The effects of selected organizational characteristics on effective performance appraisal

Material Information

The effects of selected organizational characteristics on effective performance appraisal
Davies, Jerome C
Publication Date:
Physical Description:
viii, 163 leaves : ; 28 cm


Subjects / Keywords:
Employees -- Rating of ( lcsh )
Organizational sociology ( lcsh )
Employees -- Rating of ( fast )
Organizational sociology ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 158-163).
General Note:
School of Public Affairs
Statement of Responsibility:
by Jerome C. Davies.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
61134119 ( OCLC )
LD1190.P86 1990d D38 ( lcc )

Full Text
Jerome C. Davies
B.A., University of Colorado, Boulder, 1961
M.A., University of Colorado, Boulder, 1962
M.P.A., University of Colorado, Denver, 1982
A thesis submitted to the
Faculty of the Graduate School of Public Affairs of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Graduate School of Public Affairs

This thesis for the Doctor of Philosophy
degree by
Jerome C. Davies
has been approved for the
Graduate School
of Public Affairs
Mark Emmert
Dail Sj'Neugarten

Davies, Jerome C. (Ph.D., Public Administration)
The Effects of Selected Organizational Characteristics on
Effective Performance Appraisal
Thesis directed by Professor E. Sam Overman
Considerable research has been done to ascertain
what variables promote effective performance appraisal.
Still, little has been done to evaluate the effect of
organizational characteristics on effective evaluation.
This study was designed to evaluate the effects
of selected organizational characteristics on effective
appraisal. Task and human relations dimensions are
prevalent areas of interest in organizational literature
and they were used as the basis for analysis.
Specifically, the research probed the relationship between
task structure, leader-member relations and organization
support and effective performance evaluation. Effective
performance evaluation was defined by the reactions of
subordinates to the appraisals and the psychometric
soundness of the evaluations.
The subordinate reactions of 121 randomly
selected employees in the Colorado State Personnel System
were analyzed as were the performance ratings of those
employees. Measures of task structure, leader-member
relations and supervisory perceptions of performance
improvement were provided by the 33 supervisors of the
employees. The employees assessed organization support,
attitude about appraisal and perceptions of their own

performance improvement. Employee ratings were subject to
an evaluation of rating errors in terms of their leniency-
strictness, halo effect and restriction of range. Several
findings were found to be statistically significant. The
study suggested:
. That employee age, time in the organization,
and position may impact effective appraisal.
. That the variable "organizational support" may
impact an employee's attitude about appraisal.
. That job structure may affect the amount of
perceived performance improvement including
the amount of improvement promoted by
. That more lenient ratings and ratings with
more halo effect may be found in low
structured jobs.
. That more lenient ratings may be found in
environments with higher leader-member
relations and organization support.
. That the rating given the employee may be of
little importance in predicting the employee's
attitude about appraisal.
. That the sex of the employee may be associated
with attitudes about appraisal.
. That higher job structure environments tend to
be associated with high leader-member

. That supervisors and employees do not
necessarily agree on the amount of performance
improvement the employee has demonstrated.
Several improvements were noted and future
research areas were suggested. The study results suggested
a number of future research efforts.
The form and content of this abstract are
approved. I recommend its publication.
Signed |
mber in charge of thesis

There are many I would like to thank for their
help in this project. First, I want to thank Sam Overman
for the help and inspiration he provided. I am indebted
also to the other members of my committeeMark Emmert,
Dennis Donald, Eileen Tynan, and Dail Neugarten for their
efforts in the project.
Three people in the Personnel System were
especially helpfulEd Farrell in the Personnel Department
and Les Canges in the Health Department were lifesavers in
helping me through the analysis of my data. Beth Coronado
in the Personnel Department did an excellent job in
producing the final copy.
Finally, I want to thank my wife for her
patience and support throughout this effort.

INTRODUCTION........................................... 1
Background and Purpose of the Study.................. 1
The Research Problem................................. 3
Approach and Organization of the Dissertation...... 5
REVIEW OF THEORY AND LITERATURE........................ 9
Effective Performance Appraisal...................... 9
Organizational Characteristics...................... 30
Background. . ................................... 31
Discussion.......................................... 42
Summary............................................ 87
Research Approach................................... 88
Research Question and Hypotheses.................... 95
METHODOLOGY.......................................... 102
Research Design.................................... 102
Data Coding and Analysis...,....................... Ill
FINDINGS............................................. 122
Descriptive Variables and Dependent Variables...... 122
Independent Variables and Subordinate

Independent Variables and Psychometric Soundess... 128
Additional Findings................................ 136
Summary of Findings................................ 137
APPRAISAL.......................................... 141
Conclusions........................................ 141
Future Research................................... 151
Signficance of Study............................... 155
A. Supervisory Questionnaire..................... 165
B. Employee Questionnaire........................ 172
C. Appraisal Recording Form...................... 177
D. Research Inivitation Letter................... 179

Research Model

4-1 Correlations Between Descriptive
and Dependent Variables....................... 123
4-2 Relationship Between Independent
Variables and Subordinate Reaction............ 126
4-3 Relationship Between Independent
Variables and Psychometric Soundness.......... 129
Ratings on Common Job Factors

Background and Purpose of the Study
Performance appraisal is a highly important
human resources management activity. It is receiving an
increasing amount of attention in many public
jurisdictions. Organizations understand the psychological
impact which appraisal has on its members. Additionally,
organizations are aware of the importance appraisal can be
in identifying employee strengths and deficiencies and
providing a blueprint for improving performance. Then too,
many organizations key important administrative decisions
from appraisals so that retention rights, disciplinary
actions, pay increases and promotions may be decided in
part by the results of performance evaluation. Given the
importance of appraisal, it is not surprising that a great
deal of research has been conducted to discover what
contributes to effective performance evaluation. Much of
the work has focused on how various individual
characteristics contribute to effective appraisal, or on
how various rating formats effect good evaluation. In

looking at individual variables some studies have
concentrated on how appraisal is affected by supervisory
characteristics (Bayroff et al., 1954; Cascio and Valenzi,
1977; Kirchner and Reisberg, 1962; Schneier, 1977; Taylor
et al., 1959, for example). Still other studies have been
done on ratee characteristics (Bigoness, 1976; Elmore and
LaPointe, 1974, 1975; Rosen and Jerdee, 1973; Schmitt and
Hill, 1977, for example). In looking at rating formats,
some studies have been done on the use of graphic scales
(Barrett, et al., 1958; Ryan, 1958, for example),
behaviorally anchored scales (Landy and Guion, 1970; Maas,
1965; Smith and Kendall, 1963, for example), forced choice
formats (ZaVala, 1965, for example), rating dimensions
(Kavanagh, 1971, for example), and other formats.
Other areas have also been examined to discover the
determinants of good performance evaluation. The context
of the rating process has also been evaluated and studies
have been done to ascertain whether ratings differ when
done for research rather than administrative purposes
(Sharon, 1970, for example), when associated with job
difficulty (Svetlik, et al., 1964, for example), and when
influenced by pay incentive systems (Rothe, 1949, for
example). Rating process variables have also been reviewed
and studies have been done on rater training, rater
anonymity and related matters (Bernardin and Walter, 1977;
Latham, et al., 1975, for example). And, finally, some
studies have been done on treating the results of

appraisals in different ways, for example, in reducing the
number of rating dimensions through factor analysis
(Grant, 1955, for example).
Landy and Farr (1980) have suggested that a
large number of factors contribute to the effectiveness of
performance appraisal. Their model includes those
dimensions which they feel can influence performance
appraisal (Landy and Farr, 1980, p. 94). They depict
organizational characteristics, the characteristics of
positions being evaluated, and the purpose for the rating
as three parts of the context of the appraisal process.
These features may influence the effectiveness of
appraisal as may other elements in their model the
characteristics of the rater and ratee, (including the
ability of the rater to make valid observations and
judgments), the rating instrument, and the administrative
rating process.
The Research Problem
Much of the research on performance appraisal
has focused on how rater and ratee characteristics and the
rating instrument effect appraisal. However, the problem
is there are other critical elements which have received
less attention. This study is concerned with the effects
of certain organizational characteristics on effective
appraisal. Specifically, the research question is: What
are the effects of organizational characteristics on

performance appraisal? The area seems most appropriate
for research given Landy and Farr's observation that: "It
is appalling to note how little systematic research has
addressed the impact of . organizational
characteristics on performance ratings" (Landy and Farr,
1980, p.95).
The purpose of this study is to examine the effect
certain organizational characteristics have on performance
evaluation. The research problem is to test what the
relationship is between selected organizational variables
and effective performance appraisal. It is very
appropriate to examine this area given the importance of
performance evaluation and the fact that so little has
been done to explore how it is effected by the
organization's environment.
The discussion to follow will include a detailed
explanation of the variables in the study. Briefly, the
organizational characteristics chosen were task structure,
leader-member relations and organization support. Task
structure is defined as the extent to which jobs are
circumscribed, leader-member relations concerns the extent
to which there is harmony in the work group and
organizational support is defined by the extent to which
employees feel the organization is understanding and
helpful in areas of concern to them. Effective
performance appraisal has been operationalized in terms of
two primary constructs. First, effective appraisal is

evaluated by the subordinate's reactions to the evaluation
and two variablesattitudinal response and performance
improvementare measured. Secondly, effective appraisal
is evaluated by the accuracy and validity of ratings which
is evidenced through the psychometric characteristics of
the appraisals. Three psychometric variablesleniency-
strictness, halo effect and restriction of rangeare
Approach and Organization of the Dissertation
This dissertation is comprised of four
additional chapters. Chapter II provides a review of the
literature and includes a discussion of the approach to be
taken in the study. In brief, the following approach will
be used. First, effective performance appraisal will be
defined and operationalized for this study. As an
examination of the literature will show, effective
appraisal may be examined by analyzing the psychometric
soundness of the ratings and the subordinate's response to
the evaluation.
Organizational characteristics will then be
identified based on a review of pertinent literature.
Since it was impossible to structure a research project
with so many variables, a delimitation was necessary.
Organization models suggest that an environment consists
of two primary areasthe task and human relations
dimensions. Methods of measuring these dimensions are

discussed and the research question posed is to ascertain
the relationship between effective appraisal and task and
human relations characteristics of an organization. Task
and human relations dimensions are in turn operationalized
as discussed above so that task environment is evaluated
by the amount of task structure found, and human relations
is measured by leader-member relations and perceived
organizational support. The hypothesis is made that high
task environments will be associated with high
psychometric soundness and rater/ratee reaction and high
human relations environments will be associated with high
rater/ratee reaction. The research problem and suggested
relationships are depicted in the model below.
Figure 1.1
Research Model
Organization Effective Performance
Characteristics Appraisal
Human Relations ^ Rater/Ratee
Environment Reaction
Chapter III provides a review of the
methodological approach used in the study. The discussion
describes the research design, the independent and
dependent variables, the sample selection process, the

instrumentation, data collection and data analysis. The
data analysis description includes a discussion of the
descriptive and inferential statistics chosen and applied
along with a justification for their use.
Chapter IV describes the research findings.
First, a discussion is provided on the relationship
between the descriptive variables and the dependent
variablespsychometric soundness and subordinate
reaction. The second section includes a discussion of the
findings on the relationship between the independent
variables and subordinate reaction. The third section
describes the findings on the relationship between the
independent variables and psychometric soundness. Another
section focuses on additional findings of note in the
study. The final section of this chapter includes a
summary of the research findings with specific reference
to the research hypotheses.
Chapter V presents conclusions. The chapter
highlights the significant findings and summarizes some
conclusions to be reached from them. Suggestions are then
offered on future research which would be desirable and on
some suggested improvements in the research design for
future efforts. A final discussion suggests the
significance of this study to performance appraisal
There are several benefits seen in this
research. Measuring the extent to which appraisal is

effective has involved the development of a comprehensive
method which includes objective and subjective indicators.
This evaluation method is a contribution to the field of
performance appraisal research. In addition, the research
has added to the body of knowledge on what contributes to
effective appraisal and has focused on an area where
little research has been done. This research project has
been worthwhile in examining some key organization
dimensions and exploring their relationship to performance

This chapter is divided into three sections. The
first section includes a discussion of effective
performance evaluation. The section is, in turn, divided
into two major parts. The first describes a major aspect
of effective appraisalsubordinate reaction. The second
describes another major aspect of effective evaluation
psychometric characteristics. The second section consists
of a discussion of organizational characteristics. The
discussion is broken down into three major parts which
include background, discussion, and summary segments. The
final section of this chapterResearch Approachinvolves
a discussion of the research model and questions to be
addressed through this research.
Effective Performance Appraisal
The first subject to be covered in the
literature/theory review is effective performance
appraisal. The discussion will focus on how effective
performance appraisal can be defined and measured.
In order to define what constitutes effective or
unsuccessful appraisal, it is necessary to first examine
why performance evaluation is performed. Measures of

successful appraisal should be linked to the purpose of
the appraisals. It appears that there are two major
purposes of appraisal. First, appraisal is performed to
"drive" administrative decisions. In many organizations
(including an increasing number of public ones) decisions
on such matters as pay increases (who gets what increase),
layoff (who shall be retained in a RIF situation), and
promotion may be based in whole or in part on the
appraisals. When performance appraisal is used in this
matter its major criteria for success should be its
accuracy, fairness and validity. It is extremely
important that the performance rating be an accurate
assessment of the individual rated; otherwise the wrong
person may be retained or promoted or unfair salary
decisions will be made.
The second purpose of the appraisal is to motivate
and develop employees. Latham and Wexley see a link
between these purposes and the administrative decisions.
They say that the two most important functions of
appraisal are
...the counseling (motivation) and
development (training) of employees.
These are primary because it is on the
basis of motivation and training that
decisions are made regarding the
employee's retention, promotion,
demotion, transfer, salary increase
and termination. (Latham and Wexley,
1981, p. 4)
They suggest that motivation is tied to
several aspects of a good appraisal including:

11, that allows an employee
to learn how well he or she is doing,
goal setting that specifies what the
person should be doing, team building
that allows employees to participate
with peers and the supervisor in
solving problems that impede their
productivity and monetary incentives
that reward good performance.
Burke concurs. "The primary purpose of a
performance evaluation system is to provide a measure of
job performance that will facilitate the counseling and
development of an employee." (Burke, et al., 1978, p. 904)
If the general purposes of the rating system
include accuracy, fairness, validity, motivation and
development, the next question concerns what type of
things demonstrate these purposes have been met. A review
of the literature indicates that these purposes are
evidenced by the effect on the subordinate reaction that
results from the ratings and their psychometric soundness.
Subordinate Reaction.
The subordinate's response to the rating is an
important indicator of effectiveness. The question
becomes one of assessing how well the rating was received
and used by the employee to improve performance. Burke
discusses the dimension as follows:
The dependent variables or outcomes of
the performance appraisal process
which have been included in a number
of studies have been such factors as
satisfaction with the appraisal
interview, motivation to improve,
actual performance improvements and

satisfaction with the supervisor (Burke, et al.,
1978, p. 904).
It is important to analyze the concept of
"subordinate response" and identify its important sub
dimensions. In so doing, it appears that some of the
indicators are attitudinal, regarding the employee's
attitude toward his/her supervisor, toward the process and
toward the desire to improve (motivation). On the other
hand, one of the indicators is behavioral, i.e., did
performance actually improve. The former category can be
labeled "attitudinal response" while the latter can be
called "behavioral impact."
With regard to the effect on attitudinal response,
various measurement methods and approaches are identified
in the literature. Cummings used a semantic differential
approach in tapping employee attitudes on several
attitudinal dimensionsperceptions of goodness and
badness (general evaluative tone), constructiveness and
sensitivity and clarity and consistency of the appraisal
system (Cummings, 1973). Accordingly, respondents were
asked to respond to 22 paired items on a seven point
scale, e.g.,
The Appraisal System Is:
Some of the 22 items apply to the first constructgeneral
evaluative tonesome to the secondconstructiveness and

sensitivity and some to the thirdclarity and
consistency. Summative scores are calculated for all
items pertaining to a given construct. The summary scores
on all items would be indicative of the employee's
attitude regarding the appraisal process.
Burke has written considerably on the evaluation
of the appraisal interview. It would appear that, with
some modification, the questions asked concerning attitude
toward the interview could be broadened to cover attitudes
on the performance appraisal process. In an earlier work
with Wilcox, Burke suggests that two questions can be
formulated to evaluate attitude. Employees are asked to
respond to the questions "I felt satisfied with the
(performance interview) session," and "(after the session)
I wanted to improve my performance." Respondents record
their responses by circling one of four statements, "very
well," "fairly well," "not too well" and "not at all."
(Burke and Wilcox, 1969).
In his later work with Weitzel and Weir, Burke
enlarges his approach to measuring attitudes on the
interview. While the scoring of the questionnaire is not
described in his article, presumably some type of 5-7
point Likert scale is used to ascertain the degree of
agreement with the several questions asked. His
discussion on the measures and questions is as follows:

Outcome Measures
Seven outcomes of the performance review interview
were examined. These were:
(1) "To what extent did your last performance review
session increase your understanding of what your
supervisor expected you to achieve on the job
(your job responsibilities)?"
(2) Greater Mutual Understanding. (3 item
(I) "To what extent did your last
performance review session increase
your understanding of what your
supervisor expected you to achieve on
the job (your job responsibilities)?"
(II) "To what extent did your last
performance review session result in
better ideas in your mind about what
your supervisor expects in the way of
future job performance improvements?"
(Ill) "As a result of the interview you had,
do you feel that the mutual
understanding between you and your
supervisor has changed?" These three
items were significantly and positively
(3) Fairness of Last Performance
Appraisal. "How fair do you feel your

last performance appraisal session
(4) Satisfaction with Performance Review
Process. Think of your last
performance review session. How well
does this statement describe it? "I
felt satisfied with the session."
(5) Motivation to Improve Job Performance.
Think of your last performance review
session. How well does this statement
describe it? "I felt I wanted to
improve my performance."
(6) Actual Performance Improvements.
Think of your last performance review
session. How well does this statement
describe it? "I actively have
improved my performance since then."
(7) Overall Value of Performance Review
Interviews. "In general, how valuable
are performance reviews in your
opinion?" (Burke, et al., 1978,
It is noted that one of the questions asked was on
self perception of actual performance improvements. This
question would be more appropriately categorized under
behavioral impact than attitudinal response. Indeed, in
Burke1s initial article on the assessment of the interview

he suggests a similar question, "I have improved my
Sashkin (1982) describes a third approach to
evaluating the attitudinal response. He discusses a 6
question instrument called OPAQUE which was designed to
get a "rough assessment" of an organization's appraisal
system. Three of the questions are answered by
supervisors who do ratings and three answered by employees
who are rated. Two of the questions focus on the degree
to which the appraisal system is helpful in developing
subordinates, two questions assess the extent to which the
appraisal process is helpful in letting people know where
they stand and two of the questions assess the extent to
which the system provides information for administrative
actions (e.g., promotions, layoffs, pay). Scores are
summed to get an overall assessment on each dimension and
all six questions may be added up to get an evaluation of
the overall value of the system. Since each of the
questions is rated on a 5-point scale the maximum score
for the six question instrument is 30. Sashkin interprets
the scores in the following way:
26-30 Organization's appraisal system is doing an
excellent job
21-25 Organization's appraisal system is doing a good
11-25 Organization's appraisal system is doing an
average job

6-10 Organization's appraisal system is doing a poor
The questionnaire itself is depicted below along
with the scoring format.
Respond to the following six statements by indicating the
extent to which you agree (or disagree) that the
statements accurately describe performance appraisal in
your organization. Some statements refer to your
experiences in appraising your subordinates' performance;
others refer to your experiences in being appraised
yourself. Try to reflect as accurately as you can the
current conditions in your organization based on your
experiences. 1 2
SA=Strongly Agree A=Agree ?=Neither Agree nor
D=Disagree SD=Strongly Disagree
1. I have found my boss's appraisals to be SA A ? D SD
very helpful in guiding my own career
development progress.
2. The appraisal system we have here isof SA A ? D SD
no use to me in my efforts toward
developing my subordinates to the
fullest extent of their capabilities.

Our performance appraisal system
generally leaves me even more uncertain
about where I stand after my appraisal
than beforehand.
The appraisal system we use is very
useful in helping me to clearly
communicate to my subordinates exactly
where they stand.
When higher levels of management are
making major decisions about management
positions and promotions, they have
access to and made use of performance
appraisal records.
In making pay, promotion, transfer, and
other administrative decisions, I am
not able to obtain past performance
appraisal records that could help me to
make good decisions.
Use the following grid to determine point scores for each
item by transferring your responses onto the grid. Place
the number in the box at the bottom of each column, then
add pairs of columns as indicated.

+ + +
A + B + C=
In summary, there are several ways available to
measure the attitudinal response of employees and
supervisors to performance appraisal in their
organization. The three described were questionnaires all
of which were designed to tap attitudinal responses to
appraisal in the organization. None of these
questionnaires is necessarily the best and, indeed, a
researcher could fashion his own instrument to measure
attitudes about the progress. The instrument used in this
research is described in the next chapter on methodology.

The second area to be examined in assessing
subordinate reaction is behavioral impactactual
performance improvement. Perhaps this is the most
important measure of effective performance appraisal
because, indeed, if the appraisal has worked well it will
have identified areas for improvement and, if such
improvement takes place, the appraisal has served a very
useful purpose.
It would appear that there are three ways of
measuring actual performance improvement. First, as Burke
has done, you can ask the employee whether his performance
improved. Burke's simple question: "Think of your last
performance review session. How well does this statement
describe it? I actively have improved my performance
since then." (Burke, et al., 1978, p. 309). There are two
obvious problems with this approach. First, it assumes
that the employee is honest in his appraisal and/or is
able to discern whether performance improved. Secondly,
it assumes that the appraisal was somehow a causal agent
in producing the improved performance. Both assumptions
are very questionable. However, if one realizes the risks
involved and couches conclusions accordingly, the method
provides a reasonable approach to measuring improvement.
The second method to measure behavioral impact is
to ask the rater. Indeed, a simple question such as that
asked of the ratee can be asked of the rater. Further,
the rater could be asked, if improvement was perceived,

whether he thought it could be attributed to the appraisal
process and/or to what extent it could be attributed to
it. Finally, some independent method could be used to
assess whether improvement occurred after appraisal. In
some jobs where quantitative indicators are obvious and
routinely made (e.g., data entry operator) simple
comparisons could be made before and after on the
production levels. In less routine and quantifiable jobs
other evidence could be gathered. For example, other
managers, peers and subordinates and in some cases the
users of a service could be queried as to whether a given
employee's performance had changed following appraisal.
As with the other measures discussed in this chapter, the
research approach used in this study is discussed in the
methodology chapter.
Psychometric Soundness
This is a second area to be evaluated in assessing
the effectiveness of performance evaluations. There are
various characteristics of psychometric soundness. The
first characteristic is rating errors, and the three most
common errors appear to be leniencystrictness, halo
effect and restriction of ranges (Schneier, 1977). In
general, the presumption is that ratings will be effective
when they are not overly lenient or strict, show
discrimination between rating factors (minimize the halo
effect) and indicate that the rater has used a reasonable

range in the ratings given, i.e., has not clustered all
ratings at the same score. Psychometric soundness is
especially important to the first goal of appraisal, i.e.,
the making of correct administrative decisions. As Burke
points out, when these errors are made it, "...will result
in wrong decisions on who to promote, retain, replace."
(Burke, et al., 1978, p. 904). Kirchner and Reisberg
underscore the importance of minimizing rating errors
(Kirchner and Reisberg, 1962). In a study in which they
probed the differences in rating behavior of good and poor
supervisors they found that
The better supervisors are more discriminatory
in their more spread and
variation...less effective supervisors are
more lenient in their ratings...they tend
toward rating everybody the same (Kirschner
and Reisberg^ 1962, p. 302).
Reliability is a second psychometric
characteristic of effective ratings. The concept concerns
the extent to which the ratings are consistent and free
from time sampling efforts and rater bias. Latham and
Wexley list three approaches to ascertaining reliability
(Latham and Wexley, 1981, pp.65-66). The test-retest
method is one in which the rater rates the subordinates at
one point in time and then re-rates several months later.
Correlations are then computed to see the extent of
agreement and since some deviations are expected it is
assumed that coefficients of .70 and above are reflective
of high reliability. A second approach is to obtain

interobservor reliability. Here two or more raters
evaluate performance and correlations are computed on
their level of agreement. A coefficient of .60 or higher
is thought to be sufficient evidence for reliability
whereas lower correlations suggest that the
attitudes and biases of the rater are more likely being
rated than is the performance of the ratee. Finally,
reliability can be measured by the "split halves" method
in which the scores on some items in the evaluations are
correlated with the scores on other items (e.g., the
scores on the odd numbered rating factors are correlated
with the scores on the even numbered ones). Correlations
of .80 and above are needed to support high reliability.
The methods described are those typically used to
check on the reliability of examinations. Bayroff,
questions the use of the methods for performance appraisal
(Bayroff, et al., 1954). He suggest that test-retest, for
example, is inappropriate because performance tends
to be reasonably constant and one would expect to see high
association between ratings at different time periods.
Internal consistency checks, he submits, are inappropriate
because the halo effect is a reality and one will expect
to see high internal consistency as a result. In short,
he says that measures of reliability are suspect and if
one finds correlations supporting high consistency they
are spurious.

There is disagreement as well on what methods to
use in determining the third psychometric characteristic
validity. Latham and Wexley suggest three types of
measures (Latham and Wexley, 1981, pp. 67-69). First,
content validity can be probed by getting experts to judge
the extent to which the rating criteria used adequately
measures the behaviors considered critical for job
performance. Predictive validity can be checked when
people have been promoted and a correlation is computed
between performance evaluation scores on their old job and
their new ones. The authors indicate that the ratings of
30 or more employees must be considered and conclude that
it is typically impractical to compute correlations for
this reason. Finally, Latham and Wexley discuss construct
validity. This is defined as the extent to which persons
evaluated possess the same quality or construct presumed
to be reflected in the performance measure. The method
involves getting several different measures that logically
seem to evaluate the same construct and computing the
degree of consistency in the ratings.
Bayroff is skeptical about traditional measures of
validity (Bayroff, et al., 1954, p. 112). Typically,
independent indicators are gathered and then correlated
with the performance evaluations to see the extent of
agreement. For example, in a study in which he was
involved, 400 officer students were ranked anonymously and
then the ranks were correlated with performance

evaluations. He points out that both ratings are
obviously judgments and say that rater biases operate and
raise questions about the independence of the measures.
One answer, he suggests, has been to define validity as
agreement among raters, especially the agreement of one
rating with a consensus of ratings. Accordingly, validity
becomes very similar to reliability.
One final word should be added about validity. In
the Bayroff work on rating student officers, multiple
measures were taken of performance. Associate rankings
were not only,correlated with regular performance
evaluations, but they were also correlated with official
efficiency reports and class standing in courses at the
command and general staff college. The researcher
concluded that these multiple measures provided for more
valid ratings than did singular assessments. The
correlation of the average of a number of ratings per
ratee provided for a validity coefficient of .89 whereas
the correlation of rankings with singular measures only
resulted in a .52 coefficient (Bayroff, et al., 1954, p.
97) .
To summarize, the psychometric characteristics of
performance ratings can be evaluated in several ways.
Typically, some evaluation is made of rating errors and
assessments are usually made on the extent to which
appraisals reflect halo, leniency-strictness and
restriction of range biases. In addition, measures can be

taken on the reliability of ratings and on their validity.
There is a close relationship between these concepts and
it has been suggested that they may be measured in the
same way by looking at the association between ratings and
other independent assessments of performance. It is very
desirable to gather multiple measures of performance and
correlate them to performance appraisal when computing
consensual reliability-validity measures.
It should be noted that the psychometric soundness
of ratings can be effected by a myriad of variables.
Taylor, for example, looked at the correlations between
leniency and strictness as they related to supervisory
characteristics (Taylor, et al., 1959). He compared an
individual's self-rating with the estimated rating s/he
felt they would get from the supervisor with the actual
rating given. He characterized supervisors on four
dimensions, i.e., their production orientation,
consideration for subordinates, job assignment skills and
exercise of authority. Among other things he found that
high production oriented supervisors were inclined to rate
less favorably than high consideration supervisors when
compared to self ratings. It would appear that the
supervisor's orientation and skill in the dimensions
described effect the rating process.
As mentioned above, Kirchner and Reisberg got
independent ratings of supervisors in terms of their
effectiveness as described by higher level managers. They

found the more skilled and competent supervisors showed
more spread and variation in ratings and were less likely
than their poorer counterparts to rate everybody nearly
the same (Kirchner and Reisberg, 1962, p. 302).
Bayroff found that rater competency was very
important in the validity of ratings. He concluded that
"...high raters (as judged by test scores, rankings by
independent assessors) gave more valid ratings." (Bayroff,
et al., 1954, p. 107). Indeed, he found that different
rating formats didn't result in significantly different
validity and concluded that the rater is more important
than the format in determining validity.
Schneier (1977) found that the "cognitive
complexity" of the rater was important in determining the
psychometric soundness of the ratings. In looking at
rater performance with behavior expectation scales,
Schneier hypothesized that the rater's ability to use them
effectively would be determined by cognitive complexity,
i.e., the way in which the rater organizes and integrates
his/her thoughts and the relative number of dimensions
raters use to describe what they perceive. He referred to
the ability to discriminate between dimensions as
differentiation and the ability to discriminate within
each dimension as articulation. He further hypothesized
that the match between the complexity of the rating format
and the cognitive skill of the rater would be critical and
suggested that a high degree of compatibility would

produce "...less leniency, less halo and less restriction
of range" (Schneier, 1977, p. 543). Many of Schneier's
individual hypotheses were supported and he concluded that
the cognitive complexity of the rater interacts with the
rating format to effect the psychometric characteristics
of ratings.
Cascio and Valenzi (1977) hypothesized that
certain rater characteristics would be important in
determining the psychometric soundness of appraisals. For
example, they felt that less experienced raters would be
more severe in their ratings than more experienced ones.
While they found statistical significance for some of
their hypotheses they concluded that none of the findings
had practical significance. Still, the statistical
findings reflected correlations between certain
characteristics of raters and the impact on ratings.
The discussion above has focused on how certain
characteristics of the rater may impact the psychometric
properties of the ratings. Rater characteristics are but
one dimension that may effect psychometric properties.
For example, Grey and Kipnis found that a contrast effect
was very important in determining rating levels (Grey and
Kipnis, 1976). They looked at the extent to which the
supervisor's ratings are impacted by the number of non
compliant subordinates s/he had in a work unit. They
found that supervisors tended to give higher ratings to
non compliant workers when there were a lot of them in the

work unit. When there were but a few, the ratings were
much less favorable. In short, when a supervisor could
contrast a few recalcitrants to others, the ratings became
more severe.
Zammuto found an interaction between rater
differences and organizational differences and their
impact on ratings (Zammuto, et al., 1982). Organizational
and unit differences accounted for leniency-strictness
errors by providing raters with a different "model of
performance" from which to rate. Henderson suggests that
organizational differences can have a profound impact on
ratings and lists the following characteristics as
Organization design (including the number of
levels, use of project management, task forces
and matrix organization)
Number of employees
Number of different divisions, departments (work
- Geographical distribution of work units
Different kinds of work performed
Organization philosophy and value system
- Managerial span of control
Use of appraisal rating information
Influence and status of certain positions
(Henderson, 1984, p. 54).

The discussion above concludes the presentation on
the psychometric characteristics of ratings. The features
of most utility have been identified and incorporated into
a research approach that can be used to assess the
psychometric soundness of a group of performance
ratings. The approach is described in the methodology
chapter of this paper.
In summary, this discussion has suggested that
effective appraisal can be measured by subordinate
reaction and psychometric characteristics of the ratings.
It is important to note here that this definition of
effective appraisal is not an absolute one, but is the
operational definition chosen for this study. One could
argue that either of the two dimensions is the true
measure of effective appraisal or that some other
dimension(s) is more appropriate. Given the literature
review performed for this study, the two dimensions seemed
most appropriate because they align with the basic
purposes of appraisal. As mentioned above, psychometric
characteristics are critical when the purpose of appraisal
is for administrative decisions such as pay, promotion and
retention. Subordinate reaction is critical when the
purpose of appraisal is for employee coaching, improvement
and development.
Organizational Characteristics
Organizational characteristics are those salient

features of the organization that may affect the way in
which the organization functions, and how members within
it behave. Organizational research efforts often involve
an examination of the relationship between specified
characteristics of an organization and other variables
such as performance and productivity (Campbell, et al.,
1984) .
The Research Area described in this paper focuses
on the relationship between certain organization
characteristics and effective performance appraisal.
Accordingly, this section of the paper is designed to
provide a discussion of organizational characteristics, to
describe how various writers have identified them and to
examine how they can be measured.
The first part of this discussion will focus on a
typology for grouping organizational characteristics. In
reviewing the subject, scores of characteristics were
identified. In order to make a discussion of the subject
manageable it is most desirable to find a way to group the
characteristics systematically.
A number of topologies are available to aid in the
grouping of the characteristics. Frankly, it is not
certain that a single best typology can be chosen or
necessary that an unreasonable amount of time be spent
discussing which typology is superior. What is important

is that any typology chosen be sufficiently comprehensive
so that the important organizational characteristics can
be identified.
One typology is that discussed by Kast and
Rosenzweig (1970). They describe three primary schools
in the organizational literaturethe traditional, the
behavioral, and the management science and systems
approaches. In the traditional school the authors discuss
scientific management with its emphasis on task,
standardization of method and scientific selection of
workers and then turn their attention to others in the
early days of management theory. Henri Fayol, for
example, is discussed as typical of the traditional
theorists for he emphasized many of the structural and
formal aspects of the organization, e.g., division of
work, unity of command, scalar chain, unit of direction,
etc. Luther Gulick and Lyndall Urwick were part of the
traditional school and emphasized the importance of
fitting people to the organization's structure,
departmentalizing by purpose, using an appropriate span of
control, etc. Other theorists are discussed as well and
the authors cover the contributions to traditional theory
made by Mary Parker Follett, James D. Mooney and Alan C.
Reiley. Kast and Rosenzweig discuss also Max Weber's
bureaucratic model and list his six dimensions of
bureaucrat i z ation:

o division of labor based on functional
o well-defined hierarchy of authority
o system of rules covering the rights and
duties of position incumbents
o procedures to deal with work situations
o impersonality of interpersonal relations
o promotion and selection based on technical
Finally, these authors summarize the major
assumptions of traditional theory. The organization was
viewed as a mechanistic system and was planned and
controlled by management authority. The assumption was
made that man was economically driven and that the goal of
management was efficiency through structure and control.
The concepts (e.g., division of labor, unity of command)
were labeled as principles of management.
Kast and Rosenzweig next discuss the behavioral
and management science revolutions. The management
science school developed after World War II and emphasized
systematic analysis and quantitative techniques to improve
organizational performance. The behavioral science school
emphasized the human components in the organization. The
authors discuss the contributions of Elton Mayo and the
Western Electric Company experiments along with other
human relationists and conclude with a listing of the
assumptions of this school:

o the social system and formal system may have
defined different roles and norms
o people are not motivated solely by economic
factorsother social and psychological
factors are important
o the informal work group is the dominant work
o leadership should be geared to people
concerns democratic and participative
approaches are best
o worker satisfaction is most important and is
linked to efficiency and effectiveness
o managers must have good social and technical
o employees are motivated by fulfilling
certain social psychological needs
The third school in the typology is the systems
approach. The school assumes that the most meaningful way
to study organizations is as systems. There is a clear
interest in looking at the components of an organization,
how they relate to one another and, for the open systems
theorists, an interest in how the organization relates to
its environment. Often the biological systems concepts
are applied in organizational environments so that there
is an analysis of the organization's inputs,
transformation system and outputs and other phenomena such
as entropy, homeostasis and equafinality.

Kast and Rosenzweig conceptualize a structured
sociotechnical systems model. While this model is
discussed at some length later in this paper, a summary of
the basic properties will be made at this point. The
authors view an organization as a structured system com-
posed of several subsystems. The subsystems include the
goals and values of the organization, its technology, its
structure, the psycho social element and the managerial
component. These subsystems interact with each other so
the interest is on the subelements and their interaction.
Finally, the authors discuss other properties of
organizational systems. They indicate that organizations
are contrived (not natural) systems, have boundaries with
the environment, can be seen in hierarchies, and have
mechanisms which provide for feedback, adaptation and
maintenance and growth through internal elaboration.
A different typology for viewing organization
theory is offered by Ashley and Van de Ven (1983).
They identify four views of organizations with
distinctions based on the view of man's free will (i.e.,
deterministic vs voluntaristic) and on the level of
analysis, (i.e., populations or constellations of
organizations vs single organizations).
The first viewthe systems-structural
perspectivefocuses on individual organizations and
embraces structural functionalism and systems theory.
Included here are the traditional theorists such as

Gulick, Urwick, Fayol and Weber along with Merton, Blau
and Scott and structural contingency theorists such as
Lorsch and Lorsch.
The presumption is deterministic and the schools
assume that organizations are shaped by" ...impersonal
mechanisms that act as external constraints on actors."
(Ashley and Van de Ven, 1983, p. 248). Accordingly, roles
determine behavior and people are hired, trained and
managed based on the roles they play. The role of the
manager is to manage the organizational structure to
better meet the goals given the changes in the
The second school defined by Ashley and Van de Ven
is labeled strategic choice. The presumption is more
voluntaristic and the assumption is made that roles are
influenced by how people perceive them. Organizations are
products of how actors define them. Accordingly, the
interest here is on people, their interactions and their
choices in the organization. The focus is on individual
organizations and the manager's role is proactive, i.e.,
to energize forces to shape the external world. Several
schools of thought can be placed under this category
including the symbolic school and the exchange theorists
such as Blau.
A third view of organizations described by Ashley
and Van de Ven is that held by the natural selection
theorists. This school focuses more on populations or

constellations of organizations than on single
organizations. The assumption is made that environmental
resources are structured in niches whose existence and
distribution are relatively intractable to manipulation by
single organizations. Accordingly, this deterministic view
holds that choices are greatly constricted. Moreover,
organizations can't change much to adapt to the niches so
they are placed at the mercy of their environments. The
manager's role in this situation is an inactive onethe
organization is immune to management action since
environmental concerns are primary and totally
influencing. Finally, there is a collective action view
of organizations. Again, the focus is on multiple
organizations rather than single ones with the assumption
made that organizations band together for survival through
the development of a regulated and controlled social
environment that operates to reduce the effect of the
natural environment." Networks are formed to promote
collective action and the manager's role is
interactiveto work with other organizations to negotiate
reasonable terms for mutual survival.
In summary, the matrix below depicts the four
views of organizations as discussed by Ashley and Van de

View of Man's Free Will
Level of
Analysis Deterministic Voluntaristic
Single System Strategic
Organizations Structural Choice
Multiple Natural Collective
Organizations Selection Action
A third typology is provided by Richard Scott
(1982). In looking at organization theory Scott
identifies various views of organizations and places them
in one of four categories. Type I schools embrace
rational and closed models of the organization. Here
Scott places the traditional theorists such as Taylor,
Weber, Fayol and Gulick and Urwick. "All of these
theorists conceive of organizations as tools designed to
achieve preset ends and all of them ignore or minimize the
perturbations and opportunities posed by connections to a
wider environment" (Scott, 1982, p. 126). Scott lists a
number of theorists under Type IIthose who hold to
closed but natural systems models. Here one finds the
human relations theorists such as Mayo as well as later
writers such as Dalton and McGregor. All are
characterized by a naturalistic view of organizations,
i.e., a belief that people shape the roles and activities
of the organization rather than the roles forcing
prescribed behavior on the individuals within it. Still,
the models are closed.

We learn a great deal about the emergence of
informal structuresinterpersonal systems
of power, status, communication and
friendshipand the way in which they impact
on formal systems: but whether the concern
was with formal or informal systems or the
relationships between them, the focus was
primarily on the organization's internal
arrangements. (Scott, 1982, p. 130).
Type III represents an Open Rational Systems
Model. Here are placed a number of theorists, e.g., Udy,
Perrow, Blau, Pugh, etc. These individuals were part of
the school which began in the late 1950's and continued
through the mid 1970's and which recognized the
interaction of the organization to the environment but
which focused attention on the organization as a rational
system. Typically, formal structure was seen as the
dependent variable and a host of independent variables
were studied, e.g., size, technology and uncertainty, to
discover their effects on the formal structure.
"...Organizations are presumed to design their structure
rationally... and (it is assumed they) are striving to
develop the most effective and efficient structures."
(Scott, 1982, p. 130).
The Type IV model is for theorists who assume open
and natural systems. Included are Pfeffer and Salancik,
March and Olson, Hickson and Meyer and Rowan. "These
models place great emphasis on the importance of the
environment in determining the behavior and life changes
of organizations: they are clearly open systems models.

However, the assumption that organizations behave as
rational systems is strongly challenged in this work."
(Scott, 1982, p. 131). Organizational survival, for
example, is seen as more compelling than meeting
organizational goals. And, rational assumptions are
questioned in the day-to-day operations, e.g., "...the
flow of individual actions produces a flow of decisions
that is intended by no one and is not related in a direct
way to anyone's desired outcomes." (Scott, 1982, p. 132).
In summary, the following matrix depicts Scott's
view of organization theory:
Closed Open
A final model to be discussed is provided by Lee
Bolman and Terrence Deal (1984). These authors feels that
four schools are most important:
o The Rational Systems Theorists. Assume
organizations can be studied as rational
entities. Emphasize importance of
organizational goals, roles and
technologies in studying organizations.

o The Human Resource Theorists. Assume
organizations are more complex than
rational theorists do. Assumes
organizations should be studied more in
terms of human issues, e.g., how people
interact, small group norms and what the
organization does for the people within
o The Political Theorists. Views
organizations as arenas of scarce
resources. Focuses on conflict
resolution, power, constituency
development in organizations,
o The Symbolic Theorists. Views
organizations in terms of cultures within
it. Focuses on values in the
organization and how they are carried
through such methods as rituals,
ceremonies, storytellers, etc.
These authors believe that the systems school can
best be viewed in the context of the four primary groups
that they identify. "The short term payoff will be greater
if rational, human resource, political and symbolic
theorists simply import concepts from systems when they
are useful." (Bolman and Deal, 1984, p. 232).

Discussion. Given the discussion above, it is
clear that organizational characteristics can be
identified through different conceptual views of
organizations. Further, as suggested earlier, it is not
critical that the best conceptual view be adopted but that
whatever scheme is used it provides a way of identifying
the most important characteristics. Accordingly, for this
paper the decision was made to use the basic structure of
the Bolman and Deal typology but to add systems as a
separate school because it predominates the literature and
because its omission would likely eliminate the
identification of some important characteristics. By
adding the systems school, it is felt that a framework
provided will be sufficient to identify the important
characteristics; at the same time the Bolman and Deal
framework, given the systems addition, provides for a
comprehensive typology with reasonably distinct
categories. To summarize, the following five schools will
provide the frames by which characteristics will be
o The Rational Systems Theorists
o The Human Resource Theorists
o The Political Theorists
o The Symbolic Theorists
o The Systems Theorists

The rational systems theorists. Using the
modified Bolman-Deal (1984) approach, each of the
theoretical constructs can be discussed in terms of what
characteristics are the focus of research. It appears
that the greatest effort on identifying characteristics
has been expended by those in the rational systems area.
As Bolman and Deal indicate, the rational systems
theorists emphasize the importance of organizational
goals, roles and technologies in understanding
organizations. Such theorists look for ways to develop
structures that best fit an organization's purpose. The
focus is on formal organization structures.
The term rational merits some discussion in
connection with this school. Clearly, the term can be
misleading if one were to take it to mean that this school
viewed organizations as rational entities while all other
schools viewed them as irrational institutions. Scott's
distinctions, discussed above, provide some clarification
in this regard. It will be recalled that Scott contrasts
rational and naturalistic viewpoints and characterizes
some schools as closed rational and open rational while
others are closed natural and others are open natural.
With regard to the term rational he indicates that it
means the extent to which a series of actions is organized
in such a way as to lead to predetermined goals with
maximum efficiency (Scott, 1982, p.p. 57-58). He goes on
to say that rational theorists believe that actions can be

organized in such a way as to produce the desired
outcomes. Further, the model is mechanistic and assumes
elements within the organization can be manipulated to
produce desired results. In addition, it is felt that the
roles and placement of the people within the organization
dictate their behavior and that "...formalizing structures
are thus rendered independent of the participation of any
particular individual" (Scott, 1982, p. 61). The
naturalistic school, by contrast, presumes that
organizations are much less rational, that members often
influence the roles they are to play, that predetermined
goals may not be followed, and that environmental and
individual considerations may alter rational processes.
Several representative theorists will be
discussed. William Evans (1976), for example develops a
list for describing organizations from a rational systems
o Indices of organizational hierarchy
o Hierarchy of rewards
o Hierarchy of authority including
Span of control
- Number of levels of authority from
top management to line workers
Ratio of administrative personnel
to production personnel
Time span of discretion

Amount of centralization in
decision making
Amount of limitation on
management's rights
This list leads Evan to a discussion of the
important organizational characteristics:
o Organizational size
o Age of the organization
o Complexity of the task structure
o Technological complexity
o Organizational effectiveness
o Organizational autonomy
o Functional specialization
o Bureaucratization
o Centralization of decision making
o Formalization
o Departmental specialization
o Informal organization
o Innovation (Evan, 1976, p. 196)
Campbell presents a short list:
o Organization size
o Technology
o Hierarchial structure (Campbell, et al.,
1984,p. 44).
Scott reports on the approach used in the classic
Aston studies in which they identified three primary

o Structuring of activities
o Concentration of authority
o Control of workflow (Scott, 1981, pp.
Further, other writers have developed broad
organization descriptions in which rational systems are
included. Typical of the organizational characteristics
identified in the rational systems part of such models are
those mentioned by Kast and Rosenzweig, and by
Heydebrand. Kast and Rosenzweig indicate the following
o Use of technology and knowledge in
o Structuring of organizations
o Formal relationships
o Differentiation and integration of
activities (Kast and Rosenzweig, 1979,
pp. 12-13).
Heydebrand describes his own typology which
includes the following rational-systems characteristics:
o Goal and task structure including goals,
diversity of major objectives,
geographical dispersion,
variability of tasks, and organization
o Division of labor including
technological complexity and skill

structure. (Heydebrand, 1973, p. 11).
Heydebrand discusses the classic typology for the
rational systems school developed by Pugh, Hickson,
Hinings, and Turner:
o Specialization
o Standardization
o Formalization
o Centralization
o Configuration
o Flexibility (Heydebrand, 1973, pp. 441-
470) .
The discussion above provides an insight into the
kinds of organizational characteristics the rational
systems theorists are concerned with and the kinds of
research questions that might ensue. Characteristics such
as size, structure, technology, degree of centralization,
etc., can be ascertained for any organization or for
subunits within an organization. Research questions can
be posed concerning the extent to which these
characteristics are associated with such outcome measures
as organizational effectiveness, the extent of conflict
within the organization, and the degree to which the
organization satisfies the needs of its members.
Dalton and others have reviewed the results of
some 50 studies which looked at the relationship between
various structural variables and organizational
effectiveness (Dalton, et al., 1980, pp. 49-65). The

authors make the distinction between characteristics which
are structural qualities and those that are structuring
activities. In the case of the former, the interest is in
physical characteristics such as size of the organization
and subunit size, span of control, flat/tall hierarchy and
administrative intensity. In the case of structuring
activities, the focus is on such dimensions as
specialization, formalization and centralization.
Dalton concludes that there are no clear,
unequivocal conclusions which can be drawn regarding the
relationship between these independent variables and the
outcome measure organizational effectiveness. In some
instances, little research has been done; in other areas
there is considerable research but the findings are
contradictory. For example, some studies have shown no
relationship, others a positive one and still others a
curvilinear relationship between subunit size and
organizational effectiveness.
Having provided a sense of the kinds of
characteristics in which the rational systems school is
interested, it is now desirable to look at some of the
characteristics in greater detail and discuss how they can
be measured.
Some similarities and dissimilarities will be
noted in comparing the characteristics identified by the
various writers discussed earlier. Three of the four
mention size as an important characteristic, for example,

but only one specifically lists bureaucratization. In
addition, while different labels are sometimes used for a
given characteristic it would appear the same concept is
being suggested. Accordingly, the discussion below will
highlight those key characteristics which seem important
in the rational systems school and upon which there is
typically some agreement between theorists. Since
there is no ideal sequence to use in discussing the
elements the characteristics have been arranged
Age of the organization is typically defined in
terms of the number of years the organization has existed.
Other definitions could be used. For example,
organizations are viewed as going through various cycles
in their existence and age could be taken to mean the
length or time they have spent in a given cycle.
Bureaucratization is discussed at length by Hall
(1963). After examining the work of a number of the
theorists on bureaucracy he suggests that the following
six dimensions are typically used in characterizing
bureaucratic organizations:
o A division of labor based upon functional
o A well defined hierarchy of authority,
o A system of rules covering the rights and
duties of positional incumbents.

o A systems of procedures for dealing with
work situations.
o Impersonality of interpersonal relations,
o Promotion and selection for employment based
upon technical competence.
Hall suggests that organizations can be measured
by the extent to which they meet each of the above
criteria. In other words, he argues that organizations
are not bureaucratic or non-bureaucratic but are
bureaucratic to some extent based upon the degree to which
they meet the characteristics identified. In terms of
measurement, he has developed scales to measure the extent
to which organizations are bureaucratic. Questions have
been developed for all scales and he gives the following
as examples:
1. Hierarchy of authority scale: "A person
can make his own decisions without
checking with anyone else."
2. Division of labor scale: "One thing
people like around here is the variety
of work."
3. System of rules scale: "The time for
coffee breaks is strictly regulated."
4. System of procedures scale: "We are to
follow strict operating procedures at
all times."

5. Impersonality scale: "We are expected
to be courteous, but reserved at all
6. Technical competence scale: "Employees
are periodically evaluated to see how
well they are doing." (Hall, 1963, p.
35) .
As mentioned before, some of the concepts listed
in the rational systems school overlap. Thus, for
example, the division of labor is defined by the amount of
functional specialization which, in turn, is identified as
a specific characteristic by some of the other authors.
Accordingly, when functional specialization is described
later we can note that Hall's scale on division of labor
could be used to evaluate it.
Centralization is another characteristic for
evaluation. Evans defines centralization in terms of the
decision making process. Price concurs and defines it as
the degree to which power is concentrated in an
organization. In an organization with maximum centrali-
zation all decision making would reside in one person; in
a minimally centralized organization decision making would
be exercised equally by all members (Price, 1972, p. 43).
Measures of centralization are described by Aiken and
Hage (1968). Using a five point Likert scale,
participants are asked about how frequently they
participate in various decisions, e.g., hiring new staff,

and told to rate such statements as "there can be little
action here until a supervisor approves a decision" on a
five point scale anchored by definitely false to
definitely true.
A second method for measuring the extent of
centralization is offered by Whisler (1964). He looks at
objective data in the organization, and, for example,
computes the percentage of total organizational
salaries paid at various levels and thus infers the degree
of centralization. (A high ratio of management to worker
salaries would reflect high centralization).
Finally, a third method for measuring the extent
of centralization is offered by Price. The control graph
approach can be completed by organizational members or by
a researcher. The graph is illustrated below.
The graph is augmented by some specific questions
asked of managers and other employees in the organization,
"How much influence do you have at these

"How much influence do they have over you?"
"How much influence should you have?"
The results of the graph can be plotted and the
organization's centralization profile can be graphically
displayed. Each level of employee can be depicted with
regard to the amount of influence they exert in decision
making (Williams, et al., 1959).
In the discussion to follow the political/power
school will be reviewed. It is interesting to note here
that there is an obvious relationship between
centralization and power. The control graph captures the
degree of influence exercised in the organization by
various individuals. It can be argued that centralization
is a formal measure of power and the power frame includes
this dimension plus the informal exercise of power.
Organization complexity is another characteristic
which is identified in the rational systems school. As
with several other terms, this concept overlaps with other
concepts. Price defines it as
...the degree of structural differentiation
within a social system. A highly complex
organization, for example, is characterized by
many levels of authority, a large number of
occupational roles, and many subunits (divisions
and departments) (Price, 1972, p. 71).
The concept is closely aligned with that of
structure of the organization to be described below.
Further, the concept overlaps to some degree with those of
specialization, division of labor and a number of other

concepts found in the literature. Indeed, as Price points
out, the different dimensions of complexity are often
treated as separate concepts, and this confounding is so
problematic, that is it probably not desirable to treat it
as a separate topic. Still, because it is mentioned by
many authors in the field, it deserves some discussion in
this review.
Configuration is one of the characteristics
identified by Pugh and others (Heydebrand, 1973, pp. 441-
470). The construct embraces several dimensions and,
again, some of the dimensions are found in other
constructs described in this paper. The authors define
configuration as "the shape of the role structure" and
identify the dimensions as:
o Vertical span of control. The dimension is
measured by the number of positions found
between the CEO and the line employees,
o Lateral widths. The dimension is measured
by the ratio of subordinates to first line
o The percentage of employees in direct output
jobs and the percentage of jobs in staff vs.
line capacities.
Formalization is identified by Pugh as the degree
to which rules, procedures and directives are written
(Heydebrand, 1973, pp. 441-470). The concept is

measured by counting pieces of paper written about rules
and procedures. The measure captures the number of rules
and the amount of information specified for each of them.
Price adds the measurement technique used by Hugh
and Aiken (Price, 1972, pp. 107-118). They use a survey
questionnaire (a Likert scale anchored by definitely true
and definitely false at the extremes) and ask employees
such questions as:
"There is no rules manual in the
"Going through proper channels is stressed
in the organization."
Separate measures are gathered on such elements as
job modification, job description, rules observation and
specificity of job descriptions. Measures are computed
for various levels in the organizations, e.g., managers,
first line supervisors, line workers, etc.
Price critiques both the survey and documents
counting methods of measuring formalization. He suggests
that explicit norms can include more than written
Campbell and others identify size of the
organization as a common characteristic studied by
researchers (Campbell, et al., 1984, p. 44). Price
indicates that the typical measure is the number of
personnel in the organization or a subsection although
other measures such as the total assets or budget of the

organization could be used as well (Price, 1973, pp. 174-
180) .
Specialization is another characteristic
identified by the rational systems writers. Heydebrand
isolates departmental specialization as a structural
variable and says it can be measured by the number and
relative size of specialized and differentiated subunits
of an organization, e.g., production, sales, research, etc
(Heydebrand, 1973, pp. 1-30). Pugh identifies
specialization as one of his six major organizational
characteristics (Heydebrand, 1973, pp. 441-470). He feels
there are two subcomponents of the dimensiondivision of
labor and functional specialization. Functional
specialization is defined in terms of the number of
functions performed by different people. For example,
Pugh isolates 16 organizational functions and can see how
many are performed by different people or by the same
people. For example in a "generalist" organization one
might see many personnel functions performed by line
managers; in a "specialist" organization a staff personnel
office would perform many of the functions.
Standardization is defined by Pugh as the degree
to which procedures cover situations which arise in the
organization (Heydebrand, 1973, pp. 441-470). Measurement
is done through a simple process where a score of 0 is
given when no procedures exist in the organization and a
high score is given when procedures exist for most situa-

tions that arise. The concept should be distinguished
from formalization which covers the degree to which the
organization has written procedures covering various
Structure is a multifaceted concept. Heydebrand
identifies the organization's goal and task structure as
an important characteristic and includes an assessment of
the major objectives, the variability of tasks and
geographical dispersion. He identifies a separate
structural constructskill structureand lists several
o The generalist-specialist makeup of the
o The extent to which the organization
contains professional, subprofessional and
technical personnel.
o The authority structurethe way in which
control is centralized or decentralized
(Heydebrand, 1973, p. 11).
As with several other concepts, structure is a
troublesome term because it can embrace so many dimensions
and this confounding is disturbing. Elements of structure
are found in the concepts of configuration and
centralization, for example.
Technology is the last term to be discussed. Kast
and Rosenzweig define it as:

...the organization and application of knowledge
for the achievement of practical purposes. It
includes physical manifestations such as tools
and machines but also intellectual techniques and
processes used in solving problems and
obtaining desired outcomes. It converts
spontaneous and unreflected behavior into
behavior that is deliberate and rationalizing
(Kast and Rosenzweig, 1970, p. 176).
Hage and Aiken (1972) point out that the term
technology is used to cover a number of different aspects
of an organization's environment. A number of physical
factors are included, e.g., the amount of energy used, the
noise level, and degree of cleanliness, etc. Used in this
way the concept includes what Price identifies as
mechanization which he defines as the degree to which an
organization uses inanimated sources of energy. Indeed,
several researchers have developed instruments to measure
the extent to which the organization is mechanized (Price,
1973, pp. 129-137).
More typically, however, technology has been used
to identify the routineness in the work. Charles R.
Walker's report on the classic Yale Project discusses
several dimensions of technology:
o The repetitiveness of work,
o The mechanical pacing of work,
o Frequency of breaks in job routine,
o Frequency of social interactions.
Size of interacting group (Scott, et al.,
1981, pp. 194-195).

Charles Perrow views technology on a continuum
between routine and non routine. Technology is seen as
the level of applied knowledge in getting work done and
two dimensions are involved:
o Whether problems are familiar or exceptional,
o Whether problems can be solved in known
analyzable ways.
Where there are few exceptions and standardized
problem solving procedures a routine technology is
inferred. Where there are a large number of exceptions
and problem solving is varied and difficult, a non routine
technology is found (Perrow, 1967, p. 195).
Hage and Aiken have developed an instrument to
measure the degree of routineness in the technology. The
questionnaire is given to those in the organization and
includes such questions as:
"would you describe your work as being very
routine, somewhat routine, somewhat non routine
or very non routine?"
"please circle the following statements as
definitely true, true, false or definitely
"people here do the same job in the same way
every day."
"One thing people like around here is the variety
of work."

"Most jobs have something new happening every
"There is something different to do every day."
(Hage and Aiken, 1972, p. 55-72).
Hage and Aiken indicate that other terms and
concepts are allied to routineness.
Routineness of work measures how much variety
there is in work; job codification measures how
well defined the job is; rules observation
measures the enforcement of rules; job
specificity measures how concrete the job
description or procedural manual is (Hage and
Aiken, 1972, p. 59).
Additionally, Hage and Aiken see links between
technology and some of the other dimensions discussed
above. They indicate that there is a close relationship
between technology and the social structure and that
coordination is easier if technology is routinized.
Resultingly, they say that:
o Organizations with routine work are more
likely to be characterized by a
centralization of organizational
power. As technology is routinized, the
organization is coordinated via programming.
This is likely to be accompanied by
centralization of power, formalization of
roles and some lessening of
professionalization in the organi-
zation. Studies reported show a high
negative relationship between the degree of

routineness and participation in
organizational decision making,
o Organizations with routine work are more
likely to have greater formalization of
organizational goals. Routine work is
correlated highly with formalization of
regulations as represented by the presence
of rules manuals, job descriptions and a
high degree of job specificity,
o Organizations with routine work are likely
to have staff with less professional
o Organizations with routine work are more
likely to have efficiency as an important
goal (Hage and Aiken, 1972, pp. 67-72).
The discussion above has focused on a number of
factors which structural theorists have identified as
important organizational characteristics. One of the
problems noted is that the factors cross cut one another
so one is never sure what is.a major characteristic and
what is better seen as a subelement within one. Heydebrand
notes this problem and presents the results of a factor
analysis in which he tried to identify the major
characteristics. His four orthogonal factors are:
o Structuring of activities. The focus is on
how highly structured the work is. The
concept, includes the specificity of

procedures to cases and the specificity of
tasks to roles (specialization)
o Concentration of authority. Typically there
is an inverse correlation between such
concentration and specialization, i.e., the
more specialists the more likely the
authority is to be distributed,
o Line control of work flow. This measures
the extent to which line workers can control
the flow of work and there is a high
correlation with the standardization of
o Relative size of the supportive component.
This measures the size of the staff devoted
to auxiliary activities versus those in the
line operation (Heydebrand, 1973, pp. 1-30).
In concluding the discussion on the
rational-system theorists it is desirable to suggest why
so much work has been done by its proponents in the field
of organizational research. Indeed, when one compares the
amount produced to schools discussed later in the paper it
is obvious that the rational systems theorists have been
much more productive than have other schools. There
appear to be two primary reasons for the strong efforts
produced by this school. First, it has been in existence
longer than the others and hence more time has been
available to pursue research efforts. And, secondly, this

school has identified characteristics which lend
themselves more than others to traditional research
efforts. It is easier to operationalize some of the
characteristics, e.g., size, age, and formalization than
it is to measure some of the more abstract concepts
advanced by some of the schools that follow in the
The human resource theorists. The second
perspective for viewing the organization is provided by
the human resource theorists. As indicated above, Scott
contrasts the human resource school with the so-called
rational school as follows:
o Both schools believe there are rational
elements in organizations that will
influence the way in which its
members behave. The natural systems
theorists agree, for example, that
organizations adopt goals.
0 The human resource school feels, however,
that organizations are more complex than the
1 rational systems school does. Informal
systems influence how members behave and how
organizations operate. The personal
characteristics of individuals influence the
roles played by the members.

o The human resources school promotes two
important precepts: (1) People are the most
critical resource in the organization and
(2) The focus of analysis should be on the
relationship between the organization and
the people within it (Scott, 1982, p. 64).
It is more difficult to identify salient
organizational characteristics in this school than it is
in the rational model, however. The rational model
theorists label characteristics explicitly, e.g.,
formalization, configuration, specialization. In the
human resources model the important characteristics are
more implicitly stated and have to be inferred from the
precepts adhered to in the theory.
A review of the literature, however, does produce
some consensus as to what characteristics are important
for this school. Scott discusses the human relations
school and identifies the following characteristics:
o Influence of informal group processes. The
way in which informal groups operate, the
norms they produce, the degree of
cohesiveness and an analysis of the group
dynamics are important characteristics,
o The leadership styles. The type of styles
employed (e.g., democratic-laissez faire
autocratic or task oriented-team oriented)
would be reviewed as important

o Decision-making approaches. The way in
which decisions are made and the extent to
which employees are involved in the decision
making is important.
o Employee centered characteristics. This
school would be interested in what kinds of
things the organization has done to meet the
needs of its members. For example, efforts
in job enlargement and job enrichment would
be reviewed as part of the characteristics
of an organization in this regard. (Scott,
1982, pp. 36-90)
Kast and Rosenzweig develop a similar list of
characteristics (Kast and Rosenzweig, 1970, pp. 85-100).
In addition to those mentioned by Scott, these authors
would add:
o The type and effectiveness of communications.
o The programs and efforts made to motivate
Based on these identified characteristics, a
number of different analyses could be done to evaluate an
organization's human relations characteristics. Informal
group processes, leadership, decision making,
communications, and motivation would be evaluated as would
other kinds of initiatives the organization has taken to
maximize employee needs. In looking at what the

organization has done to meet employee needs a number of
indicators could be made of an organization's policies in
such areas as pay, fringe benefits, training provided,
career development opportunities afforded, degree of
flexibility in allowing for alternate work schedules,
educational leave, etc. Aside from formal policies, some
assessment could be made of more informal areas, e.g., the
humaneness of the work environment including the perceived
fairness of practices, equity of decisions, openness of
the environment, trust, etc.
Some instruments found in the literature may be
helpful in such assessments. Price reports on one
characteristicalienationdefined as "subjectively
experienced poweressness to control one's own work."
(Price, 1972, p. 85). The degree of alienation perceived
might be seen as one characteristic in an organization's
human resource environment. The classic measurement used
in the area is Pearlin's alienation questionnaire
(Pearlin, 1962, pp. 105-112). The questionnaire includes
such questions as:
"how much say or influence do people like you have
to say about how the hospital is run?"
A Lot __________
Very Much ______
Some ______
None ______
Another important characteristic is the extent

to which the organization has built a motivating
environment. Price defines motivation as the degree to
which an organization's members are willing to work. He
points out that the characteristics may be measured
indirectly, e.g., by using data on absenteeismor
directly, by asking questions (Price, 1972, pp. 137-145).
Lodahl and Keiner have developed a 20 question instrument
with such questions as:
"I live, eat, and breathe my job."
Definitely False ____
Somewhat False ______
Somewhat True _______
Definitely True _____
There are other characteristics within the
organization that can be measured in the human resources
area. There is considerable interest in management styles
and in job enrichment. The assumption is that
participative styles and job enrichment efforts will
promote employee satisfaction, and, accordingly, help meet
employee needs. Management styles can be measured by such
assessment tools as the Blake-Mouton supervisory grid
questionnaire or by Likert's job centered-employee
centered approach (Bolman and Deal, 1984, p. 97). Job
enrichment efforts can be measured by developing an
instrument to assess what means an organization has made
to design satisfactory jobs (Hackman, et al., 1979, pp.

Communication is a more peripheral measure of
relevance in the human resources school. While it is not
a direct measure of satisfaction, it is related to how
well the organization cares for its employees and informs
them about what is going on. Price discusses the
instrument developed by Georgopoulos and Mann which
measures the adequacy, amount, frequency, quality,
informality, and direction of communication (Price, 1972,
pp. 58-70). The following question on adequacy of
communication is indicative of how the questionnaire is
"How would you rate the kind of communication
which you receive from your immediate
Superior _____
Very Adequate _____
Adequate _____
Inadequate ________
The discussion above does not include a
description of employee satisfaction indices. Indeed,
there are a myriad of ways to measure the employee1s
satisfaction with the organization (Price, 1972, 156-
174). However, overall employee satisfaction is not a
characteristic of the organization-it is the product, one
would presume, of the human resources characteristics in
the organization. Thus one would expect that
organizations characterized by good pay policies, humane

rules, equitable and fair working environments, etc. would
get high marks from employees on overall satisfaction.
But the overall satisfaction is the dependent variable in
the analysis-the characteristics are the independent
From the discussion above, it is possible to
summarize the characteristics identified by the human
resource theorists.
o Informal group processes. Assessments could
be made of the way in which informal groups
operate, the norms they produce, the degree
of cohesiveness and the group dynamics,
o Leadership styles. Assessments could be
made of the types of styles employed,
o Decision-making approaches. Assessments
could be made of how decisions are made and
the extent to which employees are involved,
o Employee centered initiatives. Assessments
could be made of the kinds of things
organizations have done to meet
the needs of their members. Included would
be an examination of efforts in job
enlargement and job enrichment, flex-time,
pay equity, training opportunities, career
development efforts and other benefits.

o Communication. Assessments could be made of
the type and effectiveness of
o Employee motivation efforts. Assessments
could be made of any other activities the
organization has engaged in to promote the
motivation of its members.
The political theorists. The third school of
organization analysis is concerned with the political
dimensions of the organization. Bolman and Deal indi-
cate that in this approach the analyst sees the
organization as an arena of scarce resources.
Accordingly, conflict will naturally occur and a manager's
success will be dependent on the degree to which he/she
can successfully negotiate, resolve conflict, manage power
and build constituencies (Bolman and Deal, 1984, pp. 108-
130) .
Jeffrey Pfeffer discusses the power/political
frame in some detail. He defines power as "...the
capability of one social actor to overcome resistance in
achieving a desired outcome or result" (Pfeffer, 1982,
p.2). Power is seen as a force to get someone to do
something he/she would not have done otherwise. Further,
when power is legitimate it becomes authority and once it
is transformed into authority it is expectednot

Pfeffer discusses politics in an organizational
setting and says it "... involves how differing preferences
are resolved in conflicts over the allocation of scare
resources" (Pfeffer, 1981, p. 6). Politics, then,
involves the use of power to obtain desired outcomes. The
implication is that decisions are often governed by
political adroitness and power rather than rational
processes (as promoted by rational theorists),
standardized processes and precedent (as suggested by
bureaucratic theorists) or random, unpredictable forces
and processes (as suggested by the decision process
theorists). Political considerations can be related to
various organization phenomenon. Obviously, decision
making is one activity that can be evaluated. According
to the political model, organizations should not be viewed
as monolithic entitles but "...pluralistic and divided
into various interests, subunits and subcultures. (In
terms of decision making), "...when preferences conflict,
the power of the various social actors determines the
outcome of the decision process" (Pfeffer, 1981, p. 28).
Accordingly, decision making is governed by those things
that determine a person's position on an issue, their
power and their preferences.
The political frame can also be assessed in terms
of its impact on organization structure. Pfeffer
dismisses rational planning as the normal determinant of
structure and says the contingency explanations have not

been supported empirically. A number of factors
undoubtedly interact to produce structure but clearly
political considerations are involved. "Structure is
itself the outcome of a process in which conflicting
interests are mediated so that decisions
emerge...(it)...can be viewed as the outcome of a contest
for control and influence within an organization"
(Pfeffer, 1981, p. 36).
Pfeffer believes that organizations can be viewed
as coalitions of power elements and its structure and
process reflect how the power elements have shaped things
to their own ends. Indeed, the organization is itself the
outcome of political contests.
Organizational characteristics in this context
would include an analysis of conflict, power and
coalitions. Conflict can be analyzed in terms of what
types of conflicts arise in the organization. The special
interest would be in conflicts over scarce
resourceswhere there is a disagreement between
organizational sections, for example, over who should get
what. Power can be looked at in various ways. First,
various centers in the organization (sections, divisions,
coalitions of divisions, e.g.,) can be analyzed in terms
of " they articulate their own policy preferences
and mobilize their power to enforce their demands on
organization decision-making processes" (Bolman and Deal,
1984, p. 114).

Secondly, various power sources could be
identified for various actors and units within the
organization. Bolman and Deal identify several such
sources including authority, expertise, control of
rewards, coercive power and personal power. An
organization could be analyzed along these lines and then
characterized in terms of primary and secondary methods of
power application (e.g., this organization typically uses
coercive power.)
There are additional issues concerning power which
could be reviewed as well. Kipnis and others have
discussed the types of influence (power) strategies
managers use and have identified the factors that underlie
the use of each type (Kipnis, et al., 1984).
The strategies identified are reason,
friendliness, coalition building, bargaining, higher
authority, assertiveness and sanctions. The choice of
strategy is influenced by the purpose of the influence and
the expectation for success. Managers can be categorized
as shot gun managers (using a full range of strategies),
tacticians (using reason primarily, but other strategies
selectively as well) and bystanders (exercising little
influence). From this information organizations could be
characterized in various ways in the managerial use of
influence. The type and frequency of strategies could be
identified, the number of managers fitting the
shotgun/tactician/bystander typology could be described

and the general climate for influence could be described
as an important characteristic (Kipnis believes for
example, that an organization characterized primarily by
bystander behavior would be unhealthy).
Others have looked at power and influence and
their conceptions could be analyzed in organizations.
Caildini has identified six broad influence strategies
reciprocation, similarity, commitment and consistency,
social proof, liking, authority and scarcity (Caildini,
1984). Those seeking compliancewishing to exercise
influenceoften use one of the methods to bring it about.
The organization could be characterized by the extent to
which the various strategies are used.
Kast and Rosenzweig discuss power in the context
of the psycho-social system (Kast and Rosenzweig, 1970,
pp. 309-337). They characterize influence on a continuum
with four points: emulation, suggestion, persuasion,
and coercion. They indicate that power can be classified
into three categoriesphysical, material and symbolic.
Physical power is seen in such institutions as prisons
where the applications of force is found. Material
rewards or sanctions come primarily through money and the
monetary incentive systems. Symbolic influence is
exercised through prestige and esteem. Further, these
influence systems can be characterized as evidencing
coercive, utilitarian and normative approaches. Coercive
approaches accompany coercive power; utilitarian

approaches (using money, promotion, etc.) accompany
material power and normative approaches accompany symbolic
Given these concepts, organizations could be
characterized by the way in which power is deployed.
Ordering organizations from high to low according
to the degree coercion is used we find
concentration camps, prisons, mental
hospitals...ordering (them) according to the
degree utilitarian power is used we find
blue-collar organizations...white collar
organizations such as insurance companies, banks
and civil service...(and) normative power is
predominant in religious organizations,
ideological-political organizations, colleges,
schools... (Kast and Rosenzweig, 1970, pp. 314-
315) .
Finally, contitutencies could be analyzed in the
organization. An analysis could reveal which actors,
sections, divisions, etc. align with others on critical
decisions made within the organization. Constituency
analysis could also include an assessment of outside
influences and how outside groups align with internal
operators in the decision-making process.
The only measurements found for the political
theorists were in the area of power. Price discusses
instruments designed to measure the basis for power and
spends most of his discussion on the Bachman questionnaire
Price, 1972, pp. 145-150). Bachman identifies five bases
for power, i.e., expert, reward, coercive, referent and
legitimate. He has designed a questionnaire in which
those influenced by an individual can describe the basis

upon which the influence of power is exercised. For
example, respondents can indicate whether the influence
was effective because coercive or expert power was
applied. In addition to the measures described above, it
seems clear that a researcher could develop his/her own
instrument to measure other areas discussed. For example,
a questionnaire could be developed to measure sources of
conflict in the organization or constituencies built on
key decisions.
From the discussion above, it is possible to
summarize the characteristics identified by the political
o Conflict. Assessments could be made of
what types of conflicts arise in the
organization, how they are resolved and
who wins.
o Power. Assessments could be made of how
various actors and subunits exercise
their power and what sources of power
are used. Additionally, assessments
could be made of managerial influence
strategies that are employed and of
organization power characteristics,
e.g., coerciveness vs. democratic.
o Coalitions. Assessments could be made
of how groups align internally and how

groups align with external sources to
obtain their way and exercise influence.
The Symbolic Theorists. The next school of
organization analysis is the symbolic school. As Bolman
and Deal describe it, these theorists are interested in
understanding organizational cultures and how meaning is
ascribed to events within the organization. The focus
here is on the values the organization holds and how these
values are carried through such methods as rituals,
ceremonies, storytellers, etc. Bolman and Deal believe
that the strength of shared values in the organization is
most important.
In companies where these cultural elements are
cohesive, consistent and widely shared, people
know what is expected and what needs to be done
and are motivated and committed to doing a good
job...each individual's identity is fused with
the culture (Bolman and Deal, 1984, p. 152).
Edgar Schein enlarges on the concept of culture.
Schein defines culture as:
...the deeper level of beliefs and basic
assumptions that are shared by members of any
organization that operate unconsciously and that
define in a basic 'taken for granted' fashion an
organization's view of itself and its
environment. These assumptions and beliefs are
learned responses to a group's problems of
survival in its external environment and its
problems of internal integration (Schien, 1985,
p. 6) .
Schein feels that culture:
(1) Can be found at various levels of an organization
and, indeed, there may be different cultures

operating in the various subunits.
(2) There are levels of culture to be examined in an
organization, i.e., artifacts and creations which
include technology, art and visible and audible
behavior patterns; values which include things
testable in the physical environment and testable by
social consensus; and basic assumptions about the
organization or subunit's relationship to the
environment, nature of reality, time and space,
nature of human nature, nature of human activity and
nature of human relationships.
(3) Ultimately, the true study of culture is to uncover
the basic assumptions. Such assumptions will be
reflected in the organization's values and in its
artifacts and creations (Schein, 1985, p. 14).
Sergiovanni and Corbally see culture as "...the
system of values, symbols and shared meanings of a group
including the embodiment of these values, symbols and
meaning into material objects and ritualized practices."
(Sergiovanni and Corbally, 1984, ix). They feel that the
job of leadership entails identifying common purposes
around which all organizational members can rally.
... to build a cultural federation of
compatibility which provides enough common
identity, enough meanings and enough of a basis
for the committed action of the organization to
function in a spirited concert (Sergiovanni and
Corbally, 1984, p. 107).

Given these perspectives what organizational
characteristics would be studied from the symbolic
perspective? Clearly, the following list of items would
be included:
o Culture
o Values
o Beliefs (assumptions)
o Symbols
There are ways of measuring these variables in an
organization although traditional quantitative research
methods may be inappropriate. Bolman and Deal indicate
Very few extensive empirical investigations have
used symbolic theories of organizations as a
conceptual base. It is already clear that
investigation of symbolic phenomena is
unlikely to employ traditional conceptions of
'rigorous' social science research methods.
Easily quantified questionnaires and highly
structured experimental investigations are
ill suited to the subtle shades of analysis. If
symbolic perspectives grow and prosper...they are
likely to bring with them a revival of
traditional field work methods from anthropology
and sociology and to promote the current interest
in qualitative methods and ethnography (Bolman
and Deal, 1984, p. 223).
While little has been done to develop scientific
methods of isolating and measuring characteristics in the
symbolic school, there are a few instances where attempts
have been made. Price reports work accomplished in
measuring consensus which he defines as "...the degree of
agreement on values among the members of the social
system" (Price, 1972, pp. 79). He discusses the

instrument developed by Tagiuri which is built from the
Aliport-Vernon-Lindzey questionnaire. The questionnaire
is 12 pages long and is designed to see the relative
weight respondents place on the following values:
theoretical, economic, aesthetic, social, political and
religious. Standard deviations can be completed for an
organization to see the extent to which there is agreement
on the six values.
Siehl and Martin (1984) have looked at values and
their measurement in organizations. They define three
value categoriestechnical, financial and humanisticall
of which can include unique values within them. They then
utilize a qualitative approach by interviewing key
managers to discover the organization's important values.
Schein provides a 10 step process for assessing
organization culture:
(1) Entry and focus on surprises (what strikes
the researcher as odd when he first enters
the organization)
(2) Systematic observation and checking
(3) Find a motivated insider (to work with the
consultant or researcher)
(4) Revealing surprises, puzzlements and hunches
to the insider
(5) Joint exploration to find explanation
(6) Formalizing hypotheses

(7) Systematic checking and consolidation, e.g.,
through interview, questionnaire,
examination of documents
(8) Pushing findings to level of assumptions
(9) Get reactions and recalibrating findings
(10) Providing formal written descriptions
(Schein, 1985, pp. 116-117).
While Schein feels the research can include an
examination of myths, stories, and legends he urges
caution to their use. They may be reflective of
historical happenings but not particularly relevant to the
organization's present operations and thus should be used
as a check on proposed assumptions, but not a way of
generating assumptions.
There are disagreements on the relationship
between organization culture and climate. William Evan
(1968) discusses the connection between organizational
climate and culture. He indicates that Argyris equates
organizational climate with culture but says "... if we
take culture to mean the set of beliefs, values and norms
that constitute the blueprints for behavior, then the
concept of culture as the basis for a definition or
organizational climate seems too broad (Evan, 1968, p.
109). (Schein agrees that culture and climate are
different things. He says that culture operates at one
level below climate and largely determines it) (Schein,

1985, p. 314). Evan concludes by defining climate as a
multidimensional perception of the essential attributes or
character of an organization system. The definition is
sufficiently vague so that it is hard to determine what
characteristics would be specifically isolated if one were
to measure climate. A typical list, however, might
include those discussed by Tagiuri:
Lots of fun-Serious
Non productive-productive
Rej ecting-Accepting
Unhelpful-helpful (Evan, 1968, p. 27)
From the discussion above, it is possible to
summarize the characteristics identified by the symbolic
o Organizational culture. This term embraces the
- core benefits (assumptions)
- symbols

- climate
methods of communication of culture, e.g.,
myths, stories, characteristics and rituals.
The systems school. The last school of
organizational analysis is the systems school. As was
mentioned in the introduction to this paper, Bolman and
Deal believe that systems concepts can be incorporated
into the other schools discussed. Indeed, some systems
concepts have been a part of the previous discussion on
the four frames. But the systems school has some concepts
that are not well accounted for in the previous discussion
and which will be discussed here.
Kast and Rosenzweig view several subsystems as
important for the organization as it takes its inputs and
processes them toward outputs (Kast and Rosenzweig, 1970,
pp. 109-137). The subsystems include the goals and values
of the organization, its technology, structure,
psycho-social factors and managerial processes. As noted
above, some of these subsystems have been discussed as
part of the four other schools. The key to the systems
approach, however, is to see the relationship between and
among these subsystems. "Thus each other school's
approach to organization and management has emphasized
particular primary subsystems with little recognition for
the importance of others" (Kast and Rosenzweig, 1970, p.
137). The modern approach views the organization as a

structured, sociotechnical system and considers each of
the primary subsystems and their interaction.
The characteristics of interest in organizational
analysis would depend on the level of abstraction the
researcher wanted to enter into. For example, at a high
level of abstraction, a researcher could characterize an
organization in terms of general systems theory. Here the
focus would be on the input, throughput and output of the
organization, the processes the organization takes to
identify negative feedback from the environment so changes
can be made, how the organization maintains a steady state
(homeostatic condition), and what activities have promoted
negative entropy. At a less abstract level, the analysis
might focus on adaptive and maintenance forces and
mechanisms in the organization, i.e., what conservative
(maintenance) forces the organization has employed to keep
the organization constant over time and what adaptive
forces have been present that have allowed the
organization to respond to changing internal and external
Still other characteristics would revolve around
what Kast and Rosenzweig call the managerial subsystems.
Here the focus is on boundary management with the
assumption made that there are boundaries that separate
the organization from its environment. The key goal of
management in this situation is to "...serve as a linking
pin or boundary agent between the various subsystems to

ensure integration and cooperation" (Kast and Rosenzweig,
1970, p. 136). The most important characteristics to be
studied would be a focus on how the organization has
interacted with the environment and how subsystems
management has been handled to promote integration and
Integration and cooperation imply effective
coordination and Price discusses ways of measuring
Price discusses coordination as an important
organizational characteristic and defines the term as the
"... degree to which each of the various interdependent
parts of a social system operates according to the
requirements of the other parts and of the total system"
(Price, 1972, pp. 85-89). He indicates that few
instruments have been developed to measure this
characteristic but does discuss the questionnaire
developed by Georgopoulos and Mann which focuses on the
coordination of subunit functions, e.g.,
"To what extent do people from the various
interrelated departments make an effort to avoid creating
problems or interference with each other's duties and
To a very great extent _____
To a great extent _____
To a fair extent _____
To a small extent

From the discussion above, it is possible to
summarize the characteristics identified by the systems
o Assessments could be made of the general
systems considerations for an organization.
Included would be:
an evaluation of the input-throughput
and output of the organization
an analysis of other systems features,
e.g., homeostatic processes, processes
to promote negative entropy
an analysis of the organization's
control mechanisms
an evaluation of the adaptive and
maintenance forces operative in the
o Assessments could also be made of the
organization's subsystems. Included
would be:
an evaluation of the organization's
goals and values
- an assessment of the organization's
an assessment of the psycho-social
factors in the organization
an evaluation of the organization's
managerial processes

an assessment of subsystem integration,
i.e., the inter relationships between
the subsystems
The discussion above has concerned five analytical
approaches for identifying organizational characteristics.
It is desirable to summarize these characteristics and the
types of measurements which could be used to assess them.
The matrix at the end of this section is provided for this
This section has been designed to discuss the
concept of organizational characteristics. The isolation
and measurement of such characteristics is important in
organizational research as attempts are often made to see
their relationship to important dependent variables, e.g.,
organizational effectiveness. The characteristics
described are numerous and often quite different depending
upon the organizational theoretical framework involved.
Researchers who are interested in studying the
relationship between characteristics and dependent
measures will have to delimit the list for any manageable
research. Such delimitation should be done on a logical
and empirical basis. Logically, the researcher should
analyze which of the characteristics are most likely to
influence the dependent measure. Those most likely to be
associated logically should be chosen. Empirically, the

researcher should identify which characteristics have in
past research been found to correlate with dependent
measures. By combining logical and empirical analysis the
researcher can set up meaningful and testable hypotheses
to examine the relationship of selected organizational
characteristics to other phenomena. In the case of this
study, the analytical framework was employed and those
organization characteristics used in the research are
discussed in the section on Research Approach which
Research Approach
As discussed above, there are multiple measures of
effective performance evaluation and many organizational
characteristics which could be evaluated. In order to make
the research manageable it was necessary to limit the
scope and restrict the measures of effective appraisal and
reduce the number of characteristics examined.
With regard to organization characteristics, some
typology was needed to distill even further some of the
characteristics in order to examine the relationship to
effective appraisal. A deeper review of the literature
provided a conceptual framework with which to work. It has
been suggested that an organizational environment consists
of two primary areasthe task and the human relations
dimensions. In the context of leadership, for example, it
is suggested that leaders must be able to deal effectively

Matrix of Organizational Characteristics
and Appropriate Measures
Major Characteristics Types of Measures *
# yrs or # yrs in given cycle
Hall's scale of Bureaucratization
Aiken and Hage scale on decision-making
Whisler's index on decentralization
Control Graph
Level of authority, number of occupational roles and number of subunits
Vertical span of control, i.e., number of positions between CEO and
line employees
- Tateral widths. Ratio of subordinates to first line supervisors.
- Percentage of employees in direct output job and percentage of job in
staff vs. line capacities.
Degree to which rules are written measured by
o counting number of written instructions
o asking employees about extent of rules/procedures
# personnel or size of total assets
# and size of subunits
# functions performed by different people
Degree to which procedures exist
Goal and task structure including major objectives, variability of tasks and
geographical dispersion
The writer has been as specific as possible as to the type of measure. So, for example, if a specific measurement
device has been developed it will be listed. If one has not been found the writer has listed the type of
variable(s) that would be assessed to evaluate the characteristic and/or the general techniques to be employed.

Manor Characteristics
Communication amount and
type used
Group Processes and Norms
Leadership Styles
Types of Measurement
Skill structure including
generalist-specialist makeup
- number of professional, subprofessional, technical personnel
- authority structureextent to which decentralized
Ehysical factors, e.g., amount of energy used, noise level, degree of
Routineness of work
Hage and Aiken questionnaire
Mann questionnaire
Observation techniques and questionnaire; Schien process consultation
Interview or questionnaire on style employed use of instrument like Blake-
Mouton Leadership Style Grid; Likert questionnaire
Interview and questionnaire to assess style of decision making and extent of
employee involvement
Lodahl and Keyner questionnaire
Interview and research developed instrument to measure extend agency has
undertaken employee programs e.g., flextime, training, job enrichment
Research developed instrument to assess types arising, hew resolved and who
Sources/BassesBachman questionnaire
Strategies employedKipnis (POIS) questionnaire
Organization power characteristics (e.g., coercive vs. democratic)
Measured by research developed instrument to assess through questionnaire
and/or interview