Readings in Global Organization Design

Article #12-05-10-06

longitudinal research into methods of assessing

managerial potential
by Gillian P. Stamp

This report describes research on the prediction of executive potential. An assessment technique,

the career Path Appreciation (cPA), was developed by Gillian Stamp, based on the logic of Jaques’

Stratified Systems theory. lower and middle level managers were assessed and followed up over

periods ranging from 4 to 13 years. Predictive validities ranged from 0.7 to 0.9, strongly suggesting

that (a) development beyond the lower and middle levels of organization is heavily dependent on

conceptual ability, and (b) individual development beyond lower and middle levels should focus

strongly on enhancing conceptual skills, as opposed to knowledge-based instruction. these results

strongly support the logic of Stratified Systems theory.

Note by T. Owen Jacobs

permitted uses

Many GO Society affiliates including the author of this document have generously granted permission to the society to distribute their
creative work for educational, non-commercial use. To ensure our affiliates’ continued willingness to contribute high quality resources, the
society includes educational, non-commercial provisions to protect authors’ rights in its site registration terms-of-use agreement.
Permitted uses of this document include your personal educational use and sharing your interest with colleagues by sending them a short
excerpt together with the GO Society web site URL where they may register for a free account and download the entire document.
Any exemption from this policy requires written permission from the editors. [email protected]
Academics may request permission to use the document for educational purposes within educational institutions. Managers may request
permission to use the document within their own organization.
Any consultant or commercial interest that wishes to email, host on a web site, print or to distribute this document in any form must seek
permission by writing directly to the author who holds copyright and requesting a copy that does not include the GO Society logo, as the
society is restrained by its author agreements from granting such permission. You can find an author’s contact information in his or her profile
on the GO web site.
 UNCLASSIFIED ·
SECURITY CLASSIFICATION OF THIS PAGE
. REPORT DOCUMENTATION PAGE
Fomr Appro'lftl
OMS No. 0704-0188
1a. REPORT SECURITY CLASSIFICATION 1b. RESTRICTIVE MARKINGS
Unclassified
2a. SECURITY CLASSIFICATION AUTHORITY 3 • DISTRIBUTION I AVAJLABIUTY OF REPORT
- Approved for public release;
2b. DECLASSIFICATION I DOWNGRADING SCHeDULE
-
4. PERFORMJNG ORGANIZATION REPORT NUMBER($)
distribution unlimited.
5. MONITORING ORGANIZATION REPORT NUMBER(S)

- ARI Technical Report 819

&.. NAME OF PERFORMING ORGANIZATION 6b. OFFICE SYMBOl. 7a. NAME OF MONITORING ORGANIZATION
(If •pplbiW)
u.s.Army Research Institute for the
Brunel University - Behavioral and Social Sciences
&c. ADDRESS (City. St•N. •nd ZIP COIHJ 7b. ADDRESS (City. St•te. •ttd ZIP Cod•J
Uxbridge, Middlesex 5001 Eisenhower Avenue
UBS 3PH, England Alexandria, VA 22333-5600

Sa. NAME OF FUNDING /SPONSORING 8b. OFFICE SYMBOL 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER
ORGANIZAnQN (If •ppllabl•)
-
8c. ADDRESS (Chy. Sue.. Mtl ZIP Cod•}
- - DAJ'45-86-C-0009
10. SOURCE OF FUNDING NUMBERS
PROGRAM PROJECT TASK WORK UNIT
- ELEMENT NO.
63731
NO.~

A792
NO.
4.5.5
ACCESSION NO.
C.l
11. nru (lndW. S«Urity O.Ufic•fionJ
Longitudinal Research Into Methods of Assessing Managerial Potential
12. PERSONAL AUTHOR(S)
Stamp, Gillian P.
13a. TYPE OF REPORT 113b. TIME COVERED . • 114. OATE OF REPORT (Ye•'• Month. O.y} , S. PAGE COUNT
Final FROM Q4£86 TO Q4£8B 1988, October 48
16. SUPPUMENTARY NOTATTON

- 6

17. COSATI CODES 18. SUBJECT TERMS (Continue on reverse if necesury. •nd identify by block number)
. FJELD GROUP SUB-GROUP Executive leadership
Executive development
Assessment co2nitive skills
19. ABSTRACT (Continu• on rev•n• if neceJUty •nd identify. by block number)

This report describes research on the prediction of executive potential.· An assess-

ment technique, the Career Path Appreciation (CPA), was developed based on the logic of
Jaques' Stratified Systems Theory. Lower and middle level managers were assessed and
followed up over periods ranging from 4 to 13 years. Predictive validities ranged from
0.7 to 0.9, strongly suggesting that (a) development beyond the lower and middle levels
of organization is heavily dependent on conceptual ability, and (b) individual develop- 1
~ent beyond lower and middle levels should focus strongly on enhancing conceptual skills,
as opposed to knowledge-based instruction. These results strongly support the logic of
Stratified Systems Theory. l
20. OISTRIBUTION/AVAILASIUTY OF ABSTRACi 21. ABSTRAcr SECURITY CLASSIFICATION
LtJ UNCLASSIFIEDIUNUMITED CJ SAME AS RPT. CJ OTIC USERS Unclassified
22a. N'AME OF RESPONSIBLE INOIVIOUAL 22b. TEl.SPHONE gnc/ude Area Code) f 22c. OFFICE SYMBOL
'T'
I.. o. Jacobs (202) 27 -9045 I ?ERI-RO
DO Form 1473. JUN 86 SECURITY CLASSIFICAnON OF THIS PAGE
UNCLASS·IFIED
i
Technical Report 819

Longitudinal Research Into Methods

of Assessing Managerial Potential

Gillian P. Stamp
Brunei Institute of Organisation and Social Studies

Executive Development Research Group

Thomas 0. Jacobs. Chief

Manpower and Personnel Research Laboratory

Newell K. Eaton, Director

u.s. Army Research Institute for the Behavioral and Social Sciences
5001 Eisenhower Avenue, Alexandria, Virginia 22333-5600

Office, Deputy Chief of Staff for Personnel

Department o1 the Army

October 1988

Army Project Number :·Aanpower and ?srscnne1

20263731A792

Approved for public release: distribution unlimited.

iii
FOREVORD

The U.S. Army Research Institute for the Behavioral and Social Sciences
(ARI) is actively pursuing a research program to identify executive-level
leadership and decision skills and to develop methodologies for assessing and
developing·these skills. This program is largely based on the Stratified
Systems Theory of Blliott Jaques.
The present report contains findings based on long-term follow-up of a
variety of lover and middle level managers who vera"" assessed using a teclmique
that purportedly measures conceptual ability. The managers were drawn from a
number of different companies and settings, and the period of follow-up ranged
from as little as 6 years to as many as 14. The remarkably high predictive
validities obtained in this research are convincingly supportive of SST and
further support the contention that the potential for development beyond the
lover and middle levels of organization rests largely on conceptual ability.
This research effort vas moni.tored by Alli's Executive Development Re-
search Group under Prosram Task 4.5.5, Bxecutive Doctrine Development, under
authority of a Letter of Instruction from DCSPBR, DA, dated 17 June 1985.
Dr. Herbert Barber, u.s. Army Var College Department of Command, Leadership,
and Management, has been briefed on this work and on possible applications,
which include developmental assessment for Var College students to aid their
subsequent growth.

+II~
BDGAll H. JOHNSON
Technic-al Director
LONGITUDINAL RESEARCH INTO METHODS OP ASSESSING MANAGERIAL POTENTIAL

EXECUTIVE SUMMARY

Requirement:
To develop a procedure to predict rate of growth of individual eapability
to haDdle increasins responsibility (managerial potential) and to test a
theory of organization structure and ~dult development.

Procedure:
Predictions of potential_ were made for 1~2 men. and women of Anglo-Saxon
and African origins vorkiq at all levels in tvo IIIUltinational and two na-
tional organizations. Ro information about predictions vas given to the indi-
viduals or to their managers. Follow-up studies were conducted over periods
ranging from 4 to 13 years. The studies iathered data from company records
about the level of responsibility at which the individual was employed and
correlated these data with the level that would have been predicted. Predic-
tions were also made for 92 people employed by an organization in a developing
country. Sixty-three percent of this group are Afriuns and 51% have not been
educated beyond primary school level. Por this group, evaluations of poten-
tial were used to predict· individual ~~omfort curves,• the rate of growth of
responsibili~ that ·is in line with the predicted rate of growth of
capabili~.

Findings:
For the whole sample, the correlation coefficient between the predicted
rate of growth of capability and level of responsibility vas .79.
By 1979, the procedure for predicting potential had been refined into a
guided interview called a Career Path Appreciation (CPA). The correlation
coefficients for follow-up of CPA are .89. This figure may be compared with
tvo other longitudinal studies of potential: one in the USA and one in the
UK, where the correlations were .44 and .66, respectively.
Follow-up on the individuals for vhom "comfort curves" were predicted has
shown that, over periods ranging from 2 to 5 years, the shape of the cur1e
predicted has been borne out in 94% of the cases.
CPA is an accurate predictor of rate of growth of capability to handle
increasing =esponsibility. !t ~equires hi~hly trained and experienesd admin-
is t=a tors Out, ~inc~ :he :n te~lieT.I :as -:s :or .Jnl.y : or 3 :1ours. :. : :s cost
affective.

vii
 CPA offers the individual the opportunity to review his or her working
life and the organization the opportunity to review the effectiveness of its
human resources and the appropriate rates for their development.
The accuracy of the predictions made by CPA provide further confirmation
of the Stratified Systems Theory of organization structure and adult
development.

Utilization of PindiQSsa
!he remarkably high predictive validities found in this research strongly
suggest that (a) development beyond the lover and middle levels of organiza-
tion is heavily dependent on conceptual ability, and (b) individual develop-·
ment beyond lover and lliddle levels should focus stroDgly on enhancing concep-
tual skills, as opposed to knowledge-based instruction. These conclusions
bave·strong implications for pedagogical approaches in the Senior Service
Colleges, and for leader development in seneral. They are supportive of cur-
rent directions in ·the Army Var COllege.· The developmental assessment will be
t~ied at the Army Var College on an experimental basis with a small group of
students duriq the next academic year.

viii
LONGITUDINAL RES~CB INTO METHODS OF ASSESSING MANAGERIAL POTENTIAL

CONTENTS

Page
INTRODUcriOH • • • • • • • • •. • • • • • ........• • • • • 1

SBCTIOH 1: PRBDICTING POTENTIAL • • • • • • • • • • • • .... 1

Meta-analyses • • • • • • • • • • • • • • • • • • • • •
Comment on Procedures for Predicting Potential • • • •
..... 1
3
The Relationship Between Procedures for Predicting
Potential and Criteria • • • • • • • • • • • • • • • • . • • • • • • 3
The Contribution of Theory ••• _• • • • • • • • • • • • • • • 4

SECTION 2: STRATIPDD SYSTEMS THEORY • • • .....• • • • ... • • • 7

Definition of Vork • • • • • • • • • • • . . . . . . . . . . . . 7
Levels of Vork • • • • • • • • • • • • •
The Array of Growth Curves • • •. • • •
........
• • • • 7
9
.
SECTION 3: TBB PROCEDURE FOR PREDICTING POTENTIAL AND THE CRITERION • • 11

Development of the Procedure • • • • • • • • • • • • • •

The Criterion • • • • • • • ·• • • • · • • • • • • • • • •
.. .. .. .. .. . . 11
15
SECTION 4: THE POUR SE'rriNGS IN VBICB DATA VERB COLLECTED .... 17
I

The Multinational Oil Company (84 respondents followed up) • 17

The Multinational Engineering Company (35 respondents followed up) 18
The Fertilizer· Company (38 respondents followed up) • • • • • • 19
The Mine in a Developing Country (25 respondents followed up) • • • • 20
SECTION 5: RESULTS AND INTERPRETATION ....... 23
The Sample • • • • • • • • • • • • • .. 23
The Data • • • • • • • • • • • • • • 23
Comparison Yith Other Long-Term Studies • • • • 30
Interpretation of the Data • • • • • • 31
REFERENCES 39

ix
CONTENTS.CContinuedl

Page
LIST OF TABLES

Table 1. Summary of five studies of validity coefficients

of procedures for predicting potential • • • • • • • • • • • • 2
2. Levels of work in civilian and militar,y organizations • • • • • 8
3. The eight levels of capability • • • • • • • • • • • • • • • • 14
4. Pollow-up on 84 respondents (Multinational oil company) • • • • 25.
5. Pollov-up on 182 respondents (4-13 year f.ollov-up) • • • • • • 26
6.
-
Follow-up on 76 respondents (4-8 year follow-up) ....... 27
·· 7. Follow-up on 59 respondents • • • • • • • • • • • • • • • • • • 28
8. Some assessment exercises predicting management potential
over 20 years • • • • • • • • • • • • • • • • • • • • • • ... 29

LIST OF FIGURES

Figure 1. Levels of work and the array of growth curves ........ 9

..... ___ -· ... -- -- ....

--------~--...... ·--------·-·-··---~
 LONGITUDINAL RESEARCH INTO METHODS· OF ASSESSING MANAGERIAL POTENTIAL

Intrgdyctign

The primary purpose of this paper is to present long-term follow-up of

predictions of individual differences in the potential to carry
responsibility for work at different levels of complexity within
organisations. The significance of predicting potential -- especially
in a rapidly changing world-- can. be simply stated: People represent a
significant short- and long-term investment for the organisation of
which they are a part. Accurate evaluation of the rate of appreciation
of that investment ensures effectiveness of the organisation, consider-
able financial savings and, not least, dignity for the individuals con-
cerned. Par example, if an organisation is likely to be called upon to
meet new challenges, an accurate picture of the current state of the
potential of its human resources a.nd of the rate at which they are
likely to grow is essential if the challenges are to be met.

Seetipn 1: Predietinl Pgtentiol

o. leto-onolxses

1. In the last twenty years there has been a very substantial number of
studies of potential. One of the best known is Bray's work with AT&T
<Bray and Grant, 1966, Bray, Campbell and Grant, 1974>. Two recent
articles provide meta-analyses of the results of many investigations of
these studies <SchDdtt, Gooding, Ice and Kirsch, 1984; Hunter and
Bunter, 1984). The implications of these studies for practice in the
U.K. are examined by Herriot <1987>.
In broad terms the meta-analyses are concerned with the relationship
between procedures for predicting potential and criteria for evaluating
success. The procedures which are most widely covered in the analysed
literature are tests of cognitive ability, work samples, <of which job
knowledge may be a sub-set>, assessment centres, biographical data, per-
sonality tests, interviews and references. The most commonly used
criteria are training success and job proficiency -- the latter eval-
uated by ratings or rate of promotion and measured by product1 vi ty or
salary.
A quick overview af ~!le .,al!di ":·r coeff!c:en:ts of i i ..,.e =i :l:l.e -:tudies
of procedures for predic~ing po~ential covered in :hese papers is su~
marized in Table l.

1
Table 1
Summary of Five Studies of Validity Coefficients of Procedures for
Predicting Potential

Herriot Dunnette Reilly & Hunter & Schmitt

(1987) (1972) Chao (1982) Hunter (1984) (1984)
Predictors

Cognitive .ssa
ability .27 .4sb - .53 .248

Work
sample .38 - - .44 .378

Job
knowledge - .51 - .78c -
Biographical
data .24 .34 .38 .37 .243
.
Assessment .63d
centres .41 -: - .33e .407

Personality
tests .15 - - - .149

Interviews .14 - .23 .14 -

References - - .17 .26 -
a. Training success.
b. Job proficiency.
c. ·Trained subjects.
d. With promotion.
e. With job proficiency.

2
b. Cgmmgnt gn Prgcedures fpr Predicting Pptential

It wiil be clear from Table 1 that assessment centres offer predic-

tions with the highest validity; but Hunter and Hunter <referring to
Cohen, et al. , 1974> point out that the correlation of . 63 is w1 th
promotion. There is a much lower correlation <.33> with actual perfor-
mnce. The findings en tests of cognitive ability as a predictor are
also not clear cut; SchDdtt <1984> suggests that the lower validity in
his studies may~ accounted for by the fact that Bunter and_Bunter had
access to extensive unpublished materials. It is also of interest that
tests of ccgni ti ve ability offer a correlation of • 55 w1 th training
success yet only .45 with job proficiency.

The relatively high validity of biodata as a procedure prompts Hunter

and Hunter to comment that its accurate use requires a l~rge sample.
They refer to one consortium where a Supervisory Profile Record <SPR>
has been developed from a data base with input · from thirty-nine
organisations. The validity of the s. P.R. for predicting supervisory
ratings is . 40, w1 th no variation over time and little across
organisations <Hunter and SchDddt, 1975, pp 1067-1068>.

As far as job knowledge is concerned, Hunter and Hunter <1984> point

out that, despite their high validity, the use of job knowledge tests is
limited by the fact that they can only be used for prediction· if the
employees <or subjects> have already been trained fer the job.

SchDdtt <1984> and Asher and·Sciarrino <1974> draw attention to the

relationship between the procedures for predicting potential and the
criteria used for evaluating success. Schmitt <1984> .refers to the
general concern that criteria should be as objective as possible; but
points out that, in the studies that he and his colleagues analyzed,
criteria tended to fall into two broad categories -- trai.ning success
and job proficiency. As outlined above, the latter may be rated by
immediate supervisors or measured by output, sales, salary, or rate of
promotion.

c The Relatignship between Procedures for Predicting Potential and

Criteria

In considering the relationship between procedures and criteria,

Schmitt <1984> follows ':iernimcn-c and Campbell's <1968> suggestion that
ie~relopmen-c of procedures ::.n1:ended to be actual job samples should
result in increased validity coefficients. Behavioural consistency
t~eory sugges-cs ~hat the closer ane can come to getting some ~asure =r
indication of performance on samples of work ac~i ·;i "ties 'tha"t peapie
:::1~h"t ·~e -=~pec-::ed ~:J ?eri~r::n :n 3. jab. ~!le ~etter ?redic~:!.~ns ~:le ~.,au~-=­
~e .!bla ::J :nake o! ~.b.e ac:ua.i ?erfor:::2nce :li i:.di ·1:d.'Jals ::n ~.:.e
ac~ivlties i~ ques-cion.

More technically, if predictor and criterion measures fall within the

sa:me content domain, the val1di ty of correlations should be !IlaXimized .
.3chmi~-: -:t ~l. <:9841 :.::npl.:.es ~::.at ':~e 1ues~:.on ct ::::!!:ta!:.~ ic~:.=.

3
underlies the validity of work samples or job try-out as good procedures
for predicting job proficiency <see Schmitt et al. , 1984, p 417 for a
summary table of average validity coefficients for predictor-~riterion
combinations; see also Bray, et al., 1966>.
The importance of procedures for predicting potential and criteria of
success being within the same content domain is underlined by Anstey in
his paper an the thirty year fallow-up of the Civil Service Selection
Board <CSSB> procedure in the United Itingdam <Anstey, 1977>. In
describing the procedure, Anstey makes it clear that the original CSSB
techniques were based an careful analysis and modelling of the work of
senior administrators. There is a suggestion that that is one of the
reasons why the Civil Service obtained probably the highest ever
validity coefficients for high grade selection in any country.
One of the overall conclusions that can be drawn from these summaries
of studies of predicting potential is that the maSt effective procedures
for predicting job proficiency at some future date are those which sa~
ple the work which is actually going to be undertaken either in the
immediate or the longer-term future. This conclusion paints to the
importance of the content validity of the tests or situations designed
to elicit the behaviours used as indications of potential.
It is, however, not possible to devise effective work-sampling proce-
dures until there has been a thorough analysis of the work to be done.
Bunter and Bunter <1984> suggest that the validity of work sampling as a
procedure could be enhanced .if it were integrated with ot~er procedures
with a reasonably high validity <tests of cognitive ability and biadata>
<see Hunter and Hunter, 1984 re. alternative predictors and Anstey,
1977 re. integrating rather than adding>.

d. The Cgntribut1gn gf Thegry

Behind content validity, there is the question of construct validity -

- the theoretical underpinning of the procedures for predicting poten-
tial, the criteria for evaluating success and the relationship between
them <see, for example, Stewart and Stewart, 1977 and Guion. 1987). At
Minimum, construct validity would <a> enhance content validity by
providing a definition of work which, by explicating the natur.e of the
predictor, would make it possible to design more effective procedures
for prediction; <b> bring predictors and criteria into a common content
domain and <c> provide an explicit statement about adult development and
individual differences.
The·meta-analyses tend to suggest that, apart from the concept of cog-
nitive ability, the predictors underlying the procedures are largely
implicit and not clearly defined. It is the procedures for predicting
poten~ial ~a~her than ~he ?redictors whic~ are af:ered ~s ~ypo~~eses :~r
tes~:ing.

A common assumption in assessment and prediction is that adults do

develop; and, further, that ~hey develop at different rates. In some
assessment settings these assumptions are made manifest in the form of

4
a Minimum Development Curve based on organizational structure and time
constraints. For example, in the United Kingdom Civil Service, a
graduate administrative entrant with a mark of 3 is predicted to be able
to work as a Principal within five· years of entering the service. This
prediction in time is relative to entrance and does not refer to the age
of the candidate.
In the absence of an explicit hypothesis about adult development and
differences in rate of growth. assessment procedures are designed on the
basis of statistical models. Tests and techniques are retained if the
correlations are good enough and there is little impetus for review.
The requirements for construct validity are met by Stratified Systems
Theory which provides the base for the development of the procedure for
predicting potential described in this paper.

5
Section 2: Stratified Systems Theory

a. Pefinitign of York

Stratified Systems Theory <SST> <see Jaques, 1975, 1982 and Evans,
1979> provides a definition of work -- "The effort to accomplish a goal
requiring the exercise of discretion within prescribed liDdts and within
a stated completion-time• • The theory postulates that the core of the
psychological experience of doing .work is •the exercise of discretion•.
The prescribed liDdts are the rules of the organisation in the form of
policies, procedures, physical controls, signals and other types of
control which are objectively set. In ·short, these controls may be
taken as an indication of the things that must be done. They state the
boundaries of the work and define the scope of the discretionary
environment.
By contrast with th~ things that must be dane, the exercise of
discretion is concerned with the choices that· must be made. The word
•discretion• comes from the past participle of the verb •to discern•,
i.e. , to distinguish, to separate apart. The word •discretion" was
chosen to convey the psychological process of mulling over a number of
courses of possible action which could be used to reach a goal,
reaching into the self to choose those <or the one) which are/is most
likely to reach the goal; then acting an that choice. In other words,
the exercise of discretion may be thought of as the imagination, formu-
lation and execution of a course of action which is nat prescribed. One
of the characteristics of discretion is that, to the extent that the
person is capable of making the choices that must be made, s/he will not
perceive that discretion 'is being exercised.
·b. Leyels of Wort
In addition to the core definition, the theory provides a model of
how work is structured in levels of increasing complex! ty within the
organization in order to take account of the complexity of the environ-
ment <see Table 2>. It follows that the scope of discretion must
increase at higher levels because the increasing complexity prohibits
the formulation of precedent, procedures and rules.
This complex! ty can be measured by considering the shortest ~i:te
period that can elapse in order to obtain feedback about choices made.
This is the minimum time that must elapse before 1t is clear that the
-=~cices :2de about -:curses of ~ct1on have Il01: oeen adequa'te to :he r:or:r
plexity of the environmen't. The consequences of grossly inappropria'te
choices '~'~i 11 become apparent fairly quick! y, but 1 t may be months or
9ven years betcre the consequences of marginally :nappropriate choices
·Jeco:::e -::ani!est. :ro:m ~his ?erspecti ~e !. t !.s ?OSSi ble to rephrase the
:e::.~1-::::=. ::::: ...,or~ 31·.ren 3.bave 3-nd -:o :-:dei:.::.e '40!"~ :!S ''::!:or~ -:::
:-'=alise 'Jur~ose :.n orac-:ice and ~a :-eview ?rac::ce :...:. :ie _.:..g.:l: --
purpose".

..,
I
Table 2
Levels of Work in Civilian and Military Organizations

Level of Civilian Military

Time-span Description Organization Organization
Work

SO yrs.
VII Strategic design, Corporation MAC OM
development, Army Board

....uco deployment of
complex systems.
CD
20 yrs. ~

...cd
~
VI. en Direct deploy- Group Corps
ment of complex
systems.
10 yrs.
v Complex systems, Subsidiary Division
encompassing
operating sys-
CD
tems and modify-
....>en ing context
5 yrs. c:CD
i...
IV Alternative General Brigade
t
(J
operating sys-
tems ~- general
Management
management
2 yrs.
III Direct operat- Unit. Battalion
ing systemS --
management of
a mutual recog-
1 yr.
.... nition unit
=
s::
.....0
II -4-l Direct operat- Section Company
~
Jot
CD
ing methods --
c..
0
supervision of
a mutual know-
1 ledge system

3 mths
I
I Direct
I

I operat- Shop Floor Squad

:.... .:.ng tasks

3
 Because the ~asurement of level of work refers tc the time scale of
review of the exercise of marginally substandard discretion, the
measurement tool· is called "the time-span of discretion".

c. The Arra1 gf Grgwth Curyes

The theory also includes a description of how work is experienced by

the person engaged in it -- of how the exercise of discretion actually
feels -- and an explicit hypothesis about rates of growth of the
individual capacity to exercise discretion; and, therefore, to do work
at levels of increasing complexity.
This hypothesis is presented in the form of an array of growth curves
which follow the sigmoidal progression characteristic of biological
growth. The array was derived originally from studies of individual
earning progressions of five hundred and twenty people and sqbsequently
.tested w1 th a further three thousand cases <see Jaques. 1968 >• J aquas
<op. cit.> then constructed the hypothesis that the regularity he had
charted in individual earning progressions reflected ·growth in the
person • s capacity to cope w1 th work of increasing complexity. The
array of curves thus represents statistical facts about pay and an
hypothesis about regularity in rate of growth over time of the capacity
to exercise discretion. ·
The relationship between the levels of work of increasing complexity
and the array of growth curves. is illustrate~ in Figure 1 where the
growth curves are grouped into modes, each mode including all cur.ves
which reach a maximum point of development in the sa~ level of work.
50 Years

20Years

10 Years

w
~ 5 Years
<
a:
~

·~
:i: 2 Years
i=

:..GE (In Yearsl

Figure 1
:..evels 'Ji ·.·lark :1nd ~~e Array :n: Grow'th C:.1rves

9
 . '

By providing a definition of work, SST <a> enhances content validity

and.allows effective work-sampling; and <b> brings predictors and cri-
teria into a common content domain. By offering an explicit hypothesis
about· rates of growth of the capacity in adults to exercise discretion,
Cand, therefore, to do work> the theory establishes a relationship over
time between specific predictors and particular criteria. For example,
the theory makes it possible to predict how lang it will be before a
particular person is likely to be capable of exercising discretion and
doing work at a particular level. Piully, the theory offers the
hypothesis that growth in •the capacity to exercise discretion• is
expressed in • the capac!ty to cape w1 th camplexi ty- • ·

10
Sectign 3; The Prgcedyre fgr Predicting Pgtential and the Criterign

The studies of .prediction of potential were undertaken in four very

different settings which will be described in detail in Section 4.
Briefly they are a multinational oil company, a multinational
engineering company, a national fertilizer company and a British-owned
multinational Bdning company operating in a developing country. ·
0 I Deyel opmep,t gf tbe Proeedure

In each setting the procedure used was an evaluation of the current

scope of the person's capacity to exercise discretion or cope with com-
plexity -- in other words, his or her •current level of capability, •
defined as •the person's capability at a particular age to generate and
to respond to complexity within the warld, within hiDVherself and in the
constant work needed to keep both in equilibriu~.
In th~ early stages of the research, the evaluation of current levei
of capability was made by means of an extended interview which was based
in the definitions and models of Stratified Systems Theory. By 1979 the
interview had been developed into a procedure called •career Path
Appreciation•.
Starting from the idea that, from a psychological perspective, work is
discretion realized in action, we decided to design a procedure which
would allow us to observe the person. at work. The definition of work
provided the basis for the content validity of the·procedure which may
be thought of as work sampling -- not in the concrete sense, but in the
sense of observing how discretion is being exercised -- or, in other
words, how the person is coping with complexity. The handling of
complexity is a generic component of. work at all levels. The predictor
is, therefore, tapping a universal ·component which is hypothesiz~d to be
highly related to advancement.
Because we wanted to set up a microcosm of the psychological exper-
ience of doing work, we decided not to use general tests of ability.
These are good predictors of crystallized intelligence and achievement
in education and training settings. But we felt that they were less
.likely to predict the fluid intelligence. which we assumed to be closer
to the idea of realizin.~ discretion in action. ie also thought that
general tests of ability were more likely to reflect "static" individual
differences rather than the growth in individual capability which we
~ought to ~redict.

In parallel with Jaques• early work on the exercise ot discre'tion.

Isaac and O'Connor <1969, 1973) had been engaged in experi~n'tal.
s'tatistical studies o:f judgment in action. Building on their ·,.fork
::saac 3.nd O'Connor. op. ,::!.-:.; ·,o~e ,iec:.ded. ":o s~ar~ ':y 'Jsi~g 3. ::1sk ·>~hie::.
ca~ cr:~::J.all:r ::-on :~e :oncep't :or::a'ticn :::.-:er~'t~r= see 3r·..:::.er.
1966). Bec~use •>~e ·,o~ere no't :ry1ng ~o measure :~e .3.bili -::-r -:o ~aar-:1. ·Ju':
to observe judgment or discretion 1n action, we decided ~a give only
minimal instructions to the respondent and not to offer rehearsal.

11
 The task required the respondent to discover and implement a predeter-
mined sorting rule. Cards carrying a symbol or combination of symbols
which vary in colour, number, size and shape, were provided. In the
pack overall there are three colours, three numbers, three size_s,. and
three shapes. At the beginning of the task four display cards were set
out. Three of these illustrated various combinations of the four
factors, and the fourth was blank. The respondent was given a pack of
162 cards, each carrying a different combination of the four factors.
The respondent was asked to place each card below the display cards in
such a way that they corresponded with a pre-set order. The goal was to
place ten consecutive cards correctly, but there were no restrictions on
time or.numher of cards. To enable solution, the adDdnistrator gave the
following information: In the case of the three cards displaying
symbols, the respondent was told whether or nat the placement was
correct. In the case of the_blank card, no feedback was given.
At the end of this. task the ·administrator initiated a .very brief
discussion with the respondents about their current work, the history of
their careers and their aspirations for the future.
With this procedure it appeared to be possible to elicit enough infor-
mation about the respondents• characteristic patterns of realizing
discretion in action to evaluate each person's level of capability --
thus, the level of work at whi~h s/he would currently be effectlve. In
the very early stages of the work this evaluation was checked against
the level of work at which the respQndent was actually employed.
Although no formal correlations wer~ made, the degree of· correspondence
was deemed satisfactory.
If the evaluation of current level of capability at a particular age
is set against the array of growth curves <see Figure 1>, it includes a
prediction about the likely current level of capability at any given
time in the future. At this stage of the development of Career· Path
Appreciation, the predictor <the evaluation of current level of capa-
bility> was made on the basis of information elicited in this way.
In the course o:f analyzing the i-nterview protocols and, in particular,
comments elicited by respondents about the way they had constructed and
tested strategies, it became clear that a ·further important source of
infor:mation about characteristic patterns of. realizing discretion in
action lay in" these comments.
Qut of mere than a hundred phrases taken from protocols of the symbol
card task, 54 were chosen for a fi~l pack of phrase cards. They were
chosen because, in the course of testing the hundred, these emerged as
~he phrases which were most open in the sense that they could ~e given a
wide range of interpretations according to the unders"tanding of the
=-~s~onden-c.1s :. .:onsequence. these ·o~ere :~e phrases '-'fhlcb. ~onsis-centl?
the :ulles-c ·1iscussions .!bcu-c :b.e wa'r t1la~ .:acl:l .:-es'Concien't
-=l:!..ci.ted
approached his ar her work. !he 54 were arranged in :1i:::1e sets ·of six
and added to the extended interview by giving them to the responden"t
before the symbol card task.

12
 The enhanced understanding of the respondent • s patterns of realizing
discretion afforded by these phrase cards, and the evident reflective
enjoyment and insight they gave, warranted their inclusion in the
procedure. · There is no attempt to score respondents• comments about the
cards; comments are analysed for content in the light of the model of
levels of capability <see Table 3>.
During the same period it became clear that more information would be
available to the adll:l.nistrator <and to the respondent> if the career
interview was mare structure4. We therefore decided to ask the respon-
dent to talk about times when a/he felt that there bad been a good
correspondence between their own capacity to exercise discretion and the
work for which they were responsible, times when they felt they had been
overstretched and times when they felt they had been underused.
Once the phrase cards and the more detailed work history had been
tested and added to the extended interview, it became clear that what
had been developed was a procedure which makes it possible for the
respondents to put their charac.teristic strategies and feelings about
work into words for the first time~ An almost universal response to an
Appreciation is the coD~JDent •I have never thought about my work like
this before•. At this point the procedure was named career Path
Appreciation.
At the beginning of the Appreciation ·the respondent is offered. each
set of phrase cards separately and asked to choose the c~ which s/he
feels reflects most closely the way s/he would approach a piece of work.
Respondents frequently comment on mere than one card, or may choose to
put the cards in some order or.to put aside those which they feel do not
relate in any way to their own approach to work. The Appreciation then
:moves on to the symbol card task, the work history and a discussion
about the respondent's aspirations for the future of his or her working
life.
The symbol card task and the phrase cards are both open to the strat-
egies -and interpretations the respondent wishes to place on them. For
example, respondents will frequently say • I take it that this card means
••.• n. It is not usually necessary for the administrator to reply; but,
if it is, an indication will be given that the respondent is free to
give his/her own meaning to the phrases. Further, the 'tQne. of the
Appreciation is. such that the respondent is encouraged to bring to bear
rational, analytic competence and the mulling and separating apart which
is the core of the exercise of discretion.
Career ?at~ A~preciation is ~ ?rocedure that allows a ?ar~icular ~ind
of work-sampling. The nature of the procedure is such that it creates a
shared setting in whic~ the respondent can play out the characteristic
ways in wb.ich s/he works. In this setting internal resources on which
!"esponaen:ts have ~een. ·.irawing ~c ·io ~b.eir :-'IOrk "::>ecome ?ar-: ~f :.!le 5har-~d.
-:x-oer:en.ce == ::ach :-es-oonden-e 3-nd aamnis-c:-3:tor. :.:1 ·Jec:=ru:!.~ 3"'0:l-
ta~eously aware of thesi resources. :he :-esponden.ts increase ~heir sk:::
for tapping them in the future. The administrator can also draw . the
attention of the respondent to these internal resources so that they may
be further explicated.

13
 Table 3
The Eight l.e~els of Capability
Vl II Trau!jfo1·ming: the i·nternational a~d/or national context and create alternative social
i mil i tu t ions .

Vll Extrapolation: from contexts at stratum VI and create connections which can sustain
tho formation and development of stratum V institutions initiated at stratum VIII.

VI Defining: generate a range of perceptions of complex stratum V systems and shape the
social, political and economic contexts in which they operate. Construct the future
rather than forecast it.

V Shaping: make relationships between previously unrelated material; create general rules and
rcJ~fine fi~lds of knowledge and experience. Engage with an open context and decide when
it !jJlOuld be closed; operate a complex .five stratum system, modify its boundaries and cope
'~itll second and third order consequences which arise.' Elements explicitly seen as inter-
Jt.;peudent; to .;hangc one part is to change the whole •
......
~ lV TJ·,msforming: retain contact with what currently exists and detach to conceptualise some-
tid ng coulpletoly different - not a modification but. a point of departure. Contrast and
Cl)Jupare alternative operating systems and alternative modes of deplC?ying or modifying
them. Maintain a patterned structure within which hypotheses are tested. ·

111 Extrapolating: extrapolate from given rules and handle ambiguity by creating new connections
'" i thin a defiued sys tern. Mould operating tasks and operating methods into a system of
J i rcct \-Jork aud fine tune that system to cope with changing trends.

1I Uefining: generate different perceptions of a given situation; organise perceptions in

alternative ways; handle ambiguity by polarising. Put together a programme of direct
operating tasks in order to accumulate knowledge about their aggregation and to change
prugrammes in the light of the given situation.
------·-·--··- -------------------------------------------------------------
Slwping: sec the world through a few focussed dimensions and engage directly with physical
ohje~ts or serve people one task at a time.
 In summary, Career Path Appreciation is a procedure which allows <a>
interpretation of the present relationship between the respondent's cur-
rent capacity to exercise discretion and ~he work for which s/he is
responsible <b> consideration of the history of that relationship <c>
consideration of the likely future of that relationship; and, for the
respondent <d> awareness of the internal resources s/he has been drawing
on to do work.
An Appreciation is concerned with the working life as the relationship
between the person and the organisation <s> in which s/he is employed.
Lofland and Lofland's <1982) description of the 1ntepsiye interview as
•a guided conversation which seeks to elicit materials of substantial
depth for use in qualitative analysis• is a useful way of thinking about
an Appreciation.
b. The eriterigp

In each case the criterion used was the level of work at which the
subject was employed at the time of follow-up. This criterion requires
some comment: The first point is that it is related not to training or
to predicted rate of promotion. but to actual performance for which the
employing organization was prepared to pay a salary.
Secondly, ideal conditions would have allowed.measurement of level of
work by means of the time-span of discretion. Unfortunately, ideal
conditions did not prevail, and such accurate measurement was not
possible. But. in the oil and the fertilizer companies, it was possible·
to use time-span a~d accurate descriptions of levels of work to match
the system for describing levels of work in use in those companies <Hay
XSL> with the model as defined by SST. In the engineering company and
the Ddne, work was structured in terms of the model.
The levels of work as measureable in time-span and as described in SST
are illustrated in Table 2.

15
 Sectign 4; The Fgur Settings in which Data were Collected

As indicated above, the studies were undertaken in four very different

settings. Each company is described in detail in order to emphasize
the very. considerable differences in the settings. The differences are
in scale of the organization, type of activity, economic and social
climate in which it is operating, and the etlmic backgrounds of
employees. Bach of the four companies conducted its awn study of the
predictive power of Career Path Appreciation, and these are described in
the section an the company.
a. The lultiJatignol Oil Cgmpon7 <84 respgndente fgllgwed up>

The original work was done in 1974 at the request of the personnel
department which felt that there was considerable common ground between
the approach made possible by SST and their awn procedures for
identifying,potential.
In the ten years that elapsed since the original evaluations of
current level of capability o£ the respondents and the follow-up in
terms of the level of work at which they were employed, ~his company had
to adopt a mere entrepreneural style and to manage the consequences of
nationalization of a number of operating companies in developing
· countries.
1. The Stud7 llp,<iertoten by the Oi 1 Cgmpon7

The original study in 1974 included 100 respondents, and it was agreed
that there would be no follow-up of specific individuals. In 1984 the
company decided that it would undertake a follow-up study. It was able
to trace 84. people still ·in its employ. Eighty-two of these were· men
· and two women. All were of Anglo-Saxon origin. , Since the study was
done purely for· research purposes, access to the original judgments was
restricted to one person and to his successor in the personnel office.
This part of the sample represents the earliest stage of the develop-
ment of· the predictor -- that is to say, the extended interview,
including only the symbol card task and a brief career history used to
elicit the material on which the evaluation of current level of
capability was ~de ..
When the company did the follow-up study, they did not appreciate that
the .:valuations ::Jade .,.,ere axplici tly of current level of capability,
~ut ~hat they carried an implicit prediction of potential. Accordingly.
their correlation on the full sample was between the explicit predictor,
that is ~o say, the evaluation of the current level of capability a~ the
initial interview and the level of wcrk achieved nine years la~er. Four
?U"i: ~ ~ ~T"-a ·,o~er~ :-e~ved bec~use ":o'th :.!le personnel d.irector and senicr
::ana~ers ~'..ldged :!:.ese pecple to be :""U'- q-r t:,e orrii""':3T"'· The correla-cicn
be-cween predic--::ed and achieved scores ·-m.s . '7!. The company co:mnen~ed
that the predictive validity of psychological testing is usually
reported as much lower than this, below say .45 <see Table 1>.

17
 A correlation was also done specifically on these members ef the sam-
ple who were under 35 at the time of the Appreciation. For that group
the correlation between pred~cted and achieved level was .84.
In completing the study reported here, results of the original evalua-
tions made for each respondent <including the four gptliere) were
extrapolated along the · appropriate curve for nine years, and that
position then correlated w1 th the level of work at which s/he was
employed in 1984. These data are presented in Table 4. The correlation
between predicted and achieved level was • '70.
In view of the fact that there can be no follow-up of specific indivi-
duals·, it is possible to note only that there appears to be a tendency
for people of mere limited potential <in terms of the predictor> to be
employed at higher levels than would be expected. Informal comment from
the company suggests that a possible explanation could be a recognized
tendency to overvalue operational skills.
b. the lultinotigpol Epsineering Cgmpop7 (35 respondents fgllgWed up>

In the course of the period of the follow-up study, this company has
had to face a decline in heavy engineering in the United Kingdom, but it
was subsequntly involved in extensive developments outside Europe,
building an international reputation in high-tech engineering. Then-in
1986 it was taken over by a rival.
Among the respondents followed up in ~his company, 5 were originally
interviewed between.19'74 and 1979 before the. refinement of the procedure
and 30 between 19'79 and 1983. The respondents included actual
graduates and those selected for an MBA course. Thirty are .men of
Anglo-Saxon origin and one is a woman. ·
The ·work had· the dual purpose of res~arch and application. The eval-
uations of current level of capability and the prediction of rate of
growth were available only to the personnel director; they were not
disclosed to the respondents or to their line managers. After 1979 the
evaluations were made by· two admdnistrators working independently.
Since there were only 35 respondents in this company. they were
included in Table 5 which displays all respondents. Those interviewed
in or aft.er 1979 are in Table 6.

1. Ibe Stydy Undertakgn by the Engineering Cpmpany

In 1987 the personnel director decided to undertake his own follcw-~n

study en 30 people interviewed between 1976 and 1984 on whom there was
sufficient inforlDation for him to judge that they could be used as :1
sample to test the predictive power of the procedure. Because he
i~cluded ~n ~dditional year in his s~udy and ~ecause he chose :a !ollow-
:Jp :Jnl:r ":!lose on whom· b.e could gather more de1:aileo. infor:sa:t·::.an. :,:~ ==
his subjects :1re the same as those of the authors, and 6 are .~radua1:es
originally interviewed in 1984.

n
l ~
 This group is all male. Twenty-six of the 30 <86. 7%> are working
capably at the level predicted; two are over-performing by one level
<one of these has just been declared redundant> and two are under-
performing.

The underperformers are described in the following cases:

Case 1: A 39-year-old metallurgist and licensed engineer. Undoubtedly

intelligent but rather a negative character; seems to exhibit a lack of
drive; obtained specialist ISC and generalist XBA degrees: operating in
specialist metallurgical field.

Case 2: A 36-year.-old computer specialist. Bas started two degree

courses but finis~ed neither although has the intelligence to do so.·

c. the fertilizer CompoUJ <38 respon4ents fgllgwed up)

In the period covered by the study this company has faced very consid-
erable difficulties, and a substantial number of its employees have left
or have been declared redundant. The company operates in a declining
field and is subject to powerful environmental lobbies. Despite these
difficulties. the company was bought in 1982 by a Norwegian company who
in~ested heavily in a new plant located in the •ortheast of the United
Kingdom.

The earliest work in this company was done·in 1976 and the bulk in
1977 as the procedure was being refined by the addition of the phrase
cards and a mere structured work history interview. All the respondents
were male ~nd of Anglo-Saxon origin. · ·

The evaluations were all made by the author as part of an extensive

procedure. They were not fed back to the respondents or to their mana-
gers nor <because of the circumstance described above> were they used
for career planning or individual development.
!he fellow-up study included interviews with 38 <61.3%) cf ~he
original respondents. In each case the criterion used was the level of
work at which they were employed in 1987 or the level at which they were
employed when t!ley ~eft the company i:f that was more than four years
after the original interview. Because there are only 38 respondents in
this company they a~e included in Table 5.

In 1981 two :nembers of the company personnel depart1:en"t decided to

~ke a study of the ~thods of identifying potential used by the company
":e"t·..,een. :976 ~nd :981. Since :hey ·.-~ere laoinn.l! !or :in .:.::mned.ia"te answer
Jn ?rad!c-t:!. 7e 7a.iue, :::ey :=uid :1ei t~er :ac::: 3.way :!:le -:ia't~ n.or '"'ai-: :=r
time "to elapse. An al "terna-ci ve :nethod was •Jsed. A s:mall ~eam ::-o: :::.e
personnel department considered each person for whom the data exis"ted
and subjectively assessed the highest level in the· organization that
~he~r we~e :ikely to reach i! there ~ere ~o constraints on ?remotion. and

19
ability alone was the criterion. Each person was considered, and the
sum of the information available from the sources listed below was
weighted and assessed:

<1> level already achievedi <2> tests <the AB5 group test of high
grade intelligence, the Watson Glaser Critical Thinking Appraisal, the
Kostick Perception and Preference Inventory, the Fineman Self-
Description Inventory York Preference and Job Climate questionnaires, an
AssesmEnt Centre ratiq; <3> career development interviews; <4> job
performance (judged by the immediate manager and mere senior managers>.
. .

Individuals were then assigned to the highest level of work at which

it was considered they had the potential to perform competently. That
level was ·called •Considered Potential•.

Considered Potential assessments were made on 223 people. All the

available data were correlated with Considered Potential and the' corre-
l~tian between the company• s ra-ting of Considered Potential and the
level of work <predicted by the extended interview which preceded the
fully developed career Path Appreciation> was .73.

The company concluded that the evaluation of potential capability

<which they saw as independent of organisation culture> was a mast
powerful and reliable method for informing decisions .involving assess-
ment of managerial potential.

d. the line in 1 PeyelgpiDS Cguntr7 <26 respondents fgllgwed up>

This company in Africa describes itself as • .t. First World company

using Pirst World technology, competing in First World markets with a
·Pirst World product; but located in a Third World country.•

The management of the mine is committed to the development of

indigenous staff, and the original invitation to work w1 th them was
specifically on this group. The work started in 1982, and the follow-up
reported here was done in 1986.

Of the 25 respondents, <15 men and 2 women, or 68%> are members of the
indigenous population which is tribal, predominantly pastoral and of
limited education <only 6 respondents <24~> have been educated beyond
primary school level. The other 8 <32~> respondents are men oi Anglo-
Saxon origin.

The evaluations of the current level of capability of t~e 17 cembers

of the indigenous population have been used to plan training and
individual development. Evaluations on the other 8 have not been so
used. Because follow-up data is available for only small ~umbers at ~he
:10men't. 3.11 25 are included in 7able 5. Then fur-+:he!" and ~ore detailed
: ::1! :Jr~"t:.an is ·~btai Jled, ::ie :iat:i ier:. ·red :r:n:n .::1sas ·.o~b.ere -: ie : '.!dg:me!l~s
~ave ~een used will oe presan'ted separa"tely.

20
 Preliminary analysis of follow-up studies of 92 more employees at the
mine is presented in Section 5. More detailed analysis of the data will
be completed by the Spring of 1988.
1. The StudJ Undertaken by the Mining Cgmpan1

In 1985 the personnel manager of the mine was asked to present a paper
on the prediction of potential at an international company conference.
In the paper he spoke of the need for the management of the company in
the future to come from the indigenous people and of the difficulty of
develop~ng people in a country with a poor education system which does
not provide a foundation of recognized and·reliable criteria of indivi-
dual assessment.
. He outlined the way in which career Path Appreciation had been used
for individual development and departmental restructuring, and concluded
that •the major benefit of using Career Path Appreciation is that it
works. It does actually give you an accurate prediction of a person's
performance potential for the future. We estimate a probability of
success of • 8, climbing all the time. The second benefit is that it is
understood and confirmed by the individuals being assessed. The third
benefit is that it is more cost effective.•

21
Sectign 5. Resylts and Interpretatign

a. The Sample

The full sample now consists of 274 people. But, in this section, the
predicted and achieved levels of only 182 people are displayed in
scattergrams aDd .fully analyzed. The details of the data on the other
92 members of the sample have not yet been fully aulyzed; but some
trends are evident in a cursory examination of the data.
Within the sample, educational qualifications range from doctorates to
primary school certificates; the age range is from 21 to 55. Respon-·
dents were currently employed at Levels I, II, III, IV, and V; 5.4~ are
women; and 211 black or brown Africans. Finally, the settings in which
the respondents were employed are significantly different from each
other <see Section 4>.
!embers of the sample were followed uy over periods ranging from 4 to
13 years. The cut-off point of four years was chosen because, that
seemed the Ddnimum necessary to indicate and established trend.
b. The Data

1. The Group of 182 People:

In presenting the data on this group, it was decided that a
correlation of the .relationship between predictor and criterion on all
182 would be made because that would include the longest period of
follow-up. But it' was also decided to do a correlation between predic-
tor and criterion on the 76 respondents where the predictor was the
refined procedure of Career Path Appreciation.
The data are presented in scattergrams wher~ the actual level of work
at which the respondent was employed at the time of follow-up is plotted
on the vertical axis and the level predicted by the extended interview
(from 1979, the Appreciation> on the horizontal axis. In order to add·to
the interpreta~ion of the scattergrams, a matrix of levels of work has
been superimposed.
Five levels of wor~. -- each divided into high, mid and low -- are
plotted on the vertical ·axis. Each level of work is assumed to require
!.:1 the individual a current level of capability <scope of exercise of
d.iscre't!.On) ~-ihic!l .:an senera-ce and :-espond to the comple~i t:.r Ji :~a\:
level. :ach corresponds ~o a commonly used category ai ~nagemen\:.

~ev~, r is ~onmanagerial work with. a time-span of up to three ~onths.

!one ~f the lon~-ter~ follow-up sample remained in this level. rev~l IT
~:; ::::-s~-~:.::::e ~naf!emen't/:t:.rs-c :evel S?ecial.:.st ·. e~ork ~·nt.b. .1 :1..:1e-·:r?an J!

specialis't work w1 tb. a ti.:ne-span ·:J:f Jne year up -:o -:·,.,o year3. · ~'f,.Q ~ -~:
is senior management/strategic staff work with a time-span of two to
five years. Leyel V is senior executive management of strategic
'usiness units with a time-~pan of five years up to ten years. and ~

23
ll is multinational corporate work with a time-span of ten years up to
twenty-five years <there are only tWo bands marked within this level>.
The horizontal axis represents the level of work which the respondent
would be expected to be able to handle in the light of the original
evaluation of current level of capability and the slope of the growth
curves; each level is siDdlarly divided into high, Ddd, and low.
Table 4 displays data on 84 respondents from the mul tinatianal ail
company who were fallowed up after 9 years. The correlation is .70. It
will be recalled that this table includes four gutliers excluded in the
company• s awn follow-up study, and that it is an extrapolation over 9
years.
Table 5 displays data an 182 respondents who were fallowed up over
pericds ranging from 4 to 13 years. It will be clear from Section 4
that the group includes men and women from different cul ~ural back-
grounds. Some are educated to the Ph. D level, and there are some who
left school at age 11. With r.egar~ to the correlation of • 79 obtained
from this data, it is important to nate that the data include
evaluations of current and future level of capability made in the
earliest stages of the research and development of Career Path
Appreciation.
Table 6 displays data on 76. respondents who were followed up over
periods ranging from 4 to 8 years; that is to say, respondents for whom
evaluations of current and future level of capability were made by using
the refined pracedure. The correlation of • 89 is assumed to ind:tcate
the effect of that refinement.
Table 7 displays data on 59 of the respondents included in Table 6.
The 59 are those who remained in the: employment of their organization.
It will be noted that the correlation here is .92. Examination of Tables
6 and 7 shows that, of the 17.who l~ft, 13 were, in terms of predicted
rate of growth, under- or over-used.
2. The Group gf 92 Pepple
The data on the other 92 members of the total sample have not yet been
put into tabular form. These people are all employees of the mining
company in a 4evelop1ng country. The material is of particular interest
because of tlie· heterogeneity of the sample in terms of race and
education.
Thir~y-four people are of Anglo-Saxon origin <36.9~>; 58 are black or
brown Africans <63~). Eighty are men and 12 are women. Of the 58,
1'7.2% are university gradua'tes, 15.5% had completed secondary school.
15.5% bad co~leted three years or less of secondary school, and 51.7%
iad !lO educ~:t:.on oeyond ?r:::ar~r SCb.Ool.
In each case ~he procedure for ?redic~ing poten~:al was ~ ~areer ?3t~
Appreciation. The evaluation of current level of capability was used to
plan individually-tailored training programs and opportunities and to
predict a comfort curve. An individual's comfort curve represents <or

24
 Tabl;.; ·I

fE_!l~~~~ ~~!!- on 84__ n:s 1londcnts (Multinational oil company)

II Ill IV v VI
_j _1:_
I
J _ ~ _ j _~ _
I
__L__ ~ }~ :; ~ _ ~ _ •
I I
__L_ ~ 1- _~ _-~ _I!, • ~ _L__ ~ _M_
• I I I
-l- .!J___ ~ _! }L -.
I 1 I
i•fl I I I I I I I 2 2 I I
vj -.,--- ··r- --
I
1

I
--r-----
I
------:-------t------~-------t-------:------
I I I I.
-----t-----1------ ----1----1
I I 1 I
t. : : : I : -~ : ~ ! .~ • :
Ill
I r
I
I
I
i
I .
I
I
;·
I
i
I •
:
I
:
1
1: 1 1
--+---~-----~----- ------~-----t-~---·------t------~----- -----t-----1------ ----1----~
v 'l•i'l II I I I 3 • I I • 2 2I I • I I
---L -- .~L
I I
. -----L-----r-------'------.t.-----·
I
------+------..I. . ••••- -----ot------1------
I I I
----1-----t
I I I

1.: I

----1---l
I
~
I
I
I
I
I
I
"t
I
I
I
I
I I
I
I
I
1
I
I
•
I
1
I
• • II ~
1

.-. I : ~
lU n: I
:
I
:
I
: • :i •
I
:
I
4
I
: 3
I
I I • I I
>
<I> ~--~~------~-----
I I I ------~-----~~----
I I ------~------~------
I I -----~-----4------
I I ----~----~
1 1
....l IV M! : : • : : _ • : I • • I I 1 1
"0 ~-- ·~1------ . . ----- ------.. . -----·----- ------,.------~---·--
I 6 -----~-----~------ ----~----4
N
1.11 ....<I>
u
1.:I :I I
I 2
~
I
I 3 I
I
:
2
I
II
2 2 I
1
I I
I
I
•
1
I
1
I I I 1
--~~--- ~
·ri
"d I I I I I I I 1 1
<I>
....
p.. II ! ! -----•------r-------.------t"'-----·
---+--- -t-
l 2 ·: 2 !
! 2 ------r------,------
~ * : ~
-----r-----,------
l
----~----~
:
Il l Ml
I I
I
I
I 2 * I
I 4
I
I
• I
I
I
I 2
I
I
I
I
1
I
1
I
-+-- -:.------:------
I I I I I
.
------:------~----- ------~-------1------ -----~-----t------ ----.&.----..1
I I I I : :
Ll : : : : I I I • I I
---,----1 I ~ I I I I I I 1

~- -·L-~---L-~--+--~--l---~--L--~--- ---~--L------l..----- -----~-----!------ ----i----4

11 M:
I I
:
I
I * I
I * I
I
I
I
I
I
I
1
I
1
1
1
1
I

-+--
L:
. ::------!------1------~------~-----
4 I I I
------·------i------ -----~-----t------ ----t----1
I 1 t 1 I 1
~---------~-----~-----~------~-----·------~------! ______ -----l-----1------ ----1----l
Actual Level at Follow-up
Cone Ltl jon (ltj - 0. 70628
5 t ,J l:t I 0 f liS l
" 2.43071
5iguifkaw.:c: 1} <. o.ouooo
 .
Table !>

Follo'~-=-~P uu 182 n.l:>pondents {4 - 13 year follow-up) 8

II III v VI IV .
-1 _!·_ J- ~- J-l!- - _L- -~ _M-- L-!! -·M - _L_
ul1 I I I * -~_f.!--~_.._-
I
~_I:_-~-~- t-!!--- !--! -I!! -.
' I I I I I I
ra I I I t I 2 2I I I . I I
VI --}---~-----~-----~------~-----t------~-----t-------~-----
Ll I I I. "I I I tlf
-----t-----1·--·--
I I
----1----1
1 * I
- - - t1- - : : :. l : : ! ! I I

Ill' j
I
I
I
I
I
:
I
;
I
;
I .
* : * : •
I I
: :
I I

v
--t---~-----~----- ------~-----t~-----------t------1r-----
f>l 1 I I I t 3 • I I •
-----t-----1------
2 I 2 tI *
----1----~
* 1 11 I
1 I I I I I IT I
--L---L-----L----- ------~-----~-----··-•••-+-·----~••••• -----·---·•4•••••• ••--1----~
I
Ll
I I
I I
I
I
I .
I
I I
I •
I
• 2 * *I
I
I
II
I
i'
·-·>
(j)
---.'- I I I I I
i
I I I

..Q)
~
H•I
I
II 'I
I
· II
I
II II
I
4 *II 7
I
* I
* ': * •;
I I * :: II I'
··:)
(j)
--+---~-----~-----
IV jl.pI I
I
I
I
------~-----~-----·------·------~----~-
*I
I I
I *
I
1412
I -----r-----,------ ----,----,
*I
I I
I
I
I
I
I
....
l·,l
.....u -LI -- -~I ---- .:~I -~---- !------..:..-----~-----
I I
------~------~------
I I
---.--~-----~------
I I
----~----~
I 1
0\
·o Ll I I 2 I 3 I 5 I 3 I 7 2 I I I I
Q)
h II l ·~ I .~ ~ I I 1. I I
c... ---4--1 I I I I I I I I I
I I I I I I I I I 1 I
li11 I I 3 I 3 I 9 11 I 2 I 2 I I I I
I I I I I I I I I l
Ill ~---~-----~--2---~--6---ii--2·--r--7---·------r--;---,------ -----r-----1------ ----1----,
~---~-----~-----~-----~--~---~--------~---·------J
1
__:________:t _____ 1~ __________ :
l ____ J
:
: * 10 : * ~ 2
I I I I I I
Ll : I I I I 1 1
- - - , - - - - 11 1 ! I I I I I I I
lf 11
---'--- I
* 1
I
7 * 5 I
I
2 4 I
I
------.. ------1------
-t------•------~------t---·---t------
I
I
I
I
I
I
1I
-----r-----y·-----
1I
----,----,
1I

1J Ml
I
: l 2
I
I 2
I
I
I
* * I
I
I
I
I
I
I
I
I
I
I
I .
I
I
I
1
I
1
1
1
'1
1
1
1
--r----.- -----:----- -~--------:------~----- ------~-------1------ -----~-----t------ t-----t----1
L', .. *
I
I
I
I
4 * I
I
I
I
I
I
I
I
I
I
I
1
I
I
I
1
1--~ -------------1..------.------------ -------------------- ------------------ ----------
~ctual Level at Follow-up
Correia! it'll (H) ::: 0.79226
Std t:r r ol· Est = 1.81508
Signi fi,:au..:e: p < u . tHHioo

a
Total :,.ampll: of 2·1:1 less 92 employees of a mining company
in a dl:\ doping l:u~111lry.

.
 Table u

II III IV V VI
_J __L_j _t.!_j_ll __ _!. __ ~ .. t-!_~_1!_. __L__ ,__J! __~_H__ r--~-~J-I__t_!j ___ 1_!_M_-.
M: I
I
I
I
I
I
I
•
I
I
I
•
I
•
I
•
I.
I
I
I
I

VI--r---~-----~-----
L: : : ------~-----t-----
I. I
------t------~-----
I I
-----t-----1·-----
I I
----~----~
1 1
1 • 1 1 I I I : I I I I
---r······ - , - - - I. I I I i
II: j I
1 I
iI :
I
:I ;I 1*
1
I*
: :
I
:
I
-.- ·-- ·-~ -----~-----~------}---:---t-----· ------t--------:-------
1
-----t-----~------ ----1----~
V 1•1 11 I I I I I 1 1 t * 1 1
I
--1; ... --\.--- ~-L-----
I
------L.-----.&.----- ,..-----+-----·+-----
I I I I I 1
-----t-------1----•-
t
-----t-----t1

L: I I
I
---~ ..... _._:....,.__
I
I
I
I'
I
I
I
I
I
I
I
I
I
I
I
•
I
I
* 1
I
I
t
•
I
t
I
1
j
I
iJI1
I
II
I
I I
I I I
I
j
I 2 I * ,; I
j ;
1 1
'I
~)
;..
Q)
I I
I
I
I I
I
I
_ . .. -- ··t- -----1------~------.------t------·
I I
I
------T-------.------
I
I
-----r-----,------ ----,----,
I
I
1
1 1
1 1
1

1-.l
.. l
IV .!:L . J ____ J _____ ~------L-----i-----· -----.. i--!---~--~--- -----~-----~------ ----~----~
....... '\J
(I)
p L:
:I
I I I
I
:
I
I
f
I
I
I
2
I
I
I
* I
I
I
* I
I
I
1
I
I
1
I
I
1
I
1
---~····I
0
·rl I I I I •I I I 1 i
'0 I I· I I I I I I I

~L-- -L -~---L----- -----l-!---~--~-- ---~--t--:---~--~--- -----~-----~------ ~----!----~

(I)
f·t
C>.

J[l J.tl
I I
:
5 :· 6 : 5
I
: 2 I
I
I
I
I I
I
I 1* I 1 1 1 1

--r---~-----r-----------~------r-----·------f------i------
7 2 * * -----~-----t------ ----t----~
1. 1 I
I
I
I
I
it
I
_jl
I
I
I
I
1
I
I
I
1
1
1
1
I I I i
-~-·
11 11 4 2
I
I
I
I
I
I
I
I
I
I
I
I
* I
I
I
I
I
I
I
I
1·
I
I
1
1
I
1
1
I
1
--L--~------~-----------~----~-~----•·------~----••1------
I I I I I I I -----r-----T------ ----y----1
I 1 1 1
I 1 ~~ II I
I
2 I
I·
2 I
I
I
I
I
I
I
I.
I
I
I
1
I
1
I
1

--:--~-:------l-------------1------~-----·------~------i------
I I • I* I I* I I I -----L-----L------1-----.l.----..l
I I I I
.I ... I I I I I I I I I I
·-------------------------------------------------------- _____ t _____ ! __________ ! ____ J
Actual Level at Follow-up
Corrdoalion (H) - 0.89224
Std t:n of Est - 1. 021:H}8
Sigui fi<:ance: p < 0.00000
 Tal.t..; 7

II III IV v VI

.--t
<IJ
>
<IJ
~

·o
<II
N j-1
~ u
•rl
'U
<IJ
H

~---L-----L-----~-----Jl-----L--~-- --~---t ______ j______ -----~-----!------ ----!----~

ll,

I II M'
I I
I
I
I *2
I
• 6
I
I
4--+-----:------ ------:~-----~----- . ------~-------1------ 5
I
I
I
•
t
I
I
•
-----~-----t------ ----t----~
I
I
I
I

L:
I
7
it : * : i
;
I
* :
-•i
II
i
I
I
•
•
I
I
I
I
I!
Ill
I
I
I 4 2 i
I
I
I
* 1
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
1
I
I
t
I
--L--~------~-----~-----~------~-----·------~------~------
I I I I I I I -----r-----T------~----,----,
I I I 1
MIl 2 2
II
I
I
I
I
I
I
I
I .
I
I
I
I I : : :
~--~-----~------
* ·I -----~------~-----·------~------i----~-
* I -----~-----t------ ----t----1
I I ;t.
:_-- - 1- - - - - -1
I
------------.!_____ ...;!., ____________ !. ______ .,! ___________ !. _____ l, ______ L.----l----1
.I I I I I I

Actual Level at Follow-up

CUITl!lation (I{) - 0.92996
Std ht· of Est .:.: 0.93108
o.ooooo
Siguil'icance: I'

a
-----·------
Till!! c 6 re::q.uudcnts
··--~ __ ... ___ who remained employed by their
.,,·,;;llti :L<Il iou::..

.
less optimistic. This shift was apparent by the fifth year and changed
little over the next 15.

Bray and Howard <1983> concluded that, •to a great extent, measureable
differences in managerial abilities and motivations that would foretell
success were present when the men first came to the original assessment
centre.• lOst of the 26 assessment dimensions had significant correla-
tions w1 th success after 8 years. Indeed JDDSt of these relationships
held even after 20 years <page 301>.
The correlations are displayed in Table 8.
Table 8

Some Assessment Exercises Predicting Janasement Pgtential over 20 years

Exercise r
SCAt& - Total .40**
Range of Interests .30**
Organizing &Planning (In-Basket) .19*
Inner Work Standards (In-Basket) .24**
Need Dominance Scale (EPPS)b .18*
Ascendance Scale (GAMIN)c .20*
Expectations Inventory .18*
Need for Advancement (Interview) .44**

Note r's = 199 - 266

*p< .055
**p < .001
a. School and College Aptitude Test
b. Edwards Personal Preference Schedule
c. Guilford-Martin Invent·ory of Factors

<b> The Anstey Study:

This iS· a 30-year :follow-up of the Civil Service Selection Board

procedure in the U.K. <Anstey 1977>. Three thousand one hundred candi-
dates for the Civil Ser"tice were assessed between 1945 and 1948: 421
between the ages of 21 and 30 were appointed. The follow-up sample was
the 301 people still in the service in 1975.

:"he predictor '~'~as <:b.e 2.rk Cor grade) at the final selec~ion board.
!his· mark represents a summation of scores from tests and interviews
concerned ·..rt. th intellectual, communication and interpersonal skills.
The cri terton was the rank achieved after 30 years. Most equalled or
3Ur?assed -:~e ?redic-:ed. :1l-:i=a"te level <Jf ·,a~or!t. 'Jut some ·~f -:~is ·::Juld
·;e 3.C::oun"ted fer JY ~n. :.=.c::-sase ::1 :!le J.umber ,Jf senior ?OS~s iur:':l~ :ie
J o-:tear peri=d..

The correlation between tbe final selection board mark and the rank
attained was r = .354. That was then corrected for selectivity to r =

29
constitutes> a rate of growth in responsibility in work which is in line
with the predicted rate of growth of his or her capability.
Vi thin this group, there was one evaluation ·of current level of
capability which was an underestimate. Apart from this case, the others
have acted in line with the prediction of rate of growth of capability
made in the Appreciation: Twenty-one percent have left the company. In
801 of those ·cases their leaving cpuld have been predicted because it
was clear from the Appreciation that they were underused. In four
individual caees, a manager made_ the decision to promote a subordinate
who <in terms of predicted rate of growth> was nat yet ready for the
~xtended responsibility. Three of these were demoted within six months
and the fourth is showing signs of considerable stress.
The overall indication from preliminary analysis of this material is
that <over a four-year period> the predicted rate of growth of capa-
bility has been borne out in 941 of the cases.

e. Cgmporisgn with gther Lons=term Studies

1. SuDmDTJ gf the studies

Two major longitudinal studies which set this research in context are
those ~f Bray <1983) and Anstey <1977>.
(o) ne BrA)' StUd.J:

In 1956 the Bell Telephone Company decided to initiate longitudinal

research into managerial lives. Four hundred and twenty-two new mana-
gers took part in a three and a half day assessment centre. Assessment
included a battery of projective tests, psychological inventories,
attitude surveys, interviews and simulations. The initial assessment of
the group took five years and included two distinct populations of white
males: recent college graduates hired as management trainees <I = 274>
and non-college employees who had worked their way up in the company
<I = 148> I

In this study the predictors were 26 assessment dimensions derived

from the total test battery. The criterion was the level of work
defined as Level III, or district <in broad terms, middle management> or
higher.
5ach person was followed up by interview every year. After a years
they were again put through an assessment centre; and, in the twent!e~h
year, through a different assessment procedure. Both intelligence and
knowledge test scores increased over the period, but there was a marked
drop in score for interpersonal skills. In ter:DS of some simula-cior.s,
:1ana~ement ~kills •,o~ere ~ot i!!rproved i"!l. the eight-?"ear period: :ind.
:i.ildeec., so:ne ·,.,ers los-c.
!he material reported from the follow-up after 20 years shows tha-c as
participants got older, their expectations became more realistic and
f

30
 . 66. Anstey suggests that this correlation may be an underestimate
because of the nature of the sample in which all the individuals were
very well .educated. Anstey also points out that it is considerably
higher than other correlations of tests· of abilities w1 th demonstrat-ed
competence at work.

2. Cgmporisop gf the Three Studies

<a> The samples: The Bray sample is the largest, with 422 members
compared with 301 in the Anstey study, and 274 in this study. The
composition of both the Bray and Anstey sample~ is less diverse than the
sample of this study. !embers of the first two samples are Anglo-Saxon
and male. The sample of this study includes a small number of women and
a fifth are African. The Bray sample includes graduates and
nongraduates. !embers of the Anstey sample were all described as •well
educated•. . The educational qualifications of members of our sample
range from very extensive high to very J...imi ted. The Bray and Anstey
studies were both done in a single organization. lembers of our sample
· ~re employed in four very different settings.
<b> The procedures for predicting potential: The Bray study used 26
different assessment dimensions which take some days and experienced
observers to determine; the Anstey study used a range of assessment
dime~ions based on five days of tests, observations and interviews
which were summarised .in a single final mark. The tests, observations,
and interviews all requi·red skilled staff. A Career ~ath Appreciation
focusses on the current capacity to generate complexity, takes only two
hours to adDdnister, but does require a highly trained a~nistrator.
<c> The criteria: The criterion in the Bray study was a particular
level of work or higher;, in the Anstey study the cri.terion was level of
work ac.hieved. The establishment of that · level was based on careful
analysis and modelling of the work of senior staff. Our criterion was
level of work achieved with the level evaluated by analysis of its cour
plexity.
<d> Time scale.of follow-up: The Anstey study extended over 30-year
period; the Bray study over 20. Our longest period to date is 13 years.
SoDe of the people have been followed up after only four years. In view
of the fact that this study set out to test an already well-for~lated
and tested hypothesis about growth in capability, the shorter time scale
can be justified. !Jhere possible. further follow-up studies will be
under~aken.: bu-c. i!l -::J.e ?rese:n-c socioeconomic cl i!!late. :. ~ :.s becoming
increasingly dif£1cult ~o ~in~ain contact w1~h people who are less and
less likely to re!llain in the employ of a single company over long
periods of time.

In a recent article Guion <1987> raised important questions about the

bivariate correlation between predictor and criterion -- the validity
~=ef!!c!ent -- in selection ~esearc~. Since selection research carr!es

31
with it an explicit or implicit concept of potential for growth, Guion's
analysis is used here as a basis for looking at the validity of the
whole process described above.
Guion pointed out that vieWs of valid! ty are changing and that the
current trend is to expect t~ be able to rely on more than a single cor-
relation coefficient. In order to consider these points in detail in
regard to the data presented in this study, I will follow his categor-
ization of the four components of the established practice: i> choice of
predictors; ii> choice of a criterion; iii) data collection, and iv>
evaluation of the predictors.
1. Chgice gf Prgpedures fgr Prediptigu

Guion <1987>, suggests that the current trend is towards predictive

procedures which are directly_ related to •work orientation•, that is to
say, a tendency to lggk to troits manifested explicitly ot work rather
than to· those shown in a broade~ ~rrar of situations.
The procedure used in this study·-- an evaluation of the respondents•
current level of capability - has been evolved from a definition of
work which includes the psychological experience of doing work and the
scale of the organizational setting in which it is being or is to be
done. As stated above, a Career Path Appreciation is a procedure for
considering and giving .meaning to the relationship between individuals
and the organizations that employ them.

Within the Appreciation, the phrase cards·give an indication of the

extent of the current and future possible contexts in which the person
can generat~ and respond to complexity. The symbol card task assesses
the person's capacity to order chaos. Both indicate the relationship
between the •cognitive map~ available to each-individual as a guide and
the actual territory of·the environment in which the person ope~ates.
Both Ashby's <1956> Law of Requisite Variety and Korzybski's <1933>
reminder that "the map is not the territory" are relevant here. They
can be drawn together into the concept of the wgrksgape of the indivi-
dual -- .that scope of responsibility for work and resources which the
individual can map. ·
The concept wgrrscape emphasizes the link between an important
postulate of SST theory and of the Appreciation. The postulate is that
levels of work are differentiated by the dimension of complexity; the
core of an At>oreciation
genera~e
. ..
and to
is an evaluation of the ,erson• s cauaci
~espond to complexity.
- - to
t""r .
Further. a Career ?ath Appreciation is, explicitly, a procedure in
which work-sampling is one of the prime components. A third point which
:..s ~oncer"!!ed 3pec::.::.cally ..,i -:i -:he ?SJchologic3.l ~xper:enca of doing
·.;or~ . .,1.::.. ·Je i!.sc~ssed..

A further point about the predictor used in this study relates to its
concurrent validity.. As Guion <1961> points out, concurrent validity is
no substitute for predictive validity, but it does give a useful indica-

32
 tion of the relationship between the predictor and actual behaviour at
work. Data on the concurrent validity of Career Path Appreciation have
become available in four settings where an organization wished to use
the procedure. But, before doing so, they wanted some indication of the
degree of felt congruence between the evaluations made and the formal or
informal managerial judgments of the respondents' current performance
and likely rate of growth.

<a> In the multinational oil company twa such correlations were made
in the course of the original study. Both were formal because the co~
· pany had in place a system for identifying potential. The first <of
.71> was between tne evaluation of current level of capability as made
in the extended interview and the level of performance as stated on the
company assessment forms. The second <of .86> was-between the evalua~
. tion of current level of capability and the actual, verbal statement of
the manager about performance. The discrepancy between these correla-
tions raised questions about the company system of assessment which it
was not possible to .pursue.
<b> In a DUltinational chemical company where work started in 1985 a
less formal study was done, but w1 th the purpose of considering the
match between evaluation and managerial judgment. In this case 37 res-
pondents were involved, and the correlation was .79.
<c> As stated in Section 4, the fertilizer company undertook its own
study of the correlation between •considered potential• and the evalua-
tion of current level of capability. They found a correlation of .73.
<d> The-British Army-Staff College <where. Caree~ Path Appreciations
have been offered for purely research purposes for the last eight years>
has done informal correlations in five of those years; they range
between .71 and .77.
An interesting point to be made about these indications of concurrent
validity is the initial attitude of the organization that the correla-
tion indicates the efficacy of their system, however formal or informal.
On further reflection the conclusion is that, although it may be
efficacious, it is not efficient simply because in-house judgments are
not held to be reliable unless they have been built up over a period of
time. Evaluation of current level of capability can be made in a Career
Path Appreciation <which lasts for two hours> .

..,
d

Guion <op. cit.> suggests that the biggest change of view about data
collection is !.n desired sample size. Researcher-s are now looking for
much larger samples than the 30 or so subjects that were previously
~c=e-o-cable. :'b.e 3ai:I?le 3i=e JI :his study ~t present :s 274-. :::1 ·1!ew
:t ~~e i:..:::.~:.:l ~y J!. :::ai:lta::l.ing c:m"tac-: ',.Jl-:i :-ss?cncie!l"CS
-::Jnsid.er:ible
i::l -:ie presen't soc!.a.l. and -:conomic .::rc~l:lS'tances '"'here -=~ange _,_
employment is the norm. a sample size of approximately 300 seems a tair
number from which to draw conclusions. The detailed data· presented here

33
 refer to 182 people with preliminary analysis of a further 92 cases.
these numbers are probably sufficient to indicate consistent trends.
3. Ghoice gf a Criterion

Guion points out that 30 years ago a single, overall criterion was
regarded as •indispensable• for selection research <l'agle, 1953>. By
the 1960s the use of multiple criteria was being urged <Dunnette, 1963,
Guion, 1961>. But, by the late 1970s. results of validity generaliza-
tion studies questioned whether fine distinctions among criteria were
worth making; and the trend <once again> favoured an overall, global
measure.
The criterion used in this study is defined by the theory which pro-
vides a well-t-ested model of ~he way in which work is organized in
levels of-increasing complexity. But there is a problem with this cri-
teria~ • . On the one hand, it is clearly job related in that the level of
work at which the respondent is employed at the point of falloW.up is
the actual, social reality of what the organization is prepared to pay
far current perceived levels of capability. On the other, one of the
difficulties of follow-up of this kind is that it was not possible to
use the time-span of discretion -- t~e instrument for measuring levels
of work which is made available by the theory. However, the level of
work can be categorized by examining the wcrk itself and directly
analyzing the level of ·complexity.
The settings in which the work was done were such that it was possible
to accommodate the ways in which work was structured within the organ-
ization with the structure of complexity provided by the model. But in
pure terms it was not possible to use a fully construct-related
criterion.
4. Construct ValiditJ

a. The array of growth curves:

SST makes it possible not only to place predictor and criterion in the
same content domain, but also <by means of the array of growth curves>
to predict the relationship between the two for a given individual at a
given time.
As we pointed out on page 4, a common assumption in all assessment is
that adults develop and that they develop at different rates. This
assump~ion may be formali:ed in a minimum development curve which
assumes a certain rate of growth. !he array of gro~h curves in the
~del makes it possible to go beyond the idea of a minimum development
curve to offer a specific prediction about the likely rate of growth of
~he capabili~7 of a ?ar~icular :!.ndividual.

!his study :!.s both an atrolication and a test o:i ::he 5lo'Oe o:i t~e ar:--:3.•1
of growth curves derived· by Jaques from data on ind{vidual earning
· progression and offered as a hypothesis about the growth of the capacity .
to exercise discretion which, at any given time, will be manifested in
~he person's cur~en~ level of capability.

34
 Details of other tests of the slope of the curves are available in the
mere extensive Technical Report on the Prediction of Potential. A more
general point may be made here about curves of this general type. In
biology there has always been considerable interest in the growth
processes of living systems. In one study quoted by Saunders <1967>, a
comparison was made between three growth curves plotted by measurements
aver a .period of time: That of an individual plant <a sunflower seed-
ling>, that of a multicellular organism <a colony of bacteria>, and that
of a population of organisms <fruit flies in a bottle>. In each case
the same form of logistic curve was fou~d to fit all three curves.
This form of logistic curve has also been used to predict the growth
patterns of populations, industries and industrial products. A number
of studies of predictive validity have been done <e. g. Lasky, 1951;
Pearl, 1922>. One study showed that the final height of a sunflower
seedling could be predicted to +1- 5S, if the data for mere than 50S of
the life cycle were used. A siDdlar study fitted the logistic curve to
the growth of railway track in Britain in the nineteenth century. The
data to 1870 <which were past the peak>· gave a ceiling prediction of
20,000 miles in 1920 whereas the actual ceiling was 23,700.
Saunders found that the logistic curve could readily be fitted to
Jaques• array of growth curvesi and hypothesised that they too would
have a very high ~egree of predictive validity. Bawever, in the absence
of a procedure which could place a person on a particular curve, he was
unable· to test his hypothesis. Of special interest are the studies
indicating that the final point of the matured curve can be predicted
with +/- 5S accuracy if the data for mere than 50S of the life cycle is
available. If we apply this to the capability growth curves <see Figure
1>, it would suggest that it would be progressively mere difficult to
predict the final point of th~ growth curves in those people whose
capability will not mature until mid-life or beyond <those whose comfort
curves fall within modes VI, VII and VIII>. This could explain the
widely accepted difficulty for managers in recognizing and acknowledging
the potential of these people in their 20s, with the consequence that
they are not given appropriate opportunities. This difficulty ma.y be
contrasted with the intuitive recognition of the potential of people
whose com£ort curves fall within modes IV and V -- the assumption being
that the curve is a straight line and the risk that they may be
overpromoted.

The correlations obtained from the data presented offer further

svidence for ~he ?r~dic~ive validity of the array of curves as the rep-
~esen~at:on ~i an intrinsic pat~er~ ~f gro~h in ~he =apabi~ity ~c gene-
rate and to respond to complexity.

35
b. The Capacity to Generate and to Respond to Complexity
The correlations reported can be partly accounted for in terms of the
coherence of the underlying theory and the predictive validity of the
array of growth curves. The coherence of the theory rests on a
definition of work that allows a common approach to the levels of co~
plexity in which work is organized, the individual capacity to generate
and to respond to these levels of camplexi ty. and the growth of that
capacity w1 thin and across l~vels. It is this coherence that has made
it possible to develop ~ more accurate procedure for predicting
potential.
Almost 20 years ago, Jaques <1970> suggested that the capacity to
generate and to respond to complexity was dependent on a constant
interplay between verbalized and nonverbalized material. He pointed out
that al thaugh •the exercise of discretion• is a profoundly falliliar
sphere of psychological activity, it is conceptually ill-defined. As a
consequence, there is no satisfactory language far it •.
People simply cannot put into words all that they are taking into
account in making the decisions that will forward their work. In that
sense, they cannot be sure that what they have decided to do will get
them where they want to go, will achieve the result they want to
achieve. They judge, think, hope -- and, indeed, pray -- that it will,
but tliey ·cannot be sure. Only tim will tell. In an attempt to convey
the experience of .the interplay between verbalized and nonverbalized
•terial and its significance for them, .people tend to use words like
•judgment, intuition, nous, skill, experience, know-how, common sense,
gut-feel, hunch, discretion, discrimination•.
Over the last 20 years there has been a steady growth of interest and
research into the processes of deci,sion-making. One of the significant
outcomes has been the distinction between •rational, logical, analytic"
styles which are readily expressed in words or numbers and •non
rational, intuitive• styles which find readier expression in symbols and
images <see, e. g., Xintzberg, 1976>. Some of this research has been
underpinned by neurophysiological evidence <see, e.g., Levy-Agresti and
Sperry, 1968 and Gordon. 1986> and some by work in artificial intelli-
gence and expert systems <see, e. g. , Barr . and Frigenbaum, 1982 and
Simon. 1979 >•

Another strand in this tapestry can be seen in what may be thought of

as "techniques• for enhancing creativity in decision-making. Many draw
explicitly on the use of symbols, metaphors and images in order to draw
on material which has not been verbalized <see, e.g., Gordon, 1961, de
Bono. 1971; and Prince. 1982; and, for a review, Rickards. 1987).
:Zl 3. :-9can"t ~r~icle ~bout t!le :-:Jle ::f '!.n.tui tion ~nd ~:motion !.:1
iec:.s1on- 2.k:.:1g. Si1:0n C.987 ~· argues :ha-: .:. t : s .:all..:1c:.aus -: o con-r:-:is~
"a.naly"tica.l." and '':!.n'tui"ti7e•• styles 3.nd :.ha"t the -3f:fec~ive Ena~er "does
not have the . luxury" of choosing between the two. He refers to
·Barnard's ·<1938> distinction between "logical• and •non logical" pro-
cesses for making decisions and quotes his comment that •By 'logical'
precesses r ~an conscious thinking which could ~e expressed in words ~r

36
 by other symbols, that is, reasoning. By 'non logical processes• I mean
those not capable of being expressed in words or as reasoning, which are
made known by a judgment, decision, or action•.

In considering contemporary views of these "two different forms of

thought•. Simon <ap. cit. > is critical of what he calls "the mere
. romantic versions• of the split-brain research. But he goes on to argue
that there is now a substantial amount of evidence far the two different
farms as essential components of the complete domain of decision making
<see, e.g., Simon, 1979; Doktor, 1978; Gordan, Charns and Sherman, 1987;
Rickards, 1987; and Taggart and Robey, 1981 >• In particular. Simon
emphasises his observation that people reach decisions w1 thout being
able to report the thought processes. that took them to their conclusion.
A considerable and grpwing trend in current thinking about decision-
making, including, e.g., Sternberg's <1985> work on •tacit knowledge",
which emphasizes.•insight•, •gut-feel•, •intuition• and the significance
of access to nonverbalized material. Career Path Appreciation --· which
is now a widely applied and well-researched procedure -- can contribute
to this approach to decision-making in three ways:

1. It can extend understanding of these processes in general terms.

2. It can make it possible for the respondent to make spontaneous
discoveries about the internal, nonverbalized resources s/he has been·
using in decision-Dating about work. This can, then, increase his/her
skill in gaining access to them. In addition, the administrator. can
offer explication of the ·way in which the respondent seems to be drawing
on internal resources, thus heightening awareness of the resources,
affirDdng use of them, and further enhancing access.
3. Because it is a predictive procedure, an Appreciation can offer an
opportunity for a better·understanding of the current and likely future
range of one's decision-making capability. Further, the underlying
theory can contribute to an extended understanding of the relationships
between levels of complexity of work and range of complexity of
decision-making in the life of organizations and of individuals.
The final point is that the construct validity of Career Path
Anpreciation rests in a coherent theory of work. The theory offers a
definition of work, a hypothesis about increasing complexity in the way
work is organized in response to extending environments and, a ~ypothe­
sis ~bout in.di vidual ·iif:ferences in ~owth in the capabil! ty to d.o '.t/Or~
~f :nc~easing co~lexity.

!~e research described i~ this ~epor~ -- in particular, the 3Ccuracy

of the predic-tions of :-a-ee of growth of poten1:1a1 -- provides :ur-.;t.er
-:onfi:-:::a"tion fQr ::!lesa ~ypctheses:
: l.) :'he iypo1:.:.es1s :ha"t :ierc: :s iisco:lt:.:.u1-:·r ·~e't·,.,een. ::~,e.!.$ ...,..; :::m-
plexity in work and in individuals. The evalua"tion of current level oi
capability is made on the premise that there are qualitative differences
between levels and that they can be observed by a well-trained admini-
31:ra'tor.

37
 <2> The hypothesis that adults do develop, that they do so at broadly
predictable rates, and that there are differences between individuals.
The confirming evidence for this comes from the follow-up data which
demcnstrates that, in a heterogeneous sample of people in four organiza-
tions operating in diverse environments and various parts of the world,
differences in rate of adult development at work are recognized by
promotion and pay, and can be predicted by an Appreciation.
In the first section af this report I suggested that a coherent theory
was required to enhance the content and the predictive validity of pro-
cedures for predicting potential. I conclude by suggesting that the
correlations obtained in these follow-up studies confirm the power of
the theory used as the basis for Career Path Appreciation. ·The
definition of work as •the exercise of discretion within prescribed
limits• made it. possible to develop a procedure based on work-sampling.
The hypothesis . about discontinuity in states of complesi ty brought
predicto~· and criterion into the same content domain and made it
possible to evaluate current capability. The array of. growth curves
provi~ed an explicit hypothesis about ·rate of individual growth.

Career Path Appreciation can be seen as an application, a confirma-

tion; and, in some ways, an extension of Stratified Systems Theory.

Gillian Stamp
February, 1988.

38
 REFERENCES
Anstey, B. <1977>. A thirty year follow-up of the CSSB procedure, with
lessons for the future. Journal gf Occupotignal Psychglgu, 50,, 149-
159.
Ashby, Y. <1956>. An intrgdugttgn tg c;ybernetics. London: Chapman &
Ball, Ltd.

Asher, J.J., & Sciarrino, J.A. <1974>. Realistic work sample tests: A
review. Persgpnel PsychglgSY, 21. 519-533.
B~rnard, C. I. <1938>. the functigns gf the exegutive. Cambridge, Xass:
Harvard University Press.
Barr, A., & Feigenbaum, B. A. <eds.> The handbggk gf artificial
intelligence. <Vol. 2>. Los Altos, California: Kaufmann.
-
Boals, D. <1985>. Data gathering techniques. Unpublished paper.
Bray, D.Y., & Grant, D.L. <1966>. The assessment centre in the measure-
ment of potential for business :management. Psn;holostcol Xanggrophs,
a.a. <17>.
Bray, D.Y., Campbell, R.J., & Grant, D.L. <1974>. Pormattye years ip
business: A long-term "AT&T study gf mgpagertal lives. lew York: John
Wiley.
Bray, D. W. , & Howard, A. , <1983>. The AT&T fongi tudinal studies of
anagers. In K. W. Schaie <Bd. > Lonsttydinal studies gf adult pyghg-
lggigal deyelgpment, 266-312. lew York: Guilford Press.

Bruner, J. <1966>. Igward a theory gf instruction• .lew York: lorton.

Cohen, B., Moses, J.L., & Byham, W.C. <1974>. The Validity of assess-·
ment centi:"es: A literature review. Pittsburgh, PA: Development
Dimensions Press.
De Bono, B. (1971>. Lateral thinking fgr management. Maidenhead,
England: McGraw-Hill.
Doktor, R. H. <1978>. Problem solving styles of executives and
management scientists. In A. Charnes, W. W. Cooper. & R. J. lTiehaus
CEds.). ~nage:ment SC"'Qnce ~pproaches to :nanpgwer planning and qrgan 1 sa-
~.; c~ ,..;ec_.; in. Amsterdam. Holland.

Dunnet'te. ;.{. D. <1963>. A note on the criterion. Joy?'"nal nf ~ppl~ed

~sychglQiY· 47, 251-254.

:::vans. ~ . .::>. <1979>. -..,~ '-!;;.r~g~T'I9'"'t .~-r '-1u'!'l~n ::ap;,c• -.,. 3radi;:r1.
3ngland: Ja.nagement ·:enter 3radford.

Gordon, W.J.J. (1961>. Synectics; The deyelgpment gf oreatiye capacity.

New York: Harper & Row.

39
Gordon, H. W. <1986>. The Cognitive Laterality Battery: Tests of
specialized cognitive function. Internatignal Jgurnal gf Jeyrgscience,
aa. 223-244.
Gordan, H.W., Charns, X.P., & Sherman, B. <1987>. Management success as
a function af performance an specialized tests. In Human Relotigns, ~.
ClO>, 671-698.
Gould, D. <1985>. An examination of leve.ls of work in academic library
technical services departments utilizing time-stratified systems theory.
Unpublished doctoral dissertation, University of Southern california,
Las Angeles.

Gray, J.L. <Ed.> <1976>. the Glocier Prg1eet: Cgncepte ond critiques.
lew York: Crane Russak and Ca., Ltd.
Guion, R.X. <19§1>. Criterion Deasurement and personnel judgments.
Personnel P§JchglQIJ, 1!, 141-149.
Guion, R. X. <1987>. Changing view for personnel selection research.
Peragppel P&JchglQIJ, 40, 199-213.
Herriot, P. <1987>. Graduate recruit~nt Getting it right.
imploxment GAzette, ~. <2>, 78-83. London: Department of Employment.
Bema, B. <1967>. The dJDam1e inter-relotigpships.between wgrk. payment.
and capocit7.. Unpu_blished doctoral .dissertation, Harvard University,
Cambridge, XA.
Hunter, J.B., and Hunter, R.F. <1984>. Validity and utility of alter-
native predictors of jab performance. Psrchglggicol Bulletin, 2§,,<1>,
72-98.
Hunter, J. E., and Schmidt, F. L. <1976>. A critical analysis of the
statistical and ethical implication of five definitions of test
.fairness. .PSJChgl ggica 1 By 11 et 1 n, Ba, 1053-1071. .
Isaac, D.J., & O'Connor, B.l. <1969>. Experimental treatment of
discontinuity theory of human development. Human Relations, 22,, <5>,
427-55.

Isaac, D.J., & O'Connor, B.X. <1973>. Use of loss of skill under stress
~o -test a ~heor7 of psychological development. Human Relati pns, ~.
(4), 487-496.

Jaques, E. <1956). Measurt:a:rent of rC\sponsi bi 11 ty. London: Tavistock

Bocks .
.Jaques.
3ooks.
-· •.:!.967). -.:qui-rabi~ ?ayznen-;, :iar:ncnswor-;h.. 3ngland: ?en9:Ui!l.

Jaques, E. <1968>. , PrggressigD handbggk. Carbondale, Il: Southern

Illinois University Press.

40
Jaques, E. <1970>. Wgrk. creatiyity and spcial 1ystice. London:
Heinemann Educational Books, Ltd.

Jaques, B. <1975>. A general theory pf bureaycracy. London: Heinemann

Educational Books, Ltd.

Jaques, B. <1978>. Leyels gf ohstn,ctign in taste oDd humon octign.

lxeter, lew Hampshire: Heinemann lducational Books.

Jaques, B. <1982>. the form gf time. Iew York: Crane Russak a Co. Ltd.

Kohler. T. <1984>. The development of capability in executives and

managers. Unpublished paper.

Korzybski, A. <1933 >• Seiepce ond sopity. Science Press Printing

Company.

Lasky,. SaG. <1965>. Mineral industry futures can be predicted.

Epgipeerini ond lip,inl Jourul, 15§,, (9), 94-96.

Levy-Agresti, 1., a Sperry, R.Y. <1968>. Differential perceptual

capacities in major and minor hemispheres. Prpeeedings gf the Jotignal
Acade:JDJ gf ~ienoe. U.S. A. , §.1.,. 1151.

Lofland, J., a Lofland, L.H. <1984>. Anolystns-social settinss; Guide

tg qyalitatiye gbseryatign ..•pd analysis. Belmont, California: Wadworth
Publishing.

Xatte Blanco, I. <1975>. The uncpnscigys as infinite sets. London:

Duckworth.

McKellar, P. <1968). Experience and behavigur. Harmcndsworth, England:. .

Pelican.

Kilner, X. <1950>. On ngt being able to paint, London: Hein~mann

Educational Books, Ltd.

Xintzberg, H. <1976>. Planning· on the left side and managing on the

right. Haryard Business Review. ~. 49-58.

Iagle, B. F. <1953> • Criterion development. Personnel Psychplpgy, Q.,

.271-289.

!ys-crom, ?.C. Cl973). Equity theory and c3reer pay: A r:ompu1:er

simulation approach. In The Journal of Applied Psychology, ~. (2),
125-131.

?o:arl. :2. (1920). :'he ?opulat:!.on ?r-oblem. Gec~-aphicai ~~yiew. Oct.

:920. 535-645.

Prince, G. <1982>. Synectics. In Olsen, S. A. <Ed.) Group ;:q anning and

Prgblem Splying in Engineering Kanagement. lew York: Wiley I~trescience.

41
Reilly, R. R.. & Chao, G. T. <1982>. Validity and fairness of some
alternative employee selection procedures. Personnel Psychglggy, ~. 1-
62.
Richardson, 1. <1971>. Fair Pa7 ond Wgrk. London: Heinemann Educational
Books, Ltd.
Rickards, T. <1980>. Designing for creativity: A state of the art
review. nesiSD Studies, ~ (5), 11-16.
Rickards, T. <1987>. Closing clown: A classi~ication of crea~ive
decision-making aids. JgurpAl gf Jonoseriol PSJChplQIJ 1 a, (3), 11-16.
Saunders, A. <1967) • a.A_.sLIIIt~o~~~u.Md~J_...Q.,.f~t..,.hMe._l....t...,.fllr.lie~e~~~,y~c~~&olwe._..,g:C..._.,..f.~~~gr.wu...,r.....,.i.,..nlt.1llld11111
u~su.t...,r...,ioal•
prgduets. Unpublished paper, Cranfield Business Schaal, Cranfield,
Bngland.
Sapnders, A. Cl968). A studJ i:n moppowet plonpipg. Unpublished · XA
thesis, Cranfield Business Schaal, Cranfield, Bngland.
SchDdtt, 1., Gooding, R.Z., Joe, R.A., & Kirsch, X. <1984>. leta-anal-
yses of validity studies published between 1964 and 1982 and the
investigation of study characteristics. Personnel pgyehglggJ, az, <3>,
40'7-422.
Simon, B.A. C1979>. The sciences of the artificial. 2nd Bd., ·
Cambridge, lass: The lit. Press.
Simon. H. A. <1987>. llaking management decisions. the Academy gf
lanosement ixecutiye, 1. · <2>, 57-64.
Sternberg, R.J. <1985>. Beyond IQ: A triarchic theory gf humon intelli-
~· Cambridge, Mass: cambridge University Press.
Stewart, A•• & Stewart, V. <1977>. Tpmorrgw•s Manager Tgda)'. London:
Institute of Personnel Management.
Taggart, W., & Robey. D.· <1981>. Minds and managers: On the dual nature
of human information processing and management. Academy pf Kanagement
Reyiew, d, <2>, 187-195.

'iernimcut. P. F., & Campbell, J. <1968>. Signs, samples, and criteria.

Jour~al qf ipplied ~sychology, 2a, 372-376.

'iijnberg, W. J. <1965>. Capacity and Income· Amsterdam, Helland: !he

Department of Research and Documentation of the General Employers•
Association.

42
 GO Society purpose and values statement GO Board Members

Michelle Malay Carter, USA

To support the organizing of work in a responsible, fair and healthy manner
in which people are led in a way that enables them to exercise their Barry Deane, Australia
capabilities. Don Fowke, GO Chairman, Canada
Azucena Gorbaran, Argentina
The Society believes this requires applying a systems framework* emerging
Jerry Gray, Canada
from reflective inquiry in which levels of work and capability are the initial
paradigm and growth in human awareness is the essential process. Jan Åke Karlsson, Sweden
Nancy Lee, USA
The benefits are organizational effectiveness, fulfilled people and
Ken Shepard, GO President, Canada
organizations designed for value-creation, sustainability and social well-
being. George Weber, GO Secretary
and Treasurer, Canada
Note: inspired by the work of Wilfred Brown and Elliott Jaques

GO Editorial Board
The Global Organization Design Society was founded in 2004 to establish Kenneth C. Craddock, M.P.A., M.A.
and operate a worldwide association of business users, consultants,
Décio Fabio, Portuguese Editor
and academics interested in science-based management to improve
Jerry L. Gray, Ph.D.
organizational effectiveness.
Owen Jacobs, Ph.D.
The GO Society fulfills its purpose by: Harald Solaas, Spanish Editor
• Promoting among existing users increased awareness, understanding and Ken Shepard, Ph.D
skilled knowledge in applying concepts of Levels of Work Complexity,
Web administrator
Levels of Human Capability, Accountability, and other concepts included
in Requisite Organization and/or Stratified Systems Theory. Javier Castro
[email protected]
• Promoting among potential users of the methods, appreciation of the
variety of uses and benefits of science-based management, and access Global Organization
Design Society
to resources. The GO Society supports the learning and development
32 Victor Avenue
of current and future practitioners by holding world conferences and
Toronto, ON, Canada M4K 1A8
professional development workshops, publishing books and a journal, and Phone: +1 317 644 0472
maintaining a resource-rich web site with related articles, monographs, Fax: +1-647-438-5656
books, videos, blogs, discussion groups, and surveys. Email: [email protected]
http://GlobalRO.org

GO Society Sponsors