\WWIIWWUHlWlllfilW\|\l|HWWHIH\le

)

  

   

108
087
THS

[go I679‘

" MICHIGAN STATE UNI VER

II II III III IIIIIIIIIIIIIIIIIIIIIIII

293 00543 6104
LIBRARY I

Michigan State
University

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the
thesis entitled

The effect of participation in quality circles
on job involvement, productivity and satisfaction.

presented by

Dawn Marie Eskew

has been accepted towards fulfillment
of the requirements for

Masters degree in Communication-

 

 

/

”19-4 . "

Major pro or

Date zflZ/QQ

0-7 639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

 

 

 

 

 

 

 

 

MSU

 

 

 

RETURNING MATERIALS:
Place in book drop to

 

LJBRARJES remove this checkout from
.-:—. your record. FINES will
be charged if book is
returned after the date
stamped below.
Harlan 1%
3m (0

 

 

 

THE EFFECT OF QUALITY CIRCLE PARTICIPATION ON
JOB INVOLVEMENT, PRODUCTIVITY AND SATISFACTION

By

Dawn Marie Eskew

A THESIS

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

MASTER OF SCIENCE

Department of Communication

1988

ABSTRACT

THE EFFECT OF QUALITY CIRCLE PARTICIPATION ON JOB
INVOLVEMENT, PRODUCTIVITY AND SATISFACTION.

By

Dawn Marie Eskew

The effect of participation in a quality circle on job involvement,
productivity and job satisfaction was investigated using meta analytic
techniques. Three hypotheses were tested, 1) Those who participate in quality
circle programs should become involved in the task, 2) Participation in quality
circles should lead to higher organizational productivity rates and 3)
Participation in quality circles should lead to higher levels of job satisfaction.
Forty-three studies were obtained from the literature search, only thirteen met
the criteria for inclusion. The point bi-serial correlation r was used to
calculate effect sizes (Hunter, Schmidt & Jackson; 1982). Effect sizes and
corrections for attenuation due to measurement error were computed for the
thirteen studies. A weighted average effect size was derived for each
dependent variable. The results of the meta analysis indicate that
participation in quality circles programs has no effect on job involvement or
job satisfaction. Not enough data was available to calculate effect size for

productivity.

ACKNOWLEDGMENTS

This is to acknowledge the encouragement of my parents, who have always
taught me that I can achieve anything I set out to.

I also wish to acknowledge the support and understanding of my husband,
who at times was pushed into the background so I might complete a seemingly
never ending task.

I also would like to acknowledge the help of my boss and friend, Jim, who
did not realize it at the time, but allowed me to complete this project during
working hours.

I am also indebted to the patience and guidance of Professor James Stiff
and Professor Kathy Miller. Without their help and hard work I would have

never seen the light at the end of the tunnel.

iii

TABLE OF CONTENTS

Introduction ............. 1
Methods ............... 21
Results ............... 31
Discussion .............. 37
Conclusion .............. 46
References .............. 49

iv

LIST OF TABLES

Table l ............... 8
Table 2 ............... 14
Table 3 ............... 23
Table 4 ............... 25
Table 5 ............... 33
Table 6 ............... 34
Table 7 ............... 36
Table 8 ............... 38

INTRODUCTION

Growth in American productivity has been on the decline for some years
now. Ferris & Wagner (1985) reported that productivity grew about 3% each
year during the 1960’s and 1.4% in the 1970’s. Since that time, however, the
growth rate remained stable, resulting in America to fall to the sixth position
out of the seven leading industrial nations. As a result of this lagging
productivity rate, American businesses began looking toward Japanese
management techniques to help improve productivity. One technique in
particular was the use of quality control circles. Mohrman & Novelli (1985)
state that the purpose of quality circle programs are ”...to create a
participative forum in which workers can generate solutions to workplace
problems which can enhance quality, productivity and other desired outcomes”.

Operationally, quality circles consist of 6-10 individual’s who work
together and meet approximately once each week for one hour, to discuss
topics ranging from the cosmetic makeup of the organization to productivity
concerns. For the sake of union harmony, topics dealing with the labor-
relations contracts are prohibited areas of discussion. Meaning that no
discussions can take place in the quality circles that would normally be dealt
with during negotiations or by union officials (e.g., grievances and pay rates).
The circle members do not have the power to implement the ideas they come
up with. Their suggestions typically go to a steering committee which is in
charge of the implementation of approved ideas. Although the number of
members and the frequency of meetings may vary from organization to
organization, the concept behind quality circles remains in tact; to improve the

functioning of the organization by tapping the knowledge of those who know

the most about the job, the employees.

Even though quality circles were introduced to Japan by American
statisticians, they are still considered a Japanese innovation. During the 1950’s
Dr. William Edward Deming began lecturing on statistical quality control in
Japan. This occurred as a result of the United States attempting to help
rebuild the Japanese industry. Deming’s work had such a great impact on the
Japanese society, that the Japanese Engineering Society (JES) was formally
adopted. Another American would also have a great impact on Japan, his name
was Dr. J. Juran. Drs. Deming and Juran lectured on ways to help improve
product quality and make all levels of the organization accept some form of
responsibility when dealing with the quality of the organization’s products.
Their ideas spread quickly, and soon Japanese organizations had included not
only top and middle management in the training of quality control techniques,
but also all their line workers. All members of the organization were trained
in how to utilize statistical quality control. Since its humble beginnings, the
number of'quality circles in Japan have increased dramatically. Cole (1980)
reports an astonishing figure indicating that one out of every eight Japanese
employees in 1978 were involved in a quality circle, almost four million
employees. The spread of quality circles reached the United States,
approximately twenty years after the circles were introduced to the Japanese
society.

Lockheed Missile and Space formed the first quality circle program in the
United States during the 1970’s. Lockheed reported 15 functioning circles by
the end of 1975. These circles were credited with suggestions which produced
documented savings of $2,844,000 during their first two years of existence

(Yager, 1979). This figure is six times greater than the costs associated with

3
the implementation of the suggestions and the maintenance of the circles
(Cole, 1980). Unfortunately, this success was short lived, the quality circle
program at Lockheed ended during the late 1970’s. Cole (1980) stated a
potential reason for this failure may have been due as a result of the labor
union. ”...When a strike occurred (at Lockheed) and the workers and union did
not receive what they felt was their due at the end of the strike, they
responded by reducing their participation in the circles... loss of key personnel
and failure to institutionalize quality control circles were extremely significant
in contributing to the decline of circle acrivity at Lockheed."

While Lockheed was among the first to implement quality circles in the
private sector, Norfolk Naval Shipyard was among the first in the federal
government to use them. It has been one of the most successful programs
among the federal agencies (Bryant & Kearns, 1981). In a pilot study
conducted during 1979, the shipyard instituted nine quality circles. These nine
circles were reported to save $7 million in costs. As a result of this dramatic
savings, the shipyard increased the number of circles to 44. Suggestions
offered by the new circles resulted in an additional $214,990 cost savings,
which produced a four to one return on their investment. Within three years,
Norfolk planned to increase the number of quality circles to 500.

There have been many other organizations which have implemented quality
circles (e.g., General Motors, Honeywell Corporation, Marietta Aerospace).
Lockheed and the Norfolk Naval Shipyard are two organizations which had
major impact on the proliferation of quality work circles in both the private
sector and the federal government. Since their introduction to the United
States, the number of quality circle programs have increased tremendously. In

1982, Seelye & Sween reported that of 828 responding organizations, only 115

4
reported not having any quality circle program functioning. Of the remaining
713 locations (each reporting two plants) a total of 12,424 quality circles were
identified as currently functioning. Today that figure has probably doubled in
size.

Yager (1982) states that quality circles may just be a fad of the 1980’s.
Quality circle proponents argue that it is much more than a fad, it is a ”trend
that is permanently changing managerial assumptions and practices in the
United States.” (Marks, 1986).

One potential reason the use of quality circles has continued to increase,
may be due to the apparent ease of implementing them in the organization.
There are many proposed reasons for the high failure rate of quality circles in
the United States. Mohrman & Novelli (1985) indicate that in Japan, quality
circles are an extension of the climate, whereas in the 0.5., they are
introduced into climates unfamiliar with participation and openness. Ruffner &
Ettkin (1987) also support this hypothesis. They state that quality circles may
not have been adapted adequately enough to our own culture to reap the same
benefits the Japanese organizations did. Steel & Shane (1986) indicate that the
cultural differences may not be the only reason for the high failure rate.

They note that very little empirical research has been done to measure the
effectiveness of quality circles and go on to state that much of the research
conducted is unusable as a result of flawed and often misleading evaluations.
This may, in part, be due to design problems, but may also be attributed to
factors inherent in quality circle programs. They take time to develop and
reap benefits (Steel & Shane. 1986). This is precisely why Shores & Thompson
(1986) strongly emphasize the need for evaluation. If quality circles take time

to show their true colors, then managers had better find someway to determine

5
whether they are going to work in their environment before investing heavily.
Although Steel & Shane (1986) state the problem is with lack of good research
and evaluation, Rieker & Sullivan (1981) indicate the problem lies in having no
proper forms of evaluation. This was also found to be true in a report
published by the General Accounting Office in 1981 which stated that currently
no systematic method of measurement or evaluation is in existence for
determining the impact of quality circle programs. Rieker & Sullivan (1981)
suggest it could take years and thousands of cases before the effectiveness of
quality circles could be evaluated, taking all possible variables into account.
Although, they too, believe that if given time, the benefits of quality circles
will surface on their own. Time of operation may be a factor that has
influenced the effective evaluation of quality circle programs. Some studies
compare programs that have been established quite a while (e.g., 13 months)
with relatively new programs (e.g., 6 months). The differences in effectiveness
due to the fact that one program is more advanced and established than the
other, (Steel et. al., 1985), indicate that quality circle programs may need to
be in operation for a certain length of time before any benefits can be reaped
from them.

The original purpose of quality circle programs was to offer a method of
quality control that improved and enhanced productivity. Currently, circle
programs are also credited with improving the quality of working life in the
organization’s which implement them. This new dimension, or outcome of
quality circle programs may be the reason evaluation has become so difficult.
There are many aspects that contribute to quality of work life, such as
motivation and satisfaction. Although many measures have been developed to

tap these aspects of the working environment, no one measure is consistently

6
used. It is also difficult to determine whether the measure being used is
actually tapping the concept it is suppose to measure. These are problems
that do not appear as frequently when measuring productivity.

Many glowing testimonials may be found in the literature referring to how
the implementation of a quality circle program saved the organization from
financial disaster and employee upheaval (Landon & Moulton, 1986; Marks,
1986; Perrill, 1984; Sikes, Connell & Donovan, 1980; Tortorich et.al., 1981). As
Mohrman & Novelli (1985) state, very few empirical studies have been done to
confirm these wonderful outcomes. However, even without the statistics to
back them up, quality circle programs are treated as the wonder cure for
today’s organizations.

It is time to asses the validity of this belief. This paper will examine how
quality circle participation affects job involvement, productivity and job
satisfaction. This will be done through a conceptual analysis of participation
and reviewing the pertinent literature on participative decision making. By
examining effects predicted from the participative decision making literature
we should be able to predict the effects of quality circle participation. The
literature dealing with job involvement will also be conceptually defined to
help clarify the issue. The productivity literature will be reviewed, along with
its relationship to quality circle programs. The results of a meta analysis of
all the empirical work done on quality circles will then be presented and
analyzed.

Part of the problem with the quality circle literature is understanding how
quality circles impact the workers. Since quality circles are a type of
participative decision making, it only makes sense to first look at the

participative decision making literature.

P i i i n

As a participation technique, quality circles have the capacity to improve
employee morale and performance, although as Locke & Schweiger (1979) and
Miller & Monge (1986) indicated, the support for this type of outcome is
limited. It is difficult to attribute any positive outcomes just to participation,
as there are many other factors that may be contributing to the outcomes
(Locke & Schweiger, 1979; Whyte, 1987). One potential problem is determining
what participation really is. Much attention has been paid to it, but few
attempts have been made to conceptually define it (Dachler & Wilpert, 1978;
French, Israel & As, 1960; Lowin, 1968; Rosenberg & Rosenstein, 1980; Scholl,
1976; Singer, 1974). For a listing of the conceptual definitions refer to
Table l.

The purpose of this section of the paper will be to adopt a standard
conceptual definition of participation. The analysis will proceed with an
explanation of the usefulness and meaning of the definition through analyzing
the various components making up the definition.

Participation refers to a transactional process in which employees exchange
information perceived to influence future decisions that may affect their own
job and/or the functioning of the organization. The usefulness and meaning of
this definition stem from the combination of the following components.

For instance, by indicating that the process is transactional, we can
eliminate the problems Lowin (1968) and Rosenberg & Rosenstein (1980) faced.
Lowin (1968) implies that decisions regarding activities are made, but never
identifies the decision as requiring any interaction among individuals. From
his definition, one could imply that an employee may make a decision to try

out a new procedure in his or her department. In doing so, the employee

Influence

Consulted prior to
decision

Decisions effect the
individual

Exchange information

Transactional
process

Perceptions

Table 1.

Key Components of Participation

Implies the individual has an impact on the events
occurring around him or her.

Individual’s input requested before a decision
is made, to assure s/he has maximum impact on the
decision.

Individual contributes to decisions that will have
an impact on his or her job. Provides
meaningfulness.

Information can be exchanged through a variety of
methods and modes. Allows for traditional
participation (face to face) and non-traditional
participation (written memo’s).

Indicates that all individuals involved in the
process are affected by and contribute to the
participation.

Identifies the importance of not what actually
occurs but of what the employee believes occurs.’

9

never consults anyone else, and makes that decision on his or her own.
According to Lowin’s definition, that process could be considered participation.
Rosenberg & Rosenstein (1980), however, never indicate what type of
interaction is taking place, nor how they would define it if it were included.
According to their definition, an employee suggestion-box would be sufficient
for participation. Granted, suggestions may be a small aspect of participation,
but they are not the only form of participation that should be included in the
definition. Through a transactional process all individuals involved, contribute
and get something out of the exchange. This characteristic was left out of
many of the earlier definitions of participation (Dachler & Wilpert, 1978;
Lowin, I968; Rosenberg & Rosenstein, 1980). In this definition of participation,
interaction among employees is considered essential. It also allows for input
made from employees regardless of how it is exchanged (e.g., in the form of
Opinions or suggestions or through formalized involvement programs). Singer
(1974) implies participation does not occur unless two or more individuals
engage jointly in that decision. This excludes from the definition those
individuals who are asked by their superiors to draw up a proposal" and turn it
in. The employee has input into the decision, and is making a contribution to
the final outcome. According to Singer’s definition, this is not considered
participation. The problem that Singer (1974) encounters with the limiting
component of joint decisionvmaking, poses no threat to this definition since
the exchange of information can include a wide range of interactions.
Similarly, the weakness Scholl (1976) faces, stems primarily from not taking
the definition far enough and applying it to participation.

By including an individual’s perceptions, this definition can avoid the

problems mentioned in French et. al. ’5 (1960) definition dealing with the

10
importance of actual influence needing to occur. Actual influence is not
necessary, all that is necessary is that the individual believes they have some
influence. Whyte (1987) points out that there is no way to determine whether
participation has been measured objectively. Participants may respond that
they felt they were participating, but this is a very subjective measure. The
perceptions of individuals are also much easier to obtain, since all one needs
to do is ask them what they feel is occurring. Just because people feel and
may at times actually participate, does not mean that their attitudes about
management have changed (Sorensen, Head & Stotz, 1985; Bradley & Hill, 1987),
which is one of the key outcomes of participation that Henderson & Goode
(1984) mention.

Another important component included in the definition of participation is
based on the idea of input into future decisions. This implies that the
employee’s input is canvassed before a decision is made and that the employee
has a chance to have some impact on the final outcome. This component was
alluded to by French, Isreal & As (1960) in their definition of participation.
However, unlike French, Isreal & As’ definition, this definition'does not limit
the participation to influencing only those making the decision.

In the next important component, a distinction between productive and
counter-productive behaviors that may take place in the organization is
identified. A counter-productive behavior, for instance, may involve two
employee’s whose only interaction comes through verbally assaulting each
other. This interaction cannot be considered participation since it will not
have an impact on the functioning of the organization or the job of either
individual.

Now that a basic understanding exists as to how participation will be

ll

referred to in this paper, we will move on to a general discussion of the
outcomes of participative decision making techniques (PDM) and more
specifically, quality circles.
P i ' i n if i it

As was mentioned earlier, the Locke & Schweiger (1979) review was one of
the first to analyze the effects of PDM techniques. After a review of the
empirical literature, they concluded that the information from experimental
field studies was inconclusive, since so many variables were involved.
However, laboratory, correlational field and univariate experimental field
studies resulted in slightly larger increases in satisfaction with no real
increase in productivity when compared to other types of techniques or styles.

Miller & Monge (1986) produced the second major critique of the PDM
literature. They examined all empirical work on PDM using meta analytic
procedures (Hunter, Schmidt & Jackson, 1982). Their results coincided with
those of Locke & Schweiger (1979). After conducting the meta analysis, it was
found that participation had a slightly stronger effect on satisfaction than on
productivity. Although participation still did not have as strong of an impact
on satisfaction that many indicated. Miller & Monge (1986) found that by
breaking participation down into subgroups, more information could be
obtained. When they looked at the effects of perceived participation focused
on a specific issue, a smaller effect was found for satisfaction than in the
other subgroups.

The most recent review on employee participation comes from Cotton,
Vollrath, Froggatt, Lengnick-Hall & Jennings (1988). They disagree with Locke
& Schweiger (1979) in that they believe that participation should be broken

down into a variety of forms and then tested regarding satisfaction and

12
performance levels. They break employee participation down into six
categories: participation in work decisions, consultative participation, short-
term participation, informal participation employee ownership and finally,
representative participation. They believe that a particular form of
participation may produce different results in regards to effectiveness based
upon performance and/or satisfaction. They classified quality circle programs
under the heading of consultative participation. In this type of situation the
employees engage in long term, formal and direct participation with content
focused on job issues. Unfortunately they label their results as ”inconclusive
due to the poor methods used in the few studies found".

The results found by Locke & Schweiger (1979) and Miller & Monge (1986)
are rather surprising when one considers the claims made by quality circle
proponents on how the technique enhances worker satisfaction and general
quality of working life. However, even though quality circles have been
referred to as a technique to improve employee satisfaction (Cole, 1980;
Henderson & Goode, 1984; Landon & Moulton, 1986; Sorensen, Head & Stotz,
1985; Yager, 1980), they were initially introduced to improve productivity.
This could provide an explanation for why so many fine that satisfaction does
not appear to be significantly influenced either positively or negatively, by the
implementation of quality circle programs (Bradley & Hill, 1983; Elvins, 1985;
Head, Molleston, Sorensen & Gargamo, 1986; Mohrman & Novelli, 1985; Rafaelli,
1985; Tortorich, Thompson, Orfan, Layfield, Dreyfus & Kelly, 1981). In fact,
this may provide some explanation for Marks’ (1986) conclusion that
"...participation did not...contrary to the usual claims of quality circle
proponents, affect worker attitudes...and overall job satisfaction.”

Satisfaction is another variable that is difficult to measure. Studies

13

dealing with it must be interpreted carefully, especially since there is no
standard measure of satisfaction utilized consistently. One potential reason
behind finding no significant increase in satisfaction may be due to a ceiling
effect. This may be particularly important in studies using a pre-test and
post-test design (Donovan & Jury, 1983; Stinnett & Perrill, 1982).
1912 invglvement

Job involvement has been defined conceptually by many individuals quite
differently each time. Table 2 provides six examples of the various ways
scholars have attempted to make sense of this concept. Several key
components will be analyzed and their relevance to job involvement will be
discussed. Following this, a conceptual definition that takes these key
components into account will be presented.

Most definitions of job involvement imply that an individual must believe
that his or her job is important, or that s/he is contributing in some way to
the job. Perceptions and importance are two key components that can be
partialled out from those definitions: Focusing on an individual’s perception,
implies that it does not matter whether something is actually true or not, but
whether the individual believes it to be true. This is why many quality circle
programs have been successful for a short amount of time and then ended in
failure. Members may have perceived participation to be important, only to
find out that they have had very little impact. This occurs as a result of
many programs being simply ”window dressing". An individual will evaluate the
situation and determine whether situation and determine whether conditions in
the environment and actual job are conducive to job involvement. Involvement

is dependent upon the individual’s own motivation, in which a key factor is

14

Table 2.

Table of conceptual definitions of job involvement

Allport (1943) - Degree to which an employee is
partieipating in his job and meeting sueh needs as
prestige, sell-[espeeL autonomy, and self regard.
Vroom (1962) - Degree to which employee pereeives that his job

performance is consistent with characteristics that

are eentrgl to his sell-eeneept.

French & Kahn (1962)- Degree to which employee pereeiveg that his jet;

perfprmance is central to his self-epneep; i.e., the
degree to which it affects his self-esteem.

Lodahl & Kejner(l965)- Degree to which a person is idengifieg

psychelegieally with his work, or the impertance
of work in his total self-image.

Lawler & Hall (1970)- Degree to which a person perceives his total wprk
sitpetion tp be an important part of his life and
to be eentral to him to satisfy his important
needs.

Gorn & Kangungo(l980)- Degree to which an individual is involved in a
partieplsr iOb and actively partieipsges in it and a
psyehelpgieal state of identificetipn with werk in
general relative to other activities.

15
the individual’s perception that something is important or valuable to him or
her. Of course, as Stinnett & Perrill (1982) and Steel, Mento, Dilla, Ovalle &
Lloyd (1985) state, not all participants in quality circles feel a sense of
involvement in the decision making process or even in the group. This may be
a result of what Marks, Mirvis, Hacket & Grady (1986) discovered. They found
that employees might not have joined a quality circle as a result of curiosity
and/or a desire to have impact on the organization, but rather because they
perceived the quality circle program provided them with a protection from
negative factors within the organization. Thus, the variable of importance is
introduced. If some value is identified in contributing to the job, involvement
is more likely. The job or task must have some meaning or significance to the
individual before there will be any type of involvement occurring. It makes
intuitive sense that a topic must be perceived to have some type of influence
on the individual before any commitment can be observed. The task must be
believed to contribute to the satisfaction of some type of belief or need the
individual feels is central to his or her well being and or self-esteem. This is
a reason the most effective quality circle programs are voluntary. Many times,
however, Dean (1985) found that the people who desire to participation in
quality circles are usually those who desire greater involvement in the decision
making process and have faith that the circles will work.

Many of the definitions also indicate that some form of participation is
necessary and important when defining involvement. If an individual actively
participates in certain aspects of his or her job, s/he is believed to be
expressing interest in those areas (March & Simon 1958). Saleh & Hosek
(1976) state that if the individual feels s/he is making an important

contribution, job involvement will increase. This implies the significance of

16
having an open environment where employees feel free to offer suggestions for
improvement or free to participate in a quality circle. Siegel & Ruh (1973)
also found that involvement was significantly related to participation in
decision making. Employees felt more involved in their job, when they had
influence upon their environment through the participation.

Another common aspect of most definitions are concern the individual
viewing the job as important to meeting needs central to his or her self-
esteem. Examples of these needs are presrige, self-respect, autonomy and
other variables affecting the individual’s self-concept. Saleh & Hosek (1976)
imply that this is an issue dealing with consistency. The individual needs to
recognize that the abilities required of the job are consistent with the abilities
s/he have to perform the job. As long as there is consistency, then the needs
central to the individual’s self-concept are enhanced and involvement is
increased. If an individual feels they have the capability to perform a
particular task, and it appears that by doing so something can be acquired (or
a need may be met) the individual will be more likely to become involved in
that task. This aspect of job involvement implies that there should be added
importance to the introduction of the quality circles to the employees, along
with extensive training, to assure employees feel that they are capable of
handling the added input.

This brings us to psychological identification. Kanungo (1979) states that
there are two cognitive states associated with psychological identification. The
first is the socialization process one goes through on the job. This deals with
issues such as the individual’s desire for autonomy and also the need for
achievement. A second aspect of the cognitive state focuses on the job

environment and how effectively it satisfies the individual’s needs. If the

l7
individual’s environment allows for much autonomy and autonomy is very
important to him/her, the individual’s needs will be satisfied through
involvement. When the environment matches the employee’s desires, s/he will
psychologically identify with that job or task function.

A distinction is also made between career involvement and involvement in a
task. Gorn & Kanungo (1980) state that career (general job) involvement
stems back to the protestant work ethic. Work in general refers to how
important the individual’s job is to his/her perceptions of self. For instance,
an individual will work late to finish a job even if it means doing so on
his/her own time. An individual may be totally devoted to his/her career and
enjoy every aspect of it. Involvement in a task (specific job involvement)
such as quality circles, parallels career involvement, but must be treated as a
separate variable. The reasoning is quite simple. An individual may find
particular aspects of the job satisfying, and still feel that the job in general is
not meeting any needs. This implies that involvement can occur at a more -
specific level than just the general work level. These levels must be
considered as separate entities in that~ they identify different levels of
involvement.

Now that the commonalities among the various conceptual definitions of
others have been examined, let us look at the way job involvement will be
conceptually defined in this paper. Job involvement is the degree to which an
individual perceives that participating in his or her job in general (career) or
a specific task with which s/he can psychologically identify with, is important
and/or central to meeting his or her needs associated with his or her self-
concept. This definition includes all six of the components just mentioned.

By incorporating each one into the definition, a strong, clear conceptualization

18
of job involvement is developed. This definition is less susceptible to the
problems that any of the other definitions face, since it incorporates the six
strongest components mentioned by others.

Very few articles make the distinction between general (career)
involvement and involvement in a specific task (such as quality circles). This
definition should be useful when analyzing the data on the effects of quality
circles on job involvement.

mm

The original purpose of quality circles was to increase productivity in the
organization with no real consideration of morale. Statistical quality control
(SQC) is a method to measure productivity by utilizing a surrogate approach
(Sinks, Tuttle & DeVries, 1984). This is the approach that Deming introduced
to the“ Japanese organizations, that spun the quality circle revolution. Today,
there have been many glowing reports in regard to increases in productivity
(Judson, 1982; Marks, 1986; Seaton, 1984), but very few sysmmatic examinations
of the circle’s effectiveness (Bryant & Kearns, 1982; Guzzo, Jette & Katzell,
1985). One reason that it may be difficult to measure productivity, even when
using SQC methods, is that this approach does not directly measure
productivity. The use of performance criteria or attributes that have been
shown to be highly correlated to productivity is used (Sink, Tuttle & DeVries,
1984). So one must be as careful when evaluating productivity measures as
when evaluating participation measures. Both can be measured in many
different ways and effected by many different factors. Even though quality
circle programs should be using SQC methods when analyzing productivity
increases, most do not report exactly which method they are utilizing.

Mohrman & Novelli (1985) found some improvement in productivity, but could

19
not determine whether or not it was due to the quality circle program. This
may have been due to the fact that when measuring productivity with the SQC
approach, one cannot be sure they are really tapping into productivity.
Sashkin (in Ferris & Wagner, 1985) states that the effects of participation on
productivity are strongly influenced by the implementation strategy. So, if a
technique is introduced to the organization correctly and training is conducted
for all, productivity should increase by 15% or more (Sashkin, 1982). Ferris &
Wagner (1985) indicate that productivity is so dependent on proper
implementation of the program, that all quality circle programs should be
examined closely before accepting the productivity results. Since there is no
standard procedure for quality circle preparation, some organizations may be
spending time to train all the participants in problem solving and decision
making, where other organizations may only provide a general orientation.
This would also effect the response each organizations got out of its
management personnel. If participants are not convinced the program will
work, or not provided with enough training to make it work, the program most
likely will not work.

Miller & Monge (1986) tested three different models; a cognitive model in
which participation was predicted to have a stronger influence on productivity
and satisfaction for decisions about which employees had knowledge in, an
affective model, where it was proposed that participation would lead to the
attainment of higher order needs (self-expression, respect, independence) which
would lead to increases in satisfaction. The last type of model discussed was
a contingency model, in which theorists predict that participation would
affect satisfaction and productivity in different ways across individuals.

In following a cognitive model, it is predicted that when employees

20
participate in the decision making process, they will attain knowledge that will
lead to increases in productivity and that this knowledge will result in higher
levels of satisfaction. These results should be even stronger when the
employees participate in decisions in which they have specific knowledge.

The effectiveness behind a quality circle program stems from the fact that
suggestions are made from employees who know the topics and areas of
concern for their department. It assumes that the best person to make a
decision affecting the job, is the individual working in that job. By including
employees in decisions that affect them and their jobs, quality circles should
make the employees even more aware of what is taking place in their areas.
This ”feeling" of participation, however, is not enough to support a quality
circle program. The employees must be able to see their participation having
some impact on the organization or at least within their own department. An
employee’s ”perception" of participation will quickly die if no results are forth
coming. The lack of results often due to a participation program implemented
as "window dressing" maybe the reason many quality circle programs fail or
are doomed to fail.

Even though this is the case, many organizations which have implemented
quality circle programs have provided narrative reports of the increases in
satisfaction and productivity. But, few have provided statistics to document
the outcomes. If the narrative reports are a-correct representation of the
effects quality circle programs have on the producrivity and quality of work
life of those who participate in them, an analysis of the empirical work done
should support the following hypotheses.

Through participation in quality circles, individuals will be exposed to more

knowledge than they had before. This knowledge will lead to a greater sense

21
of involvement resulting in the first hypothesis:
H1: Those who participate in quality circle programs should become
involved in the task.
By participating in the quality circle the individual will have knowledge of the
results, which will lead to increased productivity as the second hypothesis
predicts:
H2: Participation in quality circles should lead to higher organizational
productivity rates.
The third hypothesis predicts that while participating in the quality circle, the
individual actually sees the increases in produCtivity, which will lead to a
greater sense of satisfaction:
H3: Participation in quality circle should lead to higher
levels of job satisfaction.
The following review was conducted to examine the relationship, if any,
between quality circle participation and job involvement, productivity and
satisfaction. The use of meta analysis (Hunter, Schmidt & Jackson, 1982)
techniques were employed to actually test the impact of quality circle programs
on these dependent variables.
METHODS
This analysis was conducted to examine the strength of the relationship, if
any, between quality circle participation and job involvement, productivity and
job satisfaction. Literature from the areas of psychology, communication,
management and industrial relations were included, along with many
government publications. The search was limited to published literature
written in English and dating from 1975 to current. Psychology Abstracts and

Social Science Citation Index computer searches were conducted to assure as

22
comprehensive review as possible. The criteria used to determine inclusion for
the literature found in the searches in the meta analysis are, (l) the item must
be a research article, (2) enough data muSt be provided to calculate a relevant
effect size, (3) the study must look at involvement, productivity and/or job
satisfaction as dependent variables, and (4) the article must use quality circle
participation as the independent variable.

A total of 43 articles were found that dealt with quality circle
participation and employee involvement, productivity and/or job satisfaction.
Thirty of the 43 articles did not meet the criteria from inclusion in this
review. Articles not used in the meta-analysis were divided into four
categories, those dealing with reviews and essays (l 1), those not providing
enough data to calculate effect size (12), studies looking at dependent
variables other than involvement, productivity or satisfaction (4), and finally,
studies not dealing with quality circle participation as an independent variable
(7). Refer to Table 3 for a list of the literature broken down into these
categories.

The first step in the meta-analysis procedure was to calculate the effect
size for each study. The point-biserial correlation r was used as opposed to
the d-statistic Glass, McGaw & Smith (1981) recommend. Hunter, Schmidt &
Jackson (1982) identify the point-biserial correlation r as the Optimal statistic
due to its robustness. Once the effect sizes were calculated, corrections for
unequal sample sizes were computed. This was necessary in six of the studies,

where there were more subjects in one or another condition (Blatchey, 1984;

23

Table 3.

Categorization of research

viw

Barrick & Alexander (1987)
Landon & Moulton (1980)
Locke & Schweiger (1979)
Maine (1984)

Marks (1986)

Mento (1982)

Miller & Monge (1986)
Perrill (1984)

Rieker & Sullivan (1981)
Sikes, Connell & Donovan (1980)
Steel & Shane (1986)

MW

Blatchey (1984)
Donovan & Jury (1983)
Griffin (1987)

Head et. al. (1987)
Marks et. al. (1986)
Rafaeli (1985)

Steel et. al. (1982)

Steel et. a1. (1985)
Stinnett & Perrill (1982)
Stinnett et. a1. (1983)

n n n v 1 l 5
Bradley & Hill (1983)
Elvins (1985)
Stohl (1984)
Stohl (1987)

Different indepengent variables
Cole & Tachiki (1983)

Cole & Tachiki (1984)

Dean (1985)

Griffin & Wayne (1984)
Guzzo et. al. (1985)

Pasmore & Friedlander (1982)
Sorensen et. al. (1985)

Ne guantifiable effec; size
Atwater (1984)

Bradley & Hill (1987)
Bryant & Kearns (1981)
Donovan & Van Horn (1980)
Horn (1982)

Hunt (1981)

Meyer & Stott (1985)
Mohrman & Novelli (1985)
Seaton (1984)

Seybolt & Johnson (1985)
Steel, Ovalle & Lloyd (1982)
Tortorich et. al. (1981)

24
Rafaeli, 1985; Steel et.al., 1985; Stinnett et.al., 1983). Those studies conducting
pre-test and post-test comparisons were considered to have equal sample sizes
(Donovan et.al., I983; Stinnett et.al., 1982). Correlations were corrected for
attenuation due to measurement error. All but two studies reported
reliabilities (Donovan et.al., 1983; Steel et.al., 1982). The average reliability of
the other studies was calculated and used in those two instances. After all of
the corrections for error were finished, employee involvement, productivity and
job satisfaction were analyzed separately, by deriving a weighted average
effect size. Calculations of true variance were then computed by subtracting
the variance due to sampling error from the actual variance. If the true
variance was more than zero, and determined to be statistically significant
using a chi-square test, potential moderating variables were be examined.

A brief review of the eleven articles used in the meta-analysis is
presented below. In this review, the type of organization the study took place
in will be discussed, along with the number of participants. The amount and
type of training participants went through is also provided. Study designs are

presented, along with their dependent variables and measurement procedures

(Table 4).
Jpn jpvplvement, nggpetivity and 19p sgtisfsetipn

The only study that discussed and presented enough statistical information
to fit under this category was Marks, Mirvis, Hackett & Grady (1986). This
research was conducted in the manufacturing department of a United States
corporation. Fifty-three subjects volunteered to participate in the quality
circles and this research. The amount of training offered to these volunteers
was not provided. A quasi-experimental field study was conducted to compare

changes in perceptions of the general quality of work life, productivity and

25

Table 4.

Breakdown of study design.

 

 

 

 

 

 

 

 

 

 

 

 

 

Study n Dep. Var. Design Type of Org
Blatchey 825 Job Involv. Exper. v Dept. of
(1984) 44-181 Job Satis. Control Defense
Donovan 200 Job Satis. Pre-test Hospital
et.al. Post-test

(1983)

Griffin 146 Job Satis. Exper. v Electronics
(1987) 73-73 Control Manufacturing
Head 84 Job Satis. Exper. v Manufacturing
et. a1. 18-12 Control Plant

(1986)

Marks 106 Job Satis. Exper. v Manufacturing
et. al. 53-53 Control Plant

(1986)

Rafaeli 760 Job Involv. Exper. v Electronics
(1985) 455-305 Job Satis. Control Manufacturing
Steel 40 -'Job Satis. Exper. v Dept. of

et. a1. 20-20 Control Defense
(1982)

Steel 107 Job Involv. Exper. v Military

et. al. 34-73 Job Satis. Control Maintenance
(1985a)

Steel 163 Job Involv. Exper. v Hospital

et. al. 99-64 Job Satis. Control

(1985b)

Stinnet 68 Job Involv. Pre-test Manufacturing
et. al. 68-0 Job Satis. Post-test Plant

(1982)

Stinnett 316 Job Satis. Exper. v Manufacturing
et. a1. 67-849 Control Plant

(1983a)

Stinnett 386 Job Satis. Exper. v Manufacturing
et. al. 137-849 Control Plant

(1983b)

 

 

 

 

 

 

26

attendance behaviors for those who participated in the quality circles to those
who did not. A pre-test was administered prior to the circles being
implemented and a post-test was administered 20 months after the pre-test.
This design was utilized to assess the impact of participation in quality circles
on job involvement (personal responsibility), productivity and satisfaction. The
Michigan Organizational Assessment Package was used to obtain the
information dealing with changes in the quality of work like (e.g., job
involvement, satisfaction). Organizational records were used to measure the
change in productivity. The measures of productivity were percentage of hours
spent on production, efficiency rate and overall productivity.
19p invplvement and 19p setisfaetien

Approximately five studies fit under this category of analyzing both the
variables, job involvement and job satisfacrion (Blatchey, 1984; Rafaeli, 1985;
Stinnett & Perrill, 1982; Steel, Mento, Dilla, Ovalle & Lloyd, (1985). The first
to be dealt with was conducted by Peter Blatchey (1984). He conducted the
study in three different Department of Defense organizations. Participating in
the quality circle group were 44 individuals, with 181 members of the
organization used as a control group. Training was stated as having previously
occurred, but there was no indication as to how much or what kind of training
actually took place. Blatchey used the three organizations in the following
manner: Organization A had a quality circle group with some members of the
organization participating and others who did not, organization B had all
members participating in the quality circle program and organization C had no
quality circle program running. Blatchey originally wanted to determine
whether organization A produced different results in satisfaction, performance,

commitment, involvement, participation or cohesion than when organization B

27

and C were compared. A comparison of organization B and organization C was
done by analyzing the results from the post-test collected from each
organization. The measures used for determining job satisfaction and job
involvement were. not identified, although alpha coefficients were provided
(satisfaction, .79; involvement, .92).

Anat Rafaeli (1985) will be the second study discussed under this category.
He studied an American electronics manufacturing plant. There was a sample
size of 760 blue-collar workers from 11 different divisions in the electronics
firm. Although it was stated that extensive training was undergone by all
participants, the specific amount or time allotted for such training was not
indicated in the article. A cross sectional type of study was conducted as 11
different divisions were utilized in the collection of data. Each quality circle
group in all eleven divisions responded to an attitude survey administered at
one point in time (the exact amount of time elapsing from the beginning of
the circles to the time of the questionnaire adminisrration was not provided).
A comparison was made between participants in the quality circles to those not
participating in them. The research was conducted to examine the effects of
actual participation in quality circle activities. More specifically, three
hypotheses that were proposed were, (1) Participation in quality circles lead to
perceptions of influence on job, (2) Participation in quality circles lead to
increased job satisfaction, and finally, (3) Quality circle involvement will bear
positive relative to perceptions of variety, autonomy and interaction with
others. The measure of job involvement (perceived influence) that was used in
this research was never identified, although an alpha of .89 was reported, and
examples of the questions indicated they were tapping involvement. The

Hoppock (1935) satisfaction measure was utilized to analyze job satisfaction.

28

The third study to fit under this category is that of Steel, Mento, Dilla,
Ovalle & Lloyd (1985). They obtained data from two Department of Defense
(DOD) organizations, a military maintenance organization and a medical facility
located on the same DOD installation. In the maintenance organization, 107
subjects participated in the research. The group leaders received a 40-hour
training course dealing with various quality circle techniques, problem analysis
and group dynamics. Group members were also encouraged to participate in a
9-hour training module on quality circle techniques. A non-equivalent control
group design was used in both the maintenance and hospital organizations. A
pre-test was administered in October of 1982 and the post-test was
administered approximately one year later to assure all aetive groups had
performed at least two management presentations. The main purpose of this
study was to examine the impact of a quality circle intervention on a wide
range of survey variables (e.g., job involvement, job satisfaction, group
cohesiveness, participation in decision making, etc_.). Job involvement was
measured using Saleh & Hosek’s (1976) Central Life Interest (CLI) which
related to overall job involvement and Self-Concept (SC), which referred to
more of a task involvement dimension. Job satisfaction was measured using
Andrew & Withey’s (1976) instrument.

One hundred-sixty-three participants were studied in the hospital
organization. This department was analyzed in exactly the same way as the
maintenance department was and for the same reasons. One major difference
between the two departments stemmed from the amount of training the
hospital quality circles received. Not all of the leaders attended the 40-hour
training course, and no training was offered to members of the quality circle.

Stinnet & Perrill (1982) provide the last study in the meta-analysis that

29

fits under this category. They conducted their research in a small
manufacturing plant of a United States electronics firm, which consisted of 68
quality circle participants. All employees were participants in the quality
circle program. A seven step plan was outlined for the quality circle
participation with training during step five. All of the members and leaders
received training in various quality circle techniques, however, the length of
training was not specified in the text. A pre-test and a post-test were
administered to determine the effect that the quality circle program had on
quality, measurement, reward, leadership, structure, participation (involvement),
communication, satisfaction, peer relations and the group process.
Approximately six months passed from the administration of the pre-test to the
post-test. Although alpha coefficients were provided for all measures used, the
actual satisfaction measure used was never provided. The authors did provide
example questions of the satisfaction measure, which appeared to indicate that
an overall measure of job satisfaction was obtained from the questionnaire.
The measure of job involvement the authors labeled, participation, asked
questions regarding the degree the individual was allowed to participate in the
quality circle and the amount of commitment felt toward the circle.
mm

Five studies provided information and statistics to calculate only job
' satisfaction (Donovan & Jury, 1983; Griffin, 1987; Steel, Lloy‘d, Ovalle &
Hendrix, 1982; Stinnett, Perrill, Boster & Borstelmann, 1983).

Donovan and Jury presented results obtained from a hospital organization
consisting of 200 participants. No mention of training was ever made, perhaps
since this was just a pilot study. A pre-test post-test comparison was made

among those who participated in the quality circle program. Approximately six

30
months was provided between the administrations of the two tests. The effect
the quality circle program had on work planning & efficiency, cooperation,
communication, influence, knowledge, task significance, feedback and
satisfaction was examined using the Job Reaction Questionnaire.

Ricky Griffin (1987) conducted a study in the manufacturing plant of a US
electronic company. A total of 73 employees participated in this research by
participating in a quality circle in one of the plants. All members participated
in a two day off site training program in problem solving before participating
in a quality circle. The pre-test was provided prior to the training, with the
first post-test being administered approximately six months later, the second,

18 months later and the last post-test administered 36 months later. The
purpose of the research was to determine whether quality circle members
would report higher levels of job satisfaction, commitment, performance and
have lower turn over than non-participants. The Minnesota Satisfaction
Questionnaire was used to measure the degree of job satisfaction the
participants in the quality circles reported.

Steel, Lloyd, Ovalle & Hendrix (1982) conducted a pilot study in the
Department of Defense with 40 members of a quality circle program. The
amount or degree of training that the quality circle participants underwent was
not stated in the article. A nonequivalent control group design was used to
conduct the research, with both a pre-test and post-test administered at a six
month interval to those participating in the quality circle program. The
survey questionnaire contained items measuring job satisfaction, work group
effectiveness, general organizational climate and supervisory effectiveness. The
questionnaire used to measure these items was not identified, nor were the

alpha coefficients.

31

Stinnett, Perrill, Boster & Borstelmann (1983) conducted a longitudinal
study in three different organizations. The first organization consisted of 67
quality circle participants, the second organization had 137 quality circle
participants and the third had 249 employees who did not participate in a
quality circle program. Training was provided in two stages, during the
changing phase, all employees were trained in problem solving and the use of
quality circle techniques. During the refreezing phase, an identification of key
persons who would receive additional training in group dynamics and
organizational development occurred. A comparison of the first organization
(67 quality circle participants) and the third organization (249 non-quality
circle participants) was conducted in the meta-analysis. The second
organization (137 quality circle participants) was also compared to the third
organization, which functioned as a control group in both comparisons. An
attitude questionnaire was provided to each organization at six month intervals.
The questionnaire consisted of measures of quality, leadership, people
development, structure and reward (satisfaction). The reliability of each
measure was supplied, although the questionnaire administered was never
identified.

RESULTS

The dependent variables examined in the meta analysis were job
involvement and job satisfaction. ,Not enough data could be collected to
calculate an effect size for productivity measures.
19p invplvement

Of the thirteen studies, five were used to estimate the relationship

32
between quality circle participation and job involvement. The total sample size
was 1323. After corrections, the correlation in the Blatchey (1984) study
remained .00, the correlation in Rafaeli (1985) went up from .09 to .10. Steel
et. al. (1985a) in the study of maintenance workers had a correlation of .18,
after the corrections for unequal sample size and measurement error were
performed, it went up to .21. A similar increase occurred for the Steel et. al.
(1985b) hospital study which had a correlation of .09 prior to corrections.
After all corrections were done, the correlation increased to .10. Corrections
for unequal sample size were not done for the Stinnett et. al. (1982) study,
however, after correcting for measurement error, the correlation increased
from .04 to .04. For a complete list of studies, individual effect sizes, the
mean weighted correlation, the observed variance among effect sizes, the error
variance due to sampling error, the true variance and the chi square, refer to
Table 5.

The correlation between quality circle participation and job involvement
ranged from .00 to .18, with the mean weighted correlation of .09. Corrections
for unequal sample sizes were performed for four of the five studies (Blatchey,
1984; Rafaeli, 1985; Steel et. al., 19853 & b). All five of the studies were
corrected for measurement error, by dividing the square root of the alpha
coefficient by the effect size.

The observed variance among effect sizes was .00.- The error" variance due
to sampling error was .00 with the true variance being -.001. As a result of
the true variance being less than zero, and the chi square not being
significant, no test for moderating variables was performed. All studies

reported high alpha coefficients, however, it is impossible to tell whether the

33

Table 5.

Job involvement correlations after corrections

 

 

 

 

 

Study r
Blatchey (1984) . .0004
Rafaeli (I985) .097
Steel et. al. (l985a) .209
Steel et. al. (1985b) .102
Stinnett 8 Perrill (1982) .044

 

 

 

 

"’ 2. 'L. t
" <7: 0’; 7,2 fl

.0875 .00865 .00378 -.00106 O

34

Table 6.

Job Involvement Measures.

 

 

 

 

 

Study Measure Alpha Coefficient
Blatchey (1984) Unknown .92
Rafaeli (1985) Unknown .89
Steel et.al. Central Life .90
(1985a 8 b) Interest (Saleh
8 Hosek, 1976)
Stinnett 8 Unknown .85

Perrill (19881

 

 

 

 

35
different measures were looking at the same concept. For a list of measures
and their alpha coefficients refer to Table 6.
ti n
Twelve of the thirteen studies were used to estimate the relationship
between quality circle participation and job satisfaction. The total sample size
was 2541. After corrections, the correlation in the Blatchey (1984) study
increased from -.00 to -.00. The correlation of .12 found in the Donovan &
Jury (1983) study, increased after corrections for measurement error to .13.
The correlation in the Griffin (1987) study remained the same (.00), even after
corrections were done. The correlation calculated in the Head et. al. (1986)
study went from .24 to .28 after corrections were done. Marks et. a1. (1986)
had a mean weighted correlation of .03, after corrections it was increased to
.03. The correlation in the Rafaeli (1985) study was increased from .04 to .04
due to corrections for unequal sample size and measurement error. Steel et.
al. (1982) had a correlation go from .13 up to .14 after corrections. In the
maintenance organization, Steel et.'al. (1985a) record a mean weighted
correlation of .30, this correlation increases to .30 after corrections. In the
hospital organization (Steel et. al. 1985b), a correlation of .086 was calculated,
after corrections for unequal sample size and measurement error, it was
increased to .10. The correlation for the Stinnett & Perrill (1982) study went
from .09 to .10 due to the correction for measurement error. Correction for
unequal sample size and measurement error were performed for both Stinnett
et. al. (1983a & b) studies. The correlation for the first one (Stinnet et. al.,
1983a), went from .00 to .00, while the correlation for the second one

(Stinnett et. al., 1983b) increased from .00 to .00. A list of studies, individual

 

36

Table 7.

Job satisfaction correlations after corrections

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Study r
Blatchey (1984) -.009
Donovan 8 Jury (1983) .134
Griffin (1987) .002
Head et. a1. (1986) .278
Marks et. al. (1986) .029
Rafaeli (1985) .043
Steel et. a1. (1982) .142
Steel et. al. (1985a) .301
Steel et. al. (1985b) .099
Stinnett 8 Perrill (1982) .100
Stinnett et. al. (1983a) .008
Stinnett et. al. (1983b) .0009
F fl; 43‘ 27:" a? .

 

.0516 .0051 .0047 .0004 .02049

37
effect sizes, the mean weighted correlation,the observed variance among effect
sizes, the error variance due to sampling error, the true variance and the chi
square, is presented in Table 7. The correlation between quality circle -
participation and job satisfaction ranged from -.00 to .30 with the mean
weighted correlation of .05. Corrections for unequal sample size were
performed on six of the twelve studies (Blatchey, 1984; Rafaeli, 1985; Steel et.
al., 1985a & b; Stinnett et. al., 1983a & b). All of the studies were corrected
for measurement error.

The observed variance among effect sizes was .00. The error variance due
to sampling error was .00 with the true variance being .00. Because the true
variance was zero, and the chi square was not significant, no moderating
variables were considered. Once again a variety of measures of job
satisfaction were used. Table 8 includes a list of studies, the job satisfaction
measure used and the alpha coefficient obtained.

DISCUSSION

The results of the meta analysis indicate that participation in quality
circle programs has no effect on either job involvement or job satisfaction.
The relationship between quality circle participation and productivity could not
be analyzed, because only one study reported enough data to calculate effect
size. .

Hyppthesis # l

The first hypothesis stated that those who participate in quality circle
programs should become more involved in their jobs. The results of the meta
analysis are not consistent with this hypothesis. The mean weighted
correlation (r - .09), even after corrections was still very small and not

significant.

38

Table 8.

Job Satisfaction Measures.

 

 

 

 

 

 

 

 

 

 

 

 

 

Alpha

Study Measure Coefficient
Blatchey (1984) Unknown .79
Donovan 8 Jury Job Reaction .80
(I983) Questionnaire
Griffin (1987) Minnesota Satisfaction .88

Questionnaire
Head et. a1. Job Diagnostic Survey .72
(1986)
Marks et. al. Michigan Organizational .82
(1986) Assessment Package
Rafaeli (19851 Hoppock (1935) .82

Satisfaction Measure
Steel et. al. Unknown .80
(1982)
Steel et. a1. Andrew 8 withey (1976) .78
(1985a) Satisfaction Measure
Steel et. a1. Andrew 8 Hithey (1976) .78
(I985b1 Satisfaction Measure
Stinnett 8
Perrill (1982) Unknown .77
Stinnett et. al. Unknown .85
(1983a)
Stinnett et. al. Unknown .85
(1983b)

 

 

 

 

39

At first it was believed that this result was due to moderating variables,
however, this hypothesis was not supported by the data. These data indicate
that contrary to popular belief, quality circle programs may not influence job
involvement.

One reason the mean weighted correlation is so low may be due to the
fact there is no standard measure of job involvement; each study used a
different measure. Although all studies reported high alpha coefficients, we
cannot be sure they are indeed measuring the same concept. It is therefore
difficult to state with any confidence they were able to tap into general job
involvement. As was established earlier, there are many conceptualizations
regarding what job involvement really is. This inconsistency may have been
reflected in the results.

Another potential explanation for the results may be a result of the initial
purpose of quality circle programs. In the beginning the only purpose of the
program was to improve sagging productivity rates. -They were not designed to
increase involvement. Many organizations have proclaimed the quality circle
program they implemented improved factors other than just productivity.
Based on the frequency of this claim many just assumed increased job
involvement was an outcome of the program. The data indicate that this is an
erroneous assumption. .

A limitation of this meta analysis is the small sample size. Since the total
sample size was only 1323, the results cannot be considered conclusive. Due
to cost and time limitations, some research may have been overlooked that
could have an impact on the outcome discussed here. However, after an
extensive literature search, less than one dozen studies could be utilized,

indicating that many studies were not of high enough quality to assess effects.

40
This suggests that more research still needs to be done on the effectiveness of
quality circles.
W

The second hypothesis stated that participation in quality circles should
lead to higher organizational productivity rates. As was noted in the results
section, not enough data was available to test this hypothesis.

This was a particularly disturbing occurrence in analyzing the data.
Productivity was the reason quality circle programs were initially implemented
and practically no one reported the results in complete enough detail to be of
use in this meta analysis.

H h i #

The third hypothesis predicts that participation in quality circles should
lead to higher levels of job satisfaction. Once again, the data do not support
the hypothesis. The mean weighted correlation after corrections was .05,
which was not significant. This result is perhaps the most startling of all.
Proponents of quality circle programs have said that morale and job
satisfaction have increased dramatically due to the implementation of the
program. The results obtained from this review do not indicate that this is
the case.

One reason these results may have been found, may be due to the fact
that very few of the positive reports provided any data to support their
claims. These types of studies were omitted from the meta analysis, possibly
skewing the results.

The results may have been due to the variety of ways job satisfaction was
measured in each study. Although there are many acceptable and reliable

measures of job satisfaction, no one is used consistently more than another.

 

41
This may have produced the discrepancy that was found in the results,
however, there was no variance observed in the estimate, indicating that this
is not a possible explanation.

By correcting the correlations for attenuation and sampling error, the
variance due to sampling error decreased to .00 and all of the observed
variance could be attributed to sampling error.

One such variable may be the culture found within an organization. The
effectiveness of the quality circle programs the Japanese companies experienced
just may not transfer to the United States. One p0tential reason for the
decline of Lockheed’s quality circle program was due to union involvement
(Cole, 1980). The program became (a leverage mechanism to be used against
one side or the other. Instead of fostering cooperation and openness, the
quality circles provided another bargaining chip for the negotiation process.

This leads into another cultural difference that may hamper the quality
circle process, the closed (and often hostile) climate that exists in many
United States organizations (Mohrman & Novelli, 1985). Many times the degree
of distrust among employees and top management or between unions and
management may be simply too great to overcome. If either side suggests
implementing a quality circle program the other becomes skeptical or weary of
the underlying motives. One side may not ”buy in" to the program and provide
the necessary support. Often a quality circle program is implemented with
indifference on the part of the top managers (Ruffner & Ettkin, 1987).

Since quality circles just group people from the same work area together,
there is a misconception they are simple to implement. Poor implementation
will lead to failure in all cases. Many organizations operate on a crisis

management mentality. This brings a quality circle program in as a "quick fix”

42
program that will cure all of the organization’s ailments. Unfortunately once
implemented in this manner, the quality circle program has no where to go but
down. This has been historically true of many organizations and perhaps one
reason the failure rate for quality circles is so high.

Implementation of quality circles must begin to gain some consistency.
There is no step by step method for implementation, however there are some
basic components that all successful programs contain. One such component is
support from all levels of the organization. This also includes having both
management and union representatives ”buy into" and support the program.
Another basic component is adequate training for all levels of the organization.
In this analysis training programs not only varied in time, but also in regards
to who received the training. Everyone who .he circle program will effect
should be included in the training whether they are top executives, middle
management or line workers.

Quality circles were meant to be implemented as a permanent structure
within the organization (Cole, 1980; Meyer & Stott, 1985; Moore, 1986), even
though, some organizations have treated them only as a temporary mechanism
for handling an immediate problem (Ruffner & Ettkin, 1987). A quality circle
program needs to function along with the main organization. Stohl (1987)
found that quality circles which cross over to the main organization have more
suggestions accepted and implemented than if they remain parallel to it. The
deeper it is embedded in the organization’s normal structure, the more
successful the program will be (Putnam, 1984). If the quality circle’s operate
parallel to the organization as most do (Lawler & Mohrman, 1985), and remain
isolated, they will be doomed to failure. They must be a permanent structure

and not a temporary "cure all” used until the organization’s problems go away.

 

43

Along with being a permanent structure, quality circles must be a
legitimate form of participative decision making. This simply means that many
organizations whose programs failed, did so as a result of not following
through with employee suggestions and ideas. Employees will participate in
quality circles until they realize they have made no impact on the
organization. At that point in time, the suggestions will stop. Since the
steering committee is responsible for carrying through the ideas generated by
the quality circle participants, it is very important to have support from all
levels of the organization and have all levels represented in the steering
committee. In the participation flow chart of many organizations, the steering
committee is made up of the CEO and top executives who have the final
decision regarding implementation of the suggestions presented from the
members of the quality circles. Quality circles implemented under this type of
structure are doomed for failure as soon as the participants realize that the
participation process is simply a "window dressing" that leeks good and is used
only to make them f_e_e1 good.

When implementing a quality circle program, organizations must take into
consideration the type of t0pic they will allow the participants to make
suggestions and decisions on. The decisions may be related to key issues, such
as productivity and quality improvement, or they may deal with purely cosmetic
issues, such as, where the photocopier or snack bar should be moved. If circle
members are making decisions that affect productivity and quality, a higher
return rate will be evident. If employees participate in decisions that they
have some knowledge in, productivity returns should be higher (Miller &
Monge, 1986). A finding consistant with the cognitive model. If employees

are making decisions regarding tapics such as job design, where their

 

44
knowledge base is, higher returns on productivity can be expected. However,
once employees are asked to make company wide decisions or have philosophic
discussions, the strength participation has will decrease. Many organizations
only allow the quality circle members discuss topics in which they really have
no interest or expertise. This may be another reason why they do not produce
the productivity and satisfaction increases that are possible.

One thing to keep in mind when analyzing the success of quality circles,
is that they were not introduced to increase satisfaction. They are a way to
improve productivity through tapping the employees knowledge of their job.
This may be the reason no real increases in satisfaction can be found in the
meta analysis. However, this brings up another potential explanation for why
the results published are so discrepant. Rieker & Sullivan (1981) state that
there is currently no proper form of evaluation to use.

Satisfaction is a difficult concept to measure. As has been shown, there
are many different and quite reliable measures that can be used to tap it.
There is no standard measure of satisfaction used in current research, which
may contribute to the various results obtained. By not having any one clear,
standard definition of satisfaction, different researchers define and measure it
in different ways. Once a single approach is taken when measuring
satisfaction, similar results may be obtained. Due to the difficulty in
measuring satisfaction, studies dealing with it should be examined carefully. In
studies using a pre test/post test comparison, a ceiling effect may exist. This
may account for the lack of positive results when measuring satisfaction.

Stohl (1987) brings up an interesting point when she suggests we change
the way we determine the effectiveness of quality circles. Instead of looking

at the true effectiveness of the circles through outcomes such as productivity,

45
perhaps we should look at the social satisfaction the employees receive from
participating in the quality circles. Since many times employees do not join
circles to have an impact on their environment, but to avoid negative factors
within the organization (Marks et. al., 1986).

Some of the limitations of past research stem from the research designs
used and comparisons made. There has n0t been enough empirical work done
in the area and those studies that have been done, are not very thorough.
Steel & Shane (1986) state that much of the small amount of empirical
research that has been conducted is unusable due to flawed designs and
misleading evaluations of the results. C0tton et. al. (1988) found that results
could not be determined to be conclusive due to the poor nature of the
methods used in tne literature found on quality circles effectiveness.

Another potential problem with the research being done currently, is the
lack of experimental vs control group designs. Most research has been a
comparison of pre test and post test results. This does not tell us as much as
a comparison among different groups.

There is no consistent length of time used between the comparison of
results regardless of the design used (pretest vs posttest or experimental vs
control group). Steel et. al. (1985) suggest that the discrepancies between
studies may be due to the differences between time comparisons. By not
taking the stage of development into consideration during the comparison of
two groups, we could be missing or losing important information.

Steel & Shane (1986) suggested that quality circle programs may take time
to develop before they produce the increases in productivity and satisfaction
levels many organizations report eventually occurring. However, Griffin (1988)

suggests that circles appeared to be more effective during the early (first two

 

46
years) stages of development and lost effectiveness as time went on. The
results obtained by Griffin (1988) appear to coincide with those found by
Lawler & Mohrman (1987). A "honeymoon effect" occurs during the early
phase of the quality circle program when a few highly motivated individuals
produce a lot of good ideas and improvements. As more and more groups are
added to the program, competition and lack of interest produce a decline in
the effectiveness of the quality circles.
CONCLUSION
The results obtained from the meta analysis should not be surprising. This
is especially true after examining the results Miller & Mongc (1986) found in
their breakdown of participation. What should be surprising is the lack of
quality research done on an organizational intervention as widely used as
quality circles. The lack of research in general and quality research in
particular is disturbing. Based on current research, it is difficult to determine
whether quality circles should be dumped as a participation technique or not
(Cotton et. al., 1988). There are many areas that need further development in
the field before that determination can be made. This indicates that before
any more suggestions regarding the continuation or elimination of quality circle
program are made, more 1111111! research needs to be done. We need to study
the effectiveness and usefulness of the program more thoroughly. Currently
there is no evidence to support quality circle use. This does not indicate
quality circles are not a useful organizational intervention, they may be.
However, until better more thorough methods of evaluating quality circles are
used, we should not pass them off as the ”wonder cure" of the eighties.
F r 'r i

Stohl (1987) brings up an important aspect that should be examined more

 

47

completely. The embeddedness of a quality circle within the main structure of
an organization may have a significant impact on the circle’s effectiveness.
Few studies report how linked the two structures are.

Another potential area for further research which may have an impact on
the effectiveness of quality circles occurs during the implementation phase. A
more in depth look at the type, duration and quality of training employees
receive prior to the program may help organizations reap the desired outcomes
of increases in productivity.

To get better comparisons, we need to look at programs at the same Stage

 

of development. By comparing programs that have been in operation for the
same amount of time, a determination of what stage is more effective and why
can be reached. Steel & Shane (1986) may be correct in stating that quality
circle take time to develop. Or, Griffin’s (1988) view may be more accurate in
reporting quality circles are more effective in the short term. By comparing
different groups at the same stage in time, it may be able to be determined
whether the novelty of quality circle programs contribute to the effectiveness,
and whether the most obvious problems are taken care of right away, resulting
in major changes at first and disinterest and unproductive groups later
(Griffin, 1988).

More work in the measurement of effectiveness needs to be done along
with a look at what factors we look at to determine effectiveness. A
determination needs to be made regarding whether one standard measure can
be adopted to examine the effects uniformly.

Quality circles may have a vast amount of untapped resources that we
have been unable to see due to the many problems found in current research.

After improving our techniques, we may find that there are more effective

 

48

forms of participation, but until the field cleans up its research, this will be

unknown.

”Ir-I

 

49
REFERENCES
ALLPORT, G. W. (1943). The ego in contemporary psychology. Psyehplegv
Review, 1Q, 451-476.
ALUTTO, J. A. & BELASCO, J. A. (1975). A Typology for Participation in
Organizational Decision Making. Adminisgretive Seienee ngrgerly,
16, 191-209.

ATWATER, L. (1984). Quality Circles in Navy Organizations: An Evaluation.

at

Navy Personnel Research and Development Center, San Diego, CA.

BARRICK, M. R. & ALEXANDER, R. A. (1987). A review of quality circle
efficacy and the existence of positive-finding bias. Personnel Psychology.
4_Q_, 579-592.

BLAIR, J. D., & WHITEHEAD, C. J. (1984). Can Quality Circles Survive in the
United States? Business flprizpns, 17-23.

BLATCHEY, P. A. (1984). An Evaluation of the Effectiveness of a Quality
Circle Intervention on Attitudinal Variables in 3 DOD Organizations. Air
Force Institute of Technology, Wright - Patterson AFB.

BRADLEY, K. 8:. HILL, S. (1983). After Japan: The quality circle transplant
and productive efficiency. British Jeurngl pf Ingestriel Reletipns, 291-311.

BRADLEY, K. & HILL, 8. (1987). Quality circles and managerial interests.
Indestrial Reimieps, 26, 1, 68-82.

BRYANT, S. & KEARNS, J. (1981). The guelity circle pregrem pf Nerlelk Naval
shim. United States Office of Personnel Management.

CAMPBELL, DUNNETT, LAWLER & WEICK (1970). Manegerisl behavior,

perfermgnee and, efleetiveness. New York: McGraw Hill.

COLE, R. E. (1980). Learning from the Japanese: Prospects and pitfalls.

Mensgemen; Review, 22-42.

 

50
COLE, R. E. & TACHIKI, D. S. (1983). A Look at US and Japanese Quality

Circles: Preliminary Comparisons. The Quality Qirele learnal, 6, 10-16.
COLE, R. E. & TACHIKI, D. S. (1984). Forging institutional links: Making

quality circles work in the U.S. Natienal Predpetivity Review, 417-429.
COTTON, J. L., VOLLRATH, D. A., FROGGATT, K. L., LENGNICK-HALL, M. L.
& JENNINGS, K. R., (1988). Employee participation: Diverse Forms and
Different Outcomes. Academy pf Management Review, L3, 8-22.
DACHLER, H. P. & WILPERT, B. W. (1978). Conceptual Dimensions and
Boundaries of Participation in Organizations: A Critical Evaluation. I-
Administrative Seience Quarterly, 23, 1-39.
DEAN, J. W. (1985). The decision to participate in quality circles. Journal of

Applieg Behavieral Science, 2._l_, 3, 317-327.

DEWAR, D. L. (1980). Implementing quality circles in your organization. IAng

.A‘R

epnferenee Transaetions, a publication of the International Association of
Quality Circles.

DEWAR, D. L. (1982). To measure or not tao measure. The Quality Qircle
.D_igs_$_t.

DONOVAN, M. & JURY, P. (1983). The Impact of Quality Circles on Aspects
of the Work Climate. Transaetiens of the 5th Annual IAQC Cpnferenee, 5,
463-470.

DONOVAN, M. & VAN HORN, B (1980). Quality Circle Program Evaluation.

r ' nn lIA nf r n 2, 96-101.
ELVINS, J. P. (1985). Communication in quality circles: Members’ perceptions

of their participation and its effects on related organizational

communication variables. Qrppp Q Organizatipn Mes, l_Q, 5, 479-507.

 

51
FRENCH, J. R., ISREAL, J. & AS, D. (1960). An Experiment of Participation in
a Norwegian Factory. mm 13, 3-19.
GLASS, G., MCGAW, B. & SMITH, M. (1981). Meta-analysis in seeial research.

Beverly Hills, Calif .: Sage Publications.
GORN, G. J. & KANUNGO, R. N. (1980). Job involvement and motivation: Are
intrinsically motivated managers more job involved?, Qrganizatienal

Behavier ang Hpmap Perfprmanee, 26, 265-277.
GRIFFIN, R. W. (1987). A Longitudinal Assessment of the Effectiveness of

Quality Circles. Aeagemy pf Management Jeernal, 191-195.
GRIFFIN, R. W. & WAYNE, S. J. (1984). A Field Study of Effective and Less-
Effective Quality Circles. Ereeeegings ef the 4th Annpal Meeting pf the

Aeagemy pf Management, 217-221.
GRIFFIN, R. W. (1988). Consequences of Quality Circles in an Industrial

Setting: A Longitudinal Assessment. Academy of Management Journal, fl,

338-358.

' ' GUZZO, R. A. , JETTE, R. D. & KATZELL, R. A. (1985). The effects of

psychologically based intervention programs on worker productivity: A
meta-analysis. Bersennel Psyehelegy, 38, 275-291.

HEAD, T., MOLLESTON, J., SORENSEN, P. & GARGANO, J. (1987). QLepp at
Qrganizatieg S_t_|,t_di_es, 11, 1, 360-373.

HENDERSON, K. & GOODE, V. (1984). Employee participation through quality
circles. Parks Q, Reereatien, _1_2, 74-96.

HORN, L. J. (1982). Effect of Quality Circles on Productivity Attitudes of
Naval Air Rework Facility Production Employees. Technical Report #M82-

120 IR # 7. Industrial College of the Armed Forces.

52
HUNT, B. (1981). Measuring Results in a Quality Circles Pilot Test. The
it 1 r 4_, 26-29.

HUNTER, J., SCHMIDT, F. & JACKSON, G. (1982). Meta-analysis; Qdmplating

W. Beverly Hills, Calif .: Sage Publications.
JUDSON, A. (1982). The awkward truth about productivity. Harvatd Easiness
Review, 93-97.

KANGUNGO, RN. (1979). The concepts of alienation and involvement

revisited. Psychelegy Bulletin, 8d, 119-138.
LANDON, D. & MOULTON, S. (1986). Quality circles: what’s in them for

employees? Persepnel Jeetnal, 65, 23-26. I
LAWLER, E. (1987). High Invelvement Management; Partieipative strategies fer

impreving prganizatienal perfermanee. Jossey-Bass Inc., San Francisco, CA.
LAWLER, E. E. & HALL, D. T. (1970). Relationship of job characteristics to

job involvement, satisfaction and intrinsic motivation. Journal of Applied
Bsy_c_h_o_legy, a, 305-312. ,

LOCKE, E. A. (1976). The Nature and Causes of Job Satisfaction. In M.D.
Dunnette (Ed.) H nd 0k f In ri lan r niz ti nal P
Chicago: Rand McNally.

LOCKE, E. A. & SCHWEIGER, D. M. (1979). Participation in Decision-Making:

One More Look. Researeh in Organizational Behavier, 1, 265-339.
LODAHL & KEJNER (1965). THe definition and measurement of job

involvement. Jpptpal pf Applied Psyehelegy, 12, 24-33.
LOWIN, A. (1968). Partieipative Decision Making: A Model, Literature
Critique and Prescriptions for Research. Qrganizatiena! Behavier and

flpman Eertprmanee, 3, 68-106.
MAINE, J. (1984). The Trouble with Managing Japanese-Style. Fertupe, 50-54.

53

MARCH, J. & SIMON, H. (1958). Qrganizatiens. New York: Wiley.

I‘ MARKS, M. L. (1986). The question of quality circles. fisyelyQJQgLIpday, 2Q,
36-46.

MARKS, M. L., MIRVIS, P. H., HACKETT, E. J. & GRADY, J. F. Jr. (1986).
Employee participation in a quality circle program: Impact on quality of
work life, productivity, and absenteeism. Jeurnal ef Applied Psyehplegy,
11, 61-69.

MENTO, A. J. (1982). Motivational Reasons why Quality Circles Work in
Organizations. jljtansaetipps ef the 4th Apnpal IAQC Cpnferepee, 284-297.

MEYER, G. W. & STOTT, R. G. (1985). Quality circles: Panacea or Pandora’s
Box? Qrganizatipnal Dynamies, 13, 34-50.

(K MOHRMAN, S. A. & NOVELLI, L. Jr. (1985). Beyond testimonials: Learning
from a quality circles programme. Jeurnal pf Occupational Behaviepr, §.
93-110.

MOORE, M. L. (1986). Designing Parallel Organizations to Support

I Organizational Productivity Programs. In Joseph J. Famularo’s, 2nd
edition, Handpeek pf Hpman Reseprees Administratien. New York:
McGraw-Hill.

PASMORE, W. & FRIEDLANDER, F. (1982). An action-research program for
increasing employee involvement in problem solving. Administrative

Seienee Qparterly, L7, 343-362.
PERRILL, N. K. (1984). New Applications of Quality Circles: Contrasting

Success Stories. Btpeeedings ef the Natippal Empleyee prership and
r ' i i
RABINOWITZ, S. & HALL, D. T. (1977). Organizational research on job

involvement. Esyehplegy Bplletin, M, 265-288.

54
RAFAELI, A. (1985). Quality circles and employee attitudes. Bersennel

W 38.. 603-615.
RIEKER, W. S. & SULLIVAN, S. J. (1981). Can the effectiveness of QC circles

be measured? The Quality Qireles Jeetnal, 11, 29-31.
RITCHIE, J. B. & MILES, R. E. (1970). An Analysis of Quantity and Quality of
Participation as Mediating Variables in the Partieipative Decision Making

Process. Persepael Esyehelegy, 23_, 347-359.
ROSENBERG, R. D. & ROSENSTEIN, E. (1980). Participation and Productivity:

An Empirical Study. Industrial and Laeer Relations Review, 32, 355-367.
RUFFNER, E. R. & ETTKIN, L. P. (1987). When a circle is not a circle. SAM

Advaneed Management Jeurnal, 9-15.
SALEH, S. D. & HOSEK, J. (1976). Job involvement: concepts and

measurement. Aeademy ef Management, 12, 213-224.

SCHERMERHORN, J. R. (1986). Productivity perspectives in management
development. Jeurnal ef Management Develepment, 5, 3-6.

SCHOLL, W. (1976). Theoretical Reflections on Influence, Power and its
Alienating Effects, and Participation. In Dachler & Wilpert (1978).
MW 23.. 1-39.

{(SEATON, I. (1984). The quality circle attitude survey: Analysis and response.
1i i l 1, 42-44.

SEELYE, H. N. & SWEEN, J. A. (1982). Quality circles in US industry: Survey
results. The Quality Citeles Jeptpal, 5, 26-29.

SEYBOLT, J. W. & JOHNSON, R. L. (1985). The Effectiveness of Quality

Circles at Tenneco: Two Years Later. JAQQ Transactiens; Seventh Annual

Cenferenee and Exhibitiep, Les Apgeles, Qalifernia, 148- 152.

55

SIEGEL, A. L. & RUH, R. A. (1973). Job involvement, participation in decision
making, personal background and job behavior. WW
and Lleman Petfertpapee, 2, 318-327.

SIKES, W., CONNELL, L. & DONOVAN, M. (1980). Learning from Experience.
Itansaetiens ef the 2nd Annual IAQQ Qenferenee, 2, 90-95.

SIMS, H. P. & DEAN, J. W. (1985). Beyond quality circles: self managing
teams. ML, 25-32.

SINGER, J. N. (1974). Partieipative Decision Making about Work. Soeielegy of
W L 347-371.

SMELTZER, L. R. & KEDIA, B. L. (1985). Knowing the ropes: Organizational
requirements for quality circles. Business Herizons, 30-34.

SORENSEN, P. F., HEAD, T. C. & STOTZ, D. (1985). Quality of work life and
the small organization: A four year case study. Group & Organizatien
Stedies, 10, 320-339.

STEEL, R. P., LLOYD, R. F., OVALLE, N. K. & HENDRIX, W. H. (1982).
Designing quality circles research. The Quality Circle Journal , 5_.

STEEL, R. P., MENTO, A. J., DILLA, B. L., OVALLE, N. K. & LLOYD, R. P.
(1985). Factors influencing the success and failure of two quality circle
programs. Jeurnal e1: Management, J_1_, 99-119.

STEEL, R. P., OVALLE, N. K. & LLOYD, R. F. (1982). Quality Circles in the
Department of Defense: Some Preliminary Findings. Preeeedings ef the
24th Annpal genference ef the Military Testing Assoeiatien, 40-45.

STEEL, R. P. & SHANE, G. S. (1986). Evaluation research on quality circles:

technical and analytical implications. Hyman Relatiens, 32, 449-468.

56

STINNETT, W. D. & PERRILL, N. K. (1982). Quality circles, attitudes, and

productivity in a circuit board factory. Annpal Qentjerenee, Intetpatjepal
Witch. Boston. MA.

STINNETT, W. D., PERRILL, N. K., BOSTER, F. J. & BORSTELMANN, S. (1983).
Employee Involvement, Attitudes, and Productivity in High Technology
Manufacturing: A Two-Year, Experimental Study. Jnternatienal

m ni in iinAnnulonfrnc.

STOHL, C. (1984). Quality Circles and the Quality of Communication. [AQQ

 

157-162.

STOHL, C. (1987). Bridging the Parallel Organization: A Study of Quality
Circle Effectiveness. Cemmenieatien Yearbeek # 19, 416-430.

TORTORICH, R., THOMPSON, P., ORFAN, C., LAYFIELD, D., DREYFUS, C. &
KELLEY, M. (1981). Martin Marietta Aerospace Michoud Division New
Orleans, Louisiana. The Qeality Circles Jeurnal, A, 25-34.

UNITED STATES GENERAL ACCOUNTING OFFICE (1981). Use ef epality
eentrel eireles in the federal gevernment, Report #FPCD-8l-31.

VAN HORN, W. L. & STINNETT, W. D. (1984). The ideal work environment:

Total employee involvement. SAM Advanced Management Jeurnal, 40-47.
VROOM, V. H. (1959). Some Personality Determinants of the Effects of

Participation. Jeernal of Abnormal Psyehelegy, :2, 322-327.
WEISSENBERG, P. & GRUENFELD, L. W. (1968). Relationship between job

satisfaction and job involvement. Jeprnal ef Applied Psyehelegy, 5_2_, 469-
473.

WHITE, D. D. & BEDNAR, D. A. (1984-84). Locating problems with quality

circles. Eatienal Predeetivity Review, 4_, 45-52.

57
WHYTE, W. F. (1987). From human relations to organizational behavior:
Reflections on the changing scene. Indestrial and Laeer Relatiens Review,
1Q, 487-500.
YAGER, E. (1979). Examining the quality control circle. Persennel Jeerpal,
682-684.
YAGER, E. (1980). Quality circle: a tool for the ’805. Training m

Qevelepment Jeernal, 60-62.

YAGER, E. (1981). The quality circle explosion. Tr inin an D v 1 m n urnal.