AN INVESTIGATION OF ORGANIZATIONAL
EFFECTIVENESS USING MULTIVARLATE ANALYSIS

THESIS FOR THE DEGREE OF M. A.
MICHIGAN STATE UNIVERSITY

B‘s’fi’ON ROBERT BGYLES
I 9' 6 2

_i HESib‘

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

 

._; w“ P J ’
.a’g.."-I-: it.” I.
1 '1‘ .1” E
l 3’ f 31‘: 1 1‘
A, 1

(
Mié “3.16311 “mm

Univ ‘5,me

 

 
  

ABSTRACT

AN INVESTIGATION OF ORGANIZATIONAL
EFFECTIVENESS USING MULTIVARIATE ANALYSIS

by Byron Robert Boyles

Industrial organizations are an extremely important
part of our way of life. The decisions made by their man-
‘s made on the basis of adequate and relevant
nd evaluated against relevant criteria. The
COMPLCA--‘ the relationships among measures of organi-
zational performance and the difficulty of identifying rele-
vant criteria make a systematic study seem worthwhile.
Correlational and multivariate analysis techniques were
employed in such a study.

Twenty-four measures of the performance of a medium-
sized manufacturing organization over a sixty-month period
were collected. Accounting and personnel records were used
in the study instead of the more usual kind of psychological‘
data. This seems appropriate because such data is the basis
for the decisions made in the industrial situation. These
measures were intercorrelated. Factor analysis was used to
Simpiify the description of data by reducing the number of
necessary variables, or dimensions. Six factors resulted.

Four or these factors were discussed and interpreted in light

 

Byron Robert Boyles

of experience with the organization under study. Two dimen-
sions were seen to account for most of the variance: (ll'Labor
effeciency and (2) Contract—production. Two additional fac-
tors were interpreted: (l) Disruptive job changes and (2) Com-
pany-brand production. Correlation tables involving the
variables loaded on each of the six factors were prepared and
discussed. These tables were found to be helpful in under-
standing the relationships among the measures of organization-
al performance of interest.

The general conclusions reached were two fold:

m
P.
'1

st, four dimensions seem to account for most of the vari-
ance among the measures of organizational performance under
study. Of particular interest is the fact that two inde-
pendent factors involving the labor force emerged. These
four factors can be seen as dimensions of effectiveness for
this organization. Secondly, correlation and multivariate
analysis techniques using accounting data and personnel data
can be useful in understanding the complex relationships
among measures of organizational performance and in identi-

fying relevant criteria against which to evaluate this

 

 

performance.
Approved zkw&/:;Q§:“/K*”>."
/// .. .
Datefiy.-’;'/Q;JL*_’,,I
'/ /;

L. . I o

 

AN INVESTIGATION OF ORGANIZATIONAL

EFFECTIVENESS USING MULTIVARIATE ANALYSIS
by

Byron Robert Boyles

A THESIS

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

MASTER OF ARTS

Department of Psychology

1962

ACKNOWLEDGMENT

I would like to express my appreciation for the
guidance, advice and encouragement given me by Dr. Carl F.
Frost and Dr. Terrence M. Allen during the planning and
execution of this thesis. My thanks also go to Mr. William B.
Eddy for his help and cooperation.

I would also like to express my appreciation to the
members of Revco, Inc. whose help and complete cooperation

made this study possible.

ii

TABLE

LIST OF TABLES . . . . . .
INTRODUCTION . . . . . . .
THE PROBLEM . . . . . . . .
HISTORY OF THE PROBLEM . .
METHOD AND PROCEDURE . . .
RESULTS . . . . . . . . . .
DISCUSSION . . . . . . . .
SUMMARY AND CONCLUSIONS . .

BIBLIOGRAPHY . . . . . . .

OF CONTENTS

iii

Page

iv’

17

21

36

4O

43

Table

10.

Intercorrelation
variables . . .

LIST OF TABLES

of organizational

Rotated factors and factor loadings

Intercorrelation
on factor I . .

Intercorrelation
on factor II . .

Intercorrelation
on factor III .

Intercorrelation
on factor IV . .

Intercorrelation
on factor V . .

Intercorrelation
on factor VI . .

of variables loaded

0 O O O O O O O O C

of variables loaded

of variables loaded

of variables loaded

of variables loaded

of variables loaded

Rotated factors and factor loadings
for production variables . . . . . .

Rotated factors and factor loadings
for people variables . . . . . . . .

iv

Page

22

26

29

31

32

33

33

34

35

35

 

INTRODUCTION

One would be hard pressed to overemphasize the im-
portance of industrial organizations to our way of life.
The work force--the bulk of the adult population--spends
more than a third of its waking hours in the organizations
by which it is employed. The ubiquitousness of organi-
zations is not their sole or principal claim for attention.
As social scientists we are interested in explaining human
behavior. Taking the social psychological point of view, we
are interested in what influences impinge upon the individ-
ual human being from his environment and how he responds to
these influences. For most adults industrial organizations
represent a major part of the environment. Moreover, we
would expect industrial organizations to have an even more
significant effect upon behavior than is suggested merely by
looking at the time spent in them. Organizations in general,
and industrial organizations in particular, are important in
enabling us to achieve our present way of life. The decisions
made by their stockholders, boards of directors and managers
are of great importance to what is popularly called the
public welfare. So much so, in fact, that the President of
the United States has recently seen fit to intervene in the
process by which these decisions are made. What criteria are
relevant to such decisions? What consideration has been

given to their validity?

The modern industrial organization is a complex
entity. Buildings, machinery, power, raw materials, and
people are all brought together in a number of specific
but different ways for a particular purpose: to produce a
saleable product or service. However, it must be done in
a way that will not lead to the destruction or depletion of
the resources of the organization. Caplow (1953) has
altered this statement into a formal postulate of criteria
of organizational success: "Organizations tend to maintain
themselves in continuous operation." However, he adds

three related criteria which appear equally essential:

I

1. "Performance of objective functions as measured
by institutionally imposed standards.

2. "The minimization of spontaneous internal conflict.

3. "Maximization of satisfaction for indiViduals.”
Caplow thus emphasizes that organizational effectiveness is
multi-dimensional——not unidimensional as is sometimes thought.

In one sense, this complexity makes the study of

organizational effectiveness seem a bit discouraging: On the
other hand, decisions will be made, courses of action will
be chosen and changed in this industrial situation which
determines the way of life we know. Therefore, the search
for relevance in this complexity is most worth while in
determining the criteria of organizational effectiveness

necessary to improve the quality of these decisions.

THE PROBLEM

The first purpose of this exploratory study was to
investigate the usefulness of correlational and factor
analytic techniques in handling the complexity of the vari-.
ables involved in developing criteria of organizational
effectiveness. This is hoped to be a first step in the
analyses of organizational data upon which to begin to build
a descriptive organization theory developed inductively from
facts rather than deductively from assumptions.

The second purpose was to identify the particular
combination of measures of organizational performance which
would best describe the effectiveness of one particular
organization.

The third purpose of the study was to investigate the
use of operational accounting data and personnel data derived
from organization records in studying the relationships be-
tween certain measures of personnel performance and the per-
formance of the organization as a whole. The procedure was
to use accounting and personnel records which are available
in surprising quantity and completeness in most modern organ-
izations rather than the more traditional kinds of psycho-

logical tests, interview or questionnaire data. 1

HISTORY OF THE PROBLEM

Current approaches to organization theory by students
of organizational concepts have tended to classify the
theories into two broad categories. The first category has
been variously labeled traditional, classical, bureancratic'
or "theory X." The other has been called modern, human
relations, democratic, participative or ”theory Y.” -

The first of these categories has its intellectual
roots in two areas. One is the early scientific management
and administration school of people like Frederick W. Taylor
and Henri Fayol. Taylor (1947) investigated the effective
use of people in industrial organizations; he set himself the
general task of organization theory: to analyze the inter-
action between the characteristics of human and the social
and task environments created by organizations. The actual
area of behavior that was considered by Taylor and his suc-
cessors in the scientific management movement was much
narrower, however. Because of the historical accidents of
their positions and training, and the specific problems they
faced in industry, Taylor and his associates studied primarily
the use of men as adjuncts to machines in the performance of
routine productive tasks.

In time and methods study, the scientific management
group was concerned with describing the characteristics of

the human organism as one might describe a relatively simple

A

machine for performing a comparatively simple task. The
goal was to use the rather inefficient human organism in
the productive process in the best way possible. These
efforts brought considerable precision of measurement into
the organization of the individual employee's production
activities. It raised and partially answered a number of
fundamental questions in human engineering. It showed
that it was feasible to specify precisely the activities
involved in routine production tasks. In this respect the
work in scientific management is more relevant to mechan-
ization and automation than to the broader psychological
aspects of human behavior in organizations. ’

The second area is the sociological theory of bureauc-
racy expounded by Max Weber (1947). His major interests in
the study of organizations appear to have been four: (1) to
identify the characteristics of an entity he labeled ”bu-
reaucracy”: (2) to describe its growth and the reasons for
its growth: (3) to isolate the concomitant social changes:

(4) to discover the consequences of bureaucratic organization
for the achievement of bureaucratic goals. To be sure, Weber
goes beyond the "machine” model in several significant ways.
In particular, he analyzes in some detail the relationship
between an official and his office. But in general, he per—
ceives bureaucracy as an adaptive device for using specialized
skills, and he is not exceptionally attentive to the character

of the human organism (March and Simon, 1958).

 

The second generation or new school of thought is
largely the product of behavioral scientists, having its
origins in the work of Elton Mayo and Kurt Lewin, and colored
to some extent by the thinking of personality theorists like
Freud, Carl Rogers and Kurt Goldstein (Katzell, 1962).

Shepard (1956) has contrasted the modern with the
classical theories in the following respects:

1. ‘Wide participation in decision-making rather than
centralized decision-making.

2. The face-to—face group rather than the individual
as the basic unit of organization.

3. Mutual confidence rather than authority as the
integrative force in organization.

4. The supervisor as the agent for maintaining intra-
group and intergroup communication rather than the
agent of higher authority.

5. Growth of members of the organization to greater

eSponsibility rather than external control of

the members” performance of their tasks.

McGregor (1960), after several years of research spon-
sored by the S.1 can Foundation, has undertaken to express the
assumptions which seem to underlie each of the two general
theoretical approaches to organizations. Though the work is
done in a conceptual, deductive way and almost no empirical
evidence is cited, it seems worth reviewing. He concludes
"theory X" is based on three assumptions:

1. "The average human being has an inherent dislike
of work and will avoid it if he can.

 

A A —- .A‘.:-.a.dbnu 0.5a

2. "Because of this human characteristic of dislike
of work, most people must be coerced, controlled,
directed, threatened with punishment to get them
to put forth adequate effort toward the achieve-
ment of organizational objectives.

3. "The average human being prefers to be directed,
wishes to avoid responsibility, has relatively
li tle ambition, wants security above all."

"Theory Y," on the other hand, is based on the following

assumptions:

1. "The expenditure of physical and mental effort
in work is as natural as play or rest.

2. "External control and the.threat of punishment
are not the only meanS'for bringing about effort
toward organizational objectives. Man will
exercise self-direction and self-control in the
service of objectives t which he is committed.

3. "Commitment to objectives is a function of the
rewards associated with their achievement.

4. "The average human being learns, under proper
conditions, not only to accept but to seek
responsibility. ‘

5. "The capacity to exercise a relatively high
degree of imagination, ingenuity, and creativity
in the solution of organizational problems is
widely, not narrowly, distributed in the popu-
lation.

6. "Under the conditions of modern industrial life,
the intellectual potentialities of the average
human being are only partially utilized.”

Likert (1961) has formulated a rather more empirical
and complete statement of the "new way.” This approach
differs from McGregor's ”theory Y" mainly in that it is based

on empirical research. He sets forth a number of principles

based on the findings of the Institute for Social Research

at the University of Michigan. Briefly these principles

are:

l.

Likert

"A preponderance of favorable attitude on the part

of each member of the organization toward all the
other members, toward superiors, toward the work,
toward the organization--toward all aspects of the
job. These attitudes are those of identification
with the organization and its objectives and a high
sense of involvement in achieving them.

"This highly motivated, cooperative orientation
toward the organization and its objectives is
achieved by harnessing effectively all the major
motivational forces which can exercise signif-
icant influence in an organization setting and
which, potentially, can be accompanied by cooper-
ative and favorable attitudes.” These motives are:
a. The ego motives. These Likert defines
as "the desire to achieve and maintain
a sense of personal worth and importance."
b. The security motives.
c. Curiosity, creativity, and the desire for
new experience.
d. The economic motives.

"The organization consists of a tightly knit,
effectively functioning social system. This is
made up of interlocking work groups with a high
degree of group loyalty among the members and
favorable attitudes and trust between superiors
and subordinates. The leadership of the organ-
ization has developed what might be called a
highly effective social system for interaction
and mutual influence."

”Measurements of organizational performance are
used primarily for se‘f-guidance rather than for
superimposed control. To tap the motives which
bring cooperative and favorable rather than hos-
tile attitudes, participation and involvement in
decisions is a habitual part of the leadership
process. This kind of decision-making calls for
the full sharing of available measurements and
information."

proposes an "integrating principle" by which the above

may be achieved.

The leadership and other processes of the organiza-
tion must be such as to ensure a maximum probability
that in all interactions and all relationships with
the organization each member will, in the light of
his background, values, and expectations, view the
experience as supportive and one which builds and
maintains his sense of personal worth and

importance. (page 103)

Thus, two expressions of the ”new way." One con-
ceptual and deductive, the second more empirical and
inductive. The authors of these theories feel them to be
better than the fold way."

The shortcomings or malfunctions of the classical
theory have been pointed out. Such things as restriction
of output, low productivity, low job satisfaction,’resistance
to change, hostility to supervisors, frustration of indi-
vidual needs are alleged to be the results of the 'old way”
(Bose, 1957: Kahn, 1956; Katz, Maccoby and Morse, 1950)°

The ”new way,” incorporating the principles of
democracy, participation and human relations, has been
proposed to correct these shortcomings (Argyris, 1957; Likert,
1961: McGregor, 1960; etc.).

There is, however, enough evidence reported to raise
some question about the advantages of the "new way” over the
old in every organizational setting. Weschler, Kahane and
Tannenbaum (1952), Morse and Reimer (1956) and Leavitt (1951)
have found in several different kinds of organizations that,

although job satisfaction may be higher under the ”new way,”

production is not and may even be lower.

DEPART WENT C)!- l-‘SYCHOLUSY
MICH—IGAN STATE L'N2VER’SCTY
EAST LANSING. MICHIGAN

10

French, Israel, and Aas (1960) found, in a Norwegian
factory, that employees0 reaction to participation management
were not uniformly favorable.

Three years ago U. S. Rubber abandoned the committee
management system and divided the committee responsibilities
up among several executives (Business Week, 1960).

Thus it can be seen that simply assuming the "new way"
to be a better method of management is a bit questionable.
This, in turn, leads us to suspect that all is not well with
modern theories of organization. As Bennis (1959) has put
it: classical organization theory has addressed itself to
”organizations without people" and human relations theory
has tended to think of "people without organizations." The
two theories just reviewed seem to fall into the second
category.

What seems to be lacking is a program of study that
will combine the people and the organization. Such a program
should make possible a positive or descriptive formulation
of organization theory developed inductively from facts
instead of a normative or prescriptive one developed from
deduction and assumptions (from Katzell, 1962).

One example of the approach to organization theory
building that seems promising is that of March and Simon
(1958). They have tried to incorporate variables describing
the whole organization. Some of these variables are rela-
tively straight-forward and operational, like "cost of

changing personnel” or "size of work group." Others are

ll

complex concepts such as "motivation to reduce conflict."
One trouble is that they list some 206 variables to be
incorporated into their theoretical treatment of organizations.

A second interesting development not yet applied in
the industrial field is the ”Systems Development" approach.
Gagne and others (1962) discuss the aspects of this approach.
Nearly all the work done thus far in this area has had to do.
with military situations and the various man-in-space pro-
grams. However, the idea of considering planning, design,
development, training, operation, and evaluation of man-
machine systems as one integrated system seems to have con-
siderable relevance to the modern industrial setting.‘

These two approaches to the problem under consider-
ation hold promise. It is hoped that this study can also be
a beginning, of a somewhat different sort, toward a better
understanding of the relationship among some of the variables
which describe the over-all effectiveness of the organization.

Thus we see the need for consideration of the whole
organization in all its complexity, and the necessity for
developing a body of facts around which to build a theoret-
ical framework;

If this is to be accomplished, the relationship
among relevant variables must be discovered. Some method
of simplifying the complexity of the situation must be
developed and particular combinations of variables or key

variables must be found.

In order to approach the problem of the development
of a criterion of organizational effectiveness, we must first
determine the goals of the specific organization. The ideal
might be to specify goals that can be quantified in a way
that would allow us to develop criterion scales. In
practice, unfortunately, scaleable criteria of organizatiOn-
a1 effectiveness are extremely difficult to obtain. -
Dichotomous measurement of criteria is rather more easily
achieved. At the very least there must be a general statement
of the goals the organization seeks to achieve.

When the goals of the organization have been

\
’

specified, measurement of these goals becomes the next
problem. ‘The methods of measurement can vary from rigorous
objectivity to subjectivity. The methods of measurement
should be as objective as the particular situation and the
particular criteria permit.

A major consideration in the efforts to measure
organizational effectiveness is the method to use for ob-
taining results; irg., the strategy employed. The extremes
are represented by Drucker on the one hand and Georgopoulos
and Tannenbaum on the other.

Drucker (1954) states, ”The search for the one
objective is essentially a search for a magic formula that
will make judgment unnecessary” (page 62). He sees the
question of measuring organizational effectiveness as
unanswerable if effectiveness is considered a unitary di-

mension. The organization usually has more than one

 

 

l3

objective and the goals which it does have are usually of
such a nature that they will not lend themselves to meas—
urement. Drucker's idea is that ”objectives are needed in
every area where performance and results directly and
vitally affect the survival and prosperity (of the organi-
zation)." Nothing is said about the relationship among
these objectives.

Georgopoulos and Tannenbaum (1957) define organi-
zational effectiveness as the " . . . extent to which an
organization as a social system given resources and means,
fulfills its objeCtives without incapacitating its means
and resources and without placing undue strain upon its
members" (page 535). Georgopoulos and Tannenbaum's main
objection to the Drucker approach is that it relies on the
selection of.a priori criteria of effectiveness that seem
intuitively right without trying to systematize them into a
broader framework. The approach reported here has avoided
a priori selection and has instead studied the relationships
among available measures.

This difference in emphasis may be attributed to
the fact that an industrial organization is an extremely
complex entity and it seems unwise to look for one and only
one criterion of organizational effectiveness (Blau, 1960).
The relevant criterion might seem to be a function of who
is doing the looking. Union leaders, government tax
auditors, employees and stockholders all are concerned with

different aspects of the organization; hence, they have

4*- L‘ .“u.-h—- #-—--~'

15

purpose of industry.

If profit is accepted as the goal of industry, it
should be a simple matter to compare organizations, or one
organization’s over time, by means of a profit index and
thus assess their effectiveness. Closer examination,
however, indicates that short-run profit figures can be
misleading and comparability of profit figures over organ-
izations'of different kinds and sizes is difficult to
achieve.

Using profit as the sole criterion of effectiveness
glosses over too many variables.) It ignores the complexity
of the problem. Various techniques have been developed to
parcel out the effect of size on profitability (Albery,1954:
Bartlett, 1952). No accounting technique has thus far been
unanimously accepted.

A somewhat similar problem exists when we try to use
productivity or volume as the sole criterion of effective-
ness. Both large and small organizations can maintain a
high rate of productivity while slipping deeper and deeper
into financial deficit. Also our culture has become wary
of organizations which seem to be gaining too large a share
of the market. Anti-trust legislation is an expression of
this feeling.

Drucker suggests the need for a number of criteria,
but does not consider the relationship among them nor how to
select them. Dent cites that profit is still considered the

most important purpose of industry. But no one measure of

p.-..-‘.~, ‘ o ‘

_-.- _...--»..—.4 . r

16

organizational performance such as profit or productivity
seems entirely satisfactory. These approaches to the problem
and that taken by Georgopoulos and Tannenbaum seem to be a
sort of "organization without people” approach. The trouble
lies in failure to consider the entire situation. As
Bennis (1959) has said, either we have "people without
organizations" or "organizations without people." What is
needed is an empirical rather than conceptual approach in-
volving the organization as a whole which will allow the
formulation of a theory of organizations inductively from the
data. This is a large order.

A beginning must be made somewhere. 'That begin-
ning might best be made with measures of organizational
"outputs." Measures of "outputs” are of interest for their
own sake, but they are also of interest as what seems to be
the only way to get at the ”process" which goes on in the
organization to produce the "outputs.” As we begin to under-
stand the relationships among the "output" variables, per-
haps we can work back through them toward an understanding
of the "process." When we know more about the ”process” in
an organization, we will be in a better position to determine
what combination of "inputs" will lead to effective and
efficient ”process" and, finally, to desired ”outputs.”

All that is a long way off: at this stage we will

just be concerned with "outputs.”

METHOD

The method here employed was to measure twenty-four
variables over a sixty-month period in the life of one
organization. The organization chosen was a medium-sized"
manufacturing firm in southern Michigan. They manufacture
a nationally distributed refrigeration product requiring
approximately four hundred and fifty production personnel.
Two general categories of production exist. One is sold
under contract and distributed under a nationally known
brand-name. The other is sold under the company's own
brand-name. Inasmuch as the firm introduces a new'model
annually, their research and development staff and the

1 engineering group are quite important to the continued suc-
ces of the product. The hourly people have been represented
by the United Auto‘Workers union for the past ten years. It
is a typical medium-sized manufacturing firm.whose records

.-:r‘-’are_£airly complete and whose managers were willing to allow
the author to investigate their operations.

It is felt that comparing one organization over time
is superior to comparing several organizations or several
departments of one organization. In this way the variation
in measures can be more confidently attributed to changes in
performance of the organization. Also the exploratory

nature of the study requires that it be kept as simple as
possible.

17

18

The time span employed (60 months. from December 1956
to November 1961) is a compromise between what might be de-
sirable (the whole life of the firm?) and the availability_
of complete records. .

The decision to use accounting and personnel records
instead of the more usual kind of psychological data was
made for several reasons. Most important of the reasons is
the fact that this kind of data has been used by this or—
ganization as a guide to operations ever since it began
operations. It is the author's belief that such data is an
Objective representation of the behavior of the organization
and the personnel. The occurrence of a grievance in the
industrial situation is as objective an indicator of behav-
'ior as the lever pressing response of a rat in an operant-
conditioning situation.

Other reasons for using this type of data are:

(a) the more usual kinds of psychological data are not a-
‘vailable for this period: (b) the gathering of such data
would very likely have an effect on the functioning of the
organization: and (c) collecting such data whether by
questionnaire. interview-or whatever, over a long period of
time, would very likely influence the variables being
investigated. . . . I '

Twenty-four measures of the performance of the
organization and of'the personnel employed by the organiza-
tion were chosen. Their selection was made won the basis of

xxtility and availability according to the Judgment of

19

accounting personnel of the firm and members of the School

of Business at Michigan State University: and the judgment

of the investigator and those with whom he is working on

.4..— o

the research project of which this thesis is a part.

The following is a list of the variables used (each

a..measure is per month}:

1.

(DUO-F005)

7.
8.
9.
10.
11.
12.
13.

14.
15.
16.
17.

18.
19.

20.

21.
22.

23.
24.

Operating‘profit“

Total units produced

Sales of units under contract

Sales dollars on the above units

Sales of units under the firms own name

Sales dollars on the above units

Units shipped by the entire industry

Man hours of direct labor

Units produced/Man hours of direct labor

Number of direct labor employees

Number of indirect labor employees

Number of salaried employees ‘
Bilt-ins/Total units produced (A measure of the
difficulty of production. Bilt-ins are the most
complex brand-name product produced.)

Actual productiOn/Scheduled production

Percent of bonus paid

Material variance (VariatiOn of material costs
around the amOunt allowed per unit.)

Direct labor variance (Variation of labor costs
around the amount allowed per unit.)

Number of Grievances Filed ‘ '
Bids for jobs that become available on the basis

iofi seniority

Factory overhead variance (Variation in overhead
costs--including indirect labor--around the amount
allowed per unit.)

Bumps (Workers transfers on the basis of seniority
as a result of job eliminations.)“

Number of people who actually change jobs on the
basis of seniority as a result of eliminated jobs.
Average hourly wage

Average hourly wage with bonus

These variables were first intercorrelated. Then a

factor analysis using the principle-axis method and a_

quartimax rotation. was performed. Correlation matrices of

reduced size were then prepared from the factors resulting

* -. '_"“r W. 'v—vVJ .

'1'" ..

_ ‘ fvw—rfi —-q—-———-

 

 

20

from the factor analysis. A factor analysis was also per-
formed involving two separate sets of the data. One set
consisted of the first fourteen variables listed above. The

second set was the last ten variables listed above.'

RESULES

The first purpose of this study was to investigate
the usefulness of correlational and factor analytic tech-
niques in handling the variables involved in developing a
better understanding of criteria of organizational
effectiveness. _ _ . .

Table l/page 22 is the correlation matrix of all
'twenty-four variables. In some cases the correlation co-
'efficients have no readily-explained.meaning and in others
they are a function of bookkeeping techniques. Also the
fact that these measures are all on the same organization
over the same period of time makes it likely that some of
the coeffecients are artificially inflated. .

The interpretation of this table is rather difficult.
-As this is the first of such studies. it is difficult to
determine the significance of a correlation ofagiven mag-
nitude. None the less, some meaning can be gotten from the
relative magnitudes.“ f _

For example, it can be seen that Operating Profit_
correlates rather highly with both contract Units and Company
Units. but not so highly‘with Industry Units. Then by look-
-ing at the correlation between Industry Units and Company
Units and comparing it with the correlation between Industry

; .
Units and Contract Units we can see where the discrepancy lies.

21

22

I

 

on- em: ovu mm «on mm om- em. as am so no cw mscomxomm3_>~usom
own as- men as «on ma an- mm- man eon man ma: mm memos,»auaom
«m . on nu ma: 5H man no mo man man do mm: - momcmao son finance
me «v as on: ma 5H- so. no as. man no: on: an «mean
we: swu No: we on on «a as mm mm «m mm ow mocmaum> ommaum>o.suouuas
on hm cw no- no ma ow mm mo Ad on on ma mean
co Ho mm ea- no no mm an «a as mm mo: ma madam mmocm>uauo
no: em: mmu,nm no mu mm- «n. no oH mo- no: an mucmaum> scans saunas
mflH.. eon mo ma ma m~ mo mo mo ma cu Hon ea moanaum>.amaumuas
om: pm. pm. an no me mo- mo: mm on an ea ma mason mo anon use
no: «on man mm «a HH do- no ow ma ma can we coauuseoum emflsemaum\amsuue
Hm ma «on man man no- ma as me- me: on: on: ma means Hmuoa\m:auuaam
a we on mm: as can on an ad we as con ua Hucaomumm coaumanm
. nu ma- me we an ac nu ma av mo Ha oouom manna uumuaecu

u mo- «n ma we we we ev on o~ o” uuuom scene pooped

n ma we we o~ mm mm be ow a snow cmx\uu«:s

,-r n ma Ha as on He an we a . nexus: «such an:

n aw ma om -mv; me- on r::s:;v;. moans suumsenn

u om cu cm on on o mumaaon handsoo

u mm ma «m an m «use: manages

u mm on mm v mumaaoo uummucoo

u eh «m n moans uumsucoo

u me ~ nudes Hence

u a venous meauauomo

«H Ha oa m m n m e m w a moanwauno

 

 

. .1

.meanmaum> Hmsoaumufismmuo mo coaumaouuoouousH .H.0Hnea

\

 

23'

2. pm. 3.. on 3- mm- me «a 2. 3 on- «a mason}. 0mm: 3.33

 

.. 3- d... 2.. mo- 2.. 3 am 2. 3.. 3: mm mamas 3.25m
. pm we- ma mm mm- ma- en- mo ma «a «menace hoe Hugues

.. he- ma. mm ms. nu. aw- mo c” an .. «dean

, u no- an: an ma on pa mu: om oocmaum> ommnum>o suouunu

. «m mm- um mu- so. an ma , «can

u an. no He- «on nu ma . _cmaam moccm>oaeo

. on oe ma man pa mocmaums Hanna pounds

I e." do «on w." mosmaum> awakens!

u an mm- ma ascom.mo uses you

n mo «a coauoaeoum omasmosumxamsuoe

n ma _ muses Hmuos\unanu~«n

. NH . Hescomumm teammaeu

. a” ouuom noses uuoufiucn

OH . uuuou nonunéooudn

te, m.. snow c~x\mu«ab
. - t - {if} m 0.03.33 mason as!

h .. . . . nuwcb aumspsH

o . -. . a. numaaon 539.80

m . .. w R335 hang—BU

e .. amazon Hummus—00

- n . muses sausages
u uses: annoy

a uamommwmcquuuomo

ma - HN on ma ma .ea ma ma ea ma . «macaques

 

W

 

cmsmmmcoo as manna

W‘-

24

These correlations do not. of course, indicate
exactly what to produce. but they do give an idea of the
relationship between the two main divisions of production
and the performance of the entire industry. Thus we have
an idea of company performance relative to the whole in-
dustry. fWhen the.industry units are high. the company may
be forced into supplying contract units at the expense.of
their own company units. .Given forecasts for the whole
industry, we might also be in a better position to determine
where it would be best to concentrate future production.
Still another use of knowledge of such relationships might
be in evaluating.the company sales organization and their
———« -effectiveness in concentrating on contract or company units

in relation to industry trends. '

These relationships are rather interesting and add
”‘information on the performance of the organization and the
“relationships among various measures of performance. But

by far the most striking feature of a study of Table l is
the fact that there is too much information there for one to
keep in mind and to use effectively. In an attempt to re-
duce the number of variables necessary to keep in mind, this
matrix was factor analysed. / '

The principal object of factor analysis is to simpli-

fy the description of data by reducing the number of neces-
sary variables..or dimensions.. In this study sixty measures

of twenty-four variables were taken. If by factor analysis

<‘

25

we find that five or six factors are sufficient to account
for all the common variance covered by the original twenty-
four variables. considerable simplification will have been
achieved.

The factor analysis was performed on the MISTIC,
Michigan State Integral Computer. The procedure was to use
one as the estimate of the communalities, extract factors by
the principle axis method and then orthogonally rotate the
factors using the Ouartimax method developed by Wrigley and
Nauhaus (1954). The criterion for choosing the number of
factors to rotate was that suggested by Kiel and Wrigley
(1962) for the Ouartimax method: that is. rotate the highest
number of possible factors which will still give at least
two variables with their highest loading on each factor.

The Ouartimax method was chosen because with the
small sample size (60 time periods) a unique analytical
solution with orthogonal rotation is preferred to other
methods which may be more effected by chance variation. In
addition, no prior classification of variables nor assump-
tions about the presence or absence of a general factor is

needed. Its aim is to find the orthogonal transformation

_which maximizes the variance of the factor contributions

(i.e. the squared factor loadings). Since the total variance

is constant, an incidental consequence of the method is to

"increase the number of zero or near-zero loadings as well as

the size of the larger loadings. To this extent the method

26

is an approach to "orthogonal simple structure” (Wrigley and
Nauhaus 1954).
Table 2. page 26. shows the six rotated factors and

the variables which loaded high on each. Factor loadings of
.40 and higher were included. Loadings of .39 are shown in
parentheses. It will be noted that factors I and II account
for a large portion of the total variance. The Eigen values
for the six rotated factors are: 6.58. 5.83. 2.10, 1.68.
1.43, and 1.00. ' - '

Table 2. Rotated factors and factor loadings.

 

 

Measures I II III IV V VI

 

1. Operating Profit -- .55 -.41 -- -.49 --
2. Total Units -- .84 -- -- -- --
3. Contract Units -- .94 -- -7 -- --
4. Contract Dollars -- .95 -- -~ -_ --
5. Company Units -- -9 -- -- -.86 --
6. Company Dollars -— -- -- —— -.87 -—
7. Industry Units -- .59 -— -_ -- --
8. Man Hours'Worked’ -- .53 -- -- -- -.53
9. Units per Man Hour -.42 .66 -- -- -- --
10. Direct Labor Force .73 .45 -- -- -- --
ll. Indirect Labor Force .83 -- -- -- -- --
12. Salaried Personnel ‘ .93 -- -- -- -- --
13. Bilt-ins/Total Units -- -.49 -- -- -.47 --
(Difficulty of Mix) . 4
l4. ActualPScheduled Prod. -- -- .54 -- -- .54
15. Percent of Bonus -.66 (.39) -- -- -- --
16. Material Variance -- --' —- .71 -- --
17. Direct Labor Variance -.77 -- -- -- -- --
18. Grievances .68 »-- -- (.39) -- -,
19. Bids‘ ' ' ' -- -- -- .70 -- (.39)
20. Factory Overhead Variance-.51 .69 -- -- -- ~-
21. Bumps ' ' .41 -- .78 -- -- --
22. Actual Job Changes .43 -- .78 -- -- --
23. Hourly'Wages " ' -.82 -- -- -— -- --

24. Hourly Wages with Bonus -.77 -- -- -- -- --

 

.'

-..———' '

’,.‘.‘.5""'
sw"

27

A discussion of the interpretable factors and a
correlation matrix of the variables loaded on each factor
follows. 2
. Factor I indicates that in months when the number of
people in the labor force (Direct. Indirect and Salaried),
Grievances and the two measures of Bumps (Bumps and Actual
dob/Changes) are all high then Units per Man Hour, Percent
of Bonus, Direct Labor Variance, Factory Overhead Variance
and both measures of Average Hourly‘Wages will be law.

These relationships can, perhaps, be explained by
considering the effect of an increase in the size of the
labor force. The people who are added in the production
work force are new. inexperienced workers who receive lower
than average wages, (Hourly Wages, Hourly Wages w/Bonus).
The disrupting effect of adding workers (Bumps, Actual Job
Changes) and perhaps a new production line plus the inef-.
feciency of the new workers acts to reduce the effeciency of
production (Units per Man Hour, Direct Labor Variance,
Factory OverheadVariance').1 This results in a decrease in
Units per Man Hour and inPercent of Bonus paid and an in-
crease in certain costs7per unit which show up as negative

(Direct'Labor‘Variance.jFactory Overhead Variance):- All of

 

1This factor includes all three components of the
labor force. and not just direct labor. The loadings of the
three variables are in reverse order of magnitude from what
we'would expect on the basis of productiOn needs. Although
the differences among the loadings are rather small. this
finding could cast some suspicion on the employment practices
cf the organization. It should at least lead to further
investigation of this area.

 

28

these changes will demand the accommodation of an increased
number of workers resulting in an increase in grievances.
Generally then, this factor is the cost of production
‘ or more precisely the direct and indirect ”cost“ of the labor
force necessary for a sustained rate of production. It 2
could also be thought of as the performance of personnel in
‘the organization as opposed.to the organization itself.
Perhaps the relationships of the.variables in
Factor I can be seen more clearly by referring to
table 3/page 29. Table 3 is a reproduction of the portion of
the original correlation matrix found to be weighted on
Factor I.
Factor II indicates that in the months when pro-
duction of Contract Units, Operating Profit. Total Units,
Industry Units. Man Hours worked. Units per Man Hour. size
of Direct Labor Force. Factory Overhead Variance, and Per:
cent of Bonus are high Bilt-ins/Total “9433,13 low. This .

-'-o..-- V‘s -- d-c’a—s ‘5‘“.-

factor could be thought of as the performance of the

9159991299993..1.!‘..,B"99“F¥99.999-3%???“ Pmductv " °PP°'°d
to the performance of the personnel.

3%1 the variable? "bl?“ axe high 99 this f¥9t9r axe-.-
related to_a high level of production of contract units. The
negative loading of Btthins/Total.U“¥F!,£°;;°V3-1£.V°,9°“'
sider that when the company has the demand to produce contract
units they do so at the expense of the production of company
units. The positive loading of industry Units points out the

. I .
relationship between the contract units which are sold

I. .
P L

 

29

7

.» mm- hm- om can me. me on- em- o4- mm «H mucom\3_omms.sauaom .np.-.
wen; He- pm me- am .oe em. as- me- as as mamas saunas .mm.-.

pm me- «n mu- «m. «m on pH mm. as mmmnmno non Hugues Ame. .

 

- we- m~ as. am. we we as own a «mean .He. .

. mm- am on we- sun uoL mm m .uu> emmsum>o seasons .3m.-.

- am. He- om ~e mm as. p mmucm>oeuo .mm. .

u oe he- mm- mm- um .m .um> gonna saunas inn ..

- on- an- pa- mm m mason mo sumo son .00.-.

. we ob mm- 4 Hancomumm emaumamm .mm. .

n ma ma- n auuom gonna uoouecen .mm. o

n no: N venom umnmq unseen Ann. v

n A .uaom mmr\muwco .Ne.-.

om m m r e m e n w a amasmaum> .mcaemoq.

.H uouomu so powwoa meanmeum> mo soevamHuouumusH .m wanna

I'lul- .IO , It. 9. II .I. .| . a a ..|.. In. s-

30' -~ ~~ -

nationally under a well known brand name and the fluctuations
of the national market. , | i
_ It is noteworthy that only Direct Labor is loaded on
this factor. .This confirms the practice of the company of
pricing contract units primarily on direct labor costs.
Factory Overhead_Variance is favorable in relation to the H
V91um9 9?-cqnt¥a°$.unét8v#:en adV§9Eé969vslxhabB°rPed by a
iapgesunuwberuef “913?: .rhe high p98¢t¢ve lpad+n9 0:.Units.
per nan_Bour_can be accounted for by the relative simplicity
of the Pr969ct¢°9 rsquissd t°.fU¥€§11 the Cont¥s9t9- .-\..-
In view of the positive loadings of Factory Overhead
Yasianceo wreak P3991“. we. Limits 99: Man. Hour. and. the
"“P¢§ltivs islazionéhifi‘téissérating P¥9fik asd_P9rce9t.9f_, H-.-
Bonus we can concinde that contract operations are advantageous
-ew ”to theorganization--at least with-respect to these

5%

considerations. h A”
. __.T§9 vaséablsa loade¢_oa.ra¢F°£.II-haY¢.b¢en taken.“
from the original correlation matrix and presented_in Table 4.

9399.3}:- This will give the reader a better idea of the

relationships among these variables} ---
..As was mentioned earlier. Factors I and II account
for a large portion of the variance in the original_cor-, . '
sslafié9a matrix- “The fellavins fa¢t°¥9.é=9: th°¥9f9Fez loaded
on fewer variables and are more difficult to interpret. _
. H Faotor III shows that in months in.which Bumps, Actual
Job Changes;and Actual/Schedaled Production are high._Operating

Profit is low. The relationship betweenBumps and Operating

31

 

 

. cw- «on we on mm mm «a an mm as .um> camaum>o auouomm .mo. 9
. eo- ma- man no: me- me- can ma- a usage Hmuos\mcauuaan .me.-.

. mo- ~m_ m~ we we on an m couch gonna unused Ame. .

- as we mm mm be cc 5 mason cmz\uuans “we...

. ma an ac an no . m ensue: mason car..mm. .

u om me me on m «use: muumsecn 1mm. .

. mm oh mm ,4 «season nomuucoo Ana. 3

u m» an .n mamas nomuucoo ~em. .

n we a nude: Hmuoa .em. .

. u a sewage mcaunumdo 1mm» .

as m a e m m. e. .m m H maamaum> amusemoa.

 

.HH uouomm so pepmoa meanmaum> mo coeumaounooueusH

. 3

I t ’

’

'0 .IC

is I,

.4 wanes

’ .o,- 4‘\Cu evIsIseIL)\s§s loll D |.. I.

32

Profit reflects the effect of the direct work force employees
accommodating themselves to new*work assignments. Reference
to the size of the correlations between Operating Profit and
Bumps (and Actual dob Changes) and between Actual/Scheduled
Piodsction as? 39NP§M(§99 “sent! J°”-9P‘“9991LP¥°P¢3¥°9 #n.
Tab}9_§uénéécasse Eye 9999 £9.119%§.¢9r 19F9r2¥9tétéon to the
Esléti9espie_§9t8999.99998;299-£§9€19;“Mrhé'-99293£'-§9 a: a
929¢%SIC.faCEeriéaxelYéaa.295999n21-9ba99§i.éaéspsnéensi°§,

§§°t9§-;1 Th°H§Bt9r9955942919B§u99999-Eh°-Y551§Plfifu199999

on Factor III are abstracted from the original matrix and
-udmnind - _H_‘ r . ‘ _.
presented below. " '

- A-v

Table 5. Intercorrelation of variables loaded 6n factor 111‘

 

 

 

(Loading) Variablet. . f ' 5 A l 2 ‘3 4

<-:4%’°29§§t1n9-?t9££t_‘- 21 ' -. _
1’7 ( 354) Actual/Scheduled Production 2 . 5991 -
““‘<'}781 Bases. ‘ - ' ~ .3 ~ -§§ 08 5’

( .78) Actual Jab Changes ‘4 134 08 97 -

 

Pactos $2 is leased 99 99¥¥ $82 996 29.-£?A¥ 2.th1rd
vaaéa??e&:- Asutbers-%9 no upper??? r9399! “9?:“9E351331

'Yaréaass_é9d 9369.9h0uld be Pb? 99$9HY9¥§‘P;§! leased en

‘e‘o—a -- '-

wvA-iolv-w 4.... '.m.

thi! f'°t9£'.ER”EFfififpf-EFM1999rPESFEE1°“‘!§1l-ffi-flfd°'
_ ?hQ;V§F§‘P¥9§,}9§d°d.99nr‘9t°r IV and their correlation

coefficients are presented below.

33

Table 6. Intercorrelation of variables loaded on factor IV.

 

. - g .

 

 

(Loading) Variable J .' éJ . 1 ... 2 -I .d 3
(.71) Material Variance f _l -

(.39) Grievances" 2 OS -

(.70) Bids 3 27 52 -

 

-MHHM- Paste! V £8 1999?9m99 QPEEat§99-?=°féEz €999snvn994t9:
CQTPaFYLD°t;§:§-9PQH91AFTt9‘/?9F§1.9933‘flerh43d§9939rD§9§“}t'
from the relationship among these variables in months when a
large number of Company Units are being produced. Just as
‘ in the months when contract units are beinggproduced} profit
is made when company units are being produced. I

Table 7, page 33 presents the intercorrelations of
the variables loaded on Factor V. An examination of this
table reveals that Company Units correlate positively‘with
Operating Profit while the ratio between Bilt-ins and Total
Units correlates negatively. This suggests that the company
product exclusive of Bilt-ins is produced profitably or

conversly, Bilt-ins are not profitable.

Table 7. Intercorrelation of variables loaded on factor V.

LA‘A-‘A 4. .1 4... AL

 

 

(Loading) Variablo . r_ .11; l: AALZ .- :3 _A4
(-.49) Operating rrds1t*? '* '1 Y - ’”‘” '
(-.86) Company Units 2 . 51 J
(-.87) Company Dollars , 3‘ 58 -. 96 -

4 -lB l9 l9 -

 

(-.47)‘Bilt-ins/Total Units

., q

34

.. . Factor V1 is another one with two and possibly three
variables loaded on it. No interpretation of this factor
will be attempted. Table 8 shows the intercorrelations among

the variables loaded on Factor VI.

Table 8. Intercorrelation of variables loaded on factor VI.

 

 

 

(Loading) variable; ; l 2 3
(-.53) Man Bourstorkedh' .-1 u l - .
( .54) Actual/Scheduled.Production 2 12 -
( .39) Bids 3 O3 01 -

 

In order to further check the reliability of the method.

a factor analysis, employing the same techniques as before.

‘was_performed on two separate portions of the data--the first

fourteen variables in one set and the last ten variables in

the second.set. The first set was taken to describe the
wv'-performance and dimensions of the organization--profit, units

of production. size of labor force. etc. The second set

describes the performance of the people in the organization--
’¥__ percent of bonus, direct labor variance, grievances, etc.

. _Although these are rather small numbers of variables
to factor analyse. the results are of interest. Inspection
9€HFebl?a9th'99_3§a FRV991? . laser‘sérse.£act9r-<i>- 9
senFsacta-Pseé‘essépe. £999?! In.) -396. -a . QWRPY'RIPd‘E‘StP“ ‘
factor (1.11) find} 92 m, touad gin-f 910.9491“; sunny .

.Table 19, page 35. is essentially a further breakdown

of the variables-loaded on factors I and III of the original

analysis.

 

Table 9. Rotated factors and factor loadings for

35'

production variables

 

 

 

 

 

 

 

Measures I II III IV
1. Operating Profit -- -.53 -.61 --
2. Total Units -- -,32 -- --
3. Contract Units -- -.93 -- --
4. Contract Dollars -- -.93 -- --
5. Company'Units --‘ -- -,39 --
6. Company Dollars -‘ -- -- -.9l --
7. Industry Units ‘ -- -.6l -- --
8. Man Hours Worked ' -‘ .46 -.51 -& --
9. Units per Man Hour -- -.74 -- .41

10. Direct Labor Force .78 -- -- --

ll. Indirect Labor Force .93 -- -- --

12. Salaried Personnel * .92 -- -- --

13. Bilt-ins/Total Units -- .49 ’5 .50

(Difficulty of Mix)

14. Actual/Scheduled Prod. -- -- -- .79

Table 10. Rotated factors and factor loadings for

. people variables-

H"Measures I 11 111 IV
'~-lS.v Percent of Bonus -- -- -— —.33

16. Material Variance .84 -— -- --

17. Direct Labor Variance -- -.63 -- -.41

18. Grievances -- .35 -- --

19. Bids .47 .72 -- --

20. Factory Overhead Variance -- -- -- -.68

21. Bumps —- -- .97 --

22. Actual Job Changes -- -- .97 --

23. Hourly‘Wages .49 -.49 -- -.4O

24. Hourly wages with Bonus -- -- -- -.82

DISCUSSION

The results of this study seem to indicate the
usefulness of correlation and factor analytic techniques in
dealing with the complex relationships among a number of
measures of organizational performance.

Industrial organizations such as the one in this
study have a wealth of quantative information compiled for
various reasons._ This information is used for tax purposes,
preparing the budget and the financial reports. With the
recent developments in medium priced digital computers and

vwwa-computer service centers who rent computer time, such organi-
zations could very well make additional use of this information.
The relatively simple correlation techniques used in this
study could be used by managerial personnel to improve their

“ understanding of the performance of their organization.-

The correlation analysis reported in this study
summarized a great deal of valuable, oft times not so obvious.
information about the relationships among these various
measures of organizational performance.) The extreme com-
plexity of the modern industrial organization. however.
leads to a problem. If we include all possible or available
measures of organization performance. the correlation matrix
becomes too complex to be of very much value. But if'we want
to understand the whole srganization. which variables can be
left out? This question cannot.be satisfactorily answered

.I
36 '

37

a priori. One solution is to apply multivariate analysis
techniques to the correlation matrix in an effort to reduce
nthe number of dimensions to a more manageable size.

. Factor analysis was used to gain this advantage. Six
factors resulted. Study of the variables loaded on each of
:.these factors leads to a better understanding of the under-
_m_1ying dimensions. Correlation matrices showing the intercor-
relations among the variables loaded on each factor help in
gaining a better understanding of the relationships among
the variables.

The first factor seems to reveal the effects of
increasing the labor force to meet increased production
demands. It would appear that the increase must be made at
the expense of efficiency. The quality of the additional
*workers plus the disruption of the routine keeps the expanded
‘work force from being as efficient as the smaller group was. .

This fact seems important when a temporary increase
in labor force is being considered.l Increased labor results
in increased production--but inefficient and costly
;production--at least during the period of adjustment. This“:
ifact seems to be one important dimension of an empirically
derived criterion of organizational effectiveness.

The second factor reflects a type of production. Just
as we would expect. an important dimension in accounting for
‘the variance among organizational variables is the level of
production. The factor is also positively loaded on Operating
Profit and Units per Man Hour. This indicates the need to

38

consider not just production or just profit but both--plus
some measure of efficiency. 80 the second dimension involves
.the'type of production.

The third factor seems to be the disruption caused by
a large number of worker transfers. The disruptive effect of
Bumps and job changes seems to result in a reduction of

“"”‘profits--probably through reduced efficiency of production.
That this factor involving personnel and job changes is
independent of factor I which accounted for most of the

tqspersonnel variance is interesting. It may indicate an area
”“Tof—particular importance for further study.

The fifth factor. the last one to be interpreted, is
ccompany-name production. This factor and factor II are both
production factors. The fact that the firm under study
Issually produces either contract or company units but not
both at the same time leads to two factors instead of only
one.

Thus we have four factors that are reasonably clear

dimensions underlying the variance among the twenty-four

measures of organizational performance. - X

1. Labor efficiency 5
2. Contract-production ‘
3. Disruptive job changes

4. Company-production

It must be remembered. however, that these results a
are just for one organization over one rather short period of
time. More work using more variables on the same organization
and, more work using similar variables on different organizations

will be necessary before any useful theory-building can be done.
‘ i

 

39

Several changes in the present study would have

improved it. More measures could have been included. The
most important of which is some measure of time. The months
could have been numbered from one to sixty and' included in
the analysis. This may have resulted in a factor of long-
run trends-~growth, perhaps. Measures of overtime, voluntary
turnover, waste or scrap. engineering changes made on the I
production line. better measures of various types of
production, etc. might have been included. On the other hand.
some measures are so similar that perhaps some could have been
left out. Two measures of the average hourly wage variable

_ and the bumps variableseem unnecessary.

Other studies of this nature would very likely be of
interest. Smaller units of an organization. divisions org ‘. \
departments, could be similarly studied and compared.

Comparisons between the departmentalidimensions and the
dimensions. for'the whole Organization could be made.

Studies of this type could be done for "good times” and for
"bad times“ and the dimensions compared. Most important, how-
ever. is to repeat this study on thefsame organization '
employing the suggested changes in variables. Then a similar

I

study should be done using appropriate variables on another

organization.

- r ' _#_..—.——

.1-

. T

_—-.‘..—-1—-'-1—.v—

_ .r—F—wgw

.‘w .—r “'1

 

-_'

.—.

 

SUMMARY AND CONCLUSIONS

The nature and history of present theories of
organization were discussed. The conclusion ofithe dis-
cussion was that modern theories of organization are
deductive and based on assumptions rather than inductive
and based on a body of accumulated facts. For this reason
they do not seem to satisfactorily describe the relationships
at work ina complex industrial organization. Nor do they
act as a guide to behavior in the industrial organization.

The lack of systematic study of the variables which
describe the entire organization was pointed out. Correl-
ational and multivariate analysis techniques were suggested
as a solution to this shortcoming. '

Twenty-four measures of the performance of a medium-
sized manufacturing organization over a sixty-month periodl
*were collected. These variables were intercorrelated.
jPactor analysis was used to simplify the description of data
by reducing the number of necessary variables, or dimensions.

Six factors resulted. Four of these factors were then

. .ciiscussed and interpreted in light of experience and familarity

_ggith the organization under study. Two dimensions were seen
1:0 account for most of the variance: (1) Labor efficiency
amnd (2) Contract-production. Two additional factors seemed
vwomth interpretation: .(1) Disruptive job changes and '

( 2) Company-name production.
40

 

 

41 ,_,

In order to assist the reader in studying the
relationships among the variables loaded on each of these
factors and to further demonstrate the usefulness of factor
analysis in studying this kind of problem. correlation tables
involving the variables loaded on each factor were prepared.
It is felt that tables such as these can be more useful in
understanding the performance of the organization than the
larger matrix involving all the variables used in the study.

The data used in this study was operational accounting
and personnel data derived from organization records. It is
felt that this kind of data describes the performance of the
organization more completely and accurately than the more
traditional psychological data.

Several changes were suggested which would have im-
;proved the study. Certain measures which were included could
Ihave been left out and several measures should have been
.included which were not.

Future studies of this kind seem most worth while.
{The same organization could be studied employing the recom-
rnended changes. Different organizations could be studied
tssing the same technique. Particular periods in the life of
an organization could be studied and the dimensions which
ennerge compared with the dimensions from another period.
Studies could also be carried out on smaller units of the
organization—edivisions or departmentssl The underlying

dimensions revealed at thistlevel could be compared with those

for the whole organization.

42

The general conclusions reached are two fold: First.
four dimensions seem to account for most of the variance
among the measures of organizational performance under study.
Of particular interest is the fact that two independent.
factors involving the labor force were found. These four
factors can be seen as criteria of effectiveness for this
organization. Secondly, correlation and multivariate
analysis techniques using accounting data and personnel data
can be useful in understanding the complex relationships
among measures of organizational performance and in

identifying relevant criteria against which to evaluate this

‘performance.

Boyles. Byron Robert

' BIBLIOGRAPHY

Albery, M. An appraisal of management. Cost and Man. ement.
1954. 27.

Argyris. C. Personality and organization: The conflict

between system and the individual. New York: Harper.
1957. '

Bartlett, R. H. Net rates versus operating profit on per
employee basis. Controller, 1952, 20. .

Bennis, W. G. Leadership theory and administrative behavior:
‘ The problem of authority. Administrative Science

Quarterly. 1959. Q.

Bose, S.K. A psychological approach to productivity .
improvement. ‘Conf. of Psychologists. Bombay, 1957.

Business Week, 22 October 1960. p. 101.

Caplow T. criteria of organizational success. Social
Forces, 1953, 32. ‘

Dent. J. K. Organization correlates of goals of business

management. -Personnel Psychology. 1959. 12. ~ ;

brucker. P. 'F. The practice of management. New York:
' ' Harper. 1954.

French. J. R. P.‘,' Jr.’, Israel. J., and has. D. An experi-
ment on a Norwegian factory. Human Relations, 1960. 13.

<3agne. Robert M. and Others. Peychological Principles in
»——- -- system development. New‘York: Holt. Rinehart and

Winston, Inc.. 1962.

IGahn, R. L. The prediction of productivity. Journal Social
Issues, 1956. 12 (2).

Baits. b., Maccoby, N.. and Morse. Nancy. Productivity super-
vision and morale in an office situation. Ann Arbor.
Michigan: Institute for Social Research. 1950._

Katzel. Raymond. a. Contrasting systems of work organization.
American Psychologist. Vol. 17. No. 2 (February. 1962).
‘ l

0
v

43

44

Kiel. D.. and wrigley. C. The effects on the factorial
solution of rotating varying numbers of factors.
Unpublished Master's thesis. Michigan State University.
1962.

Leavitt. H. J. Some effects Of certain communication
patterns on group performance. gguggal og Abnogpal
and Social Psychology. 1956. 52.

Likert. R. New patterns of management. New’York: McGraw-
' Hill.'l961.- ~ .
March. J. G. and Simon. H. 3. Organizations. New‘York:
JothWiley. 1958. '

.McGregor, Douglas. The human side of enterprise. New'York:
McGrawbHi11.:1960.fl

.Morse. Nancy C.. and Reimer. E.‘ The experimental change of”
' a major organizational variable. Journal of hbnormal

and Social Psychology. 1956. 52 . .

Neuhaus. J. O. and‘Wrigley. C. The quartimax method. British
JOurnal of Statistical Psychology. 1954. Vol. 7. ‘
pp. 81-91. ~

Shepard. H. A. Superiors and subordinates in research.
Journal of Business. 1956. 29.“ H "

Taylor. P. W; Scientific management. New’York: Harpers.
1947.

‘Weber. Max. The theory of social and economic organization.
trans: A. M. Henderson and Tolcott Parsons. New York:

Oxford University Press. 1947.

‘Weschler. I. R.. Kahane. M, and Tannenbaum. R. Job satis-
faction.'productivity and morale: A core study.

Oggupgpional nggpology. 1952.-26.‘

"gym"Mullignmnymmum”INITITH‘I'HIITIFS

 

655153