Title: CPA personality profiles and their relationship to organizational size
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00097472/00001
 Material Information
Title: CPA personality profiles and their relationship to organizational size an empirical study
Physical Description: ix, 143 leaves : ill. ; 28 cm.
Language: English
Creator: Marks, Marcus Francis
Publication Date: 1978
Copyright Date: 1978
 Subjects
Subject: Accountants -- Psychology   ( lcsh )
Accounting -- Vocational guidance   ( lcsh )
Personality and occupation   ( lcsh )
Accounting thesis Ph. D   ( lcsh )
Dissertations, Academic -- Accounting -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis--University of Florida.
Bibliography: Bibliography: leaves 136-142.
Additional Physical Form: Also available on World Wide Web
General Note: Typescript.
General Note: Vita.
Statement of Responsibility: by Marcus Francis Marks.
 Record Information
Bibliographic ID: UF00097472
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000063242
oclc - 04205061
notis - AAG8440

Downloads

This item has the following downloads:

PDF ( 5 MBs ) ( PDF )


Full Text









CPA PERSONALITY PROFILES AND THEIR RELATIONSHIP TO
ORGANIZATIONAL SIZE: AN EMPIRICAL STUDY







By

MARCUS FRANCIS MARKS


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY












UNIVERSITY OF FLORIDA


1978




























Copyright 1978

by

Marcus Francis Marks



























Dedication



To the Chairmen of my Committee and to my long suffering parents.












ACKNOWLEDGEMENTS


The author wishes to extend his gratitude to his dissertation

committee who worked very hard to provide a positive atmosphere, com-

posed of both criticism and encouragement, for the conducting of this

study. The author also wishes to thank the Florida Institute of CPAs

for their assistance in this study. In addition, he wishes to extend

his appreciation to Professors Michael Barrett and Bipin B. Ajinkya

for the positive criticisms that were given.












TABLE OF CONTENTS


Page


ACKNOWLEDGEMENTS . . . . . . . . . . . .

ABSTRACT . . . . . . . . . . . . . .

CHAPTER

I. THE STUDY . . . . . . . . . . . .

The Problem . . . . . . . . . .
Previous Related Research . . . . . . .
Criticisms of Related Studies . . . . .
Purpose of the Present Study . . . . . .
Limitations of the Study . . . . . . .
Introductory Definitions . . . . . . .
Organization of the Rest of This Study . . . .

II. CONSTRUCT, STRUCTURE, AND HYPOTHESES . . . . .

Introduction . . . . . . . . . .
Organizational Structure . . . . . . .
Size . . . . . . . . . . . .
Personality . . . . . . . . . .
Personality Coherence . . . . . . . .
Selected Personality Factors . . . . . .
Summary . . . . . . . . . . .

III. RESEARCH METHODOLOGY . . . . . . . . .


Introduction . . . . . . . .
Overview of Subject Population and Selecti
Subject Selection . . . .... .
Choice of Research Instruments . . .
Statistical Analysis . . . . .
Summary . . . . . . . .

IV. PRESENTATION OF RESULTS . . . .. ..
Introduction . . . . . . . .
Statistics on Respondent Individuals . .
Statistics on Respondent Firms and Offices
Test Results of Hypotheses . . . .
Non-Hypotheses Results . . . . .
Summary . . . . . . . .


on







Page

V CONCLUSIONS. . . . . . . . . . . 119

Introduction . . . . . . . . . 119
Overview of the Study . . . . . . . . 119
Summary of Results . . . . . . . .. 120
Explanation of the Results . . . . . .. 125
Comparisons With Prior Studies. . . . . ... 127
Implications for Future Research. . . . . ... 128


APPENDIX

A LETTER OF INTRODUCTION OF EXPERIMENTER . . .

B EXAMPLES OF QUESTIONNAIRE FORMS AND INSTRUCTIONS

C KEY TO SIZE AND OTHER STRUCTURAL MEASURES. . .

BIBLIOGRAPHY . . . . . . . . . . .

BIOGRAPHICAL SKETCH . . . . . . . . .


. . 129

. . . 130

. . 134

. . 136

. . . 143









Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirement
for the Degree of Doctor of Philosophy


CPA PERSONALITY PROFILES AND THEIR RELATIONSHIP TO
ORGANIZATIONAL SIZE: AN EMPIRICAL STUDY

By

Marcus Francis Marks

March 1978

Chairman: Lawrence J. Benninger
Cochairman: Douglas A.T. Snowball
Major Department: Accounting

Thousands of non-billable man hours and a significant percentage

of gross revenues are consumed to recruit and develop professional

public accounting firm personnel. Recently, public accounting firms

have been faced with excessive turnover rates. Several researchers

have suggested that high among the reasons is the existence of person-

ality dissonance with the firm involved. Thus, primary emphasis has

been placed on the development of a measure that matches individuals

to firms so as to reduce the personality dissonance.

This paper describes an empirical field study of the relationship

of the organizational attributes of public accounting firm or office

size to selected personality traits and long term retention of Certified

Public Accountants (CPAs). Out of the structural variables comprising

a theoretical model of the organization, the size variables of office

size and firm size were selected as being the central independent

variables. The hypotheses tested were designed to determine the rela-

tionship of six personality traits (achievement, autonomy, dominance,







harmavoidance, order, play) to either office size or firm size, sepa-

rately, for long term retained CPAs and for all CPAs regardless of

length of service with the firm or office.

There were 195 CPAs from 31 offices of 24 public accounting firms

selected at random from a restricted sample population of all Florida

public accounting firms. National public accounting firms and individual

practitioners were excluded so as to yield a sample set of firm sizes in

the middle range of the size continuum.

The Personality Research Form version A was used to develop per-

sonality profiles of the CPA subjects. A demographic questionnaire was

developed and administered to the individual respondents. Another

demographic questionnaire was developed and applied to the offices of

participating public accounting firms.

Standard PA2 factor analysis was used on the various size measures

produced by the demographic questionnaires to yield factor scores. The

personality traits and the size measures were analyzed using multiple

analysis of variance procedures.

Overall, it was found that there was a broad failure to reject the

null hypotheses regardless of whether the raw factor scores or the five

class indices based upon the factor scores were used as the measure of

the size variable. The study's main results suggested that for the six

personality traits CPAs do not differ across different size firms or

different size offices.

Subsequent analyses sought to determine whether any individual

measures of size were associated with the six personality traits. The

personality trait of achievement appeared to be related directly to

certain size measures. For all CPAs regardless of length of service


viii






with a firm or office, the size measure of the average number of con-

tinuing clients per office appeared to be related directly (at the .04

significance level) to the trait of achievement. When only the CPA

subjects who had been with a particular office for more than four years

were analyzed, the size measure of the number of partners per office

appeared to be related directly (at the .09 significance level) to the

trait of achievement. The personality trait of harmavoidance also

appeared to be inversely related to the size measure of the number of

partners per office when only CPAs who had been with an office for more

than four years were examined (at the .09 significance level).

Most of the personality trait scores developed are consistent with

findings of other CPA personality researchers. It appears that at least

two public accounting office size measures are related significantly to

the personality traits of achievement and harmavoidance.











CHAPTER I

THE STUDY



In 1970, Park E. Leathers asked the question: "Are there people

whose combinations of ability and personal qualities will lead them to

uniquely thrive in national firms, local firms, industry, or govern-

ment?" (Leathers, 1970, p. 28). His remarks were addressed to the

accounting profession, and highlight that profession's continued in-

terest in attracting and retaining talented individuals. Inferred by

his remarks is that both personal qualities and the organization's

environment are factors in determining whether an individual thrives in

his career. The present study addresses this issue; specifically it in-

volves a field study of the relationship of public accounting firm or

office size to selected personality traits and long term retention of CPAs.



The Problem


Public accounting offers career opportunities that have become

increasingly attractive to college graduates. At the University of

Washington one-half of the accounting students in 1964 envisioned their

ultimate career goal to lie in public accounting (Leathers, 1971). In

the same study it was reported that 58 percent of the graduates of

the University of Florida from 1955 through 1969 indicated that they

planned to enter a career in public accounting upon graduation. Numerous







other surveys similarly have indicated that a large proportion of

accounting program graduates enter the public accounting sector. The

profession's ability to attract members is apparent. Yet a study by

Harry I. Grossman (see Coffman, 1973, p. 2) revealed "that three-fourths

of the public accounting offices included in this survey had less than

half of their beginning employees after four years." After studying the

experience of two national accounting firms, Capin reached similar con-

clusions concerning the turnover rates (see Leathers, 1971). Public

accounting firm estimates of turnover confirm these observations. The

rate at Arthur Young after five years may run to 70 percent or more,

while at Peat, Marwick, Mitchell & Co., only 10 percent after 10 years

with the firm reach audit or tax partner (Healy, 1976). Many firms

retain only 4 to 16 percent over a 10-year time span (Sorensen

and Sorensen, 1974).

In yet another study (Kollaritsch, 1968), 1220 persons receiving a

bachelor's in accounting at Ohio State University from 1920-1967 were

surveyed. Of the accounting graduates who entered the public accounting

sector, 57 percent ultimately left the public accounting sector. Thirty-

five percent of these ended up in industrial accounting, and 5 percent

ended up in the government sector. As a majority of those who left the

public accounting sector stayed in accounting, it must be assumed that

the reason for leaving is not accounting per se.

Because of the high turnover rates pointed out above, there has

been an increasing awareness that public accounting firms are faced with

a serious problem as to the selection, placement, and retention of

people. Of these the most visible problem is that of retention.







Ellyson and Shaw (1970) have described the retention issue in accounting

as follows:


. the profession does have a problem retaining people,
and this is due in part to an inadequate selection process.
To the extent that we can increase the retention rate, we
can reduce the demand and thereby be more selective. (p. 35)


It has been estimated that the replacement cost of an entry-level

staff accountant exceeds $10,000 (Sorensen and Sorensen, 1974). The

American Institute of Certified Public Accountants (AICPA) has given

this retention problem considerable attention (see in past AICPA publi-

cations Management of An Accounting Practice Bulletin No. 20 and

Economics of Accounting Practice Bulletin No. 4) and the problem also

has been the subject of current study by that body (Ellyson and Shaw,

1970; Sorensen and Sorensen, 1974).

Thousands of non-billable man hours and a significant percentage

of gross revenues are consumed to recruit and develop professional per-

sonnel (Siegal, 1963). The problems that then arise from turnover are

1) the costs of the recruitment and development; 2) the loss of seasoned

men to work with the less qualified, to train them and supervise them;

and 3) the difficulty providing top level staff and partners for the

firm's internal growth (Ellyson and Shaw, 1970; Istvan and Wollman,

1976; Sorensen and Sorensen, 1974; White and Hellriegel, 1973). Thus,

partners complain of inexperienced personnel running their engagements

and clients complain about the continuous need to pay for the training

of new staff (Istvan and Wollman, 1976). Yet,realistically, turnover is

not completely dysfunctional. Some turnover is needed to provide avenues







for promotion; it is uncontrolled turnover that is itself bad (Istvan

and Wollman, 1976).

There are probably numerous reasons for this startling turnover

rate. High among these is the existence of personality dissonance with

the firm involved. According to Tonne (1969), people usually are fired

(or terminate) because of personality conflicts and not the lack of

necessary skills. Eighteen percent of the employees fired by a large

national accounting firm's office were terminated because of personality con-

flicts; the corresponding figure for another national accounting firm was

23 percent (Harris, 1970). Consequently, it seems that in many in-

stances it was not lack of skill but personality conflicts that were the

primary reasons for termination.

The justification for the present study may be summarized aptly as

follows: 1) A retention problem exists among CPA firms; 2) the problem

may reflect an incompatability between the nature of the firm and per-

sonality of the employees; and 3) few investigations concerning the

personality characteristics of accountants have been conducted. These

same points have been expressed well by Coffman (1973):


While the importance of selecting the right man for
the right job seems self-evident, the failure to do so
compounds the retention problems faced by the account-
ing profession and in particular the public accounting
area. The ideal personnel program of a firm is one
which recognizes the unique aspects of the nature and
personality of the firm and sets out to find the type
of persons who will be happy and will prosper in that
particular environment.

S. While the personalities of employees play a key
role in retention rates, relatively few statistical
investigations have been conducted to determine the
personality characteristics of accountants in either
the public, private or governmental areas of the
accounting profession. (p. 5)







Subsequent sections of this chapter will provide an amplification

of the research problem (and of the research objectives). Though the

paucity of empirical research in the area has been noted, an examination

of the few such studies that are available provides a useful starting

point.



Previous Related Research


Given the broad problem addressed by the present study, a signifi-

cant question is whether members of an occupational group exhibit a

consistent personality. Strong (1943) found that a characteristic

pattern of likes and dislikes differentiated many occupational groups.

According to Darley and Hagenah (1955), measured interests and occupa-

tional involvement are but a reflection of personality and contributed

towards the fulfillment of personal needs. Brown (1954) likewise stated

that "there is a semantic error in regarding an interest test as ex-

clusively that, forgetting that interests and personality are inextricable"

(p. 11). These comments suggest that an individual's personality must to

some extent be consistent with his occupational group. A significant

question is whether this need for consistency between the individual's

personality and his occupation result in a set personality pattern for

Certified Public Accountants (CPAs) as an occupational group.

When biographical data and psychological measurements are combined

with personality measures, previous researchers have shown it is possible

to base personnel selection on predictive equations with a high cor-

relation coefficient. Successful and unsuccessful salesmen have been

differentiated by Wittreich and Miner (1971) with a correlation








coefficient of 0.95 by their construction of a predictive trait pattern.

These same researchers also developed another predictive trait pattern

with a resulting correlation coefficient of 0.90 which could differ-

entiate successful salesmen and successful managers. As this study and

others (see Wittreich and Miner, 1971) have not established the validity

of predictive powers of personality trait patterns across all occupations,

personality trait patterns must be developed based upon specific work

environments.



The CPA Personality Stereotypes


Personality testing has not been studied in depth in the accounting

literature despite its demonstrated usefulness (Willis, 1975). Most of

the personality studies in accounting literature were designed to refute

or support the so-called accountants' stereotype, in particular, the

Certified Public Accountant (CPA) stereotype. Willingham and Collins (1974)

have pointed out that there is a stereotype of the personality traits

of the "average" CPA. In both the psychological and

accounting literature the CPA is stereotyped as typically emotionless,

diligent, introverted, and closure-seeking. This psychological belief

is best summarized by A. Maslow (1965):


S. the psychiatrists think of them (accountants)
being the most obsessional of any group. From what I
know of them, they also attract to the school of
accounting those who are number-bound, those who are
interested in small details, those who are tradition
bound, and the like. So much of the accountant's phi-
losophy of life ultimately boils down to a mistrust of
self.

. obsessional people in general tend to mistrust
emotions, chaos of any kind, unpredictability of any
kind, human nature in general. (p. 64)







Segal and Szabo (1964) used the Rorschach test on a group of

University of Michigan students and obtained the following findings:


a) Accountants tend to accept social norms better
than creative writers.

b) Accountants attempt to control their emotions
more than creative writers.

c) Accountants are less aware of emotionalities.

d) Creative writers are better able to show hostility,
to tolerate ambiguity and are more skillful in
handling complex emotional situations.

e) Accountants have a smaller role repertoire.
(Willingham and Collins, 1974, p. 16)

Both Maslow, and Segal and Szabo support the traditional stereotype

of accountants. Yet in later studies results have been obtained sup-

porting a contrary view. In a study by DeCoster and Rhode (1971), the

negative stereotype attached to accountants was rejected. In their

study the CPI was administered to 56 employees and partners of the eight

largest national public accounting firms in Seattle, Washington,and

discriminant analysis was applied to predict accounting firm employment

levels. It was found that CPAs scored significantly higher in soci-

ability, social presence, self-acceptance, tolerance, achievement,

capacity for status, sense of well-being, responsibility, flexibility,

and intellectual efficiency than samples of other professions. DeCoster

(1971) has hypothesized that the accounting profession is made up of

both individuals who do, and individuals who do not, fit the stereotype

(the latter group being those who fit the description provided by

DeCoster and Rhode, 1971). He felt that this dichotomy of personality

patterns caused the conflict leading to high turnover rates. Though

these studies were specifically designed to test whether accountants







fit their popular stereotype, they also have provided a general picture

of the personality traits of the public practicing CPA; in particular,

they have provided insight into the pattern of personality traits for

"Big-Eight" CPAs.



CPA Personality Studies


Starting in 1972 several personality studies were conducted that

relate either directly or indirectly to Certified Public Accountants

(CPAs). Harris (1970, 1972) found significant differences between

partners in national ("Big-Eight" public accounting firms) and account-

ing students on twelve of the eighteen scales of the California

Psychological Inventory. He concluded that the personality trait

patterns of potential employees could be compared to the personality

trait patterns of current partners to improve personnel selection

practices.

Istvan (1973) used the Predictive Index (A.S. Daniels Associates)

on 210 public accounting individuals at all levels of a national public

accounting firm and five local firms (Istvan, 1973). He found that

different patterns of personality traits existed at different work

levels. Some evidence of different personality patterns among personnel

in the different specialty areas in public accounting also was reported.

These personality patterns were based upon a four factor scale:

D = Dominance-Submission

E = Extroversion-Introversion

P = Passivity-Tension

C = Conformity-Nonconformity







The two extremes of the firms (lower level versus upper level) showed

extreme personality differences, according to Istvan's results. In-

dividuals at the lowest levels tended to have personality patterns

consisting of low dominance, low extroversion, average passivity, and

high conformity. On the other hand, partners generally showed a high

degree of dominance and extroversion coupled with a low passivity factor

and tended to be creative, stimulating, hard driving, self-starting, and

goal-oriented. Interestingly, these behavioral patterns tended towards

permanency according to Istvan's analysis, which would indicate a non-

promotability from one job level (i.e., senior to manager) to another,

as personality traits necessary for the next level job would be dif-

ferent (Istvan, 1973, 1976). Istvan's results indicated also that audit

managers scored significantly higher on measures of one personality

trait (impatience) than tax managers. From this last result he specu-

lated that different personality patterns would be exhibited by each

specialization area. This speculation was not supported by the results

of a study by Willis (1975). In using the multiple trait inventory--

the Sixteen Personality Factor (16PF) Questionnaire--Willis found no

significantly different personality patterns by specialization areas.



Job Satisfaction and Personality Studies


Along with the several studies that focus directly on assessing

the personality of CPAs or accountants have been several important

studies that have concerned relationships between job satisfaction and

personality traits. A 1969 study of job satisfaction provided support

for the hypothesis that the size of the public accounting firm had




-10-


implicit within it a certain fulfillment of a Maslow need hierarchy and,

further, that a certain personality trait pattern was needed to attain

this need fulfillment (Strawser, Ivancevich, and Lyon, 1969). Strawser

and his associates felt that the relative satisfaction of the need for

self-actualization was more available to higher level "Big-Eight"

partners while higher level non-"Big-Eight" partners had intense

autonomy needs.

A statewide firm (see Corker, 1975) has analyzed the personality

traits of their successful partners and used the norms obtained to suc-

cessfully select new personnel. These new personnel reportedly exhibited

high job satisfaction.

Finally, Willis (1975) used the personality inventory questionnaire

of the 16PF in studying job satisfaction within the offices of nine

national public accounting firms. No significant relationships between

job satisfaction and personality traits were observed. Also, no sig-

nificant differences were found among auditing, tax, or management

services specialists with respect to personality traits. It should be

noted though that the study's primary aim was to match some measure of

the level of job satisfaction of those subjects remaining more than

three years with the firm against personality traits. Also the use of

the 16PF seems to have yielded results which in part contradict the

results of many previous studies that used either the California

Psychological Inventory (CPI) or other various personality instruments.

Most of the above studies (see also Corker, 1975; Osman, 1973a;

Osman, 1973b; Osman, 1973c; and others identified in Exhibits 1.1 and

1.2), have indicated that personality trait patterns exist for CPAs as

an occupational group (and even as sub-groups within that occupational





-11-


group) that distinguishes them from the general population. Further-

more, many of the above cited studies have suggested that these per-

sonality trait patterns may be most useful in the selection of personnel

for public accounting firms.



Size and Personality


The present study is concerned specifically with the relationship

between size of the unit (or organization) and personality trait pat-

terns. However prior research offers evidence that is only tangentially

related to this issue. Porter and Lawler (1965) have described many

descriptive studies that relate significantly lower turnover ratios,

absenteeism, and higher job satisfaction with smaller units or depart-

ments. They also discussed other descriptive studies which demonstrated

parallel results for the variable of the total organization size.

Sorensen (who theorized that as the CPA firm increased in size it

became more bureaucratic) found that with increasing years of experience

in the large bureaucratic public accounting firms, CPAs became more

bureaucratic and less professionally oriented (Sorensen and Sorensen,

1972). Therefore, either the orientations changed or there was a higher

retention rate for those employees possessing bureaucratic personality

trait patterns. While Sorensen hypothesized an abrupt change, Dimarco

and Norton (1974) viewed organizations as located (and moving) along a

dimension from bureaucratic to professional. Despite this difference

in view, their findings supported Sorensen's chart of attributes of

professionalization and bureaucratization. Montagna (1968) in an

analysis of "Big-Eight" firms disclosed changes in the characteristics





-12-


of bureaucracy as measured by the organizational variables of size,

centralization, and the size of the administrative component. Porter

(1963) has found that the variable of job level within the organization

interacts with the total firm size to yield certain need satisfactions.

Strawser, Ivancevich, and Lyon (1969) who studied job satisfaction of

accountants in large and small CPA firms, responded that firm size

appeared to have some degree of influence on the relative satisfaction

of needs.

In the study by Coffman (1973), 65 accountants from selected "Big-

Eight" firms and 49 from non-"Big-Eight" firms completed the CPI in-

strument. From the results, Coffman concluded that accountants from

the "Big-Eight" firms were more dominant, outgoing, talkative, and

self-confident than non-"Big-Eight" accountants.

The foregoing studies provide indications that the personality

attributes of CPA employees bear some relationship to the size and/or

complexity of the public accounting firm. However, there is a clear

need for research efforts that (1) focus more directly on this rela-

tionship and (2) that overcome the methodological deficiencies identi-

fiable in past studies in the area. These methodological deficiencies

are addressed in the following section.



Criticisms of Related Studies


There have been no studies that are directly applicable to the

problems raised earlier in this chapter. However, the results of the

research by Istvan (1973), DeCoster and Rhode (1971), Harris (1970),

and Strawser et al. (1969) infer that individuals who are successful




-13-


in public accounting (as indicated by their position in the organiza-

tional hierarchy) have different personality trait patterns than indi-

viduals at lower hierarchical levels. In other studies (e.g., Willis,

1975) there were indications that those individuals who leave the public

accounting sector have different personality trait patterns from those

who stay. These studies offer important implications for personnel

selection procedures, yet closer examination of them highlights the

need for additional research.

The study by Istvan (1973) supposedly has demonstrated the potential

benefits of personality testing for personnel selection by public

accounting firms (see Istvan and Wollman, 1976); nevertheless, it is

open to several criticisms. First, while Istvan's sample represented

members of three national and five local firms, the data were aggregated

without regard for any variations in size or scope of operations.

Accordingly, his results are of questionable applicability to either

national or local firms. Second, Istvan presented no statistical analy-

sis of his data. Readers of his report are provided neither with the

significance levels applied nor with an opportunity to assess the

appropriateness of the analysis performed. Third, Istvan used a test

instrument not widely available, not widely known, and whose norms are

unknown. Furthermore, although the test instrument (the Predictive

Index) measures six personality traits, Istvan used only four in his

analysis (those four being dominance, extroversion, passivity, and

conformity).

Of all the previous studies, only an attitudinal study (Argyris,

1957) and some British studies (Osman, 1973a, 1973b, 1973c) have covered

a genuine range of firm sizes (from individual practitioner to




-14-


"Big-Eight"). The British studies did not differentiate their results

as to size but rather averaged the responses (Osman, 1973a, 1973b, 1973c).

Corker (1975) analyzed only one firm in terms of the personality traits

of firm members. The firm's own test instrument was employed and no

comparison was made with any other size firm. Almost all the remaining

studies have been confined to "Big-Eight" firms with no reference to any

size discrimination (DeCoster and Rhode, 1971; Ellyson and Shaw, 1970;

Harris, 1970; Harris, 1972; Sorensen, 1967; Sorensen and Sorensen, 1972;

Sorensen and Sorensen, 1974; Willis, 1975). Only three studies tried

to discriminate by size and, in all of these, the size variable was

broken into "Big-Eight" versus national public accounting firms (Coff-

man, 1973; Istvan, 1973; Strawser, Ivancevich, and Lyon, 1969). In

short, no matter in which way size is defined, there have been no valid

studies spanning more than two size groups--big and biggest. Empirical

research that gives systematic attention to the effects of the size

variable clearly is needed.

A few researchers have developed and used their own personality

measures (Corker, 1975; Hellriegel and White, 1973; Istvan, 1973). The

personality instruments employed also have included the Eysenck Per-

sonality Inventory (Osman, 1973a; Osman, 1973c), the 16PF (Willis, 1975),

the Thematic Aperception/Rorschach (Ellyson and Shaw, 1970), and the

Porter Need Satisfaction instrument (Strawser, Ivancevich, and Lyon,

1969). However, the most striking feature of the studies reported in

the literature is the widespread use of the CPI (Coffman, 1973; DeCoster

and Rhode, 1971; Harris, 1970; Harris, 1972; Osman, 1973b; Sorensen,

1967; Sorensen and Sorensen, 1972; Sorensen and Sorensen, 1974; Sorensen,

Rhode, and Lawler, 1973). Therefore,the use of another, different







-15-


70



OS.
.- I

4-
C U



u0 i-
C
-I W





*-- C )

U'
4- U



- r-
*r- *r- *r-



V) r0
--*r- C

O-


0" U
0 -*- 4-






r-


W 4-J 0
I- +-






0 rJ 0)
*r E c

0 r- -O

S- 3 nr
0.4-J 4-)
U v)
0 (A



4-) 4- =3
S- a

0. -
C n


-




U-)a

0311





















S-
CL)

S-
0
-1






-I

S-


4-)
U




u
+J -







C

D- _0
c4-







4-) c-
-o














SC)
u0






4C-
-- E



Ct



M. 0
O-





-CO






S U
070


.r0


DI 3



(- A0 O
C 4-) >-






v) E
c-



*r- O 0
( 4o




0C >,
0l.- -r





14--) .0
*r- 0 0




'4- C-- -
r O O




*r- *-



-01-
4-' r-
*- >1 E
"" : E


4-)


M
Cd

E

0
0


0
o

4-)
fC
N





0 4-
co ---
C T
r- C-
r-
* E


E 4-

0 -0
o c
CD M
4-J


OC


S.- E





WO
C S-







O
o
-0 4



VC)
O c
- E-




aE .
co


00
Er
ou
















-o
03





CO
E
m
O
S n








.0


o 0
.-- 4-) *-
4-) S- 4-)

E 4- >


'CJ -
(AWO
01> E

-r- C
- C C4
WWc

CO
' D- -- C
S- 4-'
>1 >


-0 --. X

4-i C) .-
u S-
(U 3" 4




0 4- -0



) I







Ec-
*.- W3
S- 0

U C 4-W
4-' 34-'


o -


s- E *.-

( I

CU S- 04

o =' 0)
0 4-) 0
(AW
I "






0



C)E
4r




















0I









L_)
Qca:O


I~ O


a) Sr-


"-C 4-'
WJ 00
U .-

0 I r-
(f 01 Cd

mo c

r Z *-

m 0

0 0
0o
U
WO


U 0
SC
4- ) 4-)
*r- -0 n
- 0 -
r- W l

-0 D- 0.
C Wo

0r-0- Q

or
o S-
OL
I.*- C)

0 0)

0 cW

*. E C
E U 4->
0 >
a0 S-
r Co


Cd
U.C (






*r- U
n3


co
0 0
CD-0 0
4 0 -




















tO--






E


0
= o n









C\


LE





UU
oa


Cd
0>
r--
O *r-






o4-
0




r- 4-
c(D







O'r--
O-


r--C
0 4



Qr--



OE3

U E
co
-) .








) -r-
v-
r 0




C1: T
" r-











.Un
C
0
r- 0
4-' N













rcn
0 *r-





-- 0o

4-I
Ec W
S.- 0 0






4--) VO C
*- 0) 0
2 S- U

















'-4
C-.







S-


-a


C\






-16-


I O 0)
) a>
> -4- -' *.- O
S> e X >
.>- 4 04 dj
I -r- > *r- E
U-- r -
o -- .- 4-4-) GJ > *- O





4- ) 4= .> o C
SO Y u o c-jY c- .- aj-- "



Oi- O Ln UJ 4- i- - r
r 00 0l E E L
U u"U -U

U 0--- > r -- O a> o
4- 4-3iC-D- S-
O "o CI a) m c
a N O O i >
-O UL) 4--) 4- n r s- E o 4-
Ln & m a) = o
U C-- U0 O r-- LC
n 4J 9 0 f 1: .*r- C O L e

S U Q l- o o0 - U
c Ln L CL 4 O 4-
L) O O '+* o0O ) *
C 0 *'- L *4-
0 S- u a) n a) E
O LUW w E .r-u
U c U U- > c
u = r c o n am a
,: i4->.,- n3 oo j U Q 4-,-'-- C- O
UD- --- U 0 a "- 4, +-
o -i- C- c i a a u S- El-- m r
C) ru w 0w E
o O U 4- 0--- D .r- > a) > --
4- E O r*- D S- E L- E-
I ; S- --) U O(Li- w -- 0 ro
S -> C 0 O 3 S 4- 0 > =r n3 C- 3
) (u a i-- O o) > N 0a 4- u O04-
a L .- .-- 3 .- C .- C c ~* 4- 3 U
aO = n 4l- C 0 O s- U 3 3
-u o. E E u 3 a) c
>) E-m 3 0 4-) 0 4 4-' 4- C -- 4
O 0 L SL L c Ln -
U 4-)-- C O 0 E C .0 c a)- ._ a)
U n q rQ4 C 0 OCW L Q
S- ,- m C M 4- a) o0 n
a) crn 0 = -- E C3
*r4-,- =) J 0 T 3 > --r 3 **- **-
- o M*T O +- C 4-' 4Jo c- L)
S- C X O L x 0M4- CM- 4- C
o < c-. a -- )) 4- > m n. = 4- U)
S- *I- -- L 00 lr- *a- ** 4-) *r- -C
wLJ -c (u4- =3 S- o/ uJ ca i -C 4-'
s- -O > o o0 L) 1i > 0)
o =C o 0 S.- > E E c *= c*-
3 +' S- C ) L -) C LLJ 4-' 0
m Ln 4- S. S a r m m I -CW
2 CnfJ aj cr 3 11 S- o 4 = NTa c-0r oo

I O n CJ
.-- *r- C C E 4- *r- C r- *r-
J > 0L > CO C- ( C

*r-C C- C S- Ln N 4- S- 4- U-C E Ca
C a *r- (D o =3 *r-- 0 .*- o- OC CO O- a
EE E o -1 0 24 L i> R:-- u z=
a)
4-'

C U






CC 0') >
3: 4-

a s- C-
o Oj r
C) (3) 4-
o r- -
-- o O - O- -0 L



4 U "-0
cli c() > 0 Oj wo *- 0
4J0 4-4- 0 .l U.
a) -- r CLO LO 3 O
E 0 >) k.0 cli0 o m
o 2 &.- r~. -s-m S *0
Z: (3) CG 0o-- 0C-- -3 4L'1
SJ ,: ,-. o4 ^ r^ .

c m CTt m:: M) -i








-k





-17-


u, low u
-0 >- u F. U 0)
03 .X C <0 C U
1.> c 43 u 4 o) 0>4>,0> 03




e~~1 o ur )UL
4.4 03. .* -- -. 0 44 t. 04).-4 4- 4 <- W*^<

D- o 44 0 V - 0 *4- I- C .-0
Ol 3 w tf .n 4. - e.-f l -3- -t- c
*- Cnt T- It I cn<-- Ln *r-- I**" c .0) n.C Co u n
4-4 0..-0 ZU C-3 *Va 0 : a) 0o c
.-3 L.*-** ." r.-i .s- ^ -. ** r 0 7 F '* <* i *
U o C U -ij l ij *- a oe z) u o 0 ^ *- *- ;l U AJ zL
o- U4O- C 3 c 3 CU .4 : y a J **
loE cu o i n. o o cc u c u & oi >- > c




u~ ~ I >y aj 'o^*n'-s i -'': 0 *- 4) 1 -> oi - -* a 2a
o a, a c t, -..44 a.C ~ r C C ,t 44 4 LdU > o o
**-O *= i -C J C 4 ) O L.C- -0 *O-'4-0 3-a C U -
4U C r0 -.. 4) .0 C 03 *- U -' IU -0 C 4. C 4-0
u=u o io ) *4 a 4 l> 03 c 4a> 0 3
000. > 3 t'*4-. .''L.- C *>-) 03 ~ 14. '.-4- 0 4444 C.- 444*' C..- '

ti 0 i *- C *- 3C 4.: 03 0~-0303co 0-'.--
r- u t - < L l 0 'L c* C IUU ^ C ) O u a) 2-- t--u
4 > > 0 O Oo 3 -u ci U'u 0 2 o E4 u U o *
w 4 U 4 C 3 .J .- c-.- x' w Iw s s o v a) ) w 3 a)
E 4- 0 0 >a '- i U N'- 3 CL3 o u 0 > > > >1'
03 C:. 04A-.Cj .-*... i -i > 4 -o ** Cc '- 1O4-- 'oo
4.- CL4- u i_-: 4-'.3J <-*- -' o c c x o o c -a j u 3 0: U- OU J t4 .. *-4 0 0> i03 U 4 4 (0 U* '. 4- O. '0 30 U > >
..-40 034.j -* t / >O03 t c .-400.V .40- ) 03.4 E< 4* Q 0i -c >*- .0 .0 .D0)

^- C C C 'O J O X Qj Qj i- O Q J C O-' X O O t: 'O O QJ / ^QJ Q J J
4-4-0.-U 4.-OflO.-.-' t 0 3004.0 01' 046030033.0- 0 trt 03 .4i 14









2 ^~


o- M.oti
4-)4






00
4-- )4 0
C C 44
03 4-4 0 >4


oC C

t0 C1 0 03 C > 0
















C: m mc,
4 4 44 >- a-

(^l I c1/1 Lc .?ll t^ L
OJ C: C
C'-1 (I)-^ ^ 1/ r
^' 0>4 -^ -^ .44

*- -- Iwo
+.J > C 03 03 03 4.4 ^ ^

















CO~u -CA4
4r CU .4 0. 0. 0. 04 444 4
-Q ^-. .4' 3 O -J '^ '
4i 0. (u > 03 03 03 i > >
-- C 3 4 04 3 0J 3
-^ 0. .4 -: .4- 5 -'
LU O 03 03 4. 4. 4 44--..























cc L., C3
aJ C.l l 0- 0 C
^ ^ E''-- -- r0

0'< g) .0 03 03 03 .0- .0 .0
N- 0. 0 N N N 04 04 04
^ T- i- i Eo U.4~ 03 0)U )4/ L
4 04 04.-- 04 04
C- .- .4


v.40
i nvlmc ,o
c
0 03 0




S. 1 3 3 03 0 C
4 -. 4
4-i ,- 4., .0 .09 .03 .6 .Z
3 F- 03 ,0 *
o C V V3 C 03 03
O C C C C~ Q -0

1 C W 4 0 4 U 0I 03
0 -^ C cc- C -
03 0.) 4.4u u4 44 Q6 0. t
CL.






CJ4 4.4.4 4--. 4- 0
0 4- 44-) 4 44-4 0



CO 04 44 4- 4- 4- 4.
04 c ^- -" - 4-- -" 0
> -< .- -4- .- 04 *-

4' 03 C C C .-.' .4 0.

44 4 CO 0 CO C. 03 .4*-
Q .r .r ./ '.. ko 0-"
.4 04 0. (4 44 44'-
-' ~~~ *v 03 S 5
-0(' U) < U 4. > 0.4
"F- rj n Oyc ^ c





-18-


->
-J *

E v: U,
0, o t- =) o0 m
UC < r 0)r *.-0 C 0
'0 C .-' c o Cr -0 c c ''. '- c
0- -r- C, cs 0-
0 >^ *- c- >

L CC ro0 'J L- -0 ** 0 *

C .a ,- 0 C 0 C 0 : C
VlA0 > r- CC S E C






o c > c 0 -- *i .4 a -
o c r u> C cC U 4 o
C o *>

0 o >Q CL L 0U w
C0L C L C. _r
4-. C 3 0 M U o *a
c D c r 3 > C r: C 0 O J O o Or y






















SU 0 0 0
0U 0 8 o C
COJ c O U O > M *4 a,
4-' ( Q E C7 C S 0 e' 0 4 0.-











aO..,'o C 3 >












-o C o .- o4

o ) L. 4>) m














x3 u i
o l-! TO -1 0

cl





























S L 00
v u -. > a


U'n U r ( -

































<' 0 =
V ) > S:
a, Li 04 I C W












6> W- 4) U
.0 a ) .
t- 6 a, 4. -
















r-I


~ -* - a en



0 '0 a, a,


0 -J4 0 0





LU UJ 0 UJ O










0 a-.
0, C



.3 4. 4- 4


4-4 o ct c
CI 0 0 0.




-19-


measuring instrument could provide data with which to partially support

or refute the results of the previous studies based on the CPI. Many

conflicting results have emerged from the studies previously cited; the

use of a different instrument seems appropriate for any attempt to

resolve these conflicts.

A further criticism of previous studies is that in all but one

(Willis, 1975), the statistical methods used consisted of factor-by-

factor t-tests. The use of univariate significance tests in analyzing

multivariate data may lead to erroneous conclusions. Each test focuses

on only one variable at a time, thus ignoring correlative inferences

provided by the other variables. In the case of the one study (Willis,

1975), in which data was analyzed by the use of a multivariate analysis,

the results were inconclusive. Additional research efforts that use

multivariate analysis to analyze multivariate data clearly are needed.



Summary of Criticisms


The points of criticism of these prior studies may be summarized

as relating to:

1) Improper aggregation of data by several of the previous studies.

2) Limited range, if any, of firm sizes studied.

3) Use of the same personality inventory (which has questionable
validity).

4) The use of personality inventories that were really not
applicable.

5) The use of univariate statistics instead of the proper multi-
variate statistic.

6) The general absence of a theoretical framework for the studies.




-20-


Purpose of the Present Study


The specific purpose of this study is to investigate the relation-

ship of the size of CPA firms and/or CPA offices to the personality trait

pattern of the public practicing CPAs. It is hypothesized that indi-

viduals who will best be able to adjust and stay with a particular CPA

firm or office of any specific size for over four years will have a

certain ascertainable set personality trait pattern. This personality

pattern will differ from that of an individual who has been with the

firm for less than four years. Most importantly, the pattern will differ

from the personality pattern of an individual who has been with any other

size firm or office for over four years.

If the cited deficiencies of prior research (see the previous

pages) are successfully overcome a significant contribution will have

been made by the present study. Its results will provide the profession

of accountancy and other interested parties with a personality profile

of CPAs in relationship to the size of the CPA firm or office. Though

the goals of the present study are purely descriptive, the possibility

of normative benefits exists. These personality profiles could be used

to assist students in choosing the proper size CPA firm; that is, in

choosing a firm whose member's personality set is congruent with the

student's personality set. Furthermore, the results ultimately may be

useful to the subsequent development of predictive personnel selection

equations. Chapter II will discuss studies that suggest trait measures

(when combined with biographical factors) can provide accurate pre-

dictions of job success. If successful CPAs in public accounting firms

have personality trait profiles that vary according to the size variable,





-21-


and if these personality profiles can be statistically developed, pre-

dictive equations to guide firm policies could in turn be developed.

CPA firms could apply these equations in modifying their recruiting

approach, their selection process, their motivation plans, and possibly

even their promotional practices.



Limitations of the Study


The empirical study described in following chapters involved par-

ticipation by both males and females presently holding Florida CPA

licenses and practicing within the geographical boundaries of the State

of Florida. All the subjects were in public practice. All Florida CPAs

who were in private practice with industry or governmental agencies and

institutions were excluded from this study. The CPA subjects were from

all levels of each office selected; accountants, juniors, seniors,

managers, and partners were included. Offices were selected from non-

"Big-Eight" ornon-national public accounting firms.

For the most part the processes by which samples were selected for

this study have elements of convenience involved. Accordingly, it is

stressed that the findings of this study reflect only the responses of

the subjects selected and the results have a limited generalizability

to other subjects.



Introductory Definitions


Throughout this paper several terms will be used repeatedly.

Accordingly short working definitions are provided for the following:




-22-


1) "Big-Eight": This refers to the largest eight United States

public accounting firms (traditionally selected via gross billings).

The "Big-Eight" firms are as follows: Price, Waterhouse & Co.; Arthur

Anderson & Co.; Peat, Marwick, Mitchell & Co.; Haskins & Sells; Coopers

& Lybrand; Ernst & Ernst; Arthur Young & Co.; Touche, Ross & Co.

2) Non-"Big-Eight": This refers to all other public accounting

firms which are not included in the list of "Big-Eight" public accounting

firms.

3) Certified Public Accountant (CPA): An individual who has met

the statutory requirements of the State of Florida's Board of Accoun-

tancy, is licensed thereof, and permitted by state law to title himself

a "certified public accountant" or "CPA."

4) Florida Institute of Certified Public Accountants (FICPA): This

refers to the professional organization of most of Florida's practicing

and non-practicing CPAs.

5) Job Satisfaction: This means the positive or negative aspect

of an individual's attitude or feeling toward his job (Campbell,

Dunnette, Lawler, and Weick, 1970). (For purposes of this study an

individual will be considered to have job satisfaction when the individual

stays with the same office of the same firm for four or more years.)

6) Personality: Refers to "the integration of all of an indi-

vidual's characteristics into a unique organization that determines,

and is modified by, his attempts at adaptation to his continually

changing environment" (Krech, Crutchfield, and Livson, 1974, p. 45).

7) Personality Research Form A(PRF-A): Refers to Douglas N.

Jackson's version A of the personality inventory used in this study.




-23-


8) Trait: Refers to "a persisting characteristic or dimension of

personality according to which individuals can be rated or measured"

(Hilgard, Atkinson, and Atkinson, 1971, p. 60).



Organization of the Rest of This Study


The contents of the remaining chapters can be described briefly as

follows:

A) Chapter II will present a discussion of the definitions, con-

structs, theories, and hypotheses central to this study.

B) Chapter III will present the data collection and research

methodologies comprising the operational part of this study.

C) Chapter IV will describe the data obtained and present an

analysis of the same data.

D) Chapter V will summarize the results obtained; present the

conclusions of the study; and discuss recommendations for further

research.













CHAPTER II

CONSTRUCT, STRUCTURE, AND HYPOTHESES



Introduction


Whether or not individuals hold formal membership in an organiza-

tion, organizations are a pervasive influence upon the daily lives of

all. Accordingly, the subject matter of organizational theory is of

considerable interest to a wide range of people. Significant concerns

of organizational theory are the effects of personality attributes of

members and of organizational structure upon the functions, composition,

and behavior of organizations. The present study concerns relation-

ships between (rather than effects of) personality attributes and

organizational structure. This chapter provides the framework for the

empirical study described in subsequent chapters. Five major background

areas will be discussed and subsequently integrated through a series of

theoretical constructs of an organizational model of the CPA organiza-

tion. The five areas are

A. Organizational Structure,

B. Size,

C. Personality,

D. Coherence of Personality,

E. Selected Personality Traits.

This organizational theoretical model will be used to develop

specific hypotheses as to the relationships between certain structural


-24-




-25-


elements of an organization and the personality characteristics of

members of the organization. The model will take the general form:


Pf = f(Or, C) Equation #1


Hence, Pf, the personality structure of the firm, is the function of

Or, the organizational structure of the firm, and C, the sum total of the

behavior of each individual in the firm.



The CPA


As with most other professions, CPA firms initially consisted

mainly of only one CPA where all decisions and structure centered with

this one CPA. The CPA profession grew more complex; multiple CPA part-

nerships evolved with correspondingly more complex organizational

structure (necessitated by the more complex decision structure). How-

ever, even with these more complex organizational structures, most

important decisions are made by a consensus of equals. There are still

many CPA practices that operate as delineated above as do practices

within other professions. However, the majority of the CPA profession

departed from the patterns set by the other professions. With the

expansion of the services expected of the CPA, such as auditing the

multinational business firms, the multi-office, multinational CPA firms

with several hundred partners and CPA employees came into existence.

The more partners, the more CPAs as employees, the more offices

throughout the world, and the more decisions forced a new organizational

structure on the CPAs--bureaucracy of the employed CPA. This situation


It should be noted here that behavior is used in this study only as an
surrogate measure of personality.




-26-


could be best represented by the following:


S e E
e e
$ $


S C


t t
NB B
NB "^--------------^ B


In other words, as the CPAs go from, Se self-employment to, E ,
employee, the organizational structure went from, S, simple to, C,
complex. And, consequently as the organizational structure went from
simple to complex, there was the parallel trend towards, B, bureaucra-
tization from NB, nonbureaucratization,necessary to control this more
complex organization. This relationship of professionalization and
bureaucratization will be discussed further at a later point.
Modern society is to a large extent a bureaucratic society (Kast
and Rosenzweig, 1970). The CPA profession that functions with that
society is a profession that exhibits varying degrees of bureaucratiza-
tion. It thus provides a unique subpopulation for the study of the
interrelationships of behavior, personality, environment, and organiza-
tional structure. Therefore, the subset of organizational theory
selected for discussion is that of the varying levels of bureaucratiza-
tion, and in particular, varying levels of bureaucratization within the
CPA profession.




-27-


Scope of the Study


Organizational theorists have devoted considerable attention to the

structure of organizations. Organizational structure could be likened

to the functioning of the skeletal framework of the animal: Alone it is

nothing; yet, without it, there would only be a mass of protein. Organi-

zational structure of the firm provides the skeletal framework around

which the mass of protein, the members of the firm, would interact. Many

studies have singled out the possibility that structural characteristics

of organizations might substantially affect people's attitudes towards

the organization with which they are related (e.g., see the studies

reviewed by Herman and Hulin, 1972). In the present study this organi-

zational structure will be examined as to its effect upon the personality

attributes associated with successful members of the partnership-firm.

As many studies (e.g., Sorensen and Sorensen, 1974) have shown, CPAs

remaining with a particular public accounting firm for more than four

years tend to continue to be with that firm for many years to come. For

the purpose of this study such individuals, therefore, will be con-

sidered successful, survival CPAs, who have compatible personality traits

with the firm.

To study all the various elements of organizational structure and

their impact would have been beyond the scope of this paper, and there-

fore the primary focus is upon a subset of structure--the size element

of the CPA partnership/firm structure. The inquiry centers upon the

relationship between this structural element and selected personality

characteristics of the organizational members. Before discussing this

relationship, the concepts of organizational structure and personality

will be examined individually.




-28-


Organizational Structure


Introduction


A more formal definition of structure is that it refers to the

positions and subunits of the organization and the enduring, systematic

relationship these have to one another (Porter and Lawler, 1965). There

have been many studies demonstrating a relationship between job attitudes

and organizational structure (Porter and Lawler, 1965; Herman and Hulin,

1972; Herman, Dunham, and Hulin, 1975; and others cited in the previous

chapter). Many studies (reviewed in Herman and Hulin, 1972) have pointed

out the possibility that structural characteristics of the organization

might substantially affect people's attitudes toward the organization

with which they are associated. Herman, Dunham, and Hulin (1975) have

confirmed that some structural factors account for a larger variance in

attitudes than demographic characteristics and have demonstrated also

that these effects have stability. The above discussion brings forth

the question of what are the structural variables and what affects these

structural variables.



Structural Variables


Porter and Lawler (1965) identified seven structural, independent

properties of which two concern size factors. These seven variables may

be broken into two groups; total organizational properties and subsystem

properties. Following are the seven variables:




-29-


Suborganization:

1. Organizational levels

2. Line and staff hierarchies

3. Span of control

*4. Size: subunits

Total-Organization:

*5. Size: total organizations

6. Shape: tall or flat

7. Shape: centralized or decentralized.

Originally (before Porter and Lawler, 1965), organizational levels

meant just the differentiation between managers and workers (Porter and

Lawler, 1965). Since then the term has been broadened to cover the

effects of differentiation among levels with management.

Traditionally, the line organization "is vested with the primary

source of authority and performs the major functions of the organization;

the staff supports and advises the line" (Kast and Rosenzweig, 1970,

p. 177). While traditionally these concepts of two hierarchies of line

and staff have been held, some have felt, for example McGregor (1960),

that the two are merging because of the effects of modern technology.

Span of control is defined as the number of subordinates that are

under the direct control of a particular superior within an organiza-

tion.

The shape of an organization is defined as to the number of levels

in the organization relative to the organization's total size (tall vs.

flat structure, where flat is when there are few levels relative to the

size) or from whence an organization is directed (centralized vs.




-30-


decentralized structure, where the more central the control base, the

more centralized the structure).

The two most important of the seven variables, the size variables,

are of most direct interest to the present study. Subunit size is

defined as the situation where part of the membership of the organiza-

tion is excluded from any other group of the same organization. Where

the total membership of the organization is studied and none of the

members of any group within the same organization are excluded, then

total organization size is the structural variable being studied.

As they permit comparisons between parts of an organization (or in

the alternative, with parts of another organization), the first four

variables are called suborganizational properties. On the other hand,

the last three variables require that the comparisons be made among

entities such as organizations, hence the descriptor--total organization

properties.

Since few studies have found definitive relationships between struc-

tural variables and job attitudes/behavior, cause-effect relationship

cannot be called upon to support the above defined seven organizational

structure variables. These structural variables can be referred to as

independent variables in the theoretical sense; that is, if there were

cause-effect relationships, these relationships would arise from either

variations in structure as causal agents or some other causal relation-

ship which determines both structure and attitudinal behavior.



Size


While undoubtedly all seven organizational structure variables in-

fluence the survival personality traits of a firm, this study,as




-31-


previously indicated, concentrates on the size variables. The two size

variables identified in the seven point model are the size of subunits

and the size of the total organization. Porter and Lawler (1965) have

reviewed many descriptive studies that relate significantly lower turn-

over ratios, absenteeism, and higher job satisfaction with the smaller

the subunit. Also, Porter and Lawler have discussed other positive

studies which demonstrate parallel results for the variable of total

organization size.

To describe such results as parallelism may oversimplify the

relationship. Yet if subunit size versus total organization size were

considered further, a confounding result could be projected based upon

the studies done by or cited by Porter and Lawler (1965). Graphically

presented, the following example is a partial demonstration of this

possible confounding:


A2
A







B B 2







(Example A)


In this example (Example A) the area of the circle represents the

total organization size. The lines within the circles symbolize the




-32-


subdivision of the organization into departments or offices and the

subunit's size. In respect to total organization size, Al is smaller

than A2 and B1 is smaller than B2, therefore according to Porter and

Lawler (1965) A1 and B1 should have lower turnover ratios, absenteeism,

and higher job satisfaction than A2 and B2. With respect to subunit size,

Al is smaller than A2 but B1 is larger than B2, which according to Porter

and Lawler (1965) means that Al and B2 should have lower turnover ratios,

absenteeism, and higher job satisfaction. In equation form this could

be represented as:


Total Organization A1 < A2
Size
B
yet. .

Subunit Size Al < A2

B2 < B1


It should be pointed out that the studies cited by Porter and

Lawler (1965) have focused only on situations where size of the total

organization parallels the subunit's size. Because of the parallel

sizes there is no way to interpret whether the subunit or the total

organizational size is causing the measured effects.

Thus while both total organization size and subunit size have paral-

lel results (i.e., smaller size yielding lower turnover ratios, absen-

teeism, and higher job satisfaction), if considered separately, as above,

a conflicting result occurs. It is possible that this conflict would

yield confounded or mixed turnover ratios, absenteeism, and job satis-

faction. The effects, therefore, of one type of size subunitt: e.g.,

CPA office size) could be confused by the effects of the other type of




-33-


size (total organization: e.g., total CPA firm size). As will be

discussed (later in this chapter), the present study assumes that one

type of size's effects would predominate.



Personality


Personality is that ordered complex of characteristics which dis-

tinguishes the uniqueness of an individual. That the personality of an

individual affects his behavior in work, play, and in socialization has

been well established. Field theory as developed by Lewin (1935, 1951)

indicates that the behavior of an individual, B., was a function of the

person (i.e., personality of that person), P., in a particular environ-

ment, E; that is,


Bi = f(Pi' E).


All behavior is conceived in Lewin's field theory to be a function of a

life space. This life space is made up of the individual and the environ-

ment. The relationship (Lewin, 1951) is conceived of as being dynamic,

interactive, and symbiotic between the individual and his environment.



Personality and Occupational Groups


Because this study concerns the personality attributes of accoun-

tants, an obvious question is whether members of occupational groups,

such as accountants, exhibit a consistent personality. E.K. Strong, Jr.

(1943) found that a characteristic pattern of likes and dislikes dif-

ferentiated many occupational groups. According to Darley and Hagenah

(1955), measured interests and occupational involvement are but a




-34-


reflection of personality and contribute towards the fulfillment of

personal needs. Brown (1954) likewise stated that "there is a semantic

error in regarding an interest test as exclusively that, forgetting that

interests and personality are inextricable" (p. 11). These comments

suggest that an individual's personality must to some extent be con-

sistent with his occupational group; a possibility that has not been

ignored by researchers in personnel selection.

If there is a need for consistency between the individual's per-

sonality with his occupational group, then assessment of personality

attributes of potential employees may lead to improved personnel selec-

tion. Several studies have compared personality measures and their

effectiveness in projecting successful managerial personnel (Guion and

Gottier, 1965; Nash, 1965; Hedlund, 1965). After careful examination of

many individual studies prior to 1962, Guion and Gottier concluded that

personality measures had not proven themselves as predictors of success-

ful managers. They did discover strong support for specially made keys

validated for specific situations (Guion and Gottier, p. 159). Hedlund

(1965) has reached similar conclusions as to the lack of evidence to

support the use of a personality measure (in Hedlund's case,the Minnesota

Multiphasic Personality Inventory) for managerial selection. Nash (1965)

while questioning the strength of support is somewhat more optimistic

concerning the usefulness of vocational interest measures. Nash (1966)

did obtain some successful results in his use of a special scoring key

for the Strong Vocational Interest Blank for the purpose of identifying

effective managers. His results in this later study have received

support from a similar but more extensive study by Johnson and Dunnette

(1968). The studies cited above have been concerned only with the




-35-


effectiveness of using a personality or interest measure as a sole pre-

dictor of successful managers.

In situations where biographical data and psychological measurements

are combined with personality measures, several researchers have shown it

possible to base personnel selection on predictive equations with high

correlation coefficients (Owens and Henry, 1966; Sawyer, 1966; Wittreich

and Miner, 1971; Campbell, Dunnette, Lawler, and Weick, 1974). Sawyer

(1966) made a review of 45 studies where he examined the statistical

significance of the predictive accuracies of the various methodologies.

In Sawyer's study one of the highest two "significantly superior"

strategies was the "pure statistical." An example that was given of

a pure statistical method was the "use of biographical information and

test scores in a multiple regression equation to predict managerial job

performance" (Campbell, Dunnette, Lawler, and Weick, 1974). Campbell,

Dunnette et al. (1974) noted further that their analysis of many studies

showed moderate support for the contention that the use of a person-

ality inventory and background information could be used to establish

probabilities of future success of effective managers.

Successful and unsuccessful salesmen were differentiated by

Wittreich and Miner (1971) with a correlation coefficient of .95 by

their construction of a predictive trait pattern. These same two

researchers also developed another predictive trait pattern with a

resulting correlation coefficient of .90 that differentiated successful

salesmen and successful managers (1971). These studies and others have

not established the validity of the predictive powers of personality

trait patterns across all occupations (personality trait patterns must

be developed based on specific work environments). They support the




-36-


need for consistency between an individual's personality and the per-

sonality requirements of his profession that was posited earlier. Thus

it is suggested that by use of these consistency patterns it is within

the realm of possibility to predict the success of an individual in a

particular profession. The predictive powers of these patterns would

depend upon the existence of a consistent pattern of personality dif-

ferences by profession.



Occuptional Group--Accountants


That accountants as an occupational group have a set pattern of

personality traits which are significantly different from the population

of the United States was confirmed by DeCoster and Rhode (1971) in their

various studies. Of more importance to this study is whether there is a

consistent pattern of personality differences within the profession.

In 1972, Harris found significant differences between partners in

"Big-Eight" public accounting firms and accounting students on twelve

of the eighteen scales of the CPI. Harris speculated that the person-

ality trait patterns of potential employees could be compared to the

personality patterns of current partners to improve personnel selection.

In a 1973 study, Istvan (1973) used the Predictive Index on 210 public

accounting individuals at all levels of a national public accounting

firm and five local firms. He found that different patterns of person-

ality traits existed at different work levels. The two extremes of the

firms (lower level versus upper level) showed extreme personality dif-

ferences, according to Istvan's results. Individuals at the lower

levels tended to have personality patterns consisting of low dominance,




-37-


low extroversion, average passivity, and high conformity. On the other

hand, partners generally showed a high degree of dominance and extro-

version coupled with a low passivity factor and tended to be creative,

stimulating, hard driving, self-starting, and goal oriented. These

behavioral patterns tended towards permanency according to Istvan. As

Istvan's results were not segregated as to size, it is impossible to

extend these results much further than to point out that there are

certain patterns of personality in the profession.

Several other personality and job satisfaction studies reinforce the

hypothesis that there are consistent set patterns of personality within

subgroups of the CPA profession. Two such studies are the research by

Strawser, Ivancevich, and Lyon (1969) on job satisfaction of CPAs and the

1973 study by Coffman. Strawser and his associates (1969) found support

for their hypothesis that the relative satisfaction of the need for

self-actualization was more available to higher level "Big-Eight" partners

while higher level non-"Big-Eight" partners had intense autonomy needs.

In Coffman's study (1973), it was concluded that accountants from the

"Big-Eight" firms were more dominant, outgoing, talkative, and self-

confident than non-"Big-Eight" accountants.

Though the available empirical evidence is far from definitive, the

broad conclusion that emerges is that personality trait patterns exist

that are not only distinctive to the CPA profession but also distinctive

to subgroups within the profession.



Personality Coherence


In this section three areas will be discussed and linked together

to support a basic contention that a lack of personality coherence




-38-


underlies the excessive turnover within CPA firms. The three areas

are

1. Fitness of personality to firm,

2. Perception of goal achievement-satisfaction/dissatisfaction,

3. Turnover and personality conflicts.



Personality Coherence/Incoherence


Professional practice is generally based upon the situation of an

independent practitioner. As mentioned at the beginning of this chapter

the most consistently noted departure from the traditional model of

professionalism has been the trend from self-employment to employment

in a bureaucratic organization. The salaried professional may possess

different characteristics than the independent professional who has been

described so much in the past accounting literature. With this shift

from the simple organization of the individual professional to the com-

plex organization of the employer/employeed CPA, there may be a shift

in the personality trait patterns considered acceptable to the organiza-

tion (or needed by the individual before he will accept the organization)

as the size or complexity of the organization increases.

This possible shift in personality trait patterns (as the firm's

size and complexity increases) can best be brought out by a discussion

of two forces that lead to increased compatability between personality

characteristics of employees and the firm. First, the individual's self

image (set by his personality) must be in harmony with his actions or

else he would not be able to operate effectively within the firm.

Second, role systems tend to recruit only persons who can be relied

upon to give certain reactions to the demands of the role system. If




-39-


the system recruits an individual who does not exhibit the certain

reactions demanded by the role system, then it either punishes or fails

to reward these individuals (which sooner or later leads to voluntary or

involuntary terminations). In other words,as Cohen (1961) has indicated,

all institutions perpetuate themselves by the recruitment of individuals

who are compatible. Since CPA firms are institutions, then the above

points should hold for them as well as other businesses (institutions).



Personality Coherence and Job Satisfaction


An individual will expect to achieve only a subset of his goals or

needs within the CPA firm. The decision of an individual to join the

profession or a particular firm would be dependent on the individual's

perception of the instrumentality of that profession or firm in achieving

his personal goals. This occurs because (as noted above) the person must

function in harmony with the role system and at the same time the role

system tends to perpetuate itself by recruitment of compatible persons.

When the individual perceives that his firm ceases to be his means to

achievement of a goal or goals, job dissatisfaction will occur (Schiff

and Lewin, 1974; Simon, 1957). Job dissatisfaction refers to negative

attitudes on the part of the employee towards his job; if sufficiently

strong it culminates in the individual leaving employment with a par-

ticular firm.2 Barnard has described the relationship between the

individual's perception of his job and the success of an organization

as follows:



2
Recall that in the present study, an individual who has stayed four years
with a particular firm will be assumed to be satisfied in his job.




-40-


The members of an organization, then, contribute to
the organization in return for inducements that the
organization offers them. The contributions of one
group are the source of the inducements that the
organization offers others. If the sum of the con-
tributions is sufficient, in quantity and kind, to
supply the necessary quantity and kinds of induce-
ments, the organization survives and grows; otherwise
it shrinks and ultimately disappears unless an equi-
librium is reached.
(Citing of C.I. Barnard in
Simon, 1957, p. 111)


The above demonstrates that while the individual is seeking co-

herence between his personality (influenced by subgoals) and the firm's

demands, the firm is itself seeking coherence within its structure and

the various personalities of its members. This is a partial example of

the influence of environment, E, in the previously cited equation:

C = f(P., E); Equation #3. In the following paragraph, another part

of the influence of the environment will be discussed.



Personality Coherence and Turnover


The homogeneity of the work type across CPA firms may suggest that

individuals will not shift from firm to firm. Yet several studies to be

discussed below demonstrate that there is a very high turnover rate in

CPA firms and this may be viewed as evidence of job dissatisfaction and

lack of personality coherence. In a study by Harry Grossman (Coffman,

1973), it was revealed "that three-fourths of the public accounting

offices included in this survey had less than half of their beginning

employees after four years" (p. 2 ). After studying the experience of

two national accounting firms, Capin (Leathers, 1971) reached similar

conclusions. The turnover at Arthur Young after five years may run to




-41-


70 percent or more, while at Peat, Marwick, Mitchell & Co. only 10 per-

cent after 10 years with the firm make it to audit or tax partner (Healy,

1976). Many firms retain only 4 to 16 percent over a 10-year

time span (Sorensen and Sorensen, 1974). Ellyson and Shaw (1970) have

concluded:


the profession does have a problem retaining
people, and this is due in part to an inadequate
selection process. To the extent that we can in-
crease the retention rate, we can reduce the demand
and thereby be more selective. (p. 35)


These comments lead to the question of why this turnover exists when

there is a relative homogeneity of the work type in all CPA firms.

Numerous reasons could be advanced to explain high turnover. However,

according to Tonne (1969), individuals usually terminate (voluntarily or

involuntarily) because of personality conflicts and not the lack of

necessary skills. In a study cited by LeBrone Harris (1972), 18

percent of the employees terminated by a large national accounting firm's

office were terminated because of personality conflicts; the correspond-

ing figure for another national accounting firm was 23 percent.

These are strong indications that in many instances it is not a lack of

skill but personality conflicts that is a primary reason for termination.

It is possible to reconcile the foregoing factors of high turnover

with homogeneity of work type if one accepts the hypothesis that dif-

ferent CPA firms have a different organizational structure and, as a

consequence, select individuals with certain personality traits and/or

individuals with certain personality traits select certain firms. Either

way, certain personality traits of the firm seem likely to be a function

of some organizational structural elements. The following section




-42-


addresses itself to this issue; specifically it questions the relation-

ship between organizational size and personality traits of organizational

members.



Size and Personality Traits


Except for Talacchi's very limited study (1960) and Strawser,

Ivancevich, and Lyon's study (1969), there have been few studies which

compared total organization size effects on personality traits. It is

conjectured that the reason for this is that there is difficulty finding

a sufficient span of variations and at the same time sufficient numbers

of organizations within the same industry.

In this paper, the subunit is the office of the CPA firm and the

total organization unit of size is the overall firm size. In the pre-

vious discussion (see "Size"), it was pointed out that size of the

organization and its subunits, fit into a sociological framework of

seven variables. Because of the research cited above concerning the

importance of the size variables, and because operational measures of

the size variables are obtainable, size was selected as the basis of

this study.

In examining the effect of size variables upon the certain person-

ality traits necessary for survival in a particular size of CPA firm,

the work of Sorensen (1967, 1972, 1974) provides a useful starting point.

In several of his articles Sorensen sets forth a dual model of the

organizational structure of CPA firms with the professional at the one

end and the bureaucratic at the other end (see Exhibit 2.1). The theory

behind Sorensen's dual model is that in large firm professional accounting









-43-


U

4--)

















0
4-
t.-



(U
















0
'4-






- I
x


o

0




o







Cw


a,
s-













C,)


7:












U-


S0.






































S(I-

U-



*


C
0





C
c



C0







.0 4-

t 0
N





C




C
0





-n
., 4--
0
V

4 r



SE
0 0

U C





a. L- -
.- 0 *-

N '


-c



CU
3:
0







I-
>5





0



cn








.0 C
C C
C)
2






': E
4-











C:h
mOJ
CL 0









CL)
3 -

















SOJ
0 '


















s-
Qa C





















OJ
S0
.-0 *C.




























C))
o a


























Lc
OC



4C-

04
>- 0


0:<
o OJ
= C)


4-
*- 0


C:
C



4-







o
0
0

a


c













o
0-

























o
C)
3:
0
C




0
0'












s'






























41
c.
CE











0


UC
Q.
o a


(5)
05
4-1













Cn



c


0
0




C




C)
o e

*^ 0




0 0








Co
C)
"3









0
C-'


C)


00
C
en






>0
















a
C'S


00


0a
,U


,,


C)






m
0











CL
c,
0


-c
U
C.





L
1n



0

-C


rJ
>5





0
L
U

C
0





a








0.
E








:3







Co


C->

C-'



C
S



c


4->





0
4J






C
c








0,
o-"
C)










C)
O>
VIf
1-
.0


C


V1








C
(5)


4o




Z
a)
CL








4-1
C






















(U
C)
0























4-
u
C)
a-

*4-
C



















c




















C,
0





















U
4-*















5o
I- 0








0. 0

0 0-
t- C














0
l_ 0













LC
0
*- C
0











S-~

4- '


-0
C)










0































LU
0~
rC)



4--



CU
2
1)






































OL(
c

0
>,





















w j








-3















C C
.0 4















CL)
0'








n3
00


















'5-
0














a o
u
0























.0


a'S

C.
0*





cn
















2-
VII













--C

0 u
C)























C)
'54-
V10








C0
00
-C.



QJ 0


E
C) C)
0 -

.0

r 4
2 3
m- 0

a
Q

C-'
0
7: 0



0I-
0






















tC. C
a
a
SC


SC E
C)






U
0







0
C.
at


C-'
C
C)



L>



-*



C


C
0




a C.

UJ a-


>5
*- 4-
'So

C.

0


C) 4-'
4--4-


4*-' 0.



0 0
cr



0 L
0
OS.-
-C 0

4- -
0 0

C >,

















Ct-
oo
V-) rd
0 C


















~0
SOJ









0















041
C.
















=3:C
-0
C',



























c4--
r- <












03 C)


0
0 c
Q- O
















LL
0' a
cn4










.0 i
C' 0

0 *"
ta 0







0


3 0.
4> .C
0.







.0C




CC






-0
o
















CC>
Oc
JJ




C)L














f0 C'0
Cn
0 C0






,r- T
c O
'D l

E; ^
air
C <- -
-- 0
C --
0 C >
*r- J
*- r
u L '
LJC


>5
C-C

>5
0
-I
C


C

0
0

c



0
4.1
U






C.


t-
0
C.
C







a


CS

C
0




O



C'
Q.








0
'0









>5
c
0








4-1


V,




-44-


there exists the unusual and unique combination of both the professional

and the bureaucratic organizational models. As mentioned previously,

there are two separate models which are located at the extremes of a

continuum and which conflict--the bureaucratic and the professional.

These models conflict in three major areas: a) authority; b) standard-

ization; and c) specialization (see Exhibit 2.1). Sorensen's theory is

that the larger the organization size (the firm's size or office size,

depending upon the size variable selected), the more complex the firm

structure, the more one model's characteristics dominate the other model.

It is likely that this shift from the professional end of the scale

towards the bureaucratic will cause personality conflict unless the

personality trait pattern of the individual allows him to satisfactorily

operate within this different organizational model. The reason for this

belief is because of the personality coherence mentioned earlier.

A series of research studies have sought evidence relating to this

dual theoretical model. Sorensen and Sorensen (1972) found that with

increasing years of experience in the large bureaucratic public

accounting firms, CPAs became progressively more bureaucratic and less

professionally oriented. Consequently, either the orientations of the

individuals changed or the individuals with the "bureaucratic" per-

sonality trait pattern survived. Since previously cited evidence

supports the contention that personality trait patterns are subject

to little change, it is believed that the latter not the former pre-

vails. Dimarco and Norton (1974) viewed organizations as containing a

certain proportion of each dimension from bureaucratic to professional,

not as with Sorensen an either/or situation. In other respects, their

findings supported Sorensen's chart of attributes of professions and





-45-


bureaucracies. This paper adopts the same view as Mimarco and Norton;

that is, that professionalism-bureaucratism represents a continuum.

Montagna (1968) in his analysis of "Big-Eight" firms disclosed changes

in the characteristics of bureaucracy as measured by the organizational

variables of firm size, centralization, and the size of the administra-

tive component. One of Montagna's conclusions (and the only one that

directly applies to this paper) is that as the organization of CPA firms

become larger and more complex, technically stronger patterns of

"punishment-centered" bureaucracy form which are not found in smaller

CPA organizations.

Porter (1963) has found that the variable of job level within the

organization interacts with the total firm size to yield certain need

satisfactions. As stated previously, Strawser, Ivancevich, and Lyon

(1969) who were studying job satisfaction of accountants in large and

small public accounting firms, found that firm size appeared to have

some degree of influence on the relative satisfaction of needs.

These studied, in particular those by Sorensen (1967, 1972, 1974)

and Strawser, Ivancevich, and Lyon (1969) suggest that either the

orientations of individuals changed as size changes occurred or that

those who survived were those who best fit the firm's trait pattern

expectations.

Subsequent discussion will utilize the Sorensen Dual Model in

formulating hypotheses as to the relevant personality characteristics

affected by the independent variable of size. While it is believed that

the total organization size (size of the public accounting firm) will

be the dominant independent variable against which the dependent

variables of personality will function, a special concern of the study




-46-


was the inclusion of tests of the impact of both size variables (i.e.,

subunit size and total organizational size).

Certain other structural variables cited in earlier pages (organ-

izational levels, line and staff hierarchies, span of control), also may

be related to the personality characteristics of successful CPAs. Though

the empirical study to be discussed in subsequent chapters was designed

to provide evidence relating to these variables, the primary emphasis is

upon the impact of the size variable.



Selected Personality Factors


The number of personality attributes of firm members that may be

affected by organization size are numerous: This section identifies and

discusses those personality traits likely to be affected most. These

traits were chosen on the basis of the Sorensen Dual Model, the results

of interviews with members of the accounting profession, and the results

of previously cited studies. The six personality traits selected are

achievement, autonomy, dominance, harmavoidance, order, and play.



Achievement


The personality factor labeled achievement designates a person who

aspires to accomplish difficult tasks, who maintains high standards and

is willing to work toward distant goals, who responds positively to

competition, and who is willing to put forth effort to attain excellence.

Many studies have shown that CPAs in general rank higher on this trait

than the average population. The Harris (1972), Coffman (1973),

and DeCoster and Rhode (1971) studies, all provide indications that the




-47-


achievement personality factor may be significantly affected by the

overall size of the CPA firm. Interviewees also selected this factor

as being the most likely to be affected by the size factor. As the

size of the firm increases, the achievement trait measure is expected

to reduce in strength. These expectations are also consistent with the

Sorensen model.



Autonomy


The personality factor of autonomy is illustrated by persons who try

to break away from restraints, confinement, or restrictions of any kind,

who enjoy being unattached, free, not tied to people, places or obli-

gations, and who may be rebellious when faced with restraints. Strawser,

Ivancevich, and Lyon (1969) in their study found very strong evidence

that the expectation of the interviewees is in fact a realistic one. The

interviewees also felt that the personality factor of autonomy would be

significantly affected by size of the firm, such that the larger the firm

the lower the level of autonomy. The Sorensen model strongly projects

that as the size of the CPA firm increases, the autonomy personality

factor should decrease in strength in the successful individuals in

those firms.



Dominance


The dominance factor is shown by individuals who attempt to control

their environment and to influence or direct other people; who express

their opinions forcefully; and who enjoy the role of leader and may

assume it spontaneously. Jackson (1967) found that in general all




-48-


accountants are higher in the trait of dominance than the general popula-

tion of the United States. In the interviews the trait of dominance was

selected as being possibly affected by size, namely the larger the firm,

the larger the degree of dominance associated with the successful sur-

vival individuals of the firm. Istvan's study (1973) found dominance

to be influenced significantly by firm size. However,his study of five

local and one national public accounting firm did not provide detail as

to size vis-a-vis the dominance trait except for statements in the con-

clusion. An extrapolation of the Sorensen model indicates that the

larger the firm, the higher the level of dominance that should be ex-

hibited by the members of the firm.



Harmavoidance


Those who do not enjoy exciting activities, especially if danger is

involved, who avoid risk of bodily harm, and who seeks to maximize per-

sonal safety, are exhibiting the personality trait of harmavoidance.

Jackson (1967) found that this trait of harmavoidance was significantly

higher in accountants than in the general United States population.

Again in interviews, this trait of harmavoidance was stated as being

expected to become stronger as the size of the firm increased. As the

public accounting firm gets larger in size, the more strongly will the

individual CPAs exhibit the harmavoidance personality trait according

to the projections of the Sorensen Dual Model.

The last two personality traits,order and play, are not expected

to be as pronounced in their relationship to firm size as the previous

personality traits of achievement, autonomy, dominance, and harmavoidance.




-49-


Order


The characteristics of those labelled as high in the personality

trait order, are those which are concerned with keeping personal effects

and surroundings neat and organized, who dislike clutter, confusion, lack

of organization, and who are interested in developing methods for keeping

materials methodically organized. It should also be noted that the

studies of Jackson (1967) have shown that accountants in general exhibit

stronger measures of this trait of order than the United States popula-

tion in general. In interviews order was selected as possibly being

affected by the size of firm factor in that the smaller the firm, the

stronger the trait among the firm's individual CPAs. The Sorensen Dual

Model suggests a similar set of expectations.



Play


The personality trait of play is illustrated by persons who do many

things just for fun, who spend a good deal of time participating in

games, sports, social activities, and other amusements, who enjoy jokes

and funny stories, and who maintain a light-hearted, easy-going attitude

towards life. The play personality trait was found to have a higher

incidence in the accountant population than its incidence in the general

United States population, according to the results published by Jackson

(1967). In interviews, this trait was expected to be more pronounced

among successful members of larger public accounting firms. The

Sorensen Dual Model indicates that the larger the firm, the larger is

the expected measure of play among the successful, survival members of

that firm.




-50-


Effects of Other Variables


In an earlier section it was indicated that while the size variable

is the central independent variable of interest, certain other structural

variables (i.e., span of control, organizational levels, and line and

staff hierarchies) would also be considered. It is expected that:

(a) The larger the span of control, the higher would be the measures of

achievement, dominance, and harmavoidance; (b) the more organizational

levels, the higher should be the expected measure of dominance; (c) the

less the number of organizational levels, the higher the expected levels

of autonomy and order; and (d) the smaller the ratios of the line and

staff hierarchies, the higher the measures of play and order should be.

These expectations of the other structural variables (other than size)

are based mainly upon the results of previous studies (summarized in

Porter and Lawler, 1965).

It should be pointed out that while the non-successful individuals

may not have markedly different personality trait patterns compared to

the successful individuals, it is believed that as this part of the

overall population is less stable they would be a contaminated sub-

population and thus have a less distinct "average" pattern.


Summary


The equations presented throughout this chapter provide a useful

summary of the framework for the empirical study reported in the sub-

sequent chapters. These equations are


3Implicit in equation I1 and #4 is the feedback effects of firm's influ-
ence on the environment, in general, and its employees, and the subse-
quent influence of the environment on the firm.




-51-


P = f(Or, C) Equation #1


when:

Or = f(s(Vx)) Equation #2

C = f(Pi E) Equation #3

E = f(Pf, 0e) Equation #4


where:

C = sum total of the behavior of each individual in the firm

E = the environment

i = each individual

0 = organizational structure

0 = outer world environment
e
P. = personality structure of the individual
1
Pf = personality structure of the firm

V = the seven structural variables
x

The foregoing is in the general formulation that would apply to any

organizational firm. Implicit in these equations is the assumption that

successful, survival individuals would cancel out (for the most part)

the effects of any of the few non-success individuals of the firm who

would later depart from the firm.

To apply this general formulation to the specific case to be

studied, there would have to be the assumption (supported in the body

of this chapter) that though there is the effect of the organizational

structure variables on selection of the survival personality structure

by the firm, it may not be possible to measure all of these variables.

While these variables are most probably of unequal strength and direc-

tion, they each would be a relative measure of the possible effects of




-52-


the other of the organizational variables. If this assumption is made,

then the personality of the individual CPA can be represented as

PCPA= Et where "P PA" is the personality trait structure of the in-

dividual CPA, and "t are measures of the personality traits. Likewise
x
the personality pattern necessary for the individual CPA to survive or

succeed with a particular public accounting firm may be represented by

SCPA = f(P) where "SPA is the successful CPA's personality traits

structure and "Pf" which was defined in the representations above. From

the previous equations it can be seen that "Pf" is an indirect function

of the seven organizational structure variables and the one in par-

ticular--size. This along with the assumption of the relative measures

of size previously discussed completes the set of equations.

Consequently, it is expected that the organizational structure

variable of total organization size (overall firm size) will be the

independent variable that will be linked most strongly to the dependent

variables of achievement, autonomy, dominance, harmavoidance, order, and

play according to the Sorensen model. It is expected that the larger the

total firm size the more strength would be expected in the personality

traits of dominance, harmavoidance, and play in the respective firm's

members. Achievement, autonomy, and order personality traits would be

expected to be stronger in the smaller firms. Or to put these expec-

tations in a more formal sense:

df(s) d2f(s)
Ac = f(s) -d(s) < 0
d(s) ds2

ds2
Au = f(s) df(s) < 0 d2f(s) > 0
a f(S) ds2




-53-


Od = f(s) df(s) d2f(s)> 0
df ) d-0
ds

Do=f(s df(s) d2f(s)
Do = f(s) df(s)> 0 d2f(s) < 0
F sd ds2


df(s) d2f(s)
Ha = f(s) df(s)> 0 d2 < 0


df(s) d2f(s)
Py = f(s) d(s>0 d2 < 0


where:

Ac = Achievement

Au = Autonomy

Od = Order

Do = Dominance

Ha = Harmavoidance

Py = Play

s = Size


These models are assumed to be quadratic, and time is expected to

be a constant factor after four years. The reasoning behind the above

expectations can be found in earlier discussions of this chapter. Oper-

ationalizations of the hypotheses to be tested will be discussed along

with the testing instruments in the following chapter.













CHAPTER III

RESEARCH METHODOLOGY



Introduction


The research methodology utilized by this study was selected as the

most appropriate means by which the relationship between personality

characteristics of Florida CPAs and organization size could be assessed.

Preceding chapters have identified the basic questions which underly the

study. These questions were


1) Do CPAs exhibit particular strengths of certain personality
characteristics in certain size public accounting firm offices
which differ from CPAs in larger or smaller size public
accounting firm offices?

2) Do CPAs exhibit particular strengths of certain personality
characteristics in certain size public accounting firms which
differ from CPAs in larger or smaller size public accounting
firms?


Chapter II specified certain expectations relating to these broad

questions. It was expected that those individuals who will have ad-

justed and stayed with a particular CPA firm or office of any specific

size for over four years will have an ascertainable set personality

trait pattern which will be different from a long-staying individual

at any other size CPA firm or office. While those individuals with less

than four years with the firm or office may have the same set personality

trait pattern, it was believed that the pattern would be less stable and


-54-




-55-


less ascertainable because of the contamination of the non-staying

individuals. Recall also that although size is the predominant in-

dependent variable of interest, certain other selected structural

variables are also to be studied for possible interaction effects (see

Chapter II for a discussion of these variables). This chapter describes

the research methodology employed and is organized into five sections.

The five sections concern:

A. Overview of subject population and selection

B. Subject selection

C. Choice of research instruments

D. Statistical analysis

E. Summary



Overview of Subject Population and Selection


The population studied consisted of both male and female certified

public accountants selected from two groups: a) large and small multi-

office public accounting firms, and b) large and small one office public

accounting firms. All offices were located in major urban areas within

the State of Florida. The selection process was guided largely by the

need to include a range of sizes (in terms of both variables: office

size and firm size) within the sample. The responses of CPAs from the

offices selected were expected to provide indications of similar or

dissimilar personality trait strengths across levels of office size and

levels of firm size.

All CPAs in each subject office were surveyed no matter what posi-

tions they occupied (partners, managers, staff accountants, etc.). No




-56-


"Big-Eight" public accounting firms nor any individual practitioner firms

were included in the sample. Further details of the sample selection

process and sample characteristics are provided in the following

section.

Subject Selection

Introduction

The problem of subject selection was complicated by the absence of

a single roster of the CPA population of the State of Florida. There was

no listing of Florida CPA firms that could provide the number of part-

ners, the number of offices, the location of offices, or even the number

of CPAs employed for each firm. Accordingly, several sources had to be

combined to obtain such a listing. The 1975-76 membership roster of the

Florida Institute of CPAs was combined with the Florida State Board of

Accountancy's 1977 listing of firms registered with the State. This

combination yielded the number of Florida CPAs per firm, the number of

Florida CPAs per office, the number of Florida partners per firm, the

number of Florida partners per office, and the number and location of

each firm's offices.

A number of firms were excluded from the list. First, all firms

that were "Big-Eight" or in the national top fifteen of international

CPA firms, and all the firms with only one partner were excluded. The

reasoning behind the exclusion of the larger firms was that their type

of organizational structure presents too many extraneous variables (e.g.,

shape, span of control, organizational levels, line and staff hier-

archies) other than organizational size. Restricting the sample to firms

with two or more partners was based on the assumption that to have an




-57-


organizational structure in a CPA firm requires three CPA members at a

minimum. There is an extremely low probability that firms with two

partners will have less than three CPA members.

Second, those firms which were formed less than four years previously,

or who have merged a significant sized firm into itself in the past four

years were excluded. These firms were identified from the newsletters

issued by the Florida Institute of CPAs over the preceding four years.

The four year period was chosen because success with a particular firm

by an individual was defined as that individual having been a member of

the same firm for four or more years (see Chapter II). As mentioned

previously this measure of success was based upon the results of numerous

studies which pointed out that the point of termination was normally at

the end of the third year (Coffman, 1973; Cruse, 1965; etc.), especially

in respect to Florida practices (Harris, 1970; Harris, 1972). Thus, the

personality traits of the partners remaining were assumed to be the

traits necessary for success in that size public accounting firm.

The subject population was reduced further to CPA offices located
2
in major urban cities within Florida. It was felt that the attractive-

ness of rural area practices versus urban area practices may vary

according to personality characteristics of the individual. Excluding

rural practices thus prevents the introduction of a potential con-

founding variable.




In an analysis of respondent offices and firms, only one firm had less
than three CPAs.
This study was restricted to the State of Florida for reasons of
practicality and because, of all the surrounding states, Florida has
the largest number of urban practices.




-58-


Finally, the firms and offices that were selected were chosen in a

random fashion from a series of stratified groupings of the remaining

population of Florida CPA firms. First, the remaining population was

stratified into two groups: one office firms in one group and multi-

office firms in the second group. Second, the firms were classified

within each of these groups according to the number of CPAs in the office

and/or firm. The number of CPAs was selected as the most practical and

only readily obtainable premeasure of the firm size. Third, the number

of offices constituting the subject population of the multi-office group

was reduced to only the largest and smallest offices of each firm.

There were seven pairs of large and small multi-office firms selected.

Fourth, the final subject population of single office firms was chosen

at random from two lists--larger firms and smaller firms. There were

eight large firms with only one office and eight small firms with only

one office. These single office firms were then paired large with

small in the same urban center so that the result was eight pairs of

large and small firms in the same urban cities.

The reason for stratification by both multi-office and single

office, and firm size, reflects the concerns discussed in the previous

chapter. Though firm size was expected to be the independent variable,

office size was included in the data collection so as to control for

its possible effects. By matching the large office of a firm with the

smallest office of the same firm, firm size is held constant and the



3
It should be noted that prior to the collection of data several firms
selected in the initial sample dissolved or merged with others. Re-
placement firms were selected following the same procedures used in the
initial selection process.




-59-


effects of office size can be analyzed. In comparing large single office

firms with small single office firms, firms were assumed not to differ

significantly along dimensions other than size.

It should be stressed that the subject firms (and thus the subjects

within the firms) were not chosen strictly at random but randomly from

stratified sublists. However it was felt that the trade-off of strictly

randomly selected subjects with extremely low response rates as against

nonstrictly random selected subjects with extremely high response rates

fully justified the procedure.



Procedure


The cooperation of subject firms was first elicited through a letter

(see Appendix A) indicating that the study was supported and encouraged

by the Florida Institute of CPAs. This initial contact was followed by

a telephone call explaining the purposes of the study and requesting an

interview. The investigator personally went to each office of the sub-

ject firms, presented the study, explained the materials and left the

packages of materials to be distributed to all the CPAs in the firm with

the managing partner. At this interview assurances were provided that

individual responses would be held in strictest confidentiality (unless

otherwise requested). It was explained further that at the respondant's

request a computer generated profile of his personality traits and a

norm set against which to compare those traits would be provided.

The materials packet consisted of the following:

1. A brief instruction sheet.

2. A release sheet to be signed by the sub-
ject to indicate informed consent on his behalf.




-60-


3. A one page demographic sheet concerning background infor-
mation about the individual subject (see Appendix B).

4. A reusable PRF-A test booklet with instructions included.

5. A PRF-A answer sheet used for the true-false questions on
the PRF-A booklet.

All the responses of individuals in each particular office along

with the one page demographic questionnaire sheet of the firm/office were

to be returned as a group.



Choice of Research Instruments


The study involved the administration of three questionnaires:

(1) a background questionnaire for the firm's office to fill out; (2) a

personal background questionnaire for the CPA subject to fill out; and

(3) a personality instrument for the CPA subject to fill out. The

demographic questionnaire for the firm's office and the demographic

questionnaire for the individual CPA subject are shown in Appendix B.

These questionnaires were constructed to provide background data on

both the firm's office and the subjects for the purpose of adequate

data analysis on these populations. The personality inventory was used

to measure the relative strengths of the various personality traits

under study.



Demographic Questionnaires


There were two demographic questionnaires; one for each subject and

the second one given to the CPA office. The demographic questionnaire

given to the CPA office was designed to obtain an index value of the

size variable for both the CPA firm and the CPA office. Only one




-61-


questionnaire of this type was filled out per office sampled. The

demographic questionnaire completed by each CPA who participated in the

research project was designed to provide needed data (e.g., age, sex,

rank in firm, time with firm, etc.) about the subject himself. Both

questionnaires were constructed so as to minimize time demands on those

completing them. The two questionnaires were each only one page long,

simple, and designed to be completed in a short span of time. The

questionnaire was designed to provide relevant personal data about the

subjects.

The questionnaire used by Coffman (1973) functioned as a starting

point in developing the instruments used in the present study; modifi-

cations to his questionnaires were based on interviews with various

experts from the accounting profession. The set of questionnaires were

pretested on a small random sample with minor revisions being made on

the basis of interviews with the pretest subjects and on the pretest

results themselves.



Personality Inventory


The personality inventory selected for use in this study was the

Personality Research Form A (PRF) developed during the 1960's by Douglas

N. Jackson. The Personality Research Form "is designed to yield con-

veniently a set of scores for personality traits broadly relevant to

the functioning of individuals in a wide variety of situations" (Jackson,

1968, p. 8). In other words the primary thrust of the PRF is towards

normal functioning subjects.

The framework for development of the PRF was the set of Variables

of Personality constructed by Murray (1938) in his attempt to provide a




-62-


comprehensive description of normal personality. Exhibit 3.1 lists the

set of variables included in the PRF-A. There are 14 variables of per-

sonality and one validity scale.

These 15 scale variables can be grouped into a number of super-

ordinate categories (see Exhibit 3.2) suggested "in part on the basis of

theoretical considerations and in part upon the results of a number of

factor analytic studies" (Jackson, 1967, p. 5). As certain of the groups

are explicitly based upon conceptual notions, these groupings as a whole

do not strictly define a specific factor. While for convenience these

measures can be grouped, they are distinct and have been "uniquely and

discriminantly associated with relevant criteria" (Jackson, 1967, p. 5).

Although there are five formats of the Personality Research Form,

only Form A is of interest to the present study. Form A consists of

300 true-false items which are subdivided into 15 scales with 20 items

each. The time to self-administer the Form A is between 30 to 45

minutes. Hand scorable answer sheets which were specially designed

for the PRF-A were used.

The norms for the PRF-A were developed from "separate samples of

over a thousand male and over a thousand female college students selec-

ted from initial groups approximately twice as large" (Jackson, 1967,

p. 4). Over 30 North American colleges with a wide diversity of charac-

teristics provided the pool from which the norm subjects were chosen.

In terms of summary statistics, non-college samples have conformed well

(Jackson, 1967, p. 4).

The raw scores from the PRF-A are transformed into a profile of

the subject which yields his standing in relationship to the normative

group. With a mean of 50 and a standard deviation of ten, the standard




-63-


Exhibit 3.1


(Abbreviation)


Description of High Scorer


1. ACHIEVEMENT (AC)






2. AFFILIATION (AF)




3. AGGRESSION (AG)





4. AUTONOMY (AU)






5. DOMINANCE (DO)





6. ENDURANCE (EN)





7. EXHIBITION (EX)





8. HARMAVOIDANCE (HA)


Aspires to accomplish difficult
tasks; maintains high standards and
is willing to work toward distant
goals; responds positively to compe-
tition; willing to put forth effort
to attain excellence.

Enjoys being with friends and people
in general; accepts people readily;
makes efforts to win friendships and
maintain associations with people.

Enjoys combat and argument; easily
annoyed; sometimes willing to hurt
people to get his way; may seek to
"get even" with people whom he per-
ceives as having harmed him.

Tries to break away from restraints,
confinement, or restrictions of any
kind; enjoys being unattached, free,
not tied to people, places, or obli-
gations; may be rebellious when faced
with restraints.

Attempts to control his environment,
and to influence or direct other
people; expresses opinions forcefully;
enjoys the role of leader and may
assume it spontaneously.

Willing to work long hours; does not
give up quickly on a problem; per-
severing even in the face of great
difficulty; patient and unrelenting
in his work habits.

Wants to be the center of attention;
enjoys having an audience; engages in
behavior which wins the notice of
others; may enjoy being dramatic or
witty.

Does not enjoy exciting activities,
especially if danger is involved;
avoids risk of bodily harm; seeks to
maximize personal safety.


Scale




-64-


Exhibit 3.1
(continued)


(Abbreviation)


Description of High Scorer


9. IMPULSIVITY (IM)





10. NURTURANCE (NU)






11. ORDER (OR)


12. PLAY (PL)


13. SOCIAL RECOGNITION (SR)





14. UNDERSTANDING (UN)





15. INFREQUENCY (IN)


Tends to act on the "spur of the
moment" and without deliberation;
gives vent readily to feelings and
wishes; speaks freely; may be vola-
tile in emotional expression.

Gives sympathy and comfort; assists
others whenever possible, interested
in caring for children, the disabled,
or the infirm; offers a "helping
hand" to those in need; readily per-
forms favors for others.

Concerned with keeping personal
effects and surroundings neat and
organized; dislikes clutter, confu-
sion, lack of organization; interested
in developing methods for keeping
materials methodically organized.

Does many things "just for fun";
spends a good deal of time partici-
pating in games, sports, social
activities, and other amusements;
enjoys jokes and funny stories;
maintains a light-hearted, easy-going
attitude towards life.

Desires to be held in high esteem by
acquaintances; concerned about reputa-
tion and what other people think of
him; works for the approval and
recognition of others.

Wants to understand many areas of
knowledge; values synthesis of ideas,
verifiable generalization, logical
thought, particularly when directed
at satisfying intellectual curiosity.

Responds in implausible or pseudo-
random manner, possibly due to
carelessness, poor comprehension,
passive noncompliance, confusion, or
gross deviation.


Scale




-65-


Exhibit 3.2*

Opposing scales are separated by a solid line.


A. Measures of Impulse Expression and Control

Impulsivity
Harmavoidance
Order

B. Measures of Orientation toward Work and Play

Achievement
Endurance
Play

C. Measures of Orientation towards Direction of Other People

Autonomy

D. Measures of Intellectual and Aesthetic Orientations

Understanding

E. Measures of Degree of Ascendancy

Dominance

F. Measures of Degree and Quality of Interpersonal Orientations

Affiliation
Nurturance
Exhibition
Social Recognition
Aggression

G. Measures of Test-Taking Attitudes and Validity

Infrequency











*
Douglas N. Jackson, Personality Research Form Manual, Research
Psychologists Press, Inc.: Goshen, N.Y., 1967.




-66-


(T) scores are produced when the profile is constructed. The profile

conversion involves first converting every raw score of every scale to

a cumulative score for each scale and then converting the cumulative

score to the deviates of the normal curve. This latter technique yields

a most accurate conversion (see Wessler and Loevinger, 1969).



Personality Inventory Choice


Most previous studies in related areas of interest have used the

California Personality Inventory (CPI) and the reasons underlying re-

jection of this instrument for the present study need to be discussed.

As described in Chapter I, almost all the previous studies of accountants

have used the CPI as the personality inventory. Many of these studies

had conflicting results using the CPI. One study (Willis, 1975) using

a different inventory instrument had substantially different results

from the other previous studies which used the CPI. Thus, the heavy

reliance on one test instrument and the conflicting results reported by

different studies that have used the CPI provides some justification for

the rejection of this instrument. More importantly, however, the CPI

has been subjected to serious methodological criticisms in recent

reviews.

Except for possibly the most purely theoretical purposes the CPI

was not rYecommended by most of the reviewers in Buros (Walsh, 1972;

Goldberg, 1972; Crites, 1972). They and other have pointed out many

technical difficulties, most of which would apply to the present study.

The CPI with its criterion-oriented structure was severely limited in

its psychological meaningfulness and its generality (Walsh, 1972). Its




-67-


excessive number of scales has made it difficult to interpret conceptually

(Goldberg, 1972) and, also, for several of its scales there was a strong

possibility of response set contamination (Walsh, 1972).

Further concern over the use of the CPI is raised by the fact that

the reliabilities for large and varied norm groups are quite modest

(Walsh, 1972). And most of the test-retest correlations for the CPI

scales in the CPI manual range from 0.55 to 0.84 for periods up to a

year (Gough, 1956). Finally, underlying the arguments of supporters of

the CPI is the use of circular reasoning thus weakening both its theo-

retical support and its validity (Walsh, 1972). And, in fact, Taylor

and Reitz (1968) found in their study that convergent validities for the

CPI ranged from 0.21 to 0.63. Similar problems to these for the CPI were

apparent in other alternative test instruments.

On the other hand, the PRF fitted within the time-to-take con-

straints of this study and had impressive reviews and validity results

(Anastasi, 1972; Crites, 1972; Kelly, 1972; Valentine, 1972; Wessler and

Loevinger, 1972; Wiggins, 1972). Jackson (1967) cited odd-even relia-

bilities for the PRF-A which, when combined to yield a single score,

ranged from 0.72 to 0.92. He also reported a median K-R 20 coefficient

of 0.78 for the PRF-A. Reliability coefficients, no matter what method

used, clustered close to 0.80 and none fell below 0.60.

As to both the convergent and discriminant validity of the scales

of the PRF, Jackson (1967) found both the A and B form of the PRF yielded

a median "r" of 0.52 for peer behavior ratings and a median "r" of 0.56

for self-ratings. Jackson also cited several other studies yielding

validities of similar magnitudes.




-68-


It has been pointed out by several authors that careful attention

was spent to assure the structural, external, and substantive validity

of the scales (Wiggins, 1972). The PRF-A scale has been called the "most

methodologically sophisticated personality inventories presently avail-

able" (Wiggins, 1972, p.123). Robinson and Shaver (1973) also have claimed

that the PRF "having been carefully developed and validated, stands)

among the very best multitrait scales available" (p.97). Thus, as to

psycholmetric properties, the PRF-A appeals as the most appropriate

personality trait inventory for this study.


Statistical Analysis


The statistical analysis is best viewed as two separate steps; the

factor analysis to obtain an index of size, and the multiple analysis of

variance (MANOVA) to analyze for the independent and dependent variables.



Factor Analysis


The first stage of the statistical analysis was to find an index for

each firm and for each office. Two index values were required in order

to analyze whether office size or firm size was the major independent

variable. The firm's demographic questionnaire included many possible

measures of size for the firm and the particular office studied. Because

of the many possible size index variables, the few (24 firms and 31

offices) observations, and a few missing responses per subject, it was

necessary to use the factor analysis routine from the Statistical

Package for the Social Sciences routines. This provided for the pairwise

deletion of missing values (i.e., "a case is omitted from the computation




-69-


of a given correlation coefficient if the value of either of the

variables being considered is missing," SPSS, 1975, p. 504).

In classical factor analysis it is assumed that observed correla-

tions have been produced by some common underlying systematic element

of the data set. That is, implicit in factor analysis was the presump-

tion that the common elements were the only ones contributing to the

relationships sought. The basic model could best be stated as:


Z = aF + a2F2 +..+ a F + d.U. j = 1, 2,..., n
Zj j 1 +mm" a m Jm

where

Z. = variable j in standardized form

F. = hypothetical factors

U. = unique factor for variable
J
a.. = standardized multiple-regression coefficient of variable j
S on factor i (factor loading)

d. = standardized regression coefficient of variable j on unique
J factor j


Assumptions:

r(Fi Uj) = 0 i = 1, 2,..., n; j = 1, 2,..., n; i j


r(UU)=0 j / k

(SPSS, 1975, p. 471)


There are many methods of factoring available. These various

factor methods differ from one another only in the procedures used for

estimating the communalities. The method to be selected in the present

study is principle factoring with iterations (PA2). The PA2 yields

the so called inferred factors by replacing the main diagonal elements




-70-


of the correlation matrix with communality estimates based upon the

squared multiple correlation of a variable and the rest of the variables.

The PA2 method is also selected because it uses an iteration process to

improve the estimates of communality (SPSS, 1975). After the number of

factors to be extracted from the initial correlation matrix is deter-

mined, the main elements (i.e., the diagonal ones) are replaced by the

initial estimates of the communalities. This process is repeated as new

estimates of communalities continually replace the diagonal elements.

When the differences between two successive communalities are negligible

the iteration ceases. This PA2 factoring is considered the "most

widely accepted factoring method" (SPSS, 1975, p. 480).

Because of the vast difference between the number of possible index

values and the number of observations it was necessary to slowly drop

the highest and lowest correlation coefficients each run until the factor

score coefficients could be obtained. Then when a single score matrix

could be chosen with high values it was used to obtain an ordering of

the firms. This ordering was arbitrarily broken into five size groups

(small to large); either 5 (large),4 (medium-large), 3 (medium),

2 (medium-small), or 1 (small) was assigned as the size index value for

each firm.



Multiple Analysis of Variance (MANOVA)


In the second stage of the analysis, a comparison of size measures

will be made by interpreting the patterns emerging from the plotted

profiles of the PRF scale scores for all the traits. This interpreta-

tion is to be based on visual analysis. Computer analysis and printouts




-71-


was made for each comparison group of the means, standard devia-

tions, and t-values to be connected with the different personality

characteristics. For the six selected dependent variables (achievement,

autonomy, order, dominance, harmavoidance, play) an analysis using the

MANOVA program in the Statistical Analysis System (SAS, 1976) will be

performed.
4
The linear model to be used for MANOVA is as follows:


Xab = k + (ka k) + (Xab ka)


where Xab is the dependent vector variable for the bth subject in the ath

sample (i.e., a = 1, 2,..., p, where p is the number of populations under

study); k is the vector of total sample means; k is the sample mean for

sample a. As we are not interested in the vector of total sample means

but rather in casting the linear model in terms of individual subjects'

deviations from the vector of total sample means we subtract the vector

of total sample means from both sides, hence, yielding:


Xab = (ka k) + (Xab ka)


Note that the first term represents the "hypothesis effect" (the

differences in the locations of groups) and the second term represents

the "error effect" (the deviations of subjects from their sample means).

The fundamental partition theorem of MANOVA is yielded by taking

the squares and cross products of the elements of the score vector




Some sections of the following discussion on the linear model of MANOVA
and Wilk's Lambda are taken in part from Cooley and Lohnes (1971).




-72-


and its two partitions and summing over all subjects in the total

sample:


M M M
a a a
SXab X' (k k)(k k)' + (k k)(kb k)'
a=l b=1 a=1 b=l a = b=



T = A + W


Thus with the use of mnemonics we simplify the equation to:

T = A + W. To these matrices we assign the names:

T = Total

A = Among-groups

W = Within-groups

When the null hypothesis holds, the independent estimator of the

common populations dispersions, A, is equal to each of the partitions

divided by its degrees of freedom. The group means vector's estimator

can be presented as:


S= 1 (A)
A (p- 1)


which when we let M Ma yields the estimator based on the pooled
a=1
within groups definitions. Hence giving us:


DW = M- (W)


The null hypothesis is that va = for a = 1, 2,..., p, and if it

holds the best estimator of the common populations samples means, p,

is k, the vector of total sample means:




-73-


M
Sa
k = 1 Xab
a=1 b=1


This can be represented as:


H: = a


The estimator of the research hypothesis effects is the matrix of

deviation of group means from the grand means if the null hypothesis is

rejected.

The test of the null phyothesis was formulated by Wilks in 1932.

Wilks' determinant ratio test statistic is usually denoted as Lambda (A)

and is defined as:


SIwI
ITI

It should be noted that the confidence with which we reject H increases

as ITI increases relative to IWI. The Lambda test is of the null

hypothesis of the equality of the mean vectors. If the Lambda test has

produced a rejection of the null hypothesis then the inspection of the

univariate F-ratios may suggest which of the elements of the vector

variable are contributing most to the discrimination of the group.

The statistical significance testing will utilize MANOVA to deter-

mine the reliability of differences in the scores of the comparison

groups. Note that a simple analysis of variance (ANOVA) could be used

for studies which involve a single dependent variable to be measured on

several samples or cases that are suspected of arising from somewhat

different populations. In ANOVA analysis the dependent variable must

be assumed to be normally distributed with the same variance in each




-74-


population, and the research issue is concerned mainly with the validity

of the differences among the respective population means. Since this

study's main concern is of the group differences in location in a set

of multi-dimensional measurement space, MANOVA will be selected as the

appropriate analytical tool. The central distinctive element of MANOVA

is that the dependent variable is a vector variable. Just as in ANOVA

designs there is an assumption of homogeneity of variances, in MANOVA,

the dependent vector is assumed to be multivariate normal in distribu-

tion with approximately the same dispersion (variance-covariance matrix)

for each population. A test will be made for Wilk's Lambda by use of

the SAS program package.

It should be acknowledged that truly independent tests of univariate

hypotheses will not be made. Such an analysis is very complex and the

resulting "step down F-ratios" will contain various problems of inter-

pretation (Bock, 1975).



Null Hypotheses


The operational hypotheses were as follows:


H 1 There is no significant difference in any of the six personality
o characteristics as measured by the PRF-A of CPAs practicing in
any size Florida public accounting firm, with CPAs practicing in
another size Florida public accounting firm.

H 1 There is a statistical difference in the above. The larger the
a firm size the higher the level of Dominance, Harmavoidance, and
Play. The larger the firm size the lower the Achievement trait
level. The smaller the firm size the higher the measure of
Autonomy and Order.


No tests of homogeneity were run. The MANOVA test was strong enough
that no tests were necessary (Morrison, 1967; Bock, 1975). Also visual
examination of the variance-covariance matrices showed no lack of
homogeneity.




-75-


H 2 There is no significant difference in any of the six personality
o characteristics as measured by the PRF-A of CPAs with four years
or more with a particular firm practicing in any size Florida
public accounting firm, with CPAs with four or more years with
a particular firm practicing in another size Florida public
accounting firm.

H 2 Same as H 1.
a a
H 3 There is no significant difference in any of the six personality
characteristics as measured by the PRF-A of CPAs with four years
or more with a particular firm practicing in any particular
size office of a Florida public accounting firm, with CPAs with
four or more years with a particular firm practicing in another
particular size office of another Florida public accounting firm.

H 3 There is a statistical difference in the above. The larger the
a office size the higher the level of Dominance, Harmavoidance,
and Play. The larger the office size the lower the Achievement
trait level. The smaller the office size the higher the measure
of Autonomy and Order.

H 4 There is no significant difference in any of the six personality
o characteristics as measured by the PRF-A of CPAs with a particu-
lar firm practicing in any particular size office of a Florida
public accounting firm, with CPAs with a particular firm
practicing in another particular size office of another Florida
public accounting firm.

H 4 Same as H 3.
a a

(Note that after four years with the firm no shift in responses are
expected.)


While there are no formal null hypotheses for the other three

structural variables previously discussed (span of control, line and

staff, and organizational levels), MANOVA and ANOVA tests were run on

their possible effects as independent variables and as to possible

interaction effects.

6It was expected that the larger the span of control the higher would be
the measure of achievement, dominance, and harmavoidance. The larger
the number of organizational levels, the higher the degree of dominance
that would be expected. There would be expected a higher degree of the
personality traits of play and order when there is a lower line and
staff hierarchies ratio. And finally, the lesser the number of organi-
zational levels, the stronger would be the expected measures of the
personality traits of autonomy and order.





-76-


Summary


This chapter discussed: (1) the identification of the population

to be studied, how the subject sample was selected, what instruments

were to be applied to this population, and why; (2) the application of

the instruments to the sample; (3) the methods of analysis that were to

be used on the data obtained; and (4) the research hypotheses that were

tested. In the next chapter the results of the responses of CPAs in the

31 offices (24 firms) examined will be discussed.












CHAPTER IV

PRESENTATION OF RESULTS



Introduction


In this chapter the results of the study are presented. Included

with the statistical analysis are some descriptive statistics as to the

respondent populations' composition. These descriptive statistics

provide necessary background for the discussion and conclusions pre-

sented in Chapter V. This chapter is divided into five main sections:

a) Statistics on respondent individuals

b) Statistics on respondent firms and offices

c) Test results of hypotheses

d) Non-hypotheses results

e) Summary.



Statistics on Respondent Individuals


Characteristics of the individual CPAs participating in the study

may be classified into 1) general characteristics, 2) CPA related

characteristics, and 3) personality characteristics. It should be noted

that there were 195 respondents from 31 offices, 24 firms.


General Characteristics

In Table 4.1 the individual respondents are classified according

to age groups. The individual respondents ranged in age from 22 years


-77-




-78-


of age to 82 years of age. However nearly two-thirds of the respondents

are under the age of 40. As the practice of accounting has changed

dramatically in the last 20 years, this may bias the results of this

study towards the CPAs entering practice after 1950.


Table 4.1

Respondents Based on Age

Age Number

Under 25 12
25 to 30 56
31 to 40 60
41 to 50 39
Over 50 28
195
NOTE: Ages ranged from 22 to 82


Table 4.2 shows that less than 7 percent of the respondents were

women. While there is a higher percentage of female CPAs than this seven

percent in the state of Florida, many are in private, governmental, or

education practice rather than public practice. Thus, this 7 percent

seems to be representative of the larger population from which subjects

were selected.


Table 4.2

Respondents Based on Sex


Sex Number

Male 182
Female 13
195




-79-


CPA Related Characteristics


In this section summary tables describe the essential character-

istics of the individual respondents as they relate to CPA firms. Table

4.3 presents the individual respondents in relationship to the number of

years they have held a CPA certificate. The number of years that sub-

jects had held CPA certificates ranged from one to 50 years. Just under

50 percent had held a certificate for under six years; 87

percent had held their certificate for twenty years or less.



Table 4.3

Respondents Based on Years Holding CPA Certificate


Years Number

Under 6 years 93
6 to 10 30
11 to 20 46
21 to 30 21
Over 30 5
195
NOTE: Years holding ranged from 1 to 50 years


In Table 4.4 the individual respondents are classified according to

the number of years they have been with the same firm. Eighty-nine per-

cent had been with the same firms for 20 years or less. Forty-four

percent had been with the firm for under five years. The number of

years with the same firm ranged from one to 40 years.




-80-


Table 4.4

Respondents Based on Years With Same Firm

Years Number

Under 5 years 86
6 to 10 42
11 to 20 45
21 to 30 15
Over 30 7
195
NOTE: Years with same firm ranged from
1 to 40 years


In Table 4.5 the individual respondents are tabled according to the

number of years they had been with the same office. Ninety percent had

been at the same office for 20 years or less; 70 percent had been

at the same office for 10 years or less; and 48 percent had

remained with the same office for under 5 years. The number of years

with the same office ranged from one to 40 years.


Table 4.5

Respondents Based on Years With Same Office


Years Number

Under 5 years 93
6 to 10 44
11 to 20 38
21 to 30 13
Over 30 7
195
NOTE: Years with same office ranged from
1 to 40 years


The individual respondents are classified according to their posi-

tion in the firm in Table 4.6. As would be expected because of the

selection of medium to small public accounting firms, the highest




-81-


percentage of respondents were partners (54 percent). Thirty-six percent

were in positions ranging from staff accountant to senior; in other

words, they were lower level employees.


Table 4.6

Respondents Based on Position in Firm


Rank Number

Partner 105
Manager 11
Supervisor 8
Senior 15
Junior 29
Staff Accountant 27
195


The individual respondents were then classified according to their

work specialty in the firm or office. This classification is presented

in Table 4.7. Over 71 percent of the individual respondents either

indicated no specialty at all or claimed more than two specialties.

Fifteen percent were auditing specialists. Tax specialists comprised

11 percent of the individual respondents.


Table 4.7

Respondents Based on Specialty Type


Type Number

Auditing 30
Other 4
Tax 21
Management Advisory Services 1
Mixture/No Specialty 139
195




-82-


Personality Characteristics


In Table 4.8 certain statistics on personality trait character-

istics of 1) all the respondent CPAs, 2) all CPAs with a firm for more

than four years, and 3) all CPAs with an office for more than four years

are presented. Means are provided for 14 traits; the six traits

(achievement, autonomy, dominance, harmavoidance, order, and play) of

major interest to the present study are highlighted in the tables pre-

sented. The means for all subjects ranged from a low of 44.1 for the

trait characteristic of aggression to a high of 60.6 for the trait of

endurance. The means for those CPAs with a firm for more than four

years ranged from 43.2 (for the trait of aggression) to 61.2 (for the

trait of harmavoidance). For all the CPAs with an office for more than

four years the means ranged from 42.8 (aggression) to 61.1 (harmavoid-

ance). The lowest mean score for each of the three groups occurred for

the trait of aggression. The highest mean score of both the CPAs with

an office and CPAs with the firm for more than four years was for the

trait of harmavoidance; however, the total responding CPAs attained

their highest score for the trait of endurance.

In all tables presented in this subsection, the means should be

compared to the Jackson (1967) standard (T) scores. These standard

scores have uniform means of 50 and standard deviations of 10 for all

traits. The means of trait scores for all the CPA subjects displayed

six (see Table 4.8) traits in which the CPA subjects scored higher than

the general population means; four in which they scored the same as the

general population; and five in which they scored lower means than the

general population. The all CPA subjects group's means were higher





-83-









aj i- -ino crIcrCO CUo Q o CO ~ CO
cn CO--) Wh C \J C -O r CODM W r~- U3
C I -I -I kI w r w I I I I


>-4
S- cl m i C i i m i m i i C i i \i iO rV



u r C'ID- CD.o DO k-0 r, 00 0O CD C O::I ',D

s U
o ( o CD--4 C 0 00" o ,o' CO

u- t+- ) CDm CD CDm CoDC-i co coco coo i0 m u
C I I

r5-
I- 0
c oc) m wr- CD r- m w r--m O-O-Acm 4-

S- a) CD CO i co -L f ioO -f :d- C CO e

4-)


Sa) -41-4Co mwCYCON\J0 -0nmCOr- co
LJ_ cn CO ~ h h ~(~(~~ COCO ~~CO -'^Ot~ 0
c I IQ I I I I I I I I I I I I I En

*- &- 0

in >-

3 . . . O N m -
0 u- Vn c4 m c-CCoC00Dw 00r-ioo C0 Or--m

0 Wf O4-)o LnomoeoOe
(I I ) i- -I -I --I C~ (


0- r ro
C-) _r_ a)

Sm mm COo m c-LnCOCO

a) 00o 0
( C0 S.4-C I
S U a4- c
-- o u mD m m mLmn cDc r- CD orWO O ro
U) wn o ~r-wcor r- (~ COw COwr- -~ o (-C a)
Sc I I I I I I I I I I I I I I -) -C
a)-4- r3 cooU m LnIO d- m m w-iLn l CO oDo u --)
S..*- 0
oo 0 4-
4I) EI) 4o )
S: S- L)
Mr CD M rn iiM IDO CD rCDno o ro';I- Ln o Wu
4-+j n mI M )~ n ri "DC\J (r, cr 0) D" cD0 Ln C\j4- s-

Q o o00cococococon u +
F- V) -L

U Q- 4 0 >
R 50 0 r
S- C 00-C4-Nio4mC0OLOl-'4 m C f O) ro
'-C C E
CO c c -o)-ooMDrCD C \Mn c M M --i M 4- ) i
m Lm :t :d- m "o %z- "D -:d m mo -;z- :'I- O 4--) -i- > S-
o a) a0
4-- h- C)
OLI '4-
r S- < _0 0
C 5- 00- S-
o 0 4-c) ro LI

n c- O -r E +) 4- 0 m C




LI 4-- ,> U r t or >a) Law L
Uo C E o U cU > C
S- S.- a E c C 4-' 0T-- O 4 -
CU ro > i O o -T > Ln S.X X Ln 0- % -Q I aD
Saji- aj c c s- Q rfo- 3 s--x n S.- au M =c
S- 0 S5 0 E =3 4-3 a)4- Cu -4 -03-c S- C SL.- 3 ro U -34- X
u4- c) o0C5 x r ) E 3 s-U- O a c
< << WW 2I-z CD C- V) -




-84-


than the general population means in the traits of achievement, dominance,

endurance, harmavoidance, nurturance, and order. Their means were

relatively equal to the general population means in the traits of affili-

ation, social recognition, understanding, and infrequency (a validity

testing trait). All CPA subjects group's means were lower than the

general population in the traits of aggression, autonomy, exhibition,

impulsivity, and play. Finally,as to the standard deviations, two traits

in an analysis of all subjects' responses had higher standard deviations

than the general population had while the rest of the traits' standard

deviations exhibited lower values than the general population. The

traits of exhibition and impulsivity showed higher standard deviations

for the group of all CPA subjects than the general population (see

Table 4.8).

In Table 4.9 the same statistics (the means, the standard devia-

tions, and the ranges) of the personality trait characteristics of the

respondents are presented, but are classified according to sex. For

the male subjects, the means ranged from 43.7 for the trait of aggres-

sion to 60.6 for the trait of harmavoidance. For the female subjects

the means ranged from 47.3 for the trait of impulsivity to 64.6 for the

trait of achievement. Female subjects had higher mean scores for all

the personality traits except three: These exceptions were harmavoidance,

nurturance, and order. When Tables 4.8 and 4.9 are compared, again the

female subjects had higher mean scores than all CPA respondents except

for two traits: harmavoidance and nurturance.

In Table 4.10 the subjects' responses are classified by age groups.

For the age group of 22 to 29 years of age, the means ranged from 46.3

for the trait characteristic of impulsivity to 59.5 for the trait of






-85-


CD r-- ;i- CD CD ko 00 I '-- CD r c\l ::I- r--

00 to 10 r~ '-4 t) t-0 rC k.0 *.- r1 r-.

m cJ n ^- n0 C o- (J C\J e- ) C










m C:) ( J -:- C\ g o0 t- 00 Oc
t0 r- CD .: r-. a~ r ) om r-.- c- -4 o0 C)

aC co Co o 0 1 '-4 10 C\J Q C)
-4 -4 ,-4 .-i -


L9 OC L1 Oc r-4 C) C\] CM C\! C) i-
-:1- 00 -: ) CM r- 0 C) CD 1-
1 0L t01.011 m m t ::J- LO 1 0 L0


10 LO
CI I
C"] '.


m-- r--


m CM


C'i C)
cO O
I I


r- 0I
I I

Cm CM.


LO 00


(XI CI


CO


-x
X
CU


O0


0)





4-'
U)

4-j
0)
U

--





F-I-
(J


*-



4-'
(l



CD
u


5-


0v
L/

0-


c_-
O c t
r 0 w0 (0 0
4- .- >, U U -
n E C C
)- 0 (0 -
r- 0) C C 5- .
*I- 0 =3
4-- "-' E -0 C
4- 0 = 0 C X
< CC< cm Q wL


()
U
C >l
0 4-'
-0 *.- U
.-- > C
o > cL
0 (
E 4-'
.- 0- S-
ro E 3
r -


4r-

C 0
0 C

0 C
O0


rO S



CL 0) =)


r- C'..] C) CD r-- i-. M') i. r-:- M Mn M'1
C\.] Mo 00 o 0 t..- o ^) 1.0 ri- rn 1l O

n O- O- 0) Co C) 0) C) co i r0- CD 0 0 0










C:) C') Co C". C:) r-. C) C". 1000 CoC0
10 ::j j- 10 t ~ cz 1 .0 ;:I 101 LO ;I :' Z


U
4-)
CU


*r-

S-

5-

-C
u_




-86-


0;- C) m Cm CO
bOLOVO
LNhNQN


com

I
t ,,
r-. '
LON n-


C) n

4I

o c\j


10 L








CO-
a: L-


C







CO CO

I I
n r~
0>J O


LD M


0 --A -;- CO 0

CM -I CI 10 10
10 10D 1- : D ;Z-


10 O 01


N 0 Lr

-d: c


<<<


0)



Sr






(U
4)






o-



(U














4-)
4-)



*r--
0 1





s-



(U
n




I-





c:r

O/
s-
cu
Ci.


rN L CL CM D- CO m to m m -:- 0 '-- C

10 CD 10 D C0 OH 0) 0 1 C C)
D 0D LO D LO LLD ID 1- LD -


(0
u


4-) U U 0-
S E C C 4-' 0



4- 0) 4-) E -0 S-
4- 0) 3 0 c X 3
r- c .r- L O *- .. E


- 0)
*- U
> c



3 4-)
O_ S-
E D
-- 2Z


( U
-- 0
O--
CL 7


LO C
0- 0C

CO CO


4-
0

v,




0






S-
>4-







U

CD



0
4-










0.





a.
S-

ro





0
oU





.0















4-
Ia







0













L-
ro
4-
0
















V(














4- C
LO
o












I
ro


















u- U











( U v
0)
cn






4-) o





()









.rn
0)
s-, -





3r


c
(n



cr







cU








a-
n)
















4- ) *

.- 0



C-,
I- n

n3




-87-


endurance. For the age group of 30 to 39 years of age, the means ranged

from 45.6 for the trait of play to the mean of 61.1 for the trait of

endurance. The mean scores of respondents between the ages of 40 to 49

ranged from 41.6 for the trait of aggression to 64.2 for the trait of

harmavoidance. Finally, for respondents between 50 and 82 years of age,

the means ranged from 37.9 for the trait of aggression to 66.4 for the

trait of harmavoidance. Note that in both of the two younger age groups

the highest means are for the trait of endurance. In the two older age

groups both of the lowest means are for the trait of aggression and both

of the highest means are for the trait of harmavoidance. Table 4.10

also indicates that as age increased the traits of aggression, dominance,

exhibition, and impulsivity had a parallel decrease in means; as age

increased the means of the trait of harmavoidance increased.

The preceding tables were presented to give an overview of the

respondent CPA population with reference to the personality attributes

measured in the study. In the following section statistical data are

provided relating to the characteristics of the firms and offices that

employ these subjects.



Statistics on Respondent Firms and Offices


Response Firms


The first tables concern characteristics of the firms. In Table

4.11 the response firms are classified according to the dollar volume

ranges of their billings. Forty-five percent of the response firms had

billings of less than $600,000 per annum. Over 16 percent had annual




-88-


billings of less than $150,000 while over 23 percent had annual billings

of less than $300,000.


Table 4.11

Firms Based on Dollar Volume


Dollar Volume Number

No response 1
$76,000-150,000 4
151,000-300,000 2
301,000-600,000 7
Over 600,000 17
31


Table 4.12 classifies the response firms according to the number of

partners per firm. Seventy-four percent of the firms had less than

seven partners. Thirty-two percent of the response firms had from five

to six partners. The table shows a relatively even distribution of the

subject firms as to number of partners per firm.



Table 4.12

Firms Based on Number of Partners


Number of Partners Number

2 partners 7
3-4 6
5-6 10
8-10 5
11-15 3
31


Table 4.13 provides a classification by the number of CPAs with the

firm. Fifty-two percent of the firms had less than 11 CPAs in their

firm. More than 25 percent of the response firms had less than five




-89-


CPAs. As the table shows, there is a relatively even distribution of

firms as to the number of CPAs per each of those firms.


Table 4.13

Firms Based on Number of CPAs


Number of CPAs Number

Less than 5 CPAs 8
5-10 8
11-20 12
21-31 3
31


The response firms' distribution based upon

clients per firm is present in Table 4.14. Over

had less than an average 1200 clients per year.

the response firms had an average of between 325

annum.


the average number of

64 percent of the firms

Thirty-two percent of

and 1200 clients per


Table 4.14

Firms Based on Number of Clients


Number of Clients Number

No response 2
Less than 325 8
325-1200 10
1400-2900 8
3000-7000 3
31


When the response firms are broken down by the number of profes-

sional offices in the state of Florida (see Table 4.15), it can be

observed that 80 percent have two or less offices, and that 52 percent

of the response firms have only one office. Table 4.15 shows an even




-90-


distribution between one office firms and multiple office firms of the

sample population.


Table 4.15

Firms Based on Number of Offices in Florida


Number of Offices Number

1 16
2 9
3 2
4 2
5 2
31


In general 84 percent of the responding firms considered themselves

local firms, 7 percent considered themselves statewide, and 9

percent felt that they were regional in scope. Some offices of the same

firm indicated slightly different responses as to overall firm charac-

teristics and, thus, all firm characteristics were tabled by office

response forms (that is, 31 responses for 24 firms). In the section

to follow the sample population will be discussed as to the response

office characteristics.



Response Offices


In Table 4.16 the response offices are classified according to the

number of partners per office. Sixty-eight percent of the offices had

less than five partners and 42 percent of the offices had two partners

or less. The number of partners per office ranged from one to 10.




-91-


Table 4.16

Offices Based on Number of Partners

Number of Partners Number

2 partners or less 13
3-4 8
5-6 4
8-10 6
31


Table 4.17 shows the classification of offices according to number

of CPAs employed per office. Seventy-one percent of the response offices

had less than 11 CPAs engaged at that location. Approximately 39 percent

of the responding offices had less than five CPAs.


Table 4.17

Offices Based on Number of CPAs


Number of CPAs Number

Less than 5 12
5-10 10
11-30 9
31


The response offices' distribution based upon the average number of

clients per office is presented in Table 4.18. Thirty-two percent of

the response offices averaged less than 250 clients per year; a little

more than 61 percent had less than 460 clients per year.




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs