers, "Fashion Opinion Leaders."

5Sproles, "Durable Press Information Communicator."

23

:investigated in this study: (1) favorableness toward new

sstyles, and (2) social participation.

Fashion Opinion Leaders

Favorableness Toward New Styles.--Rogers general-

;ized that opinion leaders conform more closely to the social

asystem's norms than the average member. Since opinion
Ileaders are likely to be more innovative if community

rlorms favor innovation, they could likewise be expected
-t:o have favorable attitudes toward new clothing styles

:i.f attitudes of the group favored fashion change.

NO empirical studies were found which included a

measure of "Favorableness Toward New Styles" as defined in

this study. However ,

measures similar to this variable
;i_Jr1<:1uding "Style Preference," "Venturesomeness," or "Experi-
Itlesaxutation with Clothing" might be related to favorableness
1tl<2rpvard new styles or fashion change, and might, therefore,
1:><Ea

appropriate to report in this sub-section of the litera-
‘tllslszre review.

In this category only one study was applicable.
atlil 1967, Summers found Opinion leaders enjoyed testing
Ellidfii. experimenting with new clothing styles. However, he
' Coit'lcluded that "venturesomeness" in women's clothing is
n<3": necessary to fashion opinion leadership since 22 per

\

cent of the 609 homemakers in his study indicated that

Rogers, Diffusion of Innovations, pp. 233-36.

24

-they preferred to buy classic styles that are well

accepted. 1

Social Participation.--Rogers' generalizations
‘which are relevant to a discussion of the relationship
between Opinion leadership and social participation were
somewhat conflicting. For example, Rogers stated that
diffusion research indicated that there is little over-
lapping among the different types of opinion leaders:
". . . an individual who is an opinion leader for inno-
vations is not likely also to be influential in political
affairs." However, Rogers also stated that opinion

.1eaders are more cosmopolitan, have more social partici-

;Epation, and more face-to-face contacts with others than

their "followers." He stated that although the

:111aajority of research indicates that opinion leaders par-

‘t:;chipate to a greater extent in both formal and informal

ss:<::cial activities than non-opinion leaders, they are not

necessarily the powerholders in the community.2

In the studies of Katz and Lazarsfeld and Sproles

iflaa.shion Opinion leaders as each defined them, seemed to be

characterized primarily by informal leadership roles.

Katz and Lazarsfeld found that fashion opinion leaders or

 

2Essa—41.

 

lSummers, "Fashion Opinion Leaders," p. 182.

2Rogers, Diffusion of Innovations, pp. 233-36,

 

25

"Feveryday influentials," as they were termed, served
ignformal rather than formal groups and guided rather than
led the Opinions of neighbors and friends. Women with
snide social contacts were more apt to be fashion opinion
.1eaders than women with few contacts. Katz and Lazarsfeld
aalso revealed that the more "gregarious" the woman, as
Jneasured by her self-estimate of the number of informal,
:intimate friendships and the number of memberships in
.cslubs or organizations, the more likely she was to be a
:1Eashion leader.l Data from.SprOles' study of 1,000
;]rzomemakers, as a part of the New Product Adoption and
JEZDiffusion Research Project in 1968, indicated that one
=E;.ignificant predictor of fashion opinion leadership was
the number of social activities including night club and
outdoor activities.2
Further evidence for the existence of a positive
ZIT<EeL1ationship between fashion opinion leadership and formal
as well as informal social participation was indicated in
1V1E5l:1:shall's study of 100 fraternity men. "Fashion infor-
InaaI-‘tzion leaders," the fashion opinion leaders in her study,

‘V'Gazzcre well integrated into the group as indicated by their

 

1Katz and Lazarsfeld, Personal Influence, pp. 247-

 

'7() -

2Sproles, “Durable Press Information Communi-
ca-":or, pp. 119-20.

26

high social interaction scores. They also held more
(offices than non-fashion opinion leaders.1

Summers stated that organizational and social
aactivities can promote leadership in women's fashions by:
"(l) requiring fashion involvement, (2) creating Oppor-
'tunities for visual gathering of fashion information, and
(3) involve social settings appropriate for transmitting
fashion information."2 The fashion opinion leaders in
Summers' study had a greater amount of formal and informal

social participation than non-fashion opinion leaders.3

The above literature review indicates that fashion
copinion leaders are generally characterized by active
Iparticipation in informal and formal social activities,

zxnd possess attitudes similar to those of the group.

 

lEileen Marshall, "Leadership in Men's Fashions
Associated with Selected Social Characteristics for a
Group of Fraternity Men" (unpublished Master's thesis,
Pennsylvania State University, 1964), pp. 39, 48. (Here-
inafter referred to as "Leadership in Men's Fashions.")

2Summers, "Fashion Opinion Leaders," p. 180.

31bid, p. 183.

27

Fashion Innovators

Favorableness Toward New Styles.--Laver has

 

generalized that attitudes toward clothing are altered
over time:
the same costume will be indecent 10 years before its
time, outre (daring) one year before its time, smart
at the time it is currently being worn, dowdy one
year after its time, hideous 10 years after its time,
quaint 50 years after its time and beautiful 150
years after its time.1
The separate studies of Tousignant, Johnson, and
Iiolly support Laver's idea that most individuals have un-
:iiavorable attitudes toward new styles when they are first
.jLI1trOduced but that these attitudes gradually become more
favorable over time. The seventy-six sorority women in
{IPCJusignant's study generally had an attitude of dislike
1t1<>ward new styles. Within three months time 38 per cent
had accepted the new styles, within six months an
aaauéiditional 24 per cent had also changed their attitudes,
and within a year nearly everyone had grown to like and
-£ESL<:Lcept the clothing styles. Since Tousignant found that
453;, .lack Of new styles in the women's wardrobes was not
necessarily due to unfavorableness toward new styles but

Q:l'—=‘I:en due to economic reasons, she concluded that

$53"fir—Indents' acceptance of new styles was not significantly

1James Laver, Taste and Fashion (London: George
lElial1:rap and CO., Ltd., 1937), p. 225.

 

28

reflected in their wardrobes.l Approximately one-half of

the teenagers and one-third of their mothers in Johnson's

study stated that they preferred to wear clothing styles

after many other people were wearing them. More than

three-fourths of the 112 women in Holley's study said
-they enjoyed dressing according to the latest fashion but

few women said they wanted to be the first to wear the

latest 1970 fashions. One-half‘to three—fourths of the

women desired clothing similar to that of their peers.

The findings of the three preceding studies indi-
cating that the majority of individuals at first dislike
new styles are consistent with Rogers' generalizations
that the predominant values of the "early majority" and

" late majority" of adopters are deliberateness and skepti-

cism respectively. Rogers contrasts the values of the two

8 lowly adopting categories above with that Of the

:L nnovators, characterized by venturesomeness. Innovators,

Who are by definition, the first individuals in a social

 

1Rosanna P. Tousignant, "Fashion Acceptance by
S elected College Students" (unpublished Master's thesis,
EJ’hiversity of Rhode Island, 1959), pp. 35-36, 58. (Here-
2— hafter referred to as "Fashion Acceptance.")

2Judith Johnson, "Aanxploratory Study of Dif-
fus ion of Fashion with Mothers and Teen-Age Daughters"
( unpublished Master's thesis, University of Nebraska,
1 9 6 7) , p. 53.

3Zelda Holley, "Opinions of University Women
Regarding the Relative Importance of Thermal Comfort,
corinformity and Fashion" (unpublished Master's thesis,
Kansas State University, 1970), p. 34.

29

system to adopt new ideas, have more favorable attitudes
toward new ideas and also perceive themselves deviants
from the norms of their social systems.1
Pasnak supports Rogers' proposition that inno-
vators, unlike the rest of the group, have favorable atti-
tudes toward innovations. Pasnak compared fashion inno-
'vators with non-innovators on eight clothing attitude
Ineasures and found that fashion innovators scored signifi-
czantly higher on five measures. Innovativeness was defined
by a single question concerning the university women's
.1:131ative time of adoption and by the investigator's assess-
Itteznt of the fashionableness of the reSpondent's favorite
dressy coat and dress.2
One explanation for the apparent difference
1:>¢Ertween the favorable attitudes of the fashion innovator
and the unfavorable attitudes of the majority of indi-
“L?*;i_<3uals in the above studies may be because innovators

£111.£El\7e an outside reference group with favorable attitudes

‘t::u<:>‘dard new styles. Rogers states that one Of the
‘:==<=:bzusequences of innovators' deviation from the norms

‘::"Zf3’ ‘the social system is a shift or change in reference

1 I D I
Rogers, D1ffus1on of Innovat1ons, pp. 169-70,
:l—-!E) :3”_98.

<2: 2Mary Francis Drank Pasnak, "Fashion Innovators
JQLf:>'lnnq;>ared with Non-Innovators on Clothing Attitudes, Self-
Pctualization, and Tolerance of Ambiguity" (unpublished
13‘ -»ID. dissertation, Pennsylvania State University, 1968),
p " 96. (Hereinafter referred to as "Fashion Innovators.")

30

group.1 Grindereng's finding that the "early adopters"

in her study were relatively free from the dress standards
of their friends or relatives and utilized nationally or
internationally famous women as frames of reference,2

seems to support Rogers' statement.

Social Participation.--Rogers' synthesis of
adOption-diffusion research suggests that there is a
;positive relationship between the five adopter categories
.and.the amount of social participation. Rogers stated

that innovators are more cosmopolitan and are relatively

Lless involved in social relationships at the local com-
Itnanity level than are early adopters. Rogers, noted that
jLJnnovators are likely to belong to cliques and formal
<:>1:ganizations outside their social system.3

The findings of Warden's study imply that indi-

"L72j.duals who participate in formal social organizations
may also perhaps be innovators in the wearing of new
8 tyles. Warden found approximately one-fourth of the 135

university women in her 1957 study reported that they

lRogers, Diffusion of Innovations, pp. 203-04.

2Margaret Pauline Grindereng, "Fashion Diffusion:
.2n5_ EStudy by Price Range of Style Dispersion and Style
JL'eadership" (unpublished Ph.D. dissertation, Ohio State
‘EZVIHLijersity, 1965), p. 114. (Hereinafter referred to as

aishion Diffusion.")

3Rogers, Diffusion of Innovations, pp. 169-73, 183.

31

needed special clothes they would not otherwise need in
order to belong to certain campus organizations.1

The following studies seem to support Rogers'

belief that social participation and innovativeness may be
positively related. In their separate studies, Janney,
Glickman, Perkins, Sohn, Marshall, Robertson and Myers,
and Van Staden found that innovators, as each defined
them, were likely to be leaders in other areas of group
activities.

Janney's study indicated that fads were initiated
by college women who were also members of prestige bearing
<::1iques and leaders in other types of social activities.
‘ifikamen who were insensitive to fads were in general
;i_115ensitive and unskilled in other types of social situ-

éa-tions.2

In Glickman's study of 511 fifth through tenth

grade boys, the boys who were "clothing leaders" (inno-
vators) were Often leaders in COOperative and organized
groups. Glickman concluded that clothing leadership was
iEaL sspecific kind of leadership which, though related, was

<:3L1jL.sstinct from various aspects of group leadership.3

e

1Jessie Warden, "Some Desires or Goals for Clothing
O f College Women," Journal of Home Economics, XLIX, No. 10
(December, 1957), 795-96.

2Janney, "Fads and Fashion Leadership,’ pp. 275-78.

1E5 3Albert S. Glickman, "Clothing Leadership Among
‘<:’25V=3" (unpublished Ph.D. dissertation, Ohio State

32

Marshall revealed that those fraternity men in her
study who were self-designated innovators also were more
socially integrated and held more offices and memberships
in campus organizations than non-fashion innovators.1
Innovators among the college men in Sohn's study2 and
among the college women in Perkins' study3 tended to be
leaders in other areas. Robertson and Myers reported that
innovativeness in clothing correlated somewhat with socia-

loility or an outgoing and participative temperament.4

‘Van Staden found a significant relationship between the

 

ILJEIIiversity, 1952), p. 248. (Hereinafter referred to as
" C lothing Leadership. ")

lMarshall, "Leadership in Men's Fashions," pp.
—:I:4ES-—47.

- Marjorie Ann Sohn, "Personal-Social Character-
43L— =Eii:ics of Clothing Fashion Leaders Among Fraternity Men"
(I‘lclzapublished Master's thesis, Pennsylvania State University,

3— 959). p. 49. (Hereinafter referred to as "Clothing
3E?‘ iE:.sshion Leaders.")

3Olive Perkins, "An Investigation of the Clothing
F ads and Fashions of a Group of Freshmen College Women and
‘t:;]t1L<a: Reasons for Accepting or Rejecting Some Popular Campus
<:=::1—<:>thes“ (unpublished Master's thesis, Cornell University,

1 95 8), p. 41. (Hereinafter referred to as "Clothing Fads
and Fashions.")

4Robertson and Myers, "Innovative Buying Behavior,"
9 - 166.

33

leadership role in high school and wearing the mode for

the boys (defined by the boys as prestige clothing). The

girls, however, did not perceive the mode as prestigious
but indicated that other styles were prestigious and that
wearing them was related to leadership roles.1

Contrary to the above studies, the findings of

Goodell, and Freedle, revealed that social participation

may not be related to innovativeness. For one of the two

sororities she compared, Goodell found no relationship
Imetween innovativeness, as determined sociometrically,

«or "passive leadership") and social participation. The
(aorrelation between innovativeness and social participation
dEkor the other sorority was only slight.2 Freedle's
:iJnvestigation showed that the low social participators

éaLrnong the 151 university women studied chose "style" as
their motive in the selection of their last major clothing
JE:>11rchase while the high social participators chose "fit"

over "style . " 3

 

1Francine Johanna Van Staden, "The Relationship of
-]E;’:I:eestigious Clothing to Acceptance by the Peer Group Of
5335~<ii<31escent Boys and Girls" (unpublished Master's thesis,
2i.<:higan State University, 1970), pp. 52-54. (Hereinafter
3‘=‘<E::£erred to as "Prestigious Clothing.")

2

Goodell, "Two Techniques,‘ p. 51.

3Johnnie Alice Denton Freedle, "Clothing Interest
and Social Participation as Related to Clothing Selection
and Buying Processes of the College Woman" (unpublished
Master's thesis, University of Tennessee, 1968), pp.
3 9 ~40.

34

,The majority of the studies in the literature
review above revealed that fashion innovators were likely
to be leaders in other areas of group activities. While
some research showed fashion innovators were active in
formal leadership roles; other investigators found fashion
innovators were informal leaders. A few studies indicated
that social participation may not be related to inno-
vativeness. The innovator, unlike the majority of the
group, generally had favorable attitudes toward new styles
and was among the first to adopt them. The proposition
that the innovator's actual reference group may be outside
their local friendship structure, was the explanation

given for the venturesomeness of the fashion innovator.

Fashion Dual Leaders

 

Rogers noted that "there seem to be some opinion
leaders in almost every adopter category," indicating that
some individuals may serve a dual function, that of opinion
leader and innovator. Although Rogers conceptualized the
existence of "active adopters" he did not make general-
izations concerning the characteristics of these persons
since little research includes a differentiation of this
type Of individual.1 Several researchers have investigated

both fashion Opinion leadership and fashion innovativeness

 

1Rogers, Diffusion of Innovations, p. 243.

 

35

but the majority did not report the extent to which these
two roles occurred in the same individuals. Lately,
however, a few investigators have studied the character-
istics of active adopters as each defined them. Therefore,
in this sub-section of the literature review, studies
concerned with the relationship between the variables
fashion adoption and fashion opinion leadership will be
discussed in addition to the limited number of recent
studies of fashion dual leadership.

The findings of several studies seem to support a
positive relationship between fashion opinion leadership
and fashion innovativeness. Marshall discovered that both
the "fashion leader," an innovator measured by a socio-
metric technique, and the "fashion innovator," a self-
designated innovator, were sought out by others for
information.1 Marshall, however, did not have mutually
exclusive sub-groups. In Sproles study, opinion leadership
and innovativeness, as revealed by ownership of new man-
made fibers, were significantly correlated.2 Although the

findings of both Marshall and Sproles seem to suggest the

 

lMarshall, "Leadership in Men's Fashions,"

2Sproles, "Durable Press Information Communi-
cator," p. 119.

36

possibility of the existence of fashion dual leaders,
neither researcher was concerned with this combination
type of leader.

Further evidence of a positive relationship
between fashion adoption and fashion Opinion leadership
was indicated by Schrank's research. Among the 145 uni-
versity women Schrank studied, high innovativeness was
accompanied by high opinion leadership.l Schrank investi-
gated the extent of overlap of fashion innovativeness and
fashion Opinion leadership occurring within the same
individuals. Approximately one-third of her respondents
fell into an overlapping group which she termed "fashion
dual leaders." The fashion dual leaders of her study were
those respondents who scored high on both the fashion
Opinion leadership and fashion innovativeness measures.2
Therefore, they were a mutually exclusive group which was
distinct from those individuals who scored high on only
one of the measures for innovativeness or fashion opinion
leadership. Schrank compared fashion dual leaders with
innovators and opinion leaders on time-of-adoption and
found that dual leaders were the earliest of the three
groups to adOpt. Dual leaders, in fact, scored signifi-

cantly higher on time of adoption than innovators.3

 

1

Schrank, "Fashion Innovation,‘ p. 67.

2 3

Ibido' p. 59. Ibido I pp. 70-710

37

Although Goodell, like Schrank, investigated
fashion dual leaders, Goodell did not have mutually
exclusive groups. Goodell found that "active leadership"
(fashion dual leadership as determined sociometrically)
was related to the reSpondents' scores on a style list and
the number of styles owned for both of the sororities she

studied. "Passive leadership,‘ fashion innovativeness,
was also positively correlated with self-designated
leadership, fashion opinion leadership, for both
sororities.l

Robertson and Myers' data on the buying behavior
of ninty-five California housewives showed Opinion
leadership and innovativeness correlated to some extent
for the clothing product variable. They concluded that
although opinion leadership and innovativeness may overlap,
the amount is small and not great enough to equate inno-
vators with Opinion leaders.2 Unlike Schrank and Goodell,
Robertson and Myers did not consider the possibility of
the existence of a third group, the dual leaders. The
extent of overlap in Schrank's and Goodell's studies,
however, appears to have been greater than that in
Robertson and Myers'.

While the preceeding studies seemed to show that

fashion opinion leadership and fashion adoption may be

 

1Goodell, "Two Techniques," pp. 49, 65.

2Robertson and Myers, "Innovative Buying Behavior,"

p. 167.

38

related, King's study indicates somewhat conflicting
results. King found fashion Opinion leaders were well
distributed throughout both the early and late buyer groups
but were more frequently late buyers than early buyers.1
Although King did not report the extent of overlap among
fashion opinion leadership and fashion innovativeness in
his Boston study, he pointed out the functions of each
leader-type which together seem to parallel the character-
istics of Rogers' active adopter:

The innovator is the earliest visual communicator of

the season's styles for the masses of fashion consumers.

The 'influential' [fashion opinion leader; appears to

def1ne and endorse appropr1ate standards.

The conflicting findings among the studies may

have been a result of differences in sample size or the
research techniques used. Rogers states that contradictory
evidence in the literature on whether opinion leaders are

innovators points to the need to take the social system

. 3
norms 1nto account.

Favorableness Toward New Styles.--Although no

 

research was found in which fashion dual leaders' attitudes
toward new styles were investigated, dual leaders might be

expected to have favorable attitudes since they are, by

 

lKing, "A Rebuttle," pp. 121-24.

2Ibid., p. 124.

3Rogers, Diffusion of Innovations, pp. 243-45.

 

39

definition, among the first to adopt. Schrank,l however,
found fashion dual leaders to have positive attitudes
toward conformity to the norm of their friends' dress.
Glickman2 concluded that the dual leaders in his study
also had a certain minimum amount of clothing conformity.
The dual leaders' unexpected conformity may perhaps be due
to an association with a fashionable sub-group rather

than conformity to the mode.

Social Participation.--Rogers found that both

 

Opinion leaders and innovators were characterized by active
social participation.3 Fashion dual leaders, who are by
definition innovative and influential, may also be active
social participators. Contradictory findings, however,
are shown in the following studies on dual leaders as each
investigator defined them.

Sohn found a significant difference between
"fashion (dual) leaders" and non-leaders for the number Of
organized group memberships (followers, however, belonged
to more organized groups) but not for the number of
cooperative group memberships. Fashion leaders, however,

showed a tendency to belong to more cooperative groups

 

1

Schrank, "Fashion Innovators,‘ pp. 66, 69-70.

2Glickman, "Clothing Leadership," p. 248.

3Rogers, Diffusion of Innovations, pp. 169, 182-83.

 

40

than non-leaders.1 In Goodell's study, "active leadership"
in both sorority groups was not related to the amount of
social participation.2

In summary, there were conflicting findings among
the preceding fashion research studies as well as dis-
agreement between diffusion theory and fashion research
concerning the characteristics of fashion dual leaders.

In essence, dual leaders seemed to engage in some organi-
zational participation, exhibit a minimum amount of
conformity to their friends' dress, and at the same time
express some favorableness toward new styles.

The review of the literature indicates that a com-
parison of the three types of prestigious fashion leaders
(fashion innovators, fashion opinion leaders, and fashion
dual leaders) and fashion non-leaders may reveal differ-
ences on social participation and favorableness toward
new styles. Methods used to measure fashion acceptance
and prestigious fashion leaders will be discussed in the
following section of the literature review.

Methodologies Used to Measure Fashion

Acceptance and PresEigiOus
Fashion Leaders

 

 

 

The three methods which have been used to measure

fashion innovativeness and fashion opinion leadership in

 

1Sohn, "Clothing Fashion Leaders," p. 49.

2Goodell, "Two Techniques," p. 49.

41

adoption-diffusion research studies are: (l) the socio-
metric technique, (2) the key informants or judges rating
method, and (3) the self-designating or recall technique.
The sociometric technique, as used by Glickman,l
Sohn,2 Moore,3 Marshall,4 and Goodell,5 identifies
prestigious fashion leaders by asking respondents to name
those individuals whom they would either observe or go to
for advice when selecting new clothing styles. Possible
limitations of this technique are:
1. All members of the social system must be inter-
viewed.
2. Mutually exclusive sub-groups may be impossible
to establish within a sample.
3. Bias may occur when the number of prestigious
individuals which respondents are allowed to
name are limited.
4. A large number of respondents must be questioned

in order to locate a relatively small number of
prestigious individuals.

 

lGlickman, "Clothing Leadership."

2Sohn, "Clothing Fashion Leaders."

3Kathleen Anne Moore, "Fashion Leadership Desig-
nation and Related Factors Among a Group of Adolescent
Girls" (unpublished Master's thesis, Pennsylvania State
University, 1962). (Hereinafter referred to as "Fashion
Leadership Designation.")

4Marshall, "Leadership Among Men's Fashions."

5Goodell, "Two Techniques."

6Everett M. Rogers and David G. Cartano, "Methods
of Measuring Opinion Leadership," Public Opinion Quarterly,
XXVI (Fall, 1962), 438-39. '

42

Key informants or individuals who are likely to
know the prestigious individuals may be cost- and time-
saving but the method may be biased by non-objective
ratings by the informants.1 This method was used as a
secondary means of selecting fashion innovators by
Grindereng who ranked respondents according to silhouettes
purchased.2

In the self-designating technique, a respondent
answers questions concerning his perception of the relative
prestigiousness of, and recalls the time of adoption, of
the new clothing styles. Attitudes affecting fashion be-
haviors may be inherent in the responses of the respondents.
The ability of the respondent to recall, assess, and re-
port her self-image accurately must be assumed. The
respondent, however, may be more aware of his own pres-
tigiousness than others. While the self-designation
method may be conveniently used in either questionnaires
or in interviews, the number of statements may be limited.3
This technique, however, has been utilized in several

fashionstudies.4

 

1Rogers, Diffusion of Innovations, pp. 159-65.

 

2Grindereng, "Fashion Diffusion,’ p. 100.

3Rogers, Diffusion of Innovations, p. 160.

 

4Katz and Lazarsfeld, Personal Inflgence; Rogers
and Beal, "Information Sources"? Marshall, "Leadership in
Men's Fashions"; Grindereng, "Fashion Diffusion"; Goodell,
"Two Techniques"; Sproles, "Durable Press Information

Communicator."

 

43

Goodell and Grindereng each compared two of the
techniques above and both concluded that the self-
identification method was preferred. Goodell compared the
sociometric and self-identification techniques to determine
whether the two techniques identified the same individuals
as fashion leaders. Although a low positive correlation
was found between the two sorority groups' scores on the
two measures, the self-identification technique appeared
to identify more opinion leaders than the sociometric
technique.1 Grindereng also found that self-identified
adopter categories were more successful in differentiating
between early and late adopter categories based on
silhouette purchased.2

Since a large university was selected as the social
system in this study, it was likely that few of the
respondents would know each other or that judges would
know all of the subjects. Therefore, the key informants
and sociometric techniques were not applicable. The self-
designating method, however, was well suited to a question-
naire survey and appeared to be satisfactory for the
present study.

Rogers also stated that further exploration of the

social system's norms are needed to determine the

 

1

Goodell, "Two Techniques,‘ p. 69.

2Grindereng, "Fashion Diffusion," p. 112.

44

innovativeness of individuals. He listed three approaches
to measuring the social system's norms:
(l) compute the mean of the individuals innovativeness
scores, (2) find the percentage of respondents who had
favorable attitudes toward innovators, and (3) obtain
judges ratings of the social system's norms.l
Rogers cautions that the preceding measures are only rough
indications of a social system's innovativeness and that
because to few attempts to measure the norms of,a social
system have been made, these measures have not been vali-
dated.

Snow used a method of measuring social system
norms, similar to the first method described by Rogers, in
her study of fashion attitudes. A mean attitude score of
50 was designated as a neutral attitude toward new styles.
Snow compared the sample mean and sub-groups' means with
this neutral mean.2 The use of mean scores to measure
groups' norms, however, may be misleading when the
respondent's scores on a measure are skewed. The median

would be more accurate when the distribution is assymetri-

cal.

 

1Rogers, Diffusion of Innovations, pp. 69-70.

2Janet E. Snow, "Clothing Interest of Men in Four
Selected Occupations" (unpublished Master's thesis, Texas
WOman's University, 1969), p. 36. (Hereinafter referred
to as "Clothing Interest.")

CHAPTER III

STATEMENT OF THE PROBLEM

Rogers' theory of the diffusion and adoption of
innovations was used as the basis for studying prestigious
fashion leaders and the acceptance of fashions. The main
purpose of this study was to compare the characteristics of
three types of prestigious fashion leaders among university
women (fashion opinion leaders, fashion innovators, and
fashion dual leaders) and fashion non-leaders on the follow-
ing variables: (1) social participation, and (2) favorable-
ness toward new styles. The following definitions, hy-
potheses, and assumptions were used to guide the investi-

gation.

Definition of Terms

A Social System is a population of individuals

 

who are functionally differentiated and engage in col-

lective problem-solving with respect to a common goal.1

 

lRogers, Diffusion of Innovations, p. 14.

45

46

Style is a characteristic or distinctive mode or

method of expression.1

Fashion is an innovation in styles or some element
of a style which has recently been introduced, which is
perceived as new, and which a small percentage Of the social

system has adopted at one point in time.2

Fashion Adoption represents an individual's

 

decision to make relatively greater use of new styles3 or
be relatively more innovative in dress than other indi-
viduals. A respondent's innovativeness, as determined by
the Fashion Adoption Measure, is based on the relative
time of adoption of new styles, the number of new styles

purchased or sewn, and the frequency new styles are worn.

The Fashion Adgption Process consists of the mental

 

stages through which an individual passes from first
knowledge of a fashion to a decision to adopt or reject

and to confirmation of this decision.

The Adopter Categories are the classifications of

 

members of a social system on the basis of innovativeness.

 

1Nystrom, Economics of Fashion, p. 3.

 

2Bogardus, Essentials of Social Psychology, p.
121; Rogers, Diffusion of InnovatiOns, p. 13.

 

 

3Rogers, Diffusion of Innovations, p. 17.

 

41bid., pp. l6-l9.

47

The five adopter categories are: (1) innovators, (2)
early adopters, (3) early majority, (4) late majority,

and (5) laggardS.

Fashion Opinion Leadership is the extent to which

 

an individual is relatively more able than others to
informally influence individual's attitudes or behavior
concerning clothing in a desired way with relative

frequency.

On the basis ofeach_individualis-acores_On the

Fashion Opinion Leadership and Fashion Adoption Measures

\

Wan—n»- .(,1_,’..‘HI'C~~"J “*9 “in-p94 ‘5‘).L‘

 

 

*Hk‘vr'n...,h,‘..‘ ‘

(Figure l) respondents_in this study.werevclassified as

members offlone of the.four mutually_exclusive_groups

MM
. “rat-9'."
“W‘.”

whichflfpllow:.

W

l. A Fashion Innovator is a woman who reports she

 

purchases or sews a relatively greater number
of fashions which she acquires relatively
earlier and which she wears relatively more
frequently than others. Her score on fashion
adoption is in the upper one-third of the
scoring range but her score on fashion opinion
leadership is not.

2.¥ A Fashion Opinion Leader is a woman who

 

reports she is a source of fashion advice and

 

11bid., p. 208.

Fashion Adoption

 

2/3
I
I
I
I
I
I
I
I
I
I

48

 

173

273

Fashion Opinion Leadership

Figure 1.--Four sub-categories of respondents.

a .
Fash1on

Fashion
Fashion
Fashion

ALONJH
II II II II

Dual Leaders
Innovators
Opinion Leaders
Non-Leaders

49

. . 1
1nformat1on for other women she knows. Her

score on fashion opinion leadership is in the
upper one-third of the scoring range but her
score on fashion adoption is not.

3. A Fashion Dual Leader is a woman who reports

 

that she is both a fashion innovator and a
fashion opinion leader.2 Her scores on both
measures are within the upper one-third of the
scoring ranges.

4. A Fashion Non-Leader is a woman who reports

 

that she is neither a fashion innovator nor a
fashion opinion leader.3 Her scores on
neither measure are within the upper one-third
of the scoring ranges.

The remainder of the definitions follows:

Favorableness Toward New Styles is a relatively

 

more positive attitude toward fashions as determined by
an individual's score on the Favorableness Toward New

Styles measure.

Social Participation is the extent to which an

individual takes part in the university organizational

 

1

Schrank, "Fashion Innovation," p. 36.

21bid. 31bid., pp. 36-37.

50

system including the professional, athletic, academic,
social, church, and community areas. Leadership roles
within organizations reflect a greater amount of social
participation than mere organizational membership, as
determined by a modification of Chapin's Social Par-

ticipation Scale.1

A Garment Length Categogy is one of the seven

 

mutually exclusive wardrobe inventory divisions based on
specified locations on the body. Taken together the seven
lengths can be arrayed to form a continuum on the body
from the waist to the floor. The garment length cate-
gories are further sub-divided into styles such as mini-

skirt or knickers.

Socio-Economic Status is a respondent's position

 

along a weighted continuum based on the education, income,
and occupation of the main wage-earner as determined by a

modification of the McGuire-White Index.2

Hypotheses

 

The relationship between fashion Opinion leadership

and fashion adoption was to be tested in the first

 

1Francis S. Chapin, Experimental Designs in Socio-
logical Research (rev. ed.; New York: Harper, 1947),
pp. 276-78.

 

2Carson McGuire and George D. White, "The Measure-
ment of Social Status" (unpublished research paper in
Human Development, NO. 3 [revised] Department of Educational
Psychology, The University of Texas, 1963).

51

hypotheses formulated. Although three of the research
studies1 reported in the literature indicated that a
positive relationship may exist between these two vari—
ables, two other studies2 presented somewhat conflicting
data. Rogers' belief that some opinion leaders are inno-
vators while others are not,3 may help to explain the
inconsistency among these research findings. Recent
studies on fashion4 support Rogers' belief with evidence
of the existence of fashion dual leaders, who both adopt
and influence, as well as evidence for the existence of
fashion opinion leaders who only influence and fashion
innovators who only adOpt. Therefore, a positive relation-
ship between the two variables may be predicted:
Hypothesis 1. A positive relationship will exist

between fashion opinion leadership and fashion
adoption.

 

Researchers studying fashion opinion leaders have
generally assumed that these individuals have favorable
attitudes toward new styles but have not tested their

assumption. Since fashion opinion leaders, as defined in

 

lSchrank, "Fashion Innovation," p. 67; Marshall,
"Leadership in Men's Fashions," p. 48;Sproles, "Durable
Press Information Communicator," p. 119.

2King, "A Rebuttle," p. 121-24; Glickman, "Clothing
Leadership," pp. 68-71.

3Rogers, Diffusion of Innovations, p. 243.

 

4Schrank, "Fashion Innovation"; Goodell, "Two
Techniques."

52

the present study, influence others but do not wear
relatively more new styles than others, it is possible
that fashion opinion leaders may not have favorable
attitudes toward new styles. Some may actively influence
others to reject new styles.

If the "early adopter" category actually does
contain the greatest amount of opinion leadership as the
diffusion research suggests,1 then opinion leaders are
more innovative than the majority. They, therefore, would
be likely to have more favorable attitudes toward inno-
vations than the majority. Based on the single research
study2 which investigated fasion opinion leadership and
women's attitudes toward new styles, a positive relation-
ship may be predicted between the two variables:

Hypothesis 2. A positive relationship will exist

between fashion opinion leadership and favorable-
ness toward new styles.

 

Nearly all of the research studies reviewed3
supported Rogers' generalization that fashion opinion
leaders are active formal and informal social participators.
Logically fashion Opinion leaders must have frequent

contacts or associations with others in order to carry

 

1Rogers, Diffusion of Innovations, p. 169.

2

Summers, "Fashion Opinion Leaders,‘ p. 182.

3Katz and Lazarsfeld, Personal Influence, pp. 247-
70; Sproles, "Durable Press InfOrmation CommuniEator,"
pp. 119-20; Marshall, "Leadership in Men's Fashions,"
pp. 39, 48; Summers, "Fashion Opinion Leaders," p. 180.

53

out their influential role. Therefore, the following
relationship was predicted.
Hypothesis 3. A positive relationship will exist

between fashion opinion leadership and social
participation.

 

There is also strong evidence in several fashion
research studies1 that fashion innovators are likely to be
leaders in other areas of group activities. Furthermore,
diffusion research seems to show a positive relationship
between innovativeness and the amount of cosmopolitan
social participation in cliques and formal organizations.2
Although social participation is not actually necessary
for innovators to carry out their role as innovators they
may participate for two reasons: (1) to allow others to
view their new styles, or (2) because their new styles may
create a mood favorable to participation. Although no
cause and effect relationship can be determined a positive
relationship between fashion innovativeness and social

participation may be predicted:

 

1Janney, "Fad and Fashion Leadership," pp. 275-78;
Glickman, "Clothing Leadership," p. 248; Marshall,
"Leadership in Men's Fashions," pp. 46-47; VanStaden,
"Prestigious Clothing," pp. 52-54; Perkins, "Clothing Fads
and Fashions," p. 41; Sohn, "Clothing Fashion Leaders, p. 49;
Robertson and Myers, "Innovative Buying Behavior," p. 166.

2Rogers, Diffusion of Innovations, p. 183.

54

Hypothesis 4. There will be a positive relation-
shipibetween fashion adOption and social par-
ticipation.

 

Diffusion research seemed to indicate that inno-
iveness and favorableness toward new ideas are positively
related regardless of the social system's norms on inno-
vativeness.1 The majority of the studies of fashion2
also revealed that fashion innovators had positive
attitudes toward new styles while the other respondents
in the studies generally disliked new clothing styles when
they first appeared. Thus, the following hypothesis was
predicted:

Hypothesis 5. There will be a positive relation-

ship between fashion adoption and favorableness
toward new styles.

 

None of the research found compared innovators,
Opinion leaders, dual leaders, and non-leaders on social
participation. The diffusion research, however, suggested
that both innovators and opinion leaders are more active

social participants than their "followers."3

Thus, dual
leaders, who adopt as well as influence, may also differ

from their "followers" on social participation. One

 

11bid., pp. 111, 193.

2Holley, "Opinions of University Women Regarding
the Relative Importance of Thermal Comfort, Conformity
and Fashion." p. 34; Tousignant, "Fashion Acceptance," .
pp. 35-36, 58; Johnson, "An Exploratory Study of Diffu31on
of Fashion with Mothers and Teen-Age Daughters," p. 53;
Pasnak, "Fashion Innovators," P- 96-

3Rogers, Diffusion of Innovations, pp. 169, 240.

55

fashion study1 did show that fashion dual leaders were
more likely than their "followers" to be members of
organized groups. Therefore, the following hypothesis
was proposed:
Hypothesis 6. There will be a significant differ-
ence between any of the four sub-categories
(fashion Opinion leaders, fashion innovators,

fashion dual leaders, and fashion non-leaders) on
social participation.

 

The attitudes of dual leaders, opinion leaders,
innovators, and non-leaders toward new styles have not
been compared in Rogers' synthesis of diffusion research
or in studies of fashion. Since both fashion innovators
and fashion dual leaders are, by definition, among the
first to adopt new styles, they would be expected to have
more favorable attitudes toward new styles than fashion
Opinion leaders and fashion non-leaders. Rogers general-
ized that opinion leaders are more innovative and are
likely to adopt new ideas before their "followers."2
This generalization implies that Opinion leaders would
also have more favorable attitudes toward new innovations
than their "followers." Thus, the following relationship
was predicted:

Hypothesis 7. There will be a significant differ-

ence between dual fashion leaders, fashion Opinion

leaders, fashion innovators, and fashion non-
leaders on favorableness toward new styles.

 

 

1

Sohn, "Clothing Fashion Leaders,‘ p. 50.

2Rogers, Diffusion of Innovations, p. 242.

 

56

In order to test the proposition developed in the
literature review that certain types of prestigious fashion
leaders were more emulative of the dress of the group than
other types of prestigious leaders, the two hypothesis
which follow were established. The first hypothesis is
based on the theory and research findings1 which suggest
that prestigious fashion promoters or leaders cannot be
too extreme in their deviation from the group norms on
dress or they will not be followed. Rogers also stated
that Opinion leaders in the diffusion research he reviewed
could not deviate far from the norms of their social
system and still maintain their influential position.2

No research has been found which compared the
attitudes Of fashion opinion leaders toward new styles
with the attitudes of the rest of the subjects. However,
since Rogers' generalized that Opinion leaders conform more
closely to social system norms than the average member,3
fashion opinion leaders' attitudes toward new styles would
likely be similar to the social system's norms or attitudes

toward new styles--favorable if the social system favored

new styles or unfavorable if the social system did not

 

1Jack and Shiffer, "Limits of Fashion," pp. 730-38;
Simmel, "Fashion," pp. 541-58; Flugel, Psychology of
Clothes, pp. 143, 147, 154; Young, Recurring Cycles, p. 4.

 

 

2Rogers, Diffusion of Innovations, p. 243.

 

31bid., p. 233.

57

favor new styles. Therefore, the following hypothesis was
predicted:

Hypothesis 8. There will be no significant
difference between fashion opinion leaders and
the rest of the subjects on favorableness toward
new styles.

 

Rogers generalized that "innovators are perceived
as deviants by other members of their social system" and
that they also "perceive themselves as deviant from the
norms of their social system."1 Rogers also stated that
“laggards are deviants as well as innovators." Therefore,
it is likely that individuals in the present study who
are more innovative than others, (fashion innovators and
fashion dual leaders) and more laggard (fashion non-
leaders) than others would be significantly different from
the rest of the subjects in their attitudes toward new
styles. Therefore, the following hypothesis was proposed:

Hypothesis 9. There will be a significant differ-

ence between each of the following groups and the

rest of the subjects on favorableness toward new

styles: fashion innovators, fashion dual leaders
and fashion non-leaders.

 

Assumptions

 

The following list of assumptions was necessary

to the formation of the hypotheses:

 

1Rogers, Diffusion of Innovations, pp. 200-01.

 

58

A change in styles was imminent.

Variation existed among the clothing styles
owned and worn among the university student
population.

Wearing, owning and time-of-adoption were
equal dimensions of innovativeness.

The items in the fashion-cycle ranking index
and in the style list represented items of
clothing which varied in their amount of

fashionableness.

CHAPTER IV

PROCEDURE

The discussion of the procedure for this study
includes the following sections: (1) selection of the
sample, (2) selection and development of the measures,

(3) pre-testing of the initial measures, (4) administration
of the final measures, and (5) statistical analysis of

the data.

Selection of the Sample

 

Rogers' theory specifies that a social system is
one of the crucial elements in the diffusion of new ideas,1
and furthermore he states that students at a university
exemplify a social system. ,The population selectedflfor
this study of fashion included women students, married

-.-rv~*—‘- -

apd single, whose addresses were either East Lansing,

mpgpgipg, or the Michigan State University campus. Stu-
dgpggywithmeutzof-town addresses were not included since
"they‘would likely commute to the Campus and thus be out-

sidegpheysocial system.
Fri

 

1Rogers, Diffusion of Innovations, p. 12.

59

60

Women were chosen because of the auspicious time
in relation to impending changes in styles of garments as
promoted by mass media. Also women of university age were
more likely than older women to perceive the new styles of
the 19705 as innovations since they had not, thusfar in
their lives, been exposed to the mid-calf length garments
nor short-shorts which had been a popular mode in the mid-
19403. They offered a heterogeneous population with a
potential for varying fashion attitudes and leadership
abilities. In addition, comparisons could be made with
other university women who have been respondents in
similar fashion acceptance research studies.

In order to have the maximum number of prestigious
fashion leaders required for statistical analysis a random
sample of 500 women students was selected from the fall,
1970 student enrollment records. At the time of the data
collection, the Registrar's Office stated that the
directory was the most accurate and recent record of the
desired pOpulation which could be obtained. The name of
every eligible woman in the directory was numbered
serially and the sample was selected using a 5-digit table
of random numbers. As a result of the selectivity of the
subjects, the sample is representative of only those
12,765 women who live in the campus area of Michigan

State University.

61

The sample size of five hundred was selected in
order to obtain a minimum cell size for sub-group analysis.
Innovators, the first 2 1/2 per cent to adopt innovations,
were the smallest sub-group. The sample size was enlarged
in case some questionnaires were not returned or fully
completed by the respondents.

Additional random numbers were selected and
questionnaires mailed to a second group of randomly
selected respondents to substitute for the questionnaires
that were returned marked "address unknown." The original
respondents who had moved were thereby considered
unqualified for the sample as previously defined by place
of residence.

Selection and Develppment of
the Measures

 

 

Data were collected by a means of a seven-page
questionnaire which was devised so that it could be mailed
to the respondents and self-administered (see Appendix A,
pp. 143-50). The measures which were selected for use in
the study were Schrank's Fashion Opinion Leadership
Inventory,2 a modification of Chapin's Social Participation

Scale,3 and the McGuire-White Index of Socio-Economic

 

11bid., pp. 161—62.

2Schrank, "Fashion Innovation,‘ pp. 93-94.

3Chapin, Sociological Research.

62

Status.1 An instrument to measure Favorableness Toward

New Styles was developed since a suitable measure could
not be found. A modification of Schrank's Fashion Inno-
vativeness Measure2 was used to measure Fashion Adoption.
Background information concerning year in university and
marital status was obtained from the university enrollment

records. A discussion of each of the measures follows.

Fashion Adoption

 

In many previous research studies concerning the
acceptance of fashions, innovativeness was based on the
adoption of a single item of clothing rather than a number
of items of clothing. Grindereng concluded that one item
of apparel "may not be an accurate indicator of an
individual's classification as an early or late adopter."3

In their attempts to find more valid indicators
of innovativeness, several investigators4 developed
composite style lists and fashion innovations inventories

which were based on a number of selected apparel items

determined by the investigators to be fashionable at the

 

1McGuire and White, "Social Status."

2Schrank, "Fashion Innovation," pp. 91-92.

3Grindereng, "Fashion Diffusion," p. 111.

4Tousignant, "Fashion Acceptance"; Marshall,
"Leadership in Men's Fashions"; Goodell, "Two Techniques";
Snow, "Clothing Interest"; Schrank, "Fashion Innovation."

63

time of the studies. The items in these style lists and
inventories, however, may not have been representative of
fashions as perceived by the respondents themselves.

Rogers states that an "innovation is an idea, practice or

object perceived as new by individuals. It matters little

 

whether or not it is 'objectively' new as measured by the
lapse of time since its first discovery."l

Consequently, for this study both a wardrobe in-
ventory and a style list were utilized in which garment
length categories, developed by the investigator, were
weighted and scored primarily on the basis of the
respondents' perception of new styles. This was
accomplished by tabulating the respondents' perception
of new styles prior to the scoring of the actual fashion
adoption measure (Appendix A, p. 144).

The total fashion adoption section of the question-
naire consisted of five parts. They were: (1) a fashion-
cycle ranking index, (2) the self-designating innovative-
ness statements, (3) the wardrobe inventory, (4) the wear
record, and (5) a style list which consisted of five
selected styles which were assessed according to their
(a) time-of-adoption, and (b) subsequent discontinuance.

Parts 1, 2 and 5b were not used in scoring the actual

 

lRogers, Diffusion of Innovations, p. 13.

 

64

fashion adoption measure which consisted of Parts 3, 4,
and 5a.

Responses on Part 1, the fashion-cycle ranking
index, were utilized only in determining the weights to
be assigned in scoring the fashion adoption measure
(see Chapter V, p. 70). The index was a means of assessing
which styles from among the ten on the fashion-cycle
ranking style list were perceived to be the new styles
by the respondents. The index was develOped for use in
a research study in an advanced clothing class by Pamela
Johnson, a Michigan State University graduate student.1
The index was selected because it provided a clear
distinction between the clothing mode and new styles.

In Part 2, the self-designating innovativeness
statements, served as a reliability check for the fashion
adoption measure (see Appendix A, p. 146). Grindereng,2
Goodell,3 and Schrank,4 found that the self-designating
technique was effective in identifying fashion leaders.

However, Rogers stated that innovativeness is defined in

 

1Pamela A. Johnson, "Fashion Awareness and Its
Relationship to Selected Clothing Consumption Variables"
(unpublished research paper in Clothing Consumption,
Department of Human Environment and Design, Michigan State
University, 1971).

2Grindereng, "Fashion Diffusion," p. 112.

3

Goodell, "Two Techniques,‘ p. 69.

4Schrank, "Fashion Innovation," p. 79.

65

terms of actual time of adoption rather than the indi-
vidual's perception of the time of adoption in relation
to others in his system.

Therefore, the fashion adoption measure in this
study was based primarily on actual time-of-adoption
rather than self-designating innovativeness statements.
The statements, however, were weighted and scored in order
to form a second fashion adoption measure which could be
correlated with the original measure. The first statement
assessed when the respondents believed they adopted new
styles in relation to other students and the second, how
fashionably dressed they perceived themselves to be com-

pared to other students. The statements were assigned

weights of 5-1 and 3-1 points respectively in order to
give higher scores to those individuals who indicated
they were more innovative than others.

Part 5b was used only in the explanation of the
results of the study (see Appendix A, p. 146). A detailed
discussion of each of the parts (3, 4, and 5a) of the

actual fashion adoption measure follows:

Part 3. The Wardrobe Inventopy,--The wardrobe

 

inventory chart in Part 3 of the fashion adoption measure
was developed out of a skirt length chart which was a part
of a "Manual on the Procedure for Measuring Clothing,"

which had been developed by the investigator for use in a

 

1Rogers, Diffusion of Innovations, p. 160.

 

66

regional research study on fashion acceptance (see
Appendix A, p. 145). Subsequent recommendations from other
participating universities during the planning for the
research study resulted in the expansion Of the garment
categories of the chart to include bifurcated garments
in various lengths as well as skirt lengths.

The garment length categories which were estab-
lished according to specific locations on the body were

as follows:

 

 

Garment Location

Length Category on the Body
1 about 8" above the knee
2 about 5" above the knee
3 about 2" above the knee
4 near the knee-cap
5 about 2" below the knee
6 near mid-calf
7 near ankle and below

Each garment length category was further sub-divided by
style such as skirt or bifurcate. The complete wardrobe
inventory included sixteen styles and seven length
categories. The wardrobe inventory included several styles

which were believed not to be fashionable to help disguise

67

the intent of the measure and so that the more extensive
measure could be used in future studies.

Three months prior to the date of data collection
the investigator and two graduate assistants tested the
feasibility of the seven garment length categories by
taking a systematic random fashion count of the garment
lengths worn by the Michigan State University women at
the time of winter quarter pre-registration. The count
was also made to determine which garment lengths had low
market saturation and, therefore, would be more indicative
of innovativeness. At that time, only one mid-calf length
skirt was observed.

"Fashion Quiz Cards" with a few open-ended
questions which sought the respondents' opinions on the
latest styles were also distributed at the time of the
fashion count. The styles which were mentioned on the
quiz cards by the respondents were considered later in
the selection of items for the style list and the fashion-
cycle ranking index as well as for the wardrobe inventory
style examples.

Schrank recommended that fashion magazines be
used to determine the earliest date at which an adopter
could have acquired the new style items.1 Therefore,
fashion diffusion curves, plotted from semi-annual fashion
counts of the styles in the editorial sections of Vpgp§_

magazine from 1966-1970, were utilized in establishing a

68

cut-Off date for the fashiOn adoption measure (see
Appendix B, p. 151). Vpgpg_magazine was arbitrarily
chosen since it seemed to be more representative of new
styles than other fashion magazines.

The respondents in this study were requested to
include only those garments which they had bought new or
sewn since January 1, 1968, since this was approximately
the time when mid-calf lengths were first pictured in
Vpgpg. The cut-off date also prevented the inclusion Of
old or second-hand clothing which the respondents may have
acquired but which would not have been a true innovation.
Although the cut-off date may have been set beyond the
recall ability of some respondents it allowed those
individuals who were among the first to adopt the new
styles to receive higher scores on fashion adoption.

In order to assign the prOper weights for scoring
the wardrobe inventory, new styles were determined as

follows:

1. To Obtain respondents' reactions as to what
was new, those styles on the fashion-cycle
ranking index which were perceived by the
respondents as the "hottest" or newest items
as indicated by a relatively higher percentage
of the total number of styles in that par-

ticular rank, were located.

 

lSchrank, "Fashion Innovation," p. 81.

69

2. To prevent including modal styles, those styles
in the wardrobe inventory which had lower
percentages of the total number of styles in
relation to the other styles in the inventory,
were located.

3. To eliminate declining styles, those styles in
the editorial sections of Eggpg from February 1,
1970 to February 1, 1971, which had lower
percentages of the total number of styles in
relation to the other styles in Vpgpg_at that

time, (see Appendix B, p. 152) were located.1

Styles which met at least two of the above cri-

teria were considered to be new fashions and were given
higher weights. Styles which met only one or none of the
above criteria received either correspondingly lower

weights or a zero since they were considered representative

of either the mode which was the more frequently owned
styles or those styles which were completely out-of-
fashion. The following weights were assigned to the

styles in the wardrobe inventory:

 

l"Hot" pants, or very short shorts, were not
included in the fashion counts of V0 ue magazine since
they were introduced immediately pr1or to the time of
data collection.

Weights Styles

0 points skirt--8" above the knee

0 points - skirt--5" above the knee

0 points pants--5" above the knee

0 points skirt--2" above the knee

0 points pants--2" above the knee

0 points pants--just below the knee
0 points pants--ankle-length pants
1 point skirt--knee-length

2 points skirt--just below the knee
3 points skirt--ankle-length

4 points pantsuit

5 points jumpsuit/overalls

5 points skirt--mid-calf length

5 points pants--mid-calf length

5 points knickers

6 points pants--8" above the knee

The newer styles were assigned the heavier weights
since Goodell found that the number of newer styles owned

by the respondents was more discriminating than the total
number of styles in distinguishing fashion leaders from
followers.1 However, Tousignant found that a lack of new
styles in student's wardrobes was not necessarily an indi-
cation of fashion acceptance since many of the accepting
students did not own the new styles for economic reasons.2
Hence, to prevent those individuals who had the economic
resources to acquire large numbers of modal or outdated
styles from receiving disproportionately higher scores on
fashion adoption than the innovators themselves, the modal

and outdated styles were assigned zero weights. Each

respondent's total score on the wardrobe inventory was

 

lGoodell, "Two Techniques," p. 68.

2

Tousignant, "Fashion Acceptance," pp. 58-59.

71

the sum of the weights assigned to each style that the

individual indicated she owned.

Part 4. The Wear Record.--Rogers' definition of

 

adoption contains the phrase "make full use of," which
implies that the mere purchase of a single style would
not qualify an individual as a fashion adOpter.1 For
this reason the subjects were asked to indicate the
frequency which they wore each style in the wardrobe
inventory (see Appendix A, p. 145). The responses of the
women about the extent that the garments were worn were

weighted as follows:

 

 

Weight Frequencx. Definition

5 points Frequently Approx. twice weekly
4 points Regularly Weekly

3 points Occa51onally Monthly

2 points Seldom Twice yearly

1 point Never Not at all

The weights were assigned in order to give higher scores
on the fashion adOption measure to those individuals who

were their new garments more frequently.

Part 5a. Style List and Time-of-Adoption.--To

 

measure actual time-of-adoption, a style list was developed
which included five new styles. The new styles were

chosen on the basis of the investigator's observation and
judgment of the styles worn on the Michigan State Uni-

versity campus and the styles available in the local

 

lRogers, Diffusion of Innovations, p. 17.

72

retail stores and in current issues of fashion magazines.
The styles which were judged to be newly introduced and
which only a small percentage of the women on the campus
had adopted were: (1) mid-calf length skirts, (2) mid-
calf length pants (gaucho pants), (3) ankle-length skirts,
(4) pantsuits, and (5) jumpsuits or overalls (see
Appendix A, p. 146). The respondents were requested to
indicate the month and year when they first acquired each
new style.

Weights were assigned to the style list as
follows:

1. The total number of questionnaires were rank
ordered according to the earliest time of
adoption five times: once for each of the five
styles in the style list.

2. Each time the ranked scores for each style were
divided into five sub-groups based on the per-
centages used by Rogers and others to establish
adopter categories. A comparison of some of the

percentages which have been used to establish the

five adopter categories in the adoption-diffusion

research studies follow:

73

 

Beal &
Adopter Categories Bohlenl Rogers2 Schrank3
1. innovators 2% 28 2
2. early adopters 17 13% 15
3. early majority 17 17 18
4. late majority 51 51 49
5. laggards 16 16 15

In this study a larger number of subjects were

needed in each of the categories.

The following

adopter categories and percentages, therefore,

were utilized:

 

 

Adopter Categories Percentages Weights
l. innovators 5 5 points
2. early adopters ll 4 points
3. early majority 17 3 points
4. late majority 51 2 points
5. laggards l6 1 point

3. weights from five to one were assigned

to each

style in the style list in order to weight early
adoption more heavily than late adoption. When
the date of adoption indicated by the respondent
was prior to January 1, 1968, the respondent
received a zero for that particular style to

prevent inclusion of historic garments.

 

1George M. Beal and Joe M. Bohlen. The Diffusion

 

Process (Special Report No. 18, Ames Iowa: Iowa State
Un1versity of Science and Technology, Cooperative Exten-

sion Service, 1962).

2Rogers, Diffusion of Innovations, p. 162.

 

3Schrank, "Fashion Innovation," p. 60.

74

Each respondent's score on time-of-adoption section of
the fashion adoption measure was the sum of the weights
assigned to each style that the individual indicated that
she owned.

The respondent's total score on the fashion
adoption measure was determined as follows:

1. The number of garments bought new or sewn since
January 1, 1968, in each garment length category
multiplied by the weights assigned to each style.

2. The weighted score for each style owned multiplied
by the weighted frequency worn.

3. The weighted scores on owning and on wearing for
each style summed.

4. Each of the five items in the style list was
weighted according to time-of-adoption and the
weights summed.

5. The sum of the weights on time-Of-adoption added
to the sum of the weights on owning and wearing.

A higher score on the fashion adoption measure indicated

a greater amount of fashion innovativeness.

Fashion Opinion Leadership

Schrank's Fashion Opinion Leadership Measure which
was selected for use in this study (Appendix A, pp, 147-43)
met the following criteria: (1) it was develOped from
affective and behavioral operational descriptions of

fashion Opinion leadership, (2) it had the prOper balance

75

of positive and negative statements, (3) all statements
met the critical ratio of 2.50 or above, (4) all items
met the minimum .30 SVDR value, (5) it had an internal
consistency of .90, and (5) Schrank had obtained a Pear-
son r correlation of .74 between the scores on the Goodell
fashion opinion leadership measure and her own fashion
Opinion leadership measure.1

Schrank's measure contained twenty Likert state-
ments which had five possible responses for each, forming
a continuum from definitely true to definitely false.
Responses to positively stated items were weighted from
5 to l and reSponses to negatively stated items from 1
for definitely true to 5 for definitely false. Undecided
or uncertain responses were weighted 3, as were any state-
ments which were left unanswered.

The respondent's score on the measure was the sum
of the weights for each statement. A high score on the
measure indicated a greater amount of fashion opinion

leadership.

Favorableness Toward
New Styles

The majority of the measures which have been
developed to study individuals' attitudes toward clothing
have dealt with conformity to the clothing mode and the

importance of clothing in general rather than acceptance

 

1

Schrank, "Fashion Innovation,‘ pp. 45-48.

76

of specific new styles.1 Researchers studying acceptance
Of new styles have Often used, out of necessity, one of
these more general measures of clothing conformity.

Measures of conformity to the clothing mode can
only indicate whether individuals deviate from the mode
but they cannot actually show the direction of these
individuals' deviation. Rogers' summary of diffusion
research indicates that not all individuals adopt inno-
vations at the same time and that an individual's deviation
may be in one of two directions. For example, the inno-
vator adopts before the rest of the group, the opinion
leader adopts about the same time as the rest of the group,
and the laggard "overconforms" or adopts after the rest of
the group have gone on to adOpt another innovation.2
Logically these different types of individuals also may
have different attitudes toward new clothing styles--some
individuals may have more favorable attitudes, while others
may have less favorable attitudes than the rest of the
group.

Therefore, a measure entitled "Favorableness
Toward New Styles" was developed (Appendix A, pp, 143—49)

in order to investigate the differences between the

 

1Anna M. Creekmore, Methods of Measuring Clothing
Variables (East Lansing: Michigan State University Agri-
cultural Experiment Station, June, 1971).

 

2Rogers, Diffusion of Innovations, PP- 195‘97-

77

attitudes of the following four sub-categories in this
study: (1) fashion innovators, (2) fashion opinion
leaders, (3) fashion dual leaders, and (4) fashion non-
leaders. The instrument was conceptually defined as a
measure of fashion change and could be termed a "style
preference" measure since it was concerned with style
change rather than fabric, color, or other aspects of
fashion change.

A modification of Edwards' and Kilpatrick's scale
discrimination technique was utilized to select a set of
statements which were likely to form a unidimensional
measure.l Using Edwards' summary of suggestions for the
construction of attitude statements, forty statements
believed to measure attitudes toward new styles were
developed or selected from other fashion research
studies.2 The statements measured different intensities
of attitudes toward new styles forming a Likert scale with
five possible responses for each statement ranging from

definitely true to definitely false.

 

1Allen L. Edwards, Techniques of Attitude Scale
Construction (New York: Appleton-Century-Crofts, Inc.,
1957), pp. 13-14, 210-17.

 

 

zpasnak, “Fashion Innovators"; R. Jane Rudy
Roth, "Clothing Conformity and Fraternity Men's Attitudes
Toward Current Male Fashion Trends" (unpublished Master's
thesis, Ohio State University, 1969); (hereinafter referred
to as "Fraternity Men's Attitudes"); Elizabeth Susan
Sharpe, "Development of a Clothing Interest-and-Importance
Scale" (unpublished Master's thesis, Ohio State Uni-
versity, 1963).

78

A panel of ten graduate students and instructors
in textiles and clothing assisted in determining the
intensity and direction of scoring for each statement in
order to insure a balance of positive and negative
statements in the measure. The response categories for
negative statements received weights of 1-5 and positive
statements, 5-1. The panel also analyzed the statements
for face validity and made suggestions for the revision
and elimination of some of the statements. After two pre-
tests additional corrections and further elimination of
statements were made.

The final measure of favorableness toward new
styles consisted of twenty statements. Each respondent's
score on the measure was the sum of the weights for each
statement. A high score indicated more positive or

favorable attitudes toward new styles.

Social Participation

There are at least two aspects of social par-
ticipation: (1) formal, consisting of leadership or
Officership in clubs and organizations, and (2) informal,
consisting of various dimensions including friendships,
dating, and neighborliness. Measurement of both informal
and formal social participation was judged to be beyond
the scope of the present study. Therefore, the western
Regional Research Project W-98 modification of Chapin's

Social Participation Scale was used (see Appendix A, p. 144)

79

Chapin's scale is a general scale of participation in
voluntary organizational participation of all kinds--
professional, civic, and social. "It is used when the
total participation pattern is an important variable."1
Chapin's original scale had a reliability and
validity ranging from .89 to .95 and .52 to .76 respec-
tively.2 The W-98 revisions in the scale were minor.

The five components of the scale and their weights

were as follows:

 

Weight Social Participation Components

1 point membership

2 points committee member

3 points chairman of a committee

4 points elected officer other than
president

5 points president

An individual's score on social participation was
the sum of the weights on each of the five components.
A high score indicated a greater amount of social par—

ticipation.

Socio-Economic Status

 

A modification of the McGuire-White Index of
Socio-Economic Status was included in order to obtain

descriptive background information regarding the income,

 

1Delbert C. Miller, Handbook of Research Design and
Social Measurement (New York: David McKay Company, Inc.,
1964), p. 208.

2Chapin, Sociological Research, pp. 276-78.

80

occupation, and education of the main wage-earner in the
respondent's family1 (see Appendix A, pp, 149-50);
Each respondent's score on the three factors
was multiplied by a weight determined by the McGuire-
White Index. The sum of the three weighted scores was
used to find the respondent's social class by consulting
2

a standard conversion table. A low score indicated a

high social class position.

Pre-Testing the Initial Measures

Prior to the date of data collection, two separate
pre-tests of the original questionnaire were conducted.
The two pre-tests involved purposive samples of university
classes representing a number of different majors: (l)
fifty-nine Iowa State University women in a core home
economics course, and (2) sixty-eight Central Michigan
University women in an introductory geology course. The

questionnaire took an average time of fifteen minutes to

complete.

Data from each of the pre-tests were analyzed
separately as follows:
1. Wording of the instructions and statements was

examined for clarity.

 

1McGuire and White, "Social Status."

21bid., p. 4.

81

2. Scatter plots were made to determine the
relationships among the variables.

The favorableness toward new styles measure was

refined after both pre-tests in the following manner:

1. The discriminatory power of each statement in
the measure was determined.

2. Item analysis involved simple correlations
between statement to total scores to determine
the internal consistency of the measure.
Statements with low or negative coefficients
were either revised or eliminated.

3. A split-half reliability coefficient Of .76
was computed between the general statements
and the specific statements which included
references to a particular new style. The
fairly high coefficient indicated that the

measure appeared to be somewhat unidimensional.

Administration of the Final Measures

 

Five hundred questionnaires were mailed to those
Michigan university women who had been randomly selected
as the sample during winter quarter, 1970-71. A cover
letter which explained the nature of the investigation

(Appendix A, p. 143) and a self-addressed and stamped

envelope accompanied each questionnaire.

82

Although the respondents were asked not to put
their names on the questionnaires so they would feel free
in answering the statements, each questionnaire was
identified with a three-digit number. The numbers
corresponded to the respondents' names on a master list
of names of these women which had been randomly selected

from the student directory.

Statistical Analysis of the Data

The PLOTXY computer program was used to determine
the linearity of the data and to help select the appropri-
ate statistical treatment. The resulting bivariate
distributions were not linear. A comparison of the means,
medians, and modes for each measure also revealed that the
distributions of scores were skewed. Therefore, a non-
parametric or distribution-free statistic, Chi-square,
was selected, with the larger sample size compensating
for any loss of power efficiency over a more stringent
test such as the Pearson r test.1 The use of Chi-square
was also appropriate since one of the measures, fashion
adoption, could not be considered to be an interval level
measure.

Since the Chi-square test involves nominal

classifications and a certain loss Of information about

 

. lSidney Siegel, Nonpgrametric Statistics for the
Behavorial Sciences (New York: McGraw-Hill Book CO.,
1956), pp. 19-20, 26-28.

83

subjects,1 Pearson r correlation coefficients were
computed between all possible pairs of test variables
and compared with the Chi-square coefficients to verify
the results. Both statistical tests yielded similar
significant results although the Pearson r coefficients
were slightly higher in level of significance than the
Chi-square contingency coefficients.

The Favorableness Toward New Styles measure was
subjected to item analysis techniques using the FORTAP
and FACTOR A computer programs. The former yielded a
point biserial correlation of .86 between statement to
total scores which indicated the internal consistency of
the twenty statements was fairly high. The FACTOR A
program tested the unidimensionality of the Favorableness
Toward New Styles measure as Rogers suggested in his
analysis of measures, that many of the diffusion research
measures he studied were not unidimensional even though
they were Often highly reliable and showed acceptable
validity. The same problem occurred with the Favorableness
Toward New Styles measure in this study. The results of
the Factor Analysis indicated that the measure may not be
unidimensional.

An .01 Alpha level of significance was established
for the acceptance of all hypotheses. Although Type II

error, or the probability of not rejecting the null

 

lIbid., pp. 21-23.

 

YI
In

W

84

hypothesis when it should be rejected, is more common
when this confidence level is selected, the probability
of committing this error decreases as the sample size
increases.1 Therefore, the larger sample size in this
study probably counteracted the probability of committing
Type II error.

Pearson r correlation coefficients were calculated
between each of the test variables and marital status and
year in university. Chi-square tests were computed
between socio-economic status and each of the test
variables. Chi-square tests were also used to determine
whether there were significant differences among the four
sub-categories of respondents (fashion opinion leaders,
fashion innovators, fashion dual leaders, and fashion non-
leaders) on social participation and favorableness toward
new styles as predicted in hypotheses 6-9. The four sub-
categories were established in a manner similar to that
Schrank used.2 The respondents' scores on fashion
adoption and fashion opinion leadership were rank ordered,
divided into thirds and individuals were assigned to one

of the following mutually exclusive sub-categories:

 

lIbid., pp. 8-9.

2Schrank, "Fashion Innovation," p. 59.

85

 

 

 

 

Fashion Qpinion Fashion
Sub-Categories Leadership_Scores Adoption Scores
Fashion Innovators Not upper l/3 Upper 1/3 only
Fashion Dual Leaders Upper l/3 Upper 1/3
Fashion Opinion Leaders Upper 1/3 only Not upper 1/3
Fashion Non-Leaders Lower 1/3 Lower 1/3

The number and percentage of respondents in each of the

four sub-categories which resulted were as follows:

 

Sub-Categories E g
Fashion Innovators 40 17
Fashion Dual Leaders 41 17
Fashion Opinion Leaders 47 19
Fashion Non-Leaders 115 47
Total 243 100

Post hoc Chi-square confidence intervals were used
to locate significant differences between all possible
pairs of sub—categories on favorableness toward new
styles. This statistical comparison was also used to
locate significant differences between each of the four
sub-categories and the rest of the subjects on favor-

ableness toward new styles.

CHAPTER V

RESULTS AND DISCUSSION

The data presented and the discussion have been
organized in the following manner: (1) a description of
the sample, (2) descriptive data concerning each of the
variables, (3) a comparison of the differences between
prestigious fashion leaders and fashion non-leaders on
social participation and favorableness toward new styles,
and (4) a comparison of each of the four sub-categories

with the remaining subjects on favorableness toward new

styles.

Description of the Sample
Of the more than 17,000 women enrolled at Michigan
State University during fall quarter, 1970, 12,765 met the
established criteria for the population. Two hundred and
sixty of the 500 questionnaires (Appendix A, pp. 143-50)
*which were mailed to the randomly selected university I
‘women were returned. Seventeen of those which were

returned were discarded because of missing data.

87

Therefore, the sample size was 243 or 48.6 per cent of
the original sample.

The respondents' marital status and year-in-
university were obtained from the fall quarter student
enrollment records. The records indicated that nearly

all of the women in the sample were single:

 

Marital Status Number Per Cent
Married 16 6.6
Single 227 93.4
Total 243 100.0

The records also showed that over two-thirds of the

respondents were freshmen and sophomores:

 

 

Year in University; Number Per Cent
Freshmen 96 39.5
Sophomores 69 28.4
Juniors 48 19.8
Seniors 30 12.3
Total 243 100.0

In order to determine if marital status and year-in-
university might affect the relationships between the
variables tested, Pearson r correlations between marital
status and year-in-university and all other variables
were computed. This data yielded two significant negative
correlations beyond the .05 level. Year-in-university was
negatively related to fashion Opinion leadership (-.19)
and to marital status (-.21). Since these two corre-
lations were statistically significant although both were

very low in value, their significance was probably

88

exaggerated by the large number in the sample. Therefore,
marital status and year-in-university were dropped from
further statistical analysis.

A modification of the McGuire-White Short Form
Index of Social Status1 was used to determine the
respondent's socio-economic status. A composite score
based on the main wage-earner's occupation, income, and
education indicated the socio-economic level of each
respondent. The data indicated that the range of socio-
economic scores varied from 20 (high) to 80 (low) out of
a possible 12-84. All five of the socio-economic levels

were represented:

 

 

Levels Number of Women per cent
1. Upper 9 3.7
2. Upper-Middle 116 47.7
3. Lower-Middle 70 28.8
4. Upper-Lower 40 16.5
5. Lower-Lower 8 3.3

Total 243 100.0

Since approximately 50 per cent of the respondents
were in the two top levels of socio-economic status, the
distribution of scores was positively skewed when compared
to the general population as a whole. Data falling into
this pattern might be expected for a large state uni-
versity population.

Chi-square tests were employed to determine if

socio-economic status might influence the relationships

 

1McGuire and White, "Social Status."

89

among the other variables in the study. The data presented
in Table 1, however, showed no significant relationships
among socio-economic status and social participation,
fashion adoption, fashion opinion leadership, and favor-
ableness toward new styles at the .01 level. Therefore,
socio-economic status was given no further statistical

consideration.

TABLE 1.--Chi—square values and levels of significance for
the test variables and socio-economic status.

 

Socio-economic Status

 

Test Variables
Chi Square Value p

 

Social Participation 9.06 NSa
Fashion Adoption 3.19 NS
Fashion Opinion Leadership 7.65 NS
Favorableness Toward New Styles 4.33 NS

 

aNS = not significant.

p.01 < (16.218) df = 4

In the following section each of the four separate vari-
ables will be described briefly before the relationship

among variables are discussed.

9O

Descriptive Data Concerning Each
of the Variables

 

Fashion Opinion Leadership

 

Fashion opinion leadership scores were obtained by
summing the weights for each of the twenty statements on
pages 4-5 of the questionnaire (see Appendix A, pp. 147-48).
The possible range of scores was from 20-100. The actual
scores as presented below ranged from 23 to 95, a total

range of 73. The frequency distribution was as follows:

 

Intervals Number
23-30 8
31-38 12
39-46 31
47-54 31
55-62 33
63-70 36
71-78 42
79-87 37
88-95 .453
Total 243

A comparison of the mean, median and mode on
fashion opinion leadership (Table 2) revealed that the
scores on this measure were somewhat negatively skewed,
as they were in Schrank'sl study. A negative distribution
of_scores was expected due to the nature of fashion
opinion leadership which requires a substantial group of

followers or non-influentials.

 

1Schrank, "Fashion Innovation," p. 62.

91

TABLE 2.—-Summary of the measures of central tendency and
dispersion for the test variables.

 

 

Test Standard
Variables Means Medians Modes Deviations Ranges
/

Fashion Opinion
Leadership 62.6 64.4 72 17.0 23-95
Favorableness
Toward New
Styles 54.2 53.0 48,53,49 13.5 25-90
Fashion 7
Adoption 79.6 56.7 0,28 79.9 0-573
Social
Participation 2.1 .7 0 3.2 0-19

 

Fashion Adoption

 

Scores on the Fashion Adoption Measure were
obtained by summing the weighted parts 3,4, and 5a of the
fashion adoption section of the questionnaire (see Appendix
A, pp. 145-46). The weighting scheme was devised so
that negative scores were not possible to obtain and there
was no limit on the maximum score possible. The data
indicated, on the following page, that the total actual

range of fashion adoption scores was from 0 to 537.

Two women, each with scores approximately 100 points
higher than the others in the sample contributed to the
wide range of scores. Whether these women were exception-
ally innovative or did not report accurate information

is not known. The high standard deviation (Table 2) also

92

 

Intervals Number
0 7
1-30 63
31-60 58
61-90 51
91-120 18
121-150 9
151-180 9
181-210 11
211-240 4
241-270 3
271-300 2
301-330 4
333 l
343 l
405 l
537 ._;1
Total 243

indicated that the scores on fashion adoption were widely
dispersed. The mean, median and modes as well as the
frequency distribution revealed that the scores were
negatively skewed (see Table 2). A negative distribution
of scores on fashion adoption was logical in view of the
nature of the fashion movement which is based on the
premise that few individuals adopt when new styles are
first introduced.

Since the distribution of fashion adoption scores
was bimodal and the values of the two modes (Table 2)
were quite different, two distinct groups of women may
have existed in this sample: (l).those who did not adopt
any of the new styles, and (2) those who adopted a small
number of the new styles. In order to describe the
innovativeness of the respondents in greater detail, data

will be presented from four of the sub-sections of the

93

fashion adoption section of the questionnaire which
follow: (1) fashion-cycle ranking index, (2) self-
designating innovativeness statements, (3) wardrobe

inventory, and (4) style list.

Fashion-cycle ranking index.--The respondents'

 

perception of new styles on the fashion-cycle ranking
index was the primary criterion used to weight the various
items in the wardrobe inventory. As shown in Table 3,

the largest percentage of respondents rated hot pants as
the newest style and midi skirts and gaucho pants as
fairly new styles.

All women, however, did not agree as to which
styles were the newest. In order to confirm that the
three styles above (midi skirts, gauchos, and hot pants)
were actually new styles, two additional criterion were
used (see Chapter IV for detailed explanation): (1)
frequency of styles occuring in Vogue magazine (Appendix B),
and (2) styles in the wardrobe inventory with the lower
per cent-to-total (see Table 4). While the former showed
that midis, gauchos and hot pants had been recently been
introduced; the latter indicated that they were not yet
accepted by the majority of the women in the sample and,

therefore, were new styles.

 

1Rogers, Diffusion of Innovations, p. 160.

 

94

.mpoe on» mm ©m>flwoummo

.mmHmpm 3mm mHuHmm mm Um>flmoumm

Q

.mHmum umoswc on» mm pm>Hmouomm

 

 

 

mo. AH. Hm. oo. oo. mom muHommaso
oo. oo. mo. no. moo. mom magma pom
mo. AH. oo. mm. HH. mom mumm> aoHne
Ho. mo. oao. mo. om. mom muHsmuamm
oo. om. NH. oo. oH. mom muume me2
so. om. mo. nmo. oo. mom momma onoomu
oo. om. oo. ham. no. mom muume Hon
oo. mo. omo. Ho. oH. mom maouuon HHmm
No. AH. ooh. Ho. oo. mom muonmuHcHz
AH. Hm. No. om. oo. mom mumoncx
omumpuso poo mmz mcflummz 3oz EwuH z mEmuH
m.uH co m.mcomnm>m mHuHmm ummuuom “mag mHmum
.xmocfl

mcflxcmu mHomolcoHQmMM way so mmampm may mo sumo mcflumu :wEoz

mo ommpcooummln.m mamme

95

Self-designating innovativeness statements.--The
range of the self-designating innovativeness statements

was from 0 to 8. The distribution of scores follows:

 

 

Intervals Number of WOmen Per cent
0-2 3 1
3-5 174 72
6-8 66 27
Total 243 100

The majority of the women scored in the middle
interval which indicated that these women judged themselves
to be similar to others in the group rather than more or

less innovative.

Wardrobe inventory.--The seven garment length

 

categories of clothing styles which comprised the wardrobe
inventory are shown in Table 4. Both ankle-length and
mini-length garment categories contained the largest
numbers of garments; each with well over one-third of the
total number of garments in the wardrobe inventory. Since
the number of pants and mini-skirts were similar and were
more than twice that of any of the other styles in the
inventory, both styles were therefore designated as the

mode or the accepted clothing styles for this sample.

Style list, time-of-adgption, and subsequent

discontinuance.--As previously explained in Chapter IV, a

 

style list was included in the questionnaire (Appendix, A,

p. 146) in order to measure actual time of adoption.

96

TABLE 4.--Percentage distribution and mean number of styles
bought new or sewn in total wardrobe.

 

Wardrobe Inventory Number Per cent Mean

 

Garment Length Category 1

 

 

A. Skirts (micro-mini) 588 9.1 2.4
B. Pants ("hot" pants) 223 3:4 .9
811 12.5
Garment Length Category 2
A. Skirts (mini) 1,529* 23.6 6.3
B. Pants (shorts) 807 12.4 3.3
2,336 36.0
Garment Length Category 3
A. Skirts (just-above
knee) 483 7.4 2.0
B. Pants (culottes) 144_ 2.2 .6
627 _ 9.6
Garment Length Category 4
A. Skirts (knee-length) 44 .7 .2
B. Pants (knickers) 28 ii. .1
72 1.1
Garment Length Category 5
A. Skirts (just-below
knee) l6 .3 .1
B. Pants (peddle-pushers) _8 L1, .0
24 .4
Garment Length Category 6
A. Skirts (mid-calf) 36 .6 .l
B. Pants (Gauchos) 25 L4 .1
61 1.0
Garment Length Category 7
A. Skirts (maxi) 227 3.5 .9
B. Pants 1,573* 24.2 6.5
C. Jumpsuit/overalls 219 3.4 .9
D. Pantsuit 540 8.3 2.2
2,559 — 39.4
Totals 6,490 100.0 26.6

 

*Designated modes.

97

Fashion diffusion curves were plotted for each of the five
styles in the list (Appendix B, p. 153) based on the month
and year of adoption indicated by the respondents. The
data revealed that while several women owned pantsuits and
maxis in 1968, there were no midi skirts or gaucho pants
in any of the respondents' wardrobes until June, 1969 and
September, 1969, respectively. Although pantsuits and
maxis appeared to have the most rapid diffusion rates,
jumpsuits seemed to have fairly moderate popularity
throughout the period and midis and gauchos were owned by
only a few of the respondents.

Since the respondents were asked to recall those
styles adopted over a three-year period of time, time of
adoption was followed by a subsequent discontinuance
record. The data (Table 5) revealed that pantsuits and
maxi skirts were the most frequently rejected styles.

This finding was consistent with data from the wardrobe
inventory which showed that pantsuits and maxi skirts
were not the newest styles. Perhaps these two styles may
be a declining mode or a fad. However, the percentage

of the number of women discontinuing to the total number
of women who had purchased the styles was quite small.
Reasons which were given by the women for discontinuing

wearing the styles are shown on the following page.

98’

TABLE 5.--Frequency and percentage of women discontinuing
wearing style list items to total number of
women owning style list items.

 

 

 

 

 

 

 

Number Number Per cent
Style List Women Discontinuing Discontinuing
Items Owning Wearing wearing
Mid-Calf Length:
A. Skirts 21 2 10
B. Pants 20 l 5
Ankle-Length:
A. Skirt 114 9 8
B. Pantsuit 201 8 4
C. Jumpsuit/
overalls 99 4 4
Reasons Discontinuing, Number of_Women
wearing Styles Stating Reason
"doesn't fit" 7
"dislike style" 7
"not comfortable" 4
"no occasion to wear" 4
"opposite sex dislikes" l
"boredom" _1
Total 24

Favorableness Toward
New Styles

 

The scores for each respondent on favorableness
toward new styles were the summed weights of the twenty
statements on pages 5-6 of the questionnaire (see
Appendix A, pp. 148-49). The range of scores on the favor-
ableness toward new styles measure varied from 25 to 90
out of a possible range of 20-100. The distribution of

scores follows:

99

Interval Number
25-30 6
31-35 11
36-40 23
41-45 25
46-50 36
51-55 43
56-60 26
61-65 27
66-70 13
71-75 13
76-80 11
81-85 6
86-90 __3
Total 243

The frequency distribution does not appear skewed
but a comparison of the mean, median and modes in
Table 2, indicated that the scores were skewed somewhat
in the positive direction. The skewness of the scores
could not be predicted since the attitudes of the
respondents toward new styles was likely to depend on
the group's norms on innovativeness.

Since a panel of judges rated the intensity and
polarity of the statements (for a detailed explanation see
Chapter IV) in the Favorableness Toward New Styles
Measure, a score of 60 on this measure, which had a
possible range of scores of 20 to 100, may be designated
as a neutral attitudinal position. The sample mean,
median and mode on favorableness toward new styles may then
be compared with this presumed neutral position. A

closer examination of the measures of central tendency

(Table 2) showed that these scores appeared to cluster

 

100

only slightly lower than the presumed neutral attitude
score of 60. The data seem to indicate that the norm for
the group is primarily indifference toward new clothing
styles. The current popular literature also seems to
indicate that "anything goes" in dress and that individuals

do not seem to care what others wear.

Social Participation

 

Few respondents participated in university, church,
or community organizations, clubs or groups (see
Appendix A, p. 144). A comparison of the mean, median,
and mode on social participation in Table 2 as well as an
analysis of the frequency distribution of social par-

ticipation scores indicated a highly negatively skewed

distribution:
Interval Number

0 113

l 47

2-3 27
4-5 23
6-7 18
8-9 5
10-11 5
12-13 2
14-15 1
16-17 1
18-19 __1
Total 243

One hundred and thirteen women or 46.5 per cent of the
respondents received zero scores on the social par-
ticipation measure. A negatively skewed distribution was

anticipated since the social participation measure was

101

weighed to give higher scores to those women who were
presidents and officers of clubs and organizations.

A report of the findings concerning the relation-
ships among the variables of fashion adoption, fashion
opinion leadership, social participation, and favorableness
toward new styles follows.

Relationship Between Fashion Adgption
and Fashion Opinion Leadership

Fashion adOption was hypothesized to be positively
related to fashion opinion leadership. A Chi-square test
indicated that the coefficient of contingency was highly
significant (see Table 6). This result revealed that
women who scored high on fashion adoption also scored high
on fashion opinion leadership. Therefore, the hypothesis

was accepted.

TABLE 6.--Re1ationship of women's fashion adoption and
fashion opinion leadership.

 

Fashion Adoption

 

Fashion Opinion

 

 

 

Leadership 0-36 37-78 79-537 Totals
23-53 41 25 20 86
54-71 27 23 24 74
72-95 13 29 41 83
Totals 81 77 85 243 S

2

x = 22.22 c = .29 p < .001

102

This finding which indicated that women who were
highly influential in the field of fashions also owned
and wore more new styles than others, lends support to
Schrank's1 and Marshall's2 results which showed fashion
opinion leadership to be positively related to fashion
adoption. Since Rogers3 believed that the relationship
between opinion leadership and adoption was dependent
upon social system norms, and since fashion opinion
leadership and fashion adoption were found to be positively
related in this study, the norms of the sample should have
been somewhat favorable toward innovations in clothing
styles. However, the descriptive data on the Favorable
ness Toward New Styles measure indicated that the group
norm was somewhat negative or slightly below the assumed
neutral position. Perhaps the actual neutral attitude
position was lower than the assumed score of 60 due to
unknown intervening factors such as the respondents'
interpretation of the statements in the Favorableness

Toward New Styles measure.

 

1Schrank, "Fashion Innovation," p. 67.

2

Marshall, "Leadership in Mens Fashions,’ p. 48.

3Rogers, Diffusion of Innovations, p. 245-57.

 

103

Relapionship Between Fashion Opinion
Leadergpip and Favorableness
Toward’New Styles

 

 

A positive relationship between fashion opinion
leadership and favorableness toward new styles was

hypothesized. The Chi-square contingency coefficient

(Table 7) was highly significant beyond the .01 level‘and

gave ample support for the hypothesis.

TABLE 7.--Relationship of women's fashion opinion leader-
ship and favorableness toward new styles.

 

 

 

 

 

Favorableness Fashion Opinion Leadership
Toward New
Styles 23-53 54-71 72-95 Totals
0-46 53 19 20 92
47-58 25 32 28 85
59-90 8 23 35 66
Totals 86 74 83 243
x2 - 39.63 c = .37 p < .001

This finding indicates that women who were highly
influential in the field of fashion also were very
positive in their attitudes toward new styles. This
result was congruent with the findings of the first
hypothesis which indicated that women who scored high on
opinion leadership were also innovators in wearing the
new styles. Thus, although individual's attitudes are not
always in agreement with their actual behavior, the

attitudes and behaviors of the opinion leaders in this

104

study seemed to be consistent concerning the acceptance
of new styles.

Although little research has been done concerning
the relationship between these two variables, this data
supports the single study1 which showed that women who
were opinion leaders enjoyed experimenting with the newest
styles.

The results of the two preceeding hypothesis which
show that both fashion opinion leadership and fashion
adoption were significantly related to favorableness
toward new styles indicates that the latter variable might
have affected the relationship between fashion adoption
and fashion opinion leadership. Furthermore, Rogers'
generalizations suggest that the relationship between
opinion leadership and innovativeness is dependent upon
the norms of the social system.2 Therefore, the effect
of favorableness toward new styles on the relationship
between fashion adoption and fashion opinion leadership
was eliminated by Chi-square tests. Visual inspection of
the data indicated that the contingency coefficients
obtained between the high (x2 = 4.47), medium (x2 = 9.59)
and the low (x2 = 5.25) groups on the two variables were

lower (compare with Table 6) when the effects of the

 

1Summers, "Fashion Opinion Leaders," p. 180.

2Rogers, Diffusion of Innovations, p. 245.

‘/_

105

third variable, favorableness toward new styles, was
removed. Thus, favorableness toward new styles seems to
have affected the relationship between fashion opinion

leadership and fashion adoption.

Relationship Between Fashion Opinion
Leadership and Social Participation

 

A positive relationship was hypothesized between
fashion opinion leadership and social participation. The
Chi-square contingency coefficient, however, was not
significant at the .01 level, and consequently the
hypothesis could not be confirmed (see Table 8). Although
the finding reveals that no significant relationship
existed between the two variables, the result was in the
predicted direction.

TABLE 8.--Relationship of women's fashion opinion leader-
ship and social participation.

 

Fashion Opinion Leadership

 

 

 

 

 

Social
Participation 23-53 54-71 72-95 Totals
0 27 26 26 79
l 20 12 18 50
2-19 39 36 39 114
Totals 86 74 83 243
X2 = 1.35 c = .07 P = ~35 NSa
a

N5 = not significant.

106

Based on the review of the literature which
showed that fashion opinion leadership was related to both
formal and informal social participation, the present
study does not support the majority of the research. The
social participation measure itself may not have been
refined enough to distinguish between the respondents.
On the other hand, the fashion opinion leaders in this
study may have participated in informal rather than formal
social activities and a multi-dimensional measure of
informal as well as formal social participation might have
produced different results. The non-significant relation-
ship found between fashion opinion leadership and social

participation as measured by formal organizational par-

ticipation did, however, support Katz and Lazarsfeld's
statement that: "The fact that a woman is a leader in one

area has no bearing on the likelihood that she will be a

leader in another . . ."1

Relationship Between Fashion Adoption
and Social Participation

Social participation was hypothesized to be
;positively related to fashion adoption. Contrary to the
prediction, the Chi-square contingency coefficient did
:not reach the .01 level of significance, and therefore,

failed to support the original hypothesis (see Table 9).

 

1Katz and Lazarsfeld, Personal Influence, p. 334.

107

TABLE 9.--Relationship of women's fashion adoption and
social participation.

 

Fashion Adoption

 

 

 

 

 

Social Participation 0-36 37-78 79-537 Totals
0 32 25 22 79
1 16 18 16 50
2-19 33 34 47 114
Totals 81 77 85 243
x2 = 4.87 c = .14 p = .30 Nsa
a

NS = not significant at .01 level.

This finding suggests that social participation as
measured was not related to fashion adoption.

A Pearson r test to verify the results confirmed
the non-significant relationship between the two vari-
ables. Both statistical tests, however, did show that
the relationship between fashion adoption and social
participation was in the positive direction as hypothesized.
Perhaps if the measure had included cosmopolitan social
activities in addition to the formal activities on the
campus and in the local community the relationship would
have been significant. Rogers summary of diffusion
research indicated that fashion innovators were likely to

txelong to cliques and formal organizations outside their

social system.1

 

lRogers, Diffusion of Innovations, pp. 169, 183.

108

Relationship Between Fashion Adoption and
Favorableness TowardiNew Styles

 

 

A positive relationship was predicted between
fashion adoption and favorableness toward new styles. A
Chi-square test between the two variables yielded highly
significant results beyond the .01 level, confirming the
hypothesis. Women who scored high on fashion adoption
also scored high on favorablenss toward new styles (see

Table 10,)

TABLE lO.--Relationship of women's fashion adoption and
favorableness toward new styles.

4 g- , '—

Fashion Adoption'

 

 

 

 

Favorableness

Toward New Styles 0-36 37-78 79-537 Totals

0-46 41 31 20 92

47-58 24 28 33 85

59-90 16 18 32 66

Totals 81 77 85 243
x2 H 14.75 c = .24 p < .01

This study supports the research of Pasnakl whose
findings also seemed to indicate that fashion innovators
have positive attitudes toward new styles. The data also

substantiates Rogers2 statement that innovators are

 

lPasnak, "Fashion Innovators,‘ p. 96.

2Rogers, Diffusion of Innovations, pp. 111, 169.

 

109

characterized by venturesomeness and have more favorable
attitudes toward new ideas than others.
Significance of Differences Between

Sub-Categories on Social
Participation

 

 

 

In order to test whether significant differences
existed among types of prestigious fashion leaders and
fashion non-leaders, the sample was divided into four sub-
samples based on the fashion opinion leadership and
fashion adoption measures. The scores on both measures
were divided into thirds as described in Chapter IV,
pp. 84-85, and the four mutually exclusive sub-categories
which were established were: (1) fashion innovators,

(2) fashion opinion leaders, (3) fashion dual leaders,

and (4) fashion non-leaders. Since part of the re-
spondents who scored in the upper one-third of the

fashion adoption measure were removed and labeled "fashion
dual leaders," the innovators who remained may have had
characteristics which differed from those of innovators

in other studies. The fashion dual leaders' mean score

on fashion adoption (see Table 11) was slightly higher than

the innovators' which indicates that the fashion dual

leaders were in reality the most innovative of those in
the sample. Therefore, Rogers' generalizations concerning
the characteristics of innovators may be more applicable

to the fashion dual leaders in this study.

110

TABLE ll.--Mean scores on fashion adoption for each sub-
category and the rest of the subjects.

 

Mean Scores for Mean Scores for

 

Sub-Categories Sub-Categories Rest of Subjects
Fashion Dual Leader 164.4 80.1
Fashion Innovator 161.4 81.1
Fashion Opinion Leader 43.3 120.5
Fashion Non-Leader * 35.7 123.0

 

Significant differences were hypothesized between
fashion opinion leaders, fashion innovators, fashion dual
leaders, and fashion non-leaders on social participation.
Contrary to the prediction, a Chi—square test revealed
that the differences among the four sub-categories on
social participation was not significant at the .01 level
(see Table 12). Consequently, the hypothesis was not
confirmed.

This finding indicates that any one of the four
sub-categories did not have a significantly greater amount
of social participation than the other sub-categories.
The statistical test showed that there is only one chance
in a 100 that the sub-categories may differ in their
.amount of social participation. Perhaps a multi-
<iimensional measure of social participation including
(mosmopoliteness and informal social activities as well as

the formal organizational participation would yield

111

TABLE 12.--Significance of differences between sub-
categories on social participation.

 

 

 

 

 

Social Fashion Fashion Fashion Fashion
Partici- Opinion Inno- Dual Non-
pation Leaders vators Leaders Leaders Total
0 23 16 17 58 114
l 12 8 8 22 50
2-19 12 16 16 35 79
Totals 47 40 41 115 243
x2 = 3.80 c = .12 p = .70 N53

a

NS = not significant at .01 level.

significant differences between the four sub-categories.
Since the results revealed no significant differences
among the four sub-categories no further statistical
treatment of the data was undertaken.

Significance of Differences Between Sub-

Categories on Favorableness Toward
New Styles

 

Significant differences were hypothesized between
fashion opinion leaders, fashion innovators, fashion dual
leaders, and fashion non-leaders on favorableness toward
:new styles. The Chi-square contingency coefficient
«obtained between the two variables was significant beyond
‘the:.01 level (see Table 13). Therefore, the hypothesis
1nas accepted. This finding indicated that at least one

43f the pairs of sub-categories was significantly

112

TABLE 13.--Significance of differences between sub-
categories on favorableness toward new styles.

 

 

 

 

Favorable- Fashion Fashion Fashion Fashion

ness Toward Opinion Inno- Dual Non-

New Styles Leaders vators Leaders Leaders Total

25-46 27 14 27 24 92

47-58 14 13 ll 47 85

59-90 6 l3 3 44 66

Total 47 40 41 115 243
2

x = 39.62 c = .37 p < .001

different in their attitudes toward new styles and that
there was only one chance in 100 that they were alike.

In order to identify the location of the signifi-
cant differences among the groups, post hoc Chi-square
confidence intervals1 were used. The data (Table 14)
revealed significant differences between two pairs of
groups: (1) fashion dual leaders and fashion non-leaders,
and (2) fashion opinion leaders and fashion non-leaders.

Since both fashion dual leaders and fashion
opinion leaders possess the common trait of opinion
leadership, this finding which indicates that both sub-
categories were significantly different from fashion non-

leaders is consistent with the previous finding which

 

1Leonard A. Marascuilo, "Large-Sample Multiple
<20mparisons," Psychological Bulletin, LXV, No. 5 (1966),
281-84.

113

TABLE l4.--Confidence intervals for significance of
differences between pairs of sub-categories on
favorableness toward new styles.

 

Difference Between
Sub-Categories Examined Sub-Categories Decision

 

Non-leaders vs. Dual leaders +.11<p-p<+.79 S

Non-leaders vs. Opinion
leaders +.69<p-p<+.03 S

Non-leaders vs. Innovators -.20<p-p<+.44 NS

Dual leaders vs. Opinion
leaders +.33<p-p<+.51 NS

Dual leaders vs. Innovators +.74<p-p<-.12 NS

Opinion leaders vs.
Innovators +.65<p-p<-.21 NS

 

aS = significant difference at .01 level.

NS = no significant difference at .01 level.
revealed that fashion Opinion leadership was significantly
related to favorableness toward new styles.

However, this finding appears to differ from
Rogers' conclusion, based on diffusion research findings,
that opinion leaders are "just like their followers, only
more so."1 Since a sociometric technique was 22E used to
determine the fashion opinion leadership scores in the
present study, but was used in the majority of the
adoption-diffusion research studies, the fashion non-

leaders in this study cannot be assumed to be the followers

 

lRogers, Diffusion of Innovations, p. 233.

 

114

of the fashion opinion leaders. Therefore, differences
between diffusion research findings and the finding of the
present study which showed fashion opinion leaders and
fashion non-leaders to be significantly different on
favorableness toward new styles may be due to the
selection of different methods of measuring opinion
leadership.

Significance of the Difference Between

Fashion Opinion Leaders and Ehe Rest

of the Supjects on Favorableness
Toward New Styles

 

 

 

 

No significant difference was hypothesized between
fashion opinion leaders and the rest of the subjects on
favorableness toward new styles. For this analysis
fashion dual leaders, fashion innovators, and non-leaders
were grouped and compared with fashion opinion leaders.
Post hoc Chi-square confidence intervals for differences
between groups (Table 15) indicated that no significant
differences existed between fashion opinion leaders and
the rest of the subjects. Therefore, the hypothesis was
accepted.

This finding supports Rogers'l generalization
that opinion leaders conform closely to the group's

norms and Schrank's2 finding that women who scored high

 

lRogers, Diffusion of Innovations, pp. 233-36.

2Schrank, "Fashion Innovation," p. 66.

115

TABLE 15.--Confidence intervals for significance of the
difference between fashion opinion leaders and
the rest of the subjects on favorableness
toward new styles.

 

Favorableness Fashion

 

 

 

Toward New Opinion Rest of Difference
Styles Leaders Subjects Between Decision
25-46 27 65 +.52<p-p<-.04 nsa
47-58 14 71 +.22<p-p<-.34 NS
59-90 6 60 +.ll<p-p<-.47 NS
Totals ‘ 47 196

a

NS = no significant difference at .01 level.

on fashion opinion leadership also expressed a belief in
conforming to the norm of their friends' dress.
Therefore, the fashion opinion leaders in this study,
seemed more likely to possess motives of emulation or the
desire to be like the group rather than motives of
differentiation in their attitudes toward new styles.
The data also indicate support for the proposition
developed in the literature review that influential
prestigious leaders must possess attitudes similar to
those of the group in order to carry out their role as
an influential.

Although this finding that fashion opinion leaders
are not significantly different from the rest of the

subjects on favorableness toward new styles does not seem to

 

J—

116

coincide with the previous finding that fashion opinion
leaders are significantly different from non-leaders on
the same measure, the difference may be due to the in-
creased variability in the scores which occurs when the
fashion innovators and fashion dual leaders scores are
combined with the scores of the fashion non-leaders to

form one group.

Significance of the Differences Between Each of
the Followinngub-Categories and the Rest of
the Subjects on Favorableness Toward’New
Styles: Fashion Innovators, Fashion
Dual Leaders, and Fashion Non-Leaders

 

 

 

Significant differences were predicted between
each of the following sub-categories and the rest of the
subjects on favorableness toward new styles: fashion
innovators, fashion dual leaders, and fashion non-leaders.
In order to perform the statistical analysis each of the
three sub-categories was paired with the other three
remaining sub-categories combined to form one group. Post
hoc Chi-square confidence intervals between each of the
pairs revealed partial support for the hypothesis
(Table 16). Two of the sub-categories, fashion non-
leaders and fashion dual leaders, were each found to be
significantly different from the rest of the subjects.
Innovators, contrary to the hypothesis, were not sig-
nificantly different from the rest of the subjects in
their attitudes toward new styles. Therefore, the

'hypothesis could not be fully confirmed.

Q

 

 

 

117

 

.wocoumMMHp ucmoHMHcmHm on u mzQ

.uconommwp maucmoHMHcmHm H mm
mz mH.+vmumvom.n “mom mumpmmH Hmoo omumm
mz mH.+vmnmvmm.u ummm mumpmmH Hmso mmuhv
m Ho.+vm|mvno.+ ummm mumpme Hmso ovumm
mz mm.+vmnmvma.u pmmm muoum>occH omumm
mz mo.|vm|mvao.+ ummm mnoum>ocsH mmunv
mz mo.|vmlmvmo.+ ummm muoum>occH ovumm
m mo.+vmnmvav.+ umom mumpmoHucoz Anmflzv omnmm
sz «H.1vmlmvom.+ ummm mumpmmalcoz AEoHmev mmnhv
mm vm.|vmlmvma.l ummm mumpmmaucoz A30Hv wvnmm
conHowo cwo3umm muomhbsm pmcflemxm monum
mocmHmMMHo may anommumu 3oz pumzoe
mo “mom unom mmmcmHnmuo>mm

 

.muopmmHucoc
cofismmm pom .muopmoH Hmop coflnmmw .mnoum>occfl cngm8m "mmHmum 30c pumsou
mmocmHano>mm co muoanSm 0:» mo ummu one can mmHuomouMUIQSm mcH3oHH0m
on» wo sumo comsuon moocmnoMMHp mo oUQMOflMHcmHm now mHm>HmucH wocopwmcounu.mH flames

118

This finding indicates that the attitudes of the
fashion dual leaders and fashion non-leaders, but not the
fashion innovators, appeared to differ from the group norm
on favorableness toward new styles. Possibly the Fashion
Adoption and Favorableness Toward New Styles measures were
not sufficiently refined to enable fashion innovators to
be differentiated from the rest of the subjects. Or,
fashion innovators may actually be less favorable in their

attitudes than would seem likely from their innovative

behavior.

CHAPTER VI

SUMMARY AND CONCLUSIONS

Summary

The major concern of this study was to investigate
the acceptance of new styles in women's clothing and to
compare differential characteristics of three types of
prestigious fashion leaders and fashion non-leaders on
social participation and favorableness toward new styles.
Acceptance of two recently introduced styles, "hot" or
very short pants and mid-calf length skirts and pants,
were used as a means to assess the extent to which each
of the four sub-categories of respondents above exhibited
differentiation or emulation of group norms on dress.

The literature revealed that although theorists
generally agree that fashion diffusion and the acceptance
of new styles are based on two polar motives, emulation
and differentiation, writers have generally emphasized
either emulation or differentiation rather than the
relationship between the two. Despite the majority of the
theorists' emphasis on the importance of either the group

or the fashion leader, the implication throughout the

119

120

literature is that the fashion leaders' ability to
influence others or to promote styles which are later
accepted by the group is related to the fashion leaders'
own innovativeness which in turn depends upon group
norms. Since many new styles are promoted but do not
become generally accepted it is likely that certain types
of fashion leaders may show more non-conformity or differ-
entiation in dress than group norms will tolerate.
Differences found between various types of prestigious
leaders and the rest of the subjects on favorableness
toward new styles would give some support to the proposed
theoretical relationship between the concepts of differ-
entiation and emulation.‘

Rogers' theory of the diffusion and adoption of
innovations, included generalizations concerning both the
influential and innovative characteristics of certain
typical prestigious leaders in relation to group norms on
innovativeness. Therefore, Rogers' theory seemed to be
most useful in relating the two polar concepts, emulation
and differentiation. In addition, Rogers' theory appeared
to be helpful in explaining the stage in the process of
fashion change where prestigious fashion leaders actually
transfer new styles to their followers.

The population for the study consisted of married
and single undergraduate women who lived on the university
campus, in East Lansing, or Lansing, Michigan. Since the

sample of 243 respondents was drawn randomly from the

121

12,756 women who met the criteria, the findings of this
study are applicable to the other women students living
in the university area.

The questionnaires, mailed to 500 randomly
selected women, were designed to obtain information about
fashion opinion leadership, fashion adoption, favorableness
toward new styles, social participation, and socio-economic
status. Demographic data concerning marital status and
year-in-university were obtained from the university
enrollment records printed in the student directory.

Fashion curves of the frequencies of new styles
appearing in Vogue magazine were plotted to establish a
cut-off date for the Fashion Adoption Measure. The
respondents' perceptions of new styles were used to weight
specific items in the wardrobe inventory section of the
Fashion Adoption Measure.

The scores of all 243 respondents were used in the
analysis of the first five hypotheses concerning the
relationships among the test variables. The test vari-
ables were fashion adoption, fashion opinion leadership,
social participation, and favorableness toward new styles.
Since the data were not all interval level, the distribution
of scores was skewed, and the bivariate distributions non-
linear, Chi-square tests were used to determine the sig-
nificance of the relationships among the variables.

Pearson r correlation coefficients were calculated to

verify the results and to measure the effects of demographic

122

data on the test variables. Few significant correlations
resulted in the latter case. Therefore, concern for
those intervening demographic variables measured was
dismissed. Means, medians, modes, standard deviations,
and ranges were calculated to describe the distribution

of the scores.

The subjects were divided into four mutually
exclusive sub-categories for the statistical analysis of
hypotheses six through nine. The sub-categories were
based on the respondents' scores on the Fashion Adoption
and Fashion Opinion Leadership measures. The upper one-
third of the subjects on each of these variables were
chosen as fashion innovators and fashion opinion leaders.
Women who scored in the upper one-third of p232_measures
were labeled fashion dual leaders. Fashion non-leaders
were those women who did not score in the upper one-third
of either of the two measures. Chi-square tests were used
to determine whether there were significant differences
between all possible pairs of sub-categories on social
participation and favorableness toward new styles. In
order to determine the location of the differences between
the four sub-categories of respondents on favorableness
toward new styles, post hoc Chi-square confidence levels
were used. This statistical comparison was also used to
determine the significance of the differences among each
of the four sub-categories and the rest of the subjects

on favorableness toward new styles.

123

A summary Of the proposed hypotheses and the

results are recorded as follows:

Hypothesis l.--Fashion Opinion leadership will be
positively related to fashion
adoption.

A highly significant relationship was found between

fashion opinion leadership and fashion adoption, thereby

confirming the hypothesis above.

Hypothesis 2.--Fashion Opinion leadership will be
positively related tO favorableness
toward new styles.

The contingency coefficient measuring the relation-
ship between fashion Opinion leadership and favorableness
toward new styles indicated a highly significant association
between the two variables. This finding supports the
second hypothesis.

Hypothesis 3.--Fashion Opinion leadership will be
positively related to social par-
t1c1pation.

The positive relationship found between fashion

Opinion leadership and social participation as measured by
formal participation in clubs, organizations, and groups
was not significant. Therefore, the third hypothesis could

not be confirmed.

Hypothesis 4.--Fashion adoption will be positively
related to social participation.

Although the contingency coefficient between
fashion adoption and social participation was found to be

in the positive direction, the relationship between the

124

two variables was not significant. Therefore, the

hypothesis above was not accepted.

Hypothesis 5.--Fashion adoption will be positively
related to favorableness toward new
styles.

A highly significant relationship was discovered
between fashion adoption and favorableness toward new
styles. This finding confirmed the fifth hypothesis.

Hypothesis 6.--Fashion Opinion leaders, fashion
innovators, fashion dual leaders,
and fashion non-leaders will differ

significantly from each other on
social,participation.

The differences among fashion Opinion leaders,
fashion innovators, fashion dual leaders, and fashion non-
leaders on social participation were not found to be
significant at the .01 level of confidence. Therefore,

hypothesis six could not be confirmed.

Hypothesis 7.--Fashion opinion leaders, fashion
innovators, fashion dual leaders,
and fashion non-leaders will differ
significantly from each other on
favorableness toward new styles.

Highly significant differences were revealed

between fashion Opinion leaders, fashion innovators,
fashion dual leaders, and fashion non-leaders on favor-
ableness toward new styles. Consequently, the seventh

hypothesis was supported. Post hoc Chi-square confidence

intervals for differences between groups indicated that

125

two pairs Of groups: (1) fashion dual leaders and non-

leaders, and (2) fashion opinion leaders and non-leaders,

were significantly different in their attitudes toward

new styles.

Hypothesis 8.--Fashion opinion leaders will not
differ significantly from the rest
Of the subjects on favorableness
toward new styles.

 

As predicted, no significant differences were
discovered betweeen fashion opinion leaders and the rest
of the subjects on favorableness toward new styles. Thus,
the hypothesis above was accepted.

Hypothesis 9.--Fashion innovators, fashion dual
leaders, and fashion non-leaders
will each differ significantly from

the rest of the subjects on favor-
ableness toward new styles.

 

Partial support for the above hypothesis was
Obtained since two Of the above sub-categories, fashion
dual leaders and fashion non-leaders, were each found to
be significantly different from the rest of the subjects
on favorableness toward new styles. Since, however,
innovators were not significantly different from the rest
of the subjects in their attitudes toward new styles, the
hypothesis could not be fully accepted.

In summary, the findings of this study revealed
highly significant positive relationships, for university

women living in the East Lansing or Lansing area: among

126

favorableness toward new styles, fashion adoption, and

fashion Opinion leadership. NO significant relationships,

however, were revealed between fashion adoption or fashion
Opinion leadership and social participation.

Consistent with the findings which showed a lack
of relationship between fashion adoption or fashion
Opinion leadership and social participation, there were
also no significant differences found between fashion
Opinion leaders, fashion innovators, fashion dual leaders,
and fashion non-leaders on social participation. There
was a significant difference between the four sub-
categories, however, on favorableness toward new styles.
Two Of the three sub-categories which were expected to
differ from the rest of the subjects on favorableness
toward new styles were found to be significantly different.
These sub-categories were fashion dual leaders and fashion
non-leaders. Innovators, contrary to the prediction, were
not found to be significantly different from the rest Of
the subjects on the same measure. Opinion leaders, as
predicted, were not significantly different from the rest

Of the subjects on favorableness toward new styles.

Implications

 

The findings support the major Objective of this
study revealing that a theoretical relationship may exist
between: (1) the extent tO which fashion leaders may

either emulate group norms on dress or show individual

127

distinction in dress, and (2) the group norms on inno-
vativeness in dress. The evidence indicates that both
forces, emulation and differentiation, Operate within a
social system and that the balance Of these forces seem to
exert a differential effect on the three types Of
prestigious fashion leaders and the fashion non-leaders
studied.

Motives Of differentiation appeared to be evident
in the attitudes of the fashion dual leaders and the
fashion non-leaders since these sub-categories were found
to be significantly different from the rest of the
subjects on favorableness toward new styles. The fashion
dual leaders and the fashion non-leaders represented the
extreme and Opposite ends Of the innovativeness and
favorableness continuums. Their less-conforming attitudes
toward new styles, thus, were consistent with their behavior.
Fashion dual leaders and fashion non-leaders appear to be
two different types Of non-conformists: (1) fashion dual
leaders were more innovative and more favorable toward new
styles while, (2) fashion non-leaders were less innovative
and less favorable toward new styles.

Motives Of emulation seemed to be exhibited in
the attitudes Of the fashion innovators and the fashion
Opinion leaders who were both not found to be significantly
different from the rest Of the subjects on favorableness

toward new styles. While motives Of emulation had been

128

predicted for the fashion Opinion leaders; the conforming
attitudes of the fashion innovators were unexpected.

One explanation for the fashion innovators'
conforming attitudes and less-conforming behavior may be
the availability Of new styles on the market. A fashion
innovator may be the person Blumer spoke of who ". . .
unwittingly follows a fashion . . . because Of a limitation

1 It is interesting to note that the fashion

Of choice."
non-leaders who were relatively more "laggard" in their
adoption Of new styles than the other three sub-categories,
were found to have attitudes toward new styles which were
not significantly different from the fashion innovators.
Perhaps the role of the innovator may be that of a less
deliberate fashion leader. The findings Of Schrank2 and
Goodell3 seem to indicate that the fashion dual leader,
rather than the fashion innovator, is the more active Of
the two types of innovators. The significant difference
found in this study between fashion dual leaders and the
rest Of the subjects but not between fashion innovators
and the rest of the subjects on favorableness toward new
styles, suggests that women who both adopt and influence

are more favorable toward new styles than those women who

only adopt.

 

lBlumer, "Collective Selection," p. 277.

2Schrank, "Fashion Innovation," pp. 69-70, 77.

3Goodell, "Two Techniques,’ pp. 52, 65.

129

The attitudes and behaviors of the fashion Opinion
leaders appear to have been more consistent than those Of
the fashion innovators since fashion opinion leaders, who
influence others but do not adopt new styles themselves,
were also not found to be significantly different from the
rest of the subjects in their attitudes toward new styles.
This finding appears to support the prOposition developed
in the literature review that extreme differentiation in
clothing styles worn or promoted and the influential's
perception Of his ability to influence others may be
negatively related. Negative public reaction toward new
styles may, therefore, have some controlling effect on the
type Of styles worn and promoted by fashion Opinion
leaders.

Just as Jack and Schifferl found that fashion
designers and promoters must remain within certain bounds
or they will not be followed by the woman on the street,
the results Of this study seem to indicate that the fashion
Opinion leader must also not be too different from the rest
Of the group or they will not be followed. Apparently,
neither designers who attempt to dictate extreme new styles
nor prestigious fashion leaders who wear extreme new styles
will be followed.

The implications Of these findings may be Of

interest to designers, manufacturers and retailers who

 

1Jack and Schiffer, "Limits of Fashion," p. 738.

130

must Often make expensive readjustments in production or
take costly markdowns when consumers do not react favorably
toward a new textile or apparel design. Thus, the failure
Of fashion promoters and fashion leaders to Obtain
followers may subsequently result in higher prices for
consumers. Knowledge Of consumers' attitudes toward new
styles may assist fashion promoters to provide designs
which consumers need, want and can afford.

Evidence in the present study indicates that
socio-economic class level was not significantly related
to innovativeness as the "trickle-down" theory suggests.
Instead, the data lend support to King's counter theory
of the "trickle-across" nature Of fashion. Apparently
emulation and differentiation occurred within rather than
between socio-economic levels. Since favorableness toward
new styles also was not significantly related to socio-
economic class level, the reciprocal relationships which
seemed tO exist between certain prestigious fashion leaders
and group norms were apparently not affected by socio-
economic class level.

The reciprocal relationships which evidently
existed between certain Of the prestigious fashion leaders
and the rest Of the subjects may help to explain certain
aspects Of fashion change or what Blumer terms "the
historical continuity of fashion." If group norms are
able to exert some control on the attitudes held or styles

adopted by these prestigious leaders, then abrupt shifts

131

in fashion are not likely. Blumer stated that . . .
fashion innovators always have to consider the prevailing
fashion, if for no other reason than to depart from it or
to elaborate on it."1 The implications Of these findings
may be Of value to costume historians and others studying
fashion cycles or fashion trends. Perhaps the attitudes
and behaviors Of certain historical prestigious fashion
leaders, influential in fashion change, were related to
group norms on innovativeness.

While this study was confined to the investigation
Of fashion change, the implications of the findings may
have ramifications to other aspects Of human behavior.
For example, the findings concerning the reciprocal
controlling influence Of certain prestigious leaders and
group norms may be Of value to behavioral scientists as
they study social control.

Additional findings of the present study imply
that prestigious fashion leaders‘ attitudes and behavior
concerning new styles have no relationship to their leader-
ship roles in social organizations. Presumably the
committee chairmen, Officers, and presidents among the
respondents in the sample were able to carry out their
roles whether or not they conformed to group norms on
innovativeness. Since social participation showed a

slight positive relationship to fashion adoption and

 

lBlumer, "Collective Selection," p. 283.

132

fashion Opinion leadership but the relationship among these
variables was not significant, it is possible that a
refinement Of the measure might produce significant
results. The data indicated that social participation as
measured by leadership in clubs and organizations was
quite low. Since few respondents participated in clubs
and organizations the sample norm on formal social par-
ticipation may have been neutral or somewhat negative.
Differences between innovators and Opinion leaders on
social participation might have occurred if informal and
cosmopolitan social activities as well as formal social
participation were included in the measure.

In summary, it is evident that in order for a
comprehensive theory Of fashion to evolve, the concept of
emulation and differentiation must be further investigated.
A more detailed delineation of the reciprocal nature of
these forces is needed. Recommendations for further
research will be given in the following section Of the

summary.

Recommendations

 

In order tO make finer distinctions between the
three types Of prestigious fashion leaders and the fashion
non-leaders, a more refined measure Of fashion adoption
must be devised. Refinement Of the present Fashion
.Adoption measure could be accomplished through an analysis

of each of the three component parts, wearing, owning, and

133

time of adoption. These parts should be analyzed
separately in combination with all other variables. Other
ways to distinguish between innovators and Opinion leaders
would be to: (1) study a population known to have
unfavorable attitudes toward new styles, (2) collect data
on the prestigious fashion leaders immediately after the
new styles come out, and (3) study the population over a
period Of time.

An investigation Of the type Of clothing worn by
each Of the four sub-categories Of individuals might also
reveal differences among these groups on their attitudes
and innovativeness. Perhaps the innovators in this study
scored lower on favorableness toward new styles than did
the fashion dual leaders because they prefer faddish styles
(or short-lived fashions which were not included in the
Ineasure of fashion adoption since it was limited to garment
lengths and styles rather than garment details and
(accessories which might not have historical continuity
and therefore could be considered fads. A style list
similar to the one in the fashion adoption measure, but
consisting of accessories and other small items Of
apparel, should be developed and given to a group Of
respondents in conjunction with the fashion adoption
treasure in order to distinguish between sub-categories of

respondents who are fad leaders and those who are fashion

leaders.

134

Since a self-designating technique was used in
this study to establish mutually exclusive groups, it is
not known whether the fashion dual leaders were more
active verbal influentials than the fashion Opinion leaders
because their "followers" could not be sociometrically
determined. Future studies should attempt to validate the
present self-designating fashion Opinion leadership measure

through correlation with sociometric measures of fashion

 

opinion leadership and then additional studies of the
differential characteristics of fashion opinion leaders
and fashion dual leaders may be undertaken.

Additional analysis Of the present data could be
made by categorizing the respondents into five adopter
categories as the diffusion research suggests: (l)
innovators, (2) early adopters, (3) early majority, (4)
late majority, and (5) laggards. Perhaps when the inno-
vators are defined in this manner they would differ
significantly from the rest of the subjects on favorable-
ness toward new styles. When the four sub-groups were
established fashion dual leaders were distinguished from
fashion innovators by designating as innovators, those
individuals in the ipwg£_portion of the upper one-third Of
the scores on the fashion adoption measure. The character-
istics of innovators could be more precisely profiled if
the 22223.2 1/2 per cent of a large sample were defined

as innovators.

135

In order to investigate in greater detail the
differences between the sample norms and the attitudes Of
the individuals or the relationships between the theo-
retical concepts of emulation and differentiation, a more
flexible and precise means of measuring the sample norms
must be devised. The use Of measures of central tendency
to compare the norms Of a group with a prior established
neutral score on an attitude measure is valid only if the
balance Of the positive and negative statements or
positions on the measure are equal in direction and
intensity. For more valid results in future studies, the
effects Of the sample norm on innovativeness, perhaps
should be partialed out since the data in this study
seemed tO indicate that favorableness toward new styles
was an intervening variable.

Since the respondents names and home addresses
were retained from the student enrollment records it is
possible that a longitudinal study of these respondents
could be made. Research using similar measures could
attempt to determine whether an individual's innovative or
influential status and attitudes toward fashions vary over
time or vary with different garment styles. For example,
the new styles in this study were extreme in length: one,
"hot" pants, was quite short, and the other, mid-calf
lengths, quite long, in comparison to the mode which was
ankle length pants and mini skirts. If a less extreme

new style were introduced, such as skirts just below the

136

knee, individuals might have different attitudes toward
this new style than toward "hot" pants, midi skirts, and
gaucho pants.

Since a substantial number Of individuals in two
other universities were pre-tested this data could be
analyzed and compared with the findings concerning the
respondents in this study. Variation in geographic

location may reveal differences in the relationships among

 

the test variables and differences between group norms on
favorableness toward new styles.

The significant relationships found between
favorableness toward new styles, fashion adoption, and
fashion Opinion leadership suggest that additional studies
could be made concerning the relationship of these vari-
ables tO creativity and experimentation with clothing in
general.

The positive though non-significant relationships
between social participation and the other test variables
should be investigated with other populations using a
more general measure of social participation including
informal social activities and measures of cosmopoliteness
as well as formal organizational participation.

The measures used in this study could be further
refined and used in other investigations of the acceptance
of fashions by men or by women Of other ages than the
respondents in this study. Perhaps social participation

would have been related to the other test variables if a

 

137

younger group Of respondents, for example, had been
investigated.

Since the data has been collected concerning the
respondents' accuracy Of perception of new styles in terms
Of the groups perception Of new styles on the fashion-
cycle ranking index, the relationship between initial
awareness of the new styles and adoption of new styles or

favorableness toward new styles could be investigated.

 

 

 

BIBLIOGRAPHY

 

BIBLIOGRAPHY

Books

Bogardus, Emory S. Essentials Of Social Psychology.
Los Angeles: J. R. Miller Press, 1923.

 

Chapin, Francis 8. Experimental Designs in Sociological
Research. Rev. ed. New York} Harper, 1947.

 

Creekmore, Anna M. Methods Of Measuring Clothing Variables.
East Lansing: Michigan State University, Agri-
cultural Experiment Station, 1971.

 

Edwards, Allen L. Techniques Of Attitude Scale Con-
struction. New York: Appieton-Century-Crofts,
Inc., 1957.

 

Flugel, J. C. The Psychology Of Clothes. New York:
International Universities Press, Inc., 1969.

 

Jarnow, Jeanette and Judelle, Beatrice. Inside the
Fashion Business._ New York: John Wiley and Sons,
1965.

 

 

Katz, Elihu and Lazarsfeld, Paul. Personal Influence: The
Part Played by_PeOple in the Flow Of Mass Communi-
cations. New York: The Free Press, 1955.

 

 

Laver, James. Taste and Fashion. London: George Harrap
and CO., Ltd., 1937.

 

Miller, Delbert C. Handbook of Research Design and Social
Measurement. New York: David McKay Company, Inc.,
1964.

 

 

Nystrom, Paul H. Economics of Fashion. New York: Ronald
Press, 1928.

Rogers, Everett M. Diffusion of Innovations. New York:
The Free Press, 1962.

 

Siegel, Sidney. Nonparametric Statistics for the Be-
havioral SEiences. New York: McGraw-HilIiBook
Company, 1956.

 

 

138

 

139

Young, Agnes Brooks. Recurring Cycles of Fashion. New
York: Harper and Row, 1937.

 

Whyte, William H., Jr. The Organization Man. Garden City,
N.Y.: Doubleday and Company, Inc., 1957.

 

Articles

Beal, George M. and Rogers, Everett M. "Informational
Sources in the AdOption Process Of New Fabrics."
Journal Of Home Economics, XXXXIX, NO. 8 (October,
1957), 630-34.

 

Blumer, Herbert. "Fashion: .From Class Differentiation to
Collective Selection." Sociological Quarterly, X

 

 

Bogardus, Emory S. "Social Psychology of Fads." Journal
of Applied Sociology, VIII (1923), 239-43.

 

Jack, Nancy K. and Schiffer, Betty. "The Limits of Fashion
Control." American Sociological Review, XIII
(1948), 730-38. -

 

Janney, J. E. "Fad and Fashion Leadership Among Under-
graduate Women." Journal Of Abnormal and Social
Psychology, XXXVI (1941), 275-78.

 

 

King, Charles W. "Fashion Adoption: A Rebuttal to the
'Trickle-Down' Theory." Proceedings Of the Winter
Conference Of the American Marketing Association.
Boston, Mass., December 27-28, 1963, 108-25.

 

 

Marascuilo, Leonard A. "Large-Sample Multiple Comparisons."
Psychological Bulletin, LXV, NO. 5 (1966), 281-84.

 

Robertson, Thomas S. and Myers, James H. "Personality
Correlates of Opinion Leadership and Innovative
Buying Behavior." Journal of Marketing Research,
VI (May, 1969), 164-68.

 

Rogers, Everett M. and Cartano, David G. "Methods Of
Measuring Opinion Leadership." Public Opinion
Quarterly, XXVI (Fall, 1962), 435-41.

 

 

140

Sapir, Edward. "Fashion." Encyclgpaedia Of the Social
Sciences, VI. New York: The MacMillan Company,

~

Schlater, Jean D., Magrabi, Frances M., and Eicher,
Joanne B. "Social Science Methodology." Journal
of Home Economics, LV, 6 (June, 1963), 423-27.

 

Simmel, Georg. "Fashion." Reprinted from the Inter-
national Quarterly, X (October, 1904), I30-55.
American Journai Of Sociology, LXII, NO. 6 (May,
1957), 541-58.

 

 

Summers, John O. "The Identity Of WOmen's Clothing
Fashion Opinion Leaders." Journal Of Marketing
Research, VII (May, 1970), 178-85.

 

Warden, Jessie. "Some Desires or Goals for Clothing Of
College Women." Journal of Home Economics, XLIX
No. 10 (December,_i957), 795-96.

 

Wasson, Chester R. "How Predictable Are Fashion and Other
Product Life Cycles?" Journal of Marketing, XXXII
(July, 1968), 36-43.

 

Unpublished Materials

 

Beal, George M. and Bohlen, Joe M. The Diffusion Process.
Special Report NO. 18, Ames, Iowa: Iowa State
University Of Science and Technology, Cooperative
Extension Service, 1962.

 

Freedle, Johnnie Alice Denton. "Clothing Interest and
Social Participation as Related to Clothing
Selection and Buying Processes Of the College
Woman." Unpublished Master's thesis, University
of Tennessee, 1968.

Glickman, Albert S. "Clothing Leadership Among Boys."
Unpublished Ph.D. dissertation, Ohio State Uni-
versity, 1952.

Goodell, Anne Stubenrauch. "Comparison of Two Techniques
for the Identification of Fashion Leaders."
Unpublished Master's thesis, Ohio State University,
1967.

Grindereng, Margaret Pauline. "Fashion Diffusion: A Study
by Price Range of Style Dispersion and Style ’
Leadership." Unpublished Ph.D. dissertation, Ohio
State University, 1965.

141

Holley, Zelda. "Opinions of University Women Regarding
the Relative Importance of Thermal Comfort,
Conformity and Fashion." Unpublished Master's
thesis, Kansas State University, 1970.

Johnson, Judith. "An Exploratory Study Of Diffusion of
Fashions with Mothers and Teen-Age Daughters."
Unpublished Master's thesis, University of
Nebraska, 1967.

Johnson, Pamela A. "Fashion Awareness and Its Relationship
to Selected Clothing Consumption Variables."
Unpublished research paper in Clothing Consumption,
Michigan State University, 1971.

Marshall, Eileen. "Leadership in Men's Fashions Associated
with Selected Social Characteristics for a Group of
Fraternity Men." Unpublished Master's thesis,
Pennsylvania State University, 1964.

McGuire, Carson and White, George D. "The Measurement of
Social Status." Unpublished research paper in
Human Development, No. 3, revised, The University
Of Texas, 1963.

Moore, Kathleen Anne. "Fashion Leadership Designation and
Related Factors Among a Group of Adolescent Girls."
Unpublished Master's thesis, Pennsylvania State
University, 1962.

Pasnak, Mary Francis Drake. "Fashion Innovators Compared
With Non-Innovators on Clothing Attitudes, Self-
Actualization, and Tolerance of Ambiguity."
Unpublished Ph.D. dissertation, Pennsylvania State
University, 1968.

Perkins, Olive. "An Investigation of the Clothing Fads
and Fashions of a Group of Freshmen College Women
and the Reasons for Accepting or Rejecting Some
Popular Campus Clothes." Unpublished Master's
thesis, Cornell University, 1958.

Roth, R. Jane Rudy. "Clothing Conformity and Fraternity
Men's Attitudes Toward Current Male Fashion Trends."
Unpublished Master's thesis, Ohio State University,
1969.

Schrank, Holly Lois. "Fashion Innovativeness and Fashion
Opinion Leadership as Related to Social Insecurity,
Attitudes Toward Conformity, Clothing Interest and
Socioeconomic Level." Unpublished Ph.D. disser-
tation, Ohio State University, 1970.

142

Sharpe, Elizabeth Susan. "Development of a Clothing
Interest-and-Importance Scale." Unpublished
Master's thesis, Ohio State Univer51ty, 1963.

Snow, Janet B. "Clothing Interest of Men in Four Selected
Occupations." Unpublished Master's thesis, Texas
Woman's University, 1969.

Sohn, Marjorie Ann. "Personal-Social Characteristics of
Clothing Fashion Leaders Among Fraternity Men."
Unpublished Master's thesis, Pennsylvania State
University, 1959.

Sproles, George Bryan. "A Profile Analysis of the Durable
Press Clothing Information Communicator." Un-
published Master's thesis, Purdue University, 1968.

Tousignant, Rosanna P. "Fashion Acceptance by Selected
College Students." Unpublished Master's thesis,
University Of Rhode Island, 1959.

Van Staden, Francine Johanna. "The Relationship of
Prestigious Clothing to Acceptance by the Peer
Group Of Adolescent Boys and Girls." Unpublished
Master's thesis, Michigan State University, 1970.

APPENDICES

APPENDIX A

Questionnaire

M I C H I G A N S T A T E U N I V E R S I T Y East Lansing . Michigan

 

College of Human Ecology
Department of Human Envlronment and Deslqn

February 26. 1971

Dear Student:

You have been selected from among the Michigan State University coeds to
help with our survey about university women and their clothing.

Please do not write your name on the accompanying questionnaire. The
number on the questionnaire insures your anonymity.

The questionnaire will take approximately 15 minutes to complete.

Before you answer any questions, read the directions for each section care-
fully. Do not deliberate but put down the answer which first comes to mind.

Be sure to respond to every question.

Since time is important, I must ask you to please return your completed
Questionnaire in the enclosed postage-paid envelOpe by Hatch S, 1971.

Sincerely,

Cebu-{L ”777,372,889

Carol Myers

Graduate Research Assistant

Textiles and Clothing

attachment: 7-page questionnaire

143

144

List below any school, church or community organizations, clubs or groups to
which you have belonged during the school year 1970-71. Do not include those in
your home community unless you live at home.

Indicate your position in each of those groups by placing a checkmark (V ) in
the categories on the right.

I

 

 

 

 

 

N
O
3 a g Elected officer 3
u 3 u 3 3 (other than 5,:
Name of Organization " ”I 3 3: E president) 3
‘8, E3 “0 Write name of u
:8 U 5 0 position. a.
l. i
2.
i:
e:
S.* [

 

 

 

 

 

 

[
tr

 

*If you need more space you may add to the back.

Please rank the following items of clothing as to where you think they cur-
rently fall on the fashion cycle at Michigan State University.

Item:

Fashion Cycle Ranking

 

A.__

Hottest
Item

Fairly
New

Everyonés
Wearing

On Its
Way Out

Outdated

 

l. knickers r

 

2. mini skirt

 

 

3. bell bottoms

 

4. midi skirt

 

5. gaucho pants
(mid-calf length)

 

6. maxi skirt

 

7. pantsuit

 

0. thigh length
vests

 

9. hot pants
(short shorts)

 

 

10. jumpsuit

 

 

 

 

 

 

145

Listed below are seven categories of garment lengths. In order for me to establish
the popularity of these lengths, please estimaig_the following;

1. The number of garments in each category which you have either bought new or
gggg_since January 1, 1968. Omit all garments which you may have received
as gifts or hand-me-downs. Also omit all outer garments such as capes or
costs.

2. giggk out the frquency which you wear each of the garment lengths. Use the
following guide;

I Frequently (approximately twice weekly)
Regularly (weekly)

Occasionally (monthly)

Seldom (twice yearly)

Never (not at all)

A

ZmO'N'd
I

Number Bought New

 

 

 

 

 

 

 

 

 

or Sewn

EZéfliEE: Since Jan. 1, '68 Frequency Horn
4. Garment Length: near knee-cap

a. skirt (knee- length suit, dress, etc.) :5 F B 0 5 N

b. pants (knickers, etc.) I F i 0 ‘5 N
l. Garment Length: about 8" above knee

a. skirt (micro-mini dress, suit, skirt, etc.) - F R 0 S N

b. pants (short shorts, hot pants, etc.) w_n“"_r F R 0 S N
2. Garment Length: about 5” above knee

a. skirt (mini skirt, dress, suit, etc.) F R 0 S N

b. Pants (shorts, culottes, etc.) F R 0 S N
3. Garment Length: about 2” above knee

.a. skirt (just above-the-knee dress, suit, etc.) F R 0 S N

b. pants (culottes. etc.) F R 0 S N
4. Garment Length: near knee-cap

a. skirt (knee-length suit, dress, etc.) F R O S N

b. pants (knickers, etc.) F R 0 S N
5. Garment Length: about 2” below knee

a. skirt (just below-the-knee dress, suit, etc.) F R 0 S N

b. pants (peddle-pushers, etc.) . F R O S N
6. Garment Length: near mid-calf

a. skirt (midi-length dress, suit, skirt, etc.) F R. O N

b. pants (gauchos, etc.) F l O S N
7. garment Iennth: near ankle and below

a. skirt (maxi- length dress, suit, slzirt, etc. ) __. ' F R O S, N

b. pants (slacks, etc.) F R O S N

c. jumpsuit, overalls, etc. _ F R O S N

d. pantsuit (ensemble purchased to go together) __ F R O S N

I.--

146

8. Please indicate approximately when you first acquired the following garment
lengths and whether of not you have discontinued wearing any of the lengths.
If you have never owned any of the lengths, check the last column only.

 

 

 

When First Acquired Discontinued Own None
wearing
a. Mid-calf length
l.) skirt .___. ————
’(month 8 year)
L)pmms

 

(month 6 year)

b. AHpkle-length
l.) skirt

 

(month & year)

2 .) pantsuit

 

(month 6 year)

3.) jumpsuit, overalls

 

(month 6 year)

9. If you indicated that you discontinued wearing any of the garment lengths above,
please indicate why below;

 

 

 

10. Make one checkmark below to indicate when you adopt new clothing styles:

Much earlier than most students

Somewhat earlier than most students

About the same time as most students

Somewhat later than most students

Much later than most students

11. Make one checkmark below to indicate how you believe you are dressed compared to
most students on the campus:
More fashionably dressed
About the same

Less fashionably dressed

Please read the following statements about clothing. Rate each according to the
extent to which you believe the statement is true or not true.

guide; black out your response.

l:

DT Definitely True

147

PT 3 Partially True, more true than false

U a Undecided, Uncertain

PF = Partially False, more false than true

DF Definitely False

Use the following

 

10.

11.

12.

13.

I generally don't pass along fashion
information to Others.

Fashion holds a low priority as a topic
of conversation among my friends.

Others consult me for information about
the latest fashion trends.

I believe I am a very good source of
advice about fashion.

People talk too much about fashion.
I never borrow or lend fashion magazines.

My friends ask for my opinions about

-new styles.

I am more likely than most of my friends
to be asked for advice about fashion.

I do more listening than talking during
conversations about fashion.

When it comes to fashion, I am among the
least likely of my friends to be thought
of as an advice giver.

It is important to share one's opinions
about the new styles with others.

My friends don't think of me as a
knowledgeable source of information
about fashion trends..

I recently convinced someone to change
an aspect of her appearance to something
more fashionable.

DT

DT

DT

DT

DT

DT

DT

DT

DT

DT

DT

DT

DT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PF

PF

PF

PF

PF

PF

PF

PF

PF

PF

PF

PF

PF

DF

DF

DF

DF

DF

DF

DF

DF

DF

DF

DF

DF

DFJ

148

DT - Definitely True

PT = Partially True, more true than false
U = Undecided, Uncertain

PF = Partially False, more false than true

DP = Definitely False

 

14.

15.

16.

17.

18.

19.

20.

_.- w‘"

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

I believe in sharing with others what I
know about trends in fashion.

I enjoy discussing fashion.

People bypass me as a source of advice about fashion.

I dislike discussing clothes and fashion.

I like to help others make decisions about fashion.

I am never first to be asked for an opinion about
a current style.

I enjoy being asked about fashion trends.
1 h‘ "h-«s...

I look forward to the changes in women's
fashions each season.

 

I like the gaucho (mid-calf pants) look.

Wearing the latest fashions would make me
feel conspicuous and uncomfortable.

I dislike the new longer skirt lengths.

Wearing the newest fashions is stimulating and
exciting.

I believe midi (mid-calf length) skirts are
never going to be pOpular.

I enjoy being the first to wear a new clothing
style.

Women's fashions change too often for my liking.
I think long skirts make women look dowdy.

I think that the mid-calf lengths look ridiculous
on most women.

I prefer to wear clothing which stays in style
for several years.

I think long skirts are comfortable to wear.

DT

DT

DT

DT

DT

DT

DT

DT

DT

DT

DT

DT

DT

DT

'DT

DT

DT

DT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PT

PF

PF

PF

PF

PF

PF

PF

PF

PF

PF

PF

PF

PF

PF

PF

PF

PF

PF

PF

DF

DF

DF
DF

DF
DF
DF
DF
DF
DF
DF

DF
DF

DF
DF
DF

DF

DF

DF

149

 

 

 

 

 

 

 

 

 

 

 

DT I Definitely True
PT = Partially True, more true than false
U = Undecided, Uncertain
PF = Partially False, more false than true
DF 8 Definitely False
33. I readily accept new fashion trends. . DT PT U PF DF
34. I plan never to buy a midi skirt. . DT PT U PF DF
35. I think the new longer lengths make women appear DT PT U PF DF
sore feminine.
36. I prefer skirts which are above-the-knee. DT PT U PF DF
37. I believe maxi skirts are impractical. DT PT U PF DF
38. Hearing the latest fashions is important to me. DT PT U PF DF
39. I do not think long skirts are economical. DT PT U PF DF
40. I like to be considered one of the most DT PT U PF DF
fashionably dressed coeds.
BACKGROUND INFORHATION
1. Please indicate the main wage earner in your family.
father
mother
other (please specify)
(example; stepfather, uncle, brother)
2. Please indicate the source of income for the major wage earner in your family.
a) wages, hourly wages (weekly paycheck)
b) profits and fees from a business or profession
c) salary paid on a monthly basis
d) social security or unemployment insurance
e) odd jobs, irregular work, seasonal work
f) if other, please explain
3. Please explain in detail what the main wage earner does at work. Please explain

 

specifically type of work. Examples; salesman in a clothing store, waiter, man-
ages 20 other workers in an office, works on the assembly-line, owns and manages
a small store with six employees.

 

 

 

 

 

150

Does any other person contribute to the financial support of your family?

yes no

 

If yes, please explain who (mother, brother, uncle, etc.).

 

If yes, please explain in detail the type of work done by this person.

._ _. ¢-——--

Please indicate the source of income for the second person who contributes to your
family's financial support.

 

a) wages, hourly wages (weekly paycheck)

b) profits and fees from a business or profession
c) salary paid on a monthly basis

d) social security or unemployment insurance

e) odd jobs, irregular work, seasonal work

f) if other, please explain

 

 

 

Please indicate the highest level of education achieved by each of the following.

father

 

mother

main wage earner (if other than mother or father)

 

a) finished 7th grade or lower
b) finished 8th grade

c) finished 9th grade

d) finished 10th or 11th grade
e) graduated from high school

f) one to three years of c llege
g) college graduate

h) graduate school after college
i) don't know

If the main wage earner is a college graduate, what is the highest degree he t Ids?

 

Please glance back over the questionnaire

to make sure you have not omitted any items.

APPENDIX B

Fashion Curves

 

Pantsuits
Midi skirts ----------
Maxi skirts ..._....__
Jumpsuits .___.___a_
Gaucho pants..........

CD

5* 201

O

>

tH

O

U)

C:

O

'3

o 15

CD

U)

t—i

as

H

H

O

4.)

:3

m 10 Coo-00..

C‘.

H

U)

4.)

C

(D

5

LI

8 5

1H

O

H

{25

2 /
o

 

 

 

3/ 9/ 3/ ' 9/ 3/ 9/ 3/ 9/ 3/ 9/
66 66 67 67 68 68 69 69 70 7O

Semi-annual Months Reviewed

Figure B-l.--Fashion curves for styles appearing in

151

152

 

Pantsuits
Midi Skirts ————————————
Maxi skirts...i__,,,

 

Jumpsuits . . .
Gaucho pants............

     

 

 

 

(D

D

81

',>
“4

O

U) 301

‘1 '\
.3 ’—---\ H

U I, \ l \

O \ '\

o I, ‘ 1 "

\

H I \ I ‘ :- ‘ \

g l ‘ 1" :‘ ' 1

u ‘ O .

L4 20~ / \ , \ :1 t \

4) ’ \ I ‘ : ' '

-.-l I \ I \ . z ' \

a I \‘ I ‘ .' g. ' “

I U V 2' \ "\
c ’ l t .1 ‘ x’ c
-:-i I .4 1 \ , \‘
I 1 1‘. \

'2 10. I :\ '1 \
0) P . .' I

E A I /°\. 5

m ’ ~ -

w l j .

Q4 .. ,0. .0.

o a . .

H I . . \\." '. : )
(D "fl.o ' '. : .b k
.2 O 1 ;‘ ‘L‘ . '- / I ...... ‘-‘.',O )
:3

Z

3/ 47 5/ 6/ i) d/ 6/ 16/ 11/ 12/ 1/ 2/
70 7o 70 70 7o 70 70 70 7o 70 71 71

Months Reviewed

Figure B—2.--Fashion curves for styles appearing in
Vogue, 1970-71.

153

.Haumema .msmun umfla mHsUm ooh mm>uso cosgmmm--.mam magmas

sensuous mo ease:

as as on on op as as as me we we we mm
\m .\NH \m \m \m \NH \m \s \m \NH \m \s \m

 

 

 

    

...Il...limusmm onosmu
.ll.ll.ll. mumammfioh
muuflxm axe:
uuuuuuuuuu muwflxm Hoax
.......... mumomusmm

 

 

satfiqs buridopv squspuodssg go Jaqmnu

"mmmm