ASSESSMENT CENTER METHODOLOGY WITHIN THE CENTRAL INTELLIGENCE AGENCY

• Document Type: FOIA
• Collection: FOIA Collection
• Document Number (FOIA) /ESDN (CREST): 06158506
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 57
• Document Creation Date: October 29, 2024
• Document Release Date: August 19, 2024
• Case Number: F-2013-00715
• Publication Date: June 20, 1979

Approved for Release: 2024/08/05 C06158506
��
ASSESSMENT CENTER METHODOLOGY ;
WITHIN THE
CENTRAL INTELLIGENCE AGENCY
20 June 1979
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
TABLE OF CONTENTS
SECTION PAGE
I. Introduction 1
II. Origins of AC Methodology
A. Traditional Selection Procedures 1
B. OSS Needs and Experiences 2
C. Principles of OSS AC Methodology 4
D. AC Methodology Today 6
III. Chronolog of AC Methodology in the Agency � � � 8
IV. Listing of Dimensions Identified and Measured 11
V. Commentary on Agency AC Activities to Date . . � � 13
��
VI. Criticality of Job Adhlysis to AC Validity. . . � � 17
VII. AC Activities of Other Federal Agencies 20
VIII. Extensions of AC Methodology 21
IX. Bases for the Growth of AC Applications 22
X. The Agency and Assessment Centers: An Overview
A. Potential Uses 24
B. Present Situation 26
XI. The Future of Assessment Centers in the Agency . 31
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
MEMORANDUM FOR: Charles A. Bohrer, M.D.
Director, Office of Medical Services
FROM: Bernard L. Mooney, Ph.D.
DCI Fellow
Center for the Study of Intelligence
SUBJECT: Assessment Center Methodology within the
Central Intelligence Agency
I. Introduction
The comments which follow are an response to your request
both for documentation of the recent history �f Assessment
Center (AC) methodology within the Central Intelligence
Agency as well as for first-hand perspectives on the implica-
tions of the events cited in the.chronolog. The comments
encompass:' a brief background on AC methodology; a synopsis
of AC activities in the Agency; and perspectives on the past,
present and possible future of AC methodology in the Agency.
Origins of AC Methodology
A. Traditional Selection Procedures
Practical applications of the scientific study of
individual differences (pioneered by the French psychometrician,
Binet, within his nation's educational system) were quickly
seized and improved upon by the United States Military at the
onset of World War I. It was American know-how which built
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
upon this base and developed the Army's "Alpha" and "Beta"
tests, so successfully used during World War I for scleeLing
and placement of military volunteers and draftees. These
beginnings led to tle Army General Classification Teqt
(AGCT), the Armed Forces Qualification Test (AFQT) and
finally, the Army Classification Battery (ACB). The tradi-
tion of measurirg s3ng1e traits (learning ability, mechanical
ability, etc.) was firmly engrained in the American personnel
evaluation style and particularly within the American military
tradition at the time of tbe outbreak.of World War II.
Thus, it is not surprising that elements of thA traditioo
influenced the Inittal pre-screening activities of the
Office of Strategic Services (OSS).
B. OSS Needs and Experiences
In late 1943, with OSS hardly one year old, it had
become painfully clear that the organization +acEd reJaLively
unique selection problems ... the problems ui selecting
individuals "best qualified" for assignments for which the
specific demand; were either totally unknown ox weie ginvxatyd
out of the fabrtc of arm-chair speculations. Tian, t)t
these shakily defined job specifications into qualifications
lists was not, at that time, handled either by professionals
in the field of selection nor was there any uniform selection
processing system in existence. Later, it was to be formally
noted (OSS Assessment Staff, 1948) that during October, 1943,
in a morning executive meeting with General Donovan, the idea
2
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
of forming a separate selection staff was proposed and was
positively received by all the involved staffs (specifically,
Recruiting, Planning and the Schools and Training Branch).
Before the end of November 1943, the nucleus of an Assessment
Staff was formed. (The linea] descendant of that Staff
remains with the Agency today as the Psychological Services
Staff, Office of Medical Services). Operation of the 3-day
selection program began before the end of 1943 at the Schools
and Training Branch site ... the former Willard family
(owners of the Willard Hotel in Washington, D.C.) country
estate-in-northern VirginAgt. referred to as Station "81.
("S" was synonymous for "Secret".) By war's enci; 5,391
candidates had been processed through Station "S" and
Station "W" (the latter was a second selection base opened
after Station "S" and which employed a 1-day selection program.)
Not the least among the catalysts which pushed OSS into
action in developing a systematized selection program was a
"behind-the-lines" operation which not only failed in its
objectives but also resulted in considerable loss of life to
the "team" dispatched on the mission. Dr. Donald W. MacKinnon,
a psychologist who joined the Assessment Staff dulinq its
second month of operation (and who rose to Chief of the
Station "S" Staff, remaining in that position to the end of
the war), has indicated* that an adapted version of the
(* personal conversation with Dr. MacKinnon, May 1974,
West Point, New York, International Congress on Assessment
Center Methodology)
3
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
unfortunate mission formed the basis of the plot to, tin
,
r
movie "The Dirty Dozen." The sel tion of candidate. lot
t
this particularly disastrous OSS .eration singled out
persons associated with Murder In4orporated. The rationale
for selection was simply: "It taKes a dirty mtz t.a
dirty job": (The operation involved behind-the lines.
assassinations of high-ranking military officers. with
failure of the operation and General Donovan's angry dis-
covery of the selection criteria used in structuring the
assigned team, the several staffs involved were provided
cleaimeth td .pi.Oreisni.ialize and systematixe selection
processing for the future. Unfortunately, the'iesolve
forged out of the OSS experience was forgotten. Some 17
years later, the Agency found itself once more involved with
the "Dirty Job--Dirty Man" equation--again with negative
consequences fcr the Agency.*
C. Principles of OSS AC Methodology
The fact that the criterion data (the factual
information abcut specific demands for each and e%.ery possible
assignment/missionl were inaccessible coupled it1 .14 tect
that, at time cf evaluation, no candidate wa,. untuf ,t....ideta-
tion for a pre-determined assignment combined to reshape the
thinking of the ASEeSSMent Staff regarding feasible approaches.
The Staff concluder that their efforts would be best spent in
(* "Alleged Assasination Plots Involving Foreign Leaders,"
Interim Report pf the Church Committee dated 20 November 1975,
pages 74-86)
4
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
attempting: to describe the competencies of the individual
candidate in ascomplete a manner as possible yet in terms of .
dimensions of behavior having relevance to all possible
assignments; and to place particular emphasis upon the
assets of the individual candidate as these suggested potential
for effective functioning under specified conditions.
In its effort to describe the personality of the indi-
vidual candidate as a totality, the Staff identified seven
(7) major Factors or Variables and three (3) special Factors
which were combined into the following General Qualifications
4 .
List for all OSS men and women candidates:
1. Motivation for the Assignment
2. Energy and Initiative
3. Effective Intelligence
4. Emotional Stability
5. Social Relations
6. Leadership
7. Security (caution/discretion)
8. Physical Ability
9. Observing and Reporting Skills
10. Propaganda Skills
With the Factors to be measured thus defined, the Staff
set as its next task the development of assessment procedures
designed to elicit the Factors chosen. Two constraints were
applied in developing the assesment procedures and these
represented a true breakthrough in the area of personnel
evaluations. Further, these constraints have become the
cornerstones for every successful Assessment Center Methodology
in existence today. The first constraint was that each
5
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
factor would be meabured by several distinct assessment
procedures. The second constraint was that each assessment
procedure would be designed to sample the Factors under
conditions as similar to the real-life criterion situation
as possible (i.e., the use of simulations). An average ol
six (6) differert simulations/tasks were used for each of
the ten (10) Factors and these included individual as well
as group tasks. (Psychometric test devices used along with
the simulations wen. included in the average of six measures
per Factor.)
. � � ,
Scoring and analysis"-of psychometric measures proceeded
along standard lines. Performances in the simulations were
evaluated by trained observers (all professionals in the
behavioral sciences), usually three in number. Integration
of all the data gathered took place in an evaluation session
attended by all observers after close of the 3-day evaluation
program. After presentation of all observers' comments,
consensus among observers was reached via application of a
6-point rating scal2 indicating the strength of earh ol the
ten Factors in each individual case. Finally, the qamo 6-
point scale was applied in defining an overall Factoi in
each individual case which was referred to as "Job Fitness."
D. AC Methodology Today
After one abortive attempt to translate the OSS
techniques into a practical tool for selection of sales
6
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
personnel within the structure of Macy's in 1948, nothing
further was heard of the methodology until in 1956,
Dr. Douglas W. Bray applied the techniques, at first in an
experimental program, for identifying managerial potential
among employees within AT&T. Dr. Bray saw particular
utility in applications of the techniques to the problem of
managerial selection because: in its initial stages the
technique makes no assumptions about the "ideal" pattern of
assets among managerial candidates; and, it permits observers
to develop perspectives -about the ways in which candidates
are likely to handle the problems of' new job.4emands
(demands of management) by studying candidates' actual
handling of realistic simulations of these "new" demands.
It was perhaps for this latter reason that Dr. Bray saw AC
methodology as particularly appropriate for individuals on
the verge of moving into first line management positions.
The AC metholology of Dr. Bray (in fact, of all AC activities
today) has remained, point for point, true to the OSS prototype
with but one exception. Today, the observers used during
the operation of an AC are typically experienced, knowledgeable
employees who are specifically trained in the observational
techniques and operational procedures of their organization's
AC. In AC's designed for identification of managerial
potential, the observers are usually former incumbents of
7
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
the positions against which the candidates are beina
measured ... former incumbents now one or two steps abot.e the
position under study. The chronology which follows
documents the efforts made in introducing the AT&T adapta-
tion of AC Methodology to this agency.
III. Chronolog of AC Methodology in the Agency
TIME FRAME ACTIVITY
A. 1968
B. 1968
C. 1974
D. 1974
E. 1974
(b)(3)
PSS) (D)(3)
attended a Seminar in which Dr. Bray (of AT&T)
reported on his then current findings of AC
results and performance of managers
approaching mid-career with AT&T.
and Mooney developed a series
of AC tasks which were employed in the
evaluation of on-board Carder Trainees.
(Findings indicated the techniques were
particularly useful in estimating
"career stability.")
The Behavioral and Social Sciences
Committee of OMS was queried by the then
DD/M&S: "What would you recommend as an
approach to improving overall efficiency
of Agency management function"? Using
elements of a paper produced by
Dr. Mooney ("Assessment Centers: Whys
and Wherefores")(Attachment
Dr. Tietjen (D/MS) proposed the concept
of using AC's for identification of
managerial potential.
Cont1cts/br3efings with fl/OS (H. Obt )tx)
and D/OL (J. Blake) regarding AC metho-
dology ... no interest.
Contact/briefing with D/OJCS (H. Fitzwater).
Agreement reached on Management Develop-
ment Center within OJCS focussed on
potential for functioning as Branch Chief
(GS-13 level) within Office.
(DD/M&S, H. Brownman, actively supported
D/OJCS in deciding to use AC).
8
(b)(3)
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
F.
1974
G.
1975
H.
1975
I.
1976
J.
1976
R.
1977
L.
1977
M.
1977
N.
1977
Contact/briefing with DDI (Dr. E. Proctor)
and ADDI (P. Walsh) ... no interest.
Two operations of OJCS Management Develop-
(b)(3)
ment Center. Observer/Managers trained;
candidates evaluated with separate (b)(3)
developmental profiles generated for each.
Contact/briefing with
Agreement reached on Assessment Center
focussed on identification of potential
for functioning within GS-12 level positions
viz., Chief of Base, Chief of Operations
and Chief of Engineering.
Management Development Center of ODP
(formerly OJCS) cancelled by D/ODP
Reason for cancellation was negative impact
on Office of drain of critical manpower
resoucces needed to operate Center (bsIcal
3 Observer/Managers per Cenper). Mr.
recommends DDA consider Directorate-wide
for selection of managers.
Two operations of OC AC. Observer/
Managers trained; candidates evaluated
with separate evaluation reports produced for
each.
Contact/briefing of R Career Service Board
(SST) on uses of AC's. Board recommends
consideration of AC applications within FBIS.
Contact/briefing of D/FBIS on possible use
of AC's. Paper prepared and forwarded to rills
detailing job analyses required before design
of Center. (No response/reply was ever
received from FBIS.)
� 4
(b)(3)
(b)(3)
(b)(3)
(b)(3)
(b)(3)
Request from DD/OC for design (b)(3)
of new AC focussed upon evaluation of poten-
tial for functioning at GS-14 level under
the so-called "Panel O.
Request for design of multi-focus, Directorate-
wide AC for use in DDA Upward Mobility Program
... Project AIM under DDA/EE00 (C. Jones).
9
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
(b)(3)
O.
1977
Four operations of "Panel 0" Ceptvl.
(b)(3)
Observer/Managers trained; candi -
dates evaluated with separate evaluation
reports prepared for each.
(b)(3) -
P.
1.977
of Project AIM Center.
rPour_7perations
Observer/Managers trained;
(b)(3)
candidates evaluated with s epalat% vv.tlua-
tion reports prepared for each.
Q.
1976
Cancellation of GS-12 and GS-14 ("Panel 0")
level Centers by OC. Reason for cancella-
tion was negative impact on Office of drain
of critical manpower resources needed to
operate Centers (basically, 3 observer/
managers per Center). Discussion of possible
Center for GS-09 trainees. No commatments.
R.
1978
Two operations of Project AIM Centers.
candidates evaluated with separate
(b)(3)
evaluation reports prepared, for each.
S.
1979
One operation of Project AIM Center during
May 1979; candidates evaluated with
(b)(3)
separate evaluation reports for each.
Not detailed here are the regular presentations on the
topic of AC Methodology offered by the undersigned to partici-
pants in every runring of the OTR sponsored "Management Seminar"
and "Senior Serrinar" from late 1975 through early 1978. No
effective inquiries regarding AC Methodology ever emvx,:od from
these didactic efforts so that they were terminated. i'thel
formal didactic channels employed includc an .11 t It' :.� ' � 1..)
by the undersigned which was published in the April 1911; edition
of the DDA Exchange (Attachment B).
At least as critical as a chronolog of Agency AC atilVltleS
is a listing of what AC Methodology has uncovered regarding the
behavioral dimensions determined to be critical for sur-eeqs
across the range of positions studied and analyzed.
10
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
These dimensions are presented below in Section IV (with
detailed definition of the dimensions provided in attachment C).
IV. Listing of Dimensions Identified and Measured
ODP (GS-13) OC (GS-12) OC (GS-14) PROJECT AIM
DIMENSIONS DIMENSIONS DIMENSIONS DIMENSIONS
A. AWARENESS OF (A.)
DETAIL
B.
C. CLARITY OF ORAL
COMMUNICATION
D. CLARITY OF WRITTEN (D.)
COMMUNICATION
E. �REATIVITY
F. DECISIVENESS
G. DELEGATION (G.)
CAREER AMBITION
H. DEVELOPMENT OF
SUBORDINATES
I. EFFECTIVE
COMMUNICATION
J. ENERGY (J.)
K. FACILITATES GROUP
PROCESSES
L. FLEXIBILITY
M. FORESIGHT*
N.
IMPACT
0. INITIATIVE
P.
LEADERSHIP (P.)
O. LISTENING
ABILITY
(*Dimension M., FORESIGHT: appears only in the ODP Center. Was
later more appropriately labelled "PLANNING" ability and collapsed
into X., PLANNING AND ORGANIZING.)
11
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
IV.
Listing of Dimensions Identified and Measured (continue(1)
ODP (GS-13)
OC (GS-12)
OC (GS-14) PROJLCT AIM
DIMENSIONS
DIMENSIONS
DIMENSIONS DIMENSIONS
R.
MANAGEMENT
CONTROL
S.
MOTIVATION
FOR WORK
T.
ORGANIZATIONAL**
ABILITY (as a
manager)
U.
ORGANIZATIONAL**
ABILITY (as a
person)
V.
..PERCEPTIVITY_AND
(V.)
(V.)
SENSITIVITY (re-
garding people)
:1
W.
PERSUASIVENESS
X.
PLANNING AND
(X.) (X.)
ORGANIZING
Y.
PROBLEM ANALYSIS
AND JUDGMENT
Z.
RISK-TAKING
AA.
SOCIAL
ADAPTABILITY
BE.
STRESS TOIERANCE
(BB.)
CC.
TENACITY
("Dimensions T.&U., ORGANIZATIONAL ABILITY "as a manager/as a
person" appear only in the ODP Center. Were later collapsed into
the single dimension labelled "PLANNING AND ORGANIZING" ability ...
referencing only the domain of managerial behaviors and
characteristic!..)
12
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
V. Commentary on Agency AC Activities to Date
Though not immediately obvious, there has been an evolution
in AC design characteristics from the early ODP Center to the
present. In the case of the OD? Center, the design of the
Center was focussed upon behavioral dimensions identified as
critical for success within one specific position title,
namely the position of Branch Chief. The next Center (The
CC GS-12 Center) was designed to sample behavioral dimensions
identified as critical for success within three distinct but
related positions (Chief of Base, Chief of Operations, Chief
of Engineering). Next came the Project AIM Center designed
to sample behavioral dimensions which, in varying combinations,
were identified as critical for success across some eight or
nine distinct and unrelated positions. Finally, the CC GS-14
Center was designed to sample behavioral dimensions identified
as critical to success at senior managerial levels within the
Office, regardless of position title.
Common to all Centers has been the painstaking Job
Analysis phase with subsequent identification and definition
of behavior dimensions regarded as critical for success. In
addition, every Center has retained the goal of presenting
management with a reliable "profile" for each individual,
describing each person's assets and deficiencies in terms
which are relevant to the job(s) under consideration. Thus,
no individual is presented as having "Passed" or "Failed"
13
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
the Center. Ea.:11 person is presented according to his(h(x
unique pattern of capabilities apart from any real ox
"cutoff score" 3ystem. In this fashion, AC's do not provide
decision-makers wit n a cut-and-dried decision regardino the
person. Rather AC's supply job relevant data organized in
such a fashion as to permit better understanding of the
employee in questiol as well as to permit comparisons among
several employees on behavioral dimensions critical to
success in the position under consideration.
The evolution in the design of Agency Centers reflects
the Impact of growing experience with the Meth4dology. In
other words, the range of positions included within a Center
increased both in scope as well as level following the
principle of: "...proceeding from the better known to the
less well known...." Each successive Center Design Phase
used the behavioral dimension data of previous Centers for
purifying and refining definitions of new dimensions.
Proceeding in tlis fashion, Agency AC's avoided the pitfalls
which have bese: other Centers namely, developing definxtions
of behaviorcl dImer.lon: so broad in nature that any wrhot
of performance tasks in a Center might be argued as suitable
measurement teclniqaes. For example, with a dimension called
LEADERSHIP defiled as: "effectively directs others," what is
an adequate behavior sample? It may be sufficient to infer
LEADERSHIP from the manner in which a person respond9 to
14
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
paper-and-pencil exercises such as the classic In-Basket
Task or it may be sufficient to study the person's behavior
in a group situation where he/she has the opportunity to
influence the group. Perhaps both samples of behavior are
necessary...and one type of behavior should be given greater
weight than the other in arriving at an overall estimate of
LEADERSHIP. A detailed job-analysis during the Center
Design Phase can lead to a more precise definition of LEADERSHIP
such as: is effective in getting his/her ideas accepted by
others; is effective in guiding a group or an individual
toward accomplishment of iveguired tasks." (This is the
actual definition of the dimension of LEADERSHIP which
emerged from the CC Center Design Phase.) In this instance,
it is clear that the behaviors to be evaluated must be
sampled in both a group as well as a one-to-one situation
and must involve a focus upon a task assigned to the person
which he/she accepts as a task to be accomplished.
The behavioral dimensions thus far identified through
the several Center job analysis phases, exhibit interesting
communalities. For example, two of the dimensions
(D., CLARITY OF WRITTEN COMMUNICATION and J., ENERGY) surface
as critical across all positions analyzed. (If one combines
dimensions M., T. and U. of the ODP Center into dimension X.,
PLANNING and ORGANIZING, there are actually three dimensions
common to all positions analyzed.)
15
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
Considering only the managerial positions studied, two
dimensions emerge as common viz., G., DELEGATION and V.,
PERCEPTIVITY and SENSITIVITY.
At the same time, behavioral dimensions unique to each
Center have been identified. For the ODP Center: AWARENESS
OF DETAIL, LISTENING ABILITY, PERSUASIVENESS AND TENACITY.
For the OC (GS-12) Center: CREATIVITY, IMPACT and MANAGEMENT
CONTROL. For the OC (GS-14) Center: DEVELOPMENT OF SUBORDINATES,
INITIATIVE, PROBLEM ANALYSIS AND JUDGMENT and RISKTAKING.
For the Project AIM Center: CAREER AMBITION, MOTIVATION FOR
WORK and SOCIAL ADAPTABILfTY.
Do the present 26 dimensions "capture" the essence of
the behavioral dimenions underlying most Agency positions?
Probably not: Job analysis data gathered in support of the
OC (GS-14) Center suggested that several unique behavioral
dimensions are associated with successful functioning
within OC (GS-14) positions in the Overseas environment as
Opposed to the Headquarters environment. (Unfortunately,
the small numbers t.f respondents vis-a-vis the Overmed:.
environment did no' permit reliable defInitIon o4 Iflo,t tol
dimensions.)
Thus, even with positions considered near-identical,
distinctions between overseas and headquarters were aerom-
panied by variations in behavioral demands. Furthermoxe,
there exist critical contextual variations even within
identically latelled dimensions when these dimensions are
16
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
applied to positions at different levels within an organization.
For example, dimension D., CLARITY OF WRITTEN COMMUNICATION
was both measured and evaluated for Project AIM candidates
in a context considerably different from that for OC Center
candidates or ODP candidates. In other words, even a "common"
dimension may require AC methodologies unique to the job
behavior context within which the dimension is being studied.
Progress has been made toward the goal of developing a
"catalog" of behavioral dimensions underlying Agency position
titles but obviously much remains to be done.
VI. Criticality of Job Arialysis to AC Validity,
The foregoing Agency AC experience underlines the fact
that identically named position titles cannot be assumed to
represent ,identical behavioral demands upon incumbents
unless and until behavioral job analyses actually establish
similarity. Also, identically named behavioral dimensions
cannot be assumed to be validly measured by one and the same
AC performance task unless and until behavioral job analyses
establish the similarity of job demands.
Is it then impossible to-design an effective AC to
measure potential for successful functioning in positions
generically and simply defined as, for example, first line
management or middle-management or executive level management?
The answer is that it is not impossible provided only that
one is committed to carry out the time-consuming, laborious
17
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
but necessary behivioral analysis of the demands of all
positions to be included within the AC.
"Arm-chair" analyses do not suffice. Beginninq In
1973, the Civil Service Commission in collaboration with the
Office of Manxgement and Budget commissioned the deblyn of
an AC to identify "executive-generalist potentialm(within
the Federal Service) as an element for selection into the
Federal Executive Development Program (FEDP) ... a program
intended to faster development of executive potential among
highly-promising Federal Employees at the GS-15 level.
Dimensions of behavior to be measured were identified via
"arm-chair" rather than empirical methodology. In all, 12
dimensions of behavior were determined to be critical for
successful executive-level ,functioning within the Federal
bureaucracy. Early published research findings underlined
the fact that the FEDP AC findings did significantly influence
the final decisions regarding acceptance into the program.*
Recently published follow on research, however, reveals that
only one of the "arm-chair" dimensions correlates .;i(nifiyantly
with on-the-job ratings of performance assigned to thf
selectees. In addition, it has been determined that / of
the 12 dimensions originally measured in the FEDP Center
actually represert two distinct types of on-the-job behaviors
each for a tctal of 14 job-related behaviors. It is not
(*"An Overview of the Federal Executive Development Program Ii
Assessment Center," Civil Service Commission, August 1q76
(PR 261-705))
18
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
surprising that none of these 7 FEDP Center measures have
proved predictive of any of the 14 on-the-job performance
ratings. Few, if any, of the FEDP Center tasks actually
bore any similarity to the real-life, on-the-job performance
tasks which the FEDP selectees now face.
More will be added later (Section 9.) about the Job
Analysis Phase of AC development. For now, it should be
pointed out that job analysis is not a unique characteristic
of the AC methodology. In point of fact, it should be a
prelude to any attempt to employ behavioral measures (whether
these be AC performancetasks or psychometric devices) to
predict on-the-job performance. The job analisis data which
do exist within the Agency are of the type generated by
Position Management and Cowensation Division (PMCD) of the
Office of Personnel. While PMCD data are of important use
in projecting pay-scales against job responsibilities, the
data are couched in such generalities that they cannot be
used to support efforts such as AC task design. Thus, the
overall perspective regarding adequate job analysis data
(adequate for behavioral science use) within the Agency is
hardly promising. A potential side-benefit of the year-
long DCI Fellowship Project of the undersigned may be
the opportunity to establish a base-line of behavioral
data regarding managerial functions at the executive
level within the Agency. Such a side-benefit would
represent an important advance not just for AC methodology
19
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
but for all behavioral science efforts directed toward..
early identification of employee potential for operating
effectively in positions beyond present incumbency levels.
VII.
TIME
AC Activities of Other Federal Agencies
FRAME ORGANIZATION TYPE(S) OF CENTER(S)
A.
1969*.
Internal Revenue
Service
Identification of potential
for first-level Supervisor.
B.
1973*
Civil Service
Commission &
Office of Manage-
ment and Budget
Identification of executive-
generalist potential for
selection into the Federal
Executive Development Program.
C.
1973*
Federal Aviation
Administration
Three separate Centers for
identification of supervisory,
middle-management and senior
management potential.
D.
1973*
Social Security
Center to select candidates for
Administration
2-year Management Intern Program;
Center to identify developmental
needs of upper-middle managers
who are enrolled in SSA's
Executive Fellow Program. '
E.
1974*
Housing and Urban
Development
Identification of potential for
first-level supervisory positions.
F.
(?)*
Department of Army
A-succession of Center designs
to identify leadership potential
among commissioned officeis.
G.
1977*
redkral Bureau of
Inv..stigation
By regulation, all caneldateq
for first-line ,3,..pe I V 11,01 411
evaluated in an A( de.;icinvd
assess supervisory potential.
Plans are underway for a second
Center to identify potential
for functioning at executive
levels.
(* Centers for identification of supervisory/managerial/execu-
tive potential.)
20
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
VII. AC Activities of Other Federal Agencies (continued)
TIME FRAME
H. (?)**
I. 1973**
ORGANIZATION
Department of Air
Force
Law Enforcement
Assistance
Administration
J. 1975** Bureau of
Engraving and
Printing
K. 1977**
Equal Employment
Opportunity:.
Commission
L. 1979** Department of
State
TYPE(S) OF CENTER(S)
Center to identify scientist
potential for contributing to
the mission of an R&D Command.
A series of Centers funded by
LEAA for use by regional and
local agencies in selection
of sergeants, captains and
detectives.
Center for the selection of
candidates in the Bureau's
Upward Mobility Program.
tenter for selection of candi-
dates for specialized staff
aesignments.%
�
Center (under development) for
selection of candidates for
Foreign Service Officer.
The foregoing list is ndt to be considered inclusive of
all AC activities of other Federal agencies. The list has
been derived from a review of printed materials regarding
AC's made available by the named agencies as well as from
personal information obtained by the undersigned in contacts
with private and federal groups.
VIII. 'Extensions of AC Methodology
Inasmuch as the core element of AC's is behavior known
to be associated with successful performance within a defined
realm of work activities, the extended applications of AC's
(** Centers for identification of potential for specialist
positions.)
21
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
have found great acceptance among certifying and licen..inu
agencies. Today, there exist AC's for certifying competencieh
within the field of Eddcation. (The State of Wisconsin will
accept AC findings in lieu of grade transcripts for accrediting
teachers within the State.) Within the State of New York,
the State Psychological Association requires AC finding!. for
purposes of licensing Psychologists for practice within the
state. The American Psychological Association is presently
evaluating the design of an AC which the Association hopes
to use in awarding the "Diplomate" in Psychology ... a
certification of professional excellence at the' National
level. Exploratory efforts are underway by several medical
and medical-related organizations regarding the utility of
Centers for evaluating candidtes for licensing within
certain medical specialties and within the general fields of
pharmacology and nursing.
IX. Bases for the Growth of AC Applications
With the Ecual Employment Opportunity Commission (1E0C)
proscriptions regarding the necessary characteristic.- 04 twlec-
tion techniques, the simplicity of AC methodology et u o,
appeal. Current EEOC criteria require "Content Validity" of
the selection methodology. In other words, the behaviors
measured by the selection processing techniques must be
demonstrably sinilar in content to the behaviors demanded by
the position for which the person is being considered. Since
22
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
all Centers are preceded by a job analysis phase dnd hIncv
Center performance tasks include elements of actual job
behaviors, AC's are, "de facto," in compliance with the EEOC -
defined criteria. An incidental benefit of the close rela-
tionship between actual job behaviors and AC performance demands
is the aspect of "Face Validity." In other words, AC candidates,
perceiving the obvious relationship between Center and job
behaviors, tend to be both less resistant to as well as to
be more positively oriented towards evaluation via AC
methodology.
Of great impbrtance are the EEOC criteria regarding
possibilities of "adverseampact" associated with selection
procedures. An exhaustive research effort conducted within
the Michigan Bell Telephone System-of AT&T* has established
that, when the Standards for 'Assessment Center Operations**
are followed, there exists no evidence to support the conten-
tion that AC results exhibit adverse impact vis-a-vis Black
or Female Minority Group members. While presumption (grounded
in past differential validity research) supports AC methodology
as free of adverse impact, nevertheless, as the Standards
indicate, it is incumbent upon .each Center user to maintain
adequate documentation of the Center process.
A unique characteristic of AC methodology, enjoyed by
no other selection procedure is its inclusion in a consent
(* "Determinants of Assessment Center Ratings for White and
Black Females and the Relationship of these Dimensions to Sub-
sequent Performance Effectiveness," Ph.D. Dissertation,
Huck, James R., Wayne State University, 1974.)
("" "Applying the Assessment Center Method," Moses, J. L. and
Byham, Wm. C., Pergamon Press, New York, 1977. Appendix p. 303 ff.)
23
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
agreement arrived at through the courts between the tick aNd
AT&T in the early 1970s. As the result of a class at.tion
suit brought agiinst AT&T in behalf of 1700 women emploiees,
claiming discriminatory exclusion from managerial positions,
AT&T was directed by the courts to design and operate Absess-
ment Centers for these complainants in evaluating their
potential for advancement to managerial level positions.
While many other selection procedures struggle to establish
evidence of "fairnessTM, Assessment Center methodology stands
alone as the only court-ordered/EEOC approved selection
procedure in the United ttates today. The EEOC has itself
used AC Methodology in the selection of candidates for
placement in its 22Iistrict Director positions.
X. The Agency and Assessment Centers; An Overview
A. Potential Uses
The original promise of AC Methodology still holds.
Introduced to the Agency as a technique particularly well-
suited (though not limited) to the task of identilying
managerial potential, AC's expand the range of activitios ot
behavioral sclentirfts to include focus upon the tq, f'. nq
and structure of management within the Agency. Lxp.1 nt;
the role of behavioral scientists in the Agency at a time
when human resources to support such expansion are limited
is made feasible by the key operating characteristic:, of Centers.
24
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
In other words, while Center Methodology permits expansion
of the behavioral scientists' roles in the Agency, it also
promises to accomplish their current requirements in a more
cost-effective manner. As one example, a well-designed Center
staffed by Observers drawn from either or both the Career
Training Staff and the D Career Service/Career Management
Staff (under the monitorship of one Psychologist) when
operating on a once a week basis, could essentially meet the
annual DDO-applicant evaluation processing requirements for
the Career Training Program. (A fully operational Center can
be anticipated to extend the Psychologist's assessment ser-
vices by a factor of 3 to 4.)
Note the critical phrase "fully operational Center" in
the last sentence above. Only the Project AIM Center has
provided the opportunity to document the behavioral science man-
power savings of Centers. In the May 1979 AIM Center operation,
a well-trained non-psychologist selected and combined the task
elements which constituted the Center, designed the center schedule,
made all the necessary administrative arrangements for the Center,
conducted a "refresher training" course for the Observers,
directed the operation of the Center and conducted the post-
Center Evaluation Session. Only because no specific training
has been provided in the areas of Center Report Writing and
Center Feedback to participants, these two tasks are currently
handled by Psychologists. Given the opportunity to provide
25
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
training in these two task areas, it is anticipated the
Project AIM CeLter may eventually operate under the dliec-
tion of the non-psychologist Center Administrator with a
Psychologist available to the Administrator on an as-needed,
standby basis. The Psychologist, of course, retains the
responsibility regarding continuing validation of the Center
(as detailed in Attachment A., Page 9 ). AC Methodology is
capable of both meeting (and perhaps even exceeding) current
evaluation requirements while, simultaneously conserving
behavioral science resources for investment/involvement in
developing and offering new services in new areas within the
�
Agency. (For actual Center "costs," see Appendix 1:0).
B. Present Situation
In the instances of the Centers designed to identify
Office-specific managerial potential, it has been Center-
related manpower requirements which have formed the bases
for Center can.7ellations. Both ODP and CC stated that the
manpower drain (resulting from the use of high-level Office
managers as Centex Observers) seriousl} hampered the operating
efficiency of the tespective Ufftreb thu, tequirtn,. -aticvnatton
of their Centers. (This has not proved a critical problem
with the Project AIM Center since each Observer required is
drawn from a separate DDA Office.) A direct solution in
reducing the manpower drain was proposed to and rejected by
26
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
both ODP and OC during Summer, 1976 namely, Observer Teams.
It was proposed that managers from both Offices be cross-
trained in the two Centers and assigned in Teams to a given
Center so that no single Office would have to "bear the brunt"
of the Center manpower demands. (Though it was then too
early in the OC Center design phase to consider, it later
became apparent that some melding of the ODP and OC Centers
might have proved feasible ... providing a single Center
staffed by Observers from both Offices and dedicated to
processing candidates from both Offices as co-participants.)
Rejection of the Observer Teams idea rested upon the manage-
ment conviction that Observers drawn from one Career Sub-
group were not sufficiently knowledgeable to be able to produce
valid and useful evaluations of candidates from another Career
Subgroup. In addition, there was important concern over the
implications of permitting Careerists from one Career Sub-
group (as Center Observers) to make input to decisions
affecting the treatment/handling of members of another Career
Subgroup.
The foregoing examples of inter-Career Subgroup stereo-
typing as "not sufficiently knowledgeable about us and our
operations" draws attention to a pragmatic basis for insisting
that Observer manpower be derived from the 'consumer' components
rather than from a behavioral scientist resource pool.
27
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
Career Subgroups look to behavioral scientists for ":0-teonim"
individuals and perceive themselveq as sole resource lot
"assessing" individuals. Evaluations conducted by trained
local managers Ere more likely to be accepted and actnd upon
by the Office. However, when the attempt is made (through
AC Methodology) to systematize and formalize the Subgroups'
assessment activities and to give Subgroups direct control
over and responsibility for assessment and its consequences,
the responses of Subgroups are not characteristically
enthusiastic.
�Op
��
These seemingly conflicted tendencies of subgroups to
endorse sole control of their Centers and yet to insist that
they cannot supply the requisite manpower can be better
appreciated through considering the operating characteristics
of Centers (specifically those Centers designed to foster
career development and/or to assess potential for advance-
ment to or withLn managerial ranks.)
Unlike standard assessment activities and psychomntiic
testing, Assessnent Centers are not a "one shot afiati"
representing a data source coming Iron somewheie iht
chain-of-command. A Center is very carefully implanted
within the very heart of an organization's personnel manage-
ment policies and practices. In order to merit continuance,
28
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
the Center must not only "feed into" the local personnel
system ... it must also be "fed by" the same system. It is
typically during the 'implantation' phase that the consumer
Office comes face-to-face with short-falls in its local
personnel management system. Concrete examples of such
oshort-falls" have surfaced repeatedly even during the short
history of Centers in the Agency.
Faced with the decision whether to permit employees to
"volunteer" for a Center or to require employees to be
onominated� by supervisors, one consumer Office opted for
supervisor nomination. It was quickly determined that
supervisors had never been asked to evaluate employees in
terms of their management pptential and hence there existed
no standardized system for nominating candidates. In another
instance, a consumer Office concluded that it would be
highly desirable to "track"the progress of employees in
their efforts to correct deficiencies surfaced by the Center
evaluation. It was felt that important data regarding the
individual's career ambition could be gathered by documentino
what efforts the individual put forth to increase his/her
overall readiness for advancement. However, no formal
mechanism existed to accomplish this purpose. (It was proposed
that specific goals regarding self-improvement could be included
in the person's Letter-of-Instruction ... but the proposal was
29
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
rejected since mana.4ement felt the LOI should be liwttpd to
"job-related activities".)
Even more critical than the instances cited above was
the situation where a consumer Office found itself unable to
come to grips with the reality implications of the substandard
performer in its Center ... the Center participant whose Center
evaluation reveals major deficiences in all areas deemed cri-
tical for advarcement. While the consumer Office agreed that
such a participant was destined to advance no further in the
Office" it quickly became apparent that the Office possessed
.�
no mechanism for dealing with the participant. for designing
a program to maximize use of the participant's skills at
his/her then current level or. if necessary, for assisting
the participant to transfer to another area of the Agency
where his/her present skills would be in demand.
In essence then,. while operation of the Centers themselves
do place important demands upon the manpower resources of the
consumer Office, the total manpower impact of Centers In the
Agency has ben tn point up additional needs in lora:
personnel management systems. While it may be x( utit 01 e
that ODP and 0C chose to cancel their Centers, yet at the
same time, these were prudent decisions. Granted these Offices
found it impolsible to support a personnel management system
at their "locil" levels which would fully utilize the benefits
30
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
of a Center, far better that the Office cancel the Center
than to maintain it as only a screening tool in promotion
decisions.
Though the ODP and OC Centers no longer are in operation,
residual benefits remain for the consumer Offices. Each now
possesses a small cadre of managers whose capacities for
evaluating subordinates have (according to the testimony of
the managers themselves) been altered and improved. Most
importantly, the Offices now possess behaviorally-anchored
descriptions/definitions of performance characteristics of
successful managers (at given levels within their organiza-
tions) which represent the consensus of Office-wide manage-
ment. Such definitions, if regarded as criteria for advance-
ment into or through managerial levels, can contribute much to
the Offices in systematizing and expanding their perspectives
in regard to the career potential of their employees.
XI. The Future of Assessment Centers in the Agency
Of the three Centers (one in ODP and two in OC) designed
within the Agency as career management/development tools, none
are in operation today. The explanations offered at time of
cancellation by each of the Offices are identical " ... opera-
tion of the Centers places a manpower requirement upon the
Office which the Office cannot support while simultaneously
maintaining its efficiency in accomplishing its stated
missions and goals ...." It is quite accurate to say that
31
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
had the Offices restructured and expanded their respective
personnel management systems so as to offer a "full service"
system to their employees (while using Center results to the
opt2mum), thr Officrs would indeed have been incapacitated
by the manpower requirements of so doing. Though Offices
strive to perpetuate their local practices and controls over
employees in thcir human resource management systems, Offices
are not adequately staffed to accomplish this effectively.
In essence, the experience with AC Methodology in the
Agency has brought into clear focus the cost of the "trade-
offs" involved In the decentralized personnel management
System. Only a strong, centralized system integrated with
the style, needs and realities of line management is capable
of supporting and fully utilizing AC Methodology to the
greatest benefits of management, the employee and the Agency.
While the term "personnel management system" has been used,
the system referenced must not be defined in the restrictive
sense of Office of Personnel functions alone. The centralized
system describe .1 must also include resources for employee
training an(. dceolot.,ent (also reaching jrAL the dc .1. � d
Office of Training functions.) A system of the type described
has been proposed ("The CIA Personnel Management System,"
NAPA Report dated 13 March 1979, pp. 91 ff.) which, at least
conceptually, holds promise of avoiding the "shortfalls" of
local personnel management systems (particularly the shortfalls
in manpower resources.)
32
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
The foregoing reference to the NAPA Team Recommendations
is not intended as a self-serving proposal to accomplish
"implantation" of. AC Methodology in the Agency at any cost.
It is, first, an observation that the type of system
recommended by the NAPA Team regarding executive development
answers the very problems which have proved the bane of
existence/continuance of Centers in the Agency thus far.
Second, it is an observation (based on experience of the under-
signed) that AC Methodology provides a central focus for and
a systematic approach to defining the essential elements of
_
any effectiveApersonnelloanagementidevelopmeng prpgram.
_
_ - %
wy.
Eyery Center consumer has come away from -the experience
with a alearer appreciation for: what types of behaviors are
demanded by.gpecific managerial positions; Wait the assets
and deficiencies of the present pool of candidates for
appointment to those positions are; what steps can be taken
now (in a training/development mode) to bring the pool of
candidates up to a level more compatible with present (and
even future) behavioral demands of specific manaaerial posi-
tions. These same benefit, can certainly stand on their own
merits in the type of executive development program proposed
by the NAPA Team.
33
_
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
Over and above this fact, since the executive development
program would oe starting "de novo", the data collection
mechanisms of AC Methodology represent an invaluable
resource for continual monitoring and evaluation of the
program.
Bernard L. Mooney, Ph.0
20 June 1979
s �
-
34
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
APPENDICES
A. "Assessment Centers: Whys and Wherefores"
(paper dated May, 1974).
B. "More About Assessment Centers"
(DDA Exchange article dated April, 1978).
C. Behavioral Definitions of Agency Assessment
Center Dimensions
D. Recorded Costs for Office of Communication �
Panel "0" Assessment Center (Fall, 1977).
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
APPENDIX A
Paper Entitled:
"Assessment Centers: Whys and Wherefores"
May 1974
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
What is an Assessment Center?
It is a set of procedures, rather than a place...a set
of procedures developed jointly by management and behavioral
science which are applied by management in identifying
managerial potential among employees. Originating within
OSS for evaluation of intelligence operations potential, the
original concept has been expanded and developed for in-
dustrial/business applications, largely through the efforts
of psychologists (Drs. Bray and Grant) employed by American
Telephone and Telegraph within the Bell System subdivision.
Beginning as early as 1956, Bray and Grant, working alongside
Bell managers, devised a series of situational, job-related
problems which were presented to candidates for advancement
into or within the Bell managerial structure. Systematic
observations of the candidates' behaviors in the face of
these job-related problems were recorded and evaluated as to
their efficiency, originality and utility. These behavioral
evaluations or ratings have since been studied against the
criterion of the given candidates progress through the pro-
-motional structure of the Bell System.
Reports of the initial findings regarding the success
of Assessment Center procedures in predicting future managerial
�
success did not appear in professional literature until Bray
and Grant had followed the first Assessment Center candidates
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
for about 10 years (first formal report of findings appeared
in 1966). The exchange of ideas among behavioral scientists
engaged in industrial/business psychology, of course, far
antedated the Bell System Management Progress report of 1966.
At the present time, for example, organizations such as
Sohio, Sears, Penneys and IBM, all have developed and operate
their own "custom-tailored" Assessment Centers. Conservative
estimates suggest that since 1956, over 100,000 persons have
been processed through Assessment Centers designed specifically
for identification of managerial ability. (These numbers do
not include persons processed through Centers designed to:
(a) identify creative abilities; (b) identify sales potential;
(c) identify candidates for advanced ("war college") military
training by foreign governments. Likewise, these figures do
not include the numbers of "on-board" and "applicant" Career
1
Trainees processed through the Psychological Services Staff's
(PSS's) assessment center designed to point up: (a) career
directions; (b) long range potential.
Bow does the Assessment Center work?
It may be helpful to comment first on a comparison of
Assessment Center procedures with traditional behavioral
science (psychological) assessment procedures.
First, the Assessment Center approach places the candidate
into a problem situation in which he must act (behave) so as to
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
handle same. Unlike traditional procedures, Assebt.ment Center
procedures ar3 of exceptionally wide bandwidth i.e., the range
of behaviors possible far outstrip the more narrow bandwidth
procedures demanding either sole]Ar speed or general intelligence
or verbal facility or eye-hand coordination, etc. The candidate
is placed in a position which demands he display more global
samples of his behavior than do traditional techniques.
Next, the Assessment Center procedures are developed by
behavioral scientists so as to simulate job-situat3ons defined
as "stumbling blocks" or "stepping-stones" to managerial success
by successful managers within the organization concerned. In
other words, after close consultation with management, behavioral
scientists design situational tasks which parallel those both
"par for the course" and guaranteed to "test the mettle" of
managers in the organization. In essence, the Assessment Center
tasks are miniature life situations faced by the organization's
managers in their day-to-day operations. In this sense, the
Assessment Center procedures are akin to the training techniques
for commercial airlines pilots...you put the candidate pilot
in a realistic but simulated situation (where his worst per-
formance costs neither lives nor a multimillion-dollar aircraft)
in order to determine the reasonabldness of advancing him to
the real-life situation.
Finally, many of the traditional evaluation techniques are
constrained by the need to identify THE RIGHT ANSWER from among
THE WRONG ANSWERS in order to generate a quantifiable score.
-3-
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
In the Assessment center approach, attempts to resolve the
problem situation are not matched against a "school solution"
...the candidate's attempts may be judged to range from highly
successful to highly unsuccessful...providing a clearly more
detailed description of the candidate's performance than the
simple "Right" vs "Wrong" dichotomy.
In this same vein, it should be noted that descriptions
of Assessment Center candidates expand rather than constrict
the range of possible dispositions of the candidates. Given
a representative number of job-related tasks, it is the rare
candidate who comes through the Center as either 100% or 0%.
Candidates come through identified as to specific strengths
and deficiencies. Thus it is that the Assessment Center,
properly used by an organization, does not proceed to replace
or convert to "rubber stamps", the organization's ongoing
mechanisms for advancement of employees. Instead, the Assess-
ment Center provides such mechanisms (viz, promotion panels)
with an additional, vital source of data to assist in decision-
making.
Why use an Assessment Center?
The most obvious reason is that the Assessment Center works:
The less obvious reason is that, given the research support
necessary to document the validity and utility of the Center
in any given organization, the highest levels of management
are constantly up-to-date vis-a-vis the make-up of the managerial
�
-4-
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
character of the group; are alerted to sources for new input;
and are in a position to input new elements to the managerial
structure in the face of projected requirements and demands
for the future. la epsence, the Center offers highest levels
of management, the capability for prediction and direction of
the character and style of the organization.
Lest the foregoing sound overly optimistic, let's look
at the "box-score" for the Assessment Center approach.
One way of checking the box-score is to ask whether the
Center offers advantages over previous procedures. The answer
is YES in the range of magnitude from 10% to 30% improvement
in successful prediction of who will "make it" and "how far"
in the organization.
Next, in a unique sort of arrangement set up by Drs. Bray
and Grant in the Bell System, conclusions from the Center can
be held back from management. Later, Center predicted success
can be compared with actual success in the organization. The
time elapsed in the Bell System study (from Assessment Center
to roughly ten years performance in the organization) draws focus
on Center predictions regarding capacity to reach "middle-
management" positions. Here, the box-score shows that of can-
didates described by the Center as having middle-management
potential, 2 out of 3 did realize their potential. Of all those
described by the Center as deficient in such potential, only
1 out of 3 wexe actually advanced to middle-management positions.
-5-
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
What of those studies wherein Assessment Center results
are directly communicated to management? One of the better,.
more recent and representative studies in this area has come
out of IBM. Using 1016 employees, considered reasonable
candidates for advancement into managerial positions in the
time period 1965 to 1970, IBM used the following summary
rating of managerial potential to describe Assessment Center
findings:
Rated Level Description
1 Executive Potential
2 High Level Potential
3 Second Line Potential
4 First Line Potential
Remain Non-Management
Of all candidates (1086) processed by the IBM Center, the
following Rated Levels were assigned:
Rated Level
Percentage (of 1086)
1
4%
2
16%
3
24%
4
28%*
5
�
(*Over 50% of all candidates were rated as incapable of pro-
gressing beyond first line management...a fact having impli-
cations for preselection of Center candidates to be discussed
under the "When" section.)
-6-
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
Granted that the foregoing ratings were availdhle to
management, it should not be surprising that the di.cl!,Ions
about promotion to First-Line (First-Level) manag e:heflt WOre
the following
'
Rated Level
PercentLge Fronoted-
to First-Lint
1
87%
2
48%
3
42%
4
29%
5
24%
There is a suggestion in the above data of the "Crown
Prince Effect' i.e., if you are rated high by the Center, your
future in the organization is guaranteed. Further data re-
garding promotions after this initial promotion into management
has been secured and looks like this:
Rated Level
Percentage Promoted Beyond
First-Line M,Ipagement
1
34%
2
32%
3
27%
4
13%
5
7%
Thus, after .irst Line promotion, later promotions tend
to "level off' for the three highest Assessment Center ratings.
It would appear that while later promotions are less influenced
by the "halo effect" of earlier Center ratings and more determined
by factors such as actual on-the-job performance, nonetheless,
-7-
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
1.
those identified by the Center as having higher potential
actually do perform at higher levels than lower rated candidates.
What about the "kiss-of-death" effect i.e., if a person
receives a low Assessment Center rating, is he doomed? The
following figures on separations among the Center candidates
by IBM seems to answer this question:
Rated Level
Percentage of Separations
Among Candidates
1
0%
2
2%
3
3%
4
3%
5
2%
Obviously, separations are evenly spread across all cate-
gories. (Note that given the small number of candidates rated
at Level 1p the loss of even one person would be equal to 2% of
the group!) Thus low ratings do not unreasonably prejudice the
candidates career. Remember that Bray and Grant found that,
after about 10 years, management had promoted to given levels
33% of those people rated by the Center as incapable of ad-
vancing to those given levels. (Note that Bray and Grant used
only the first Center prediction of ten years earlier, unrefined
by data regarding training received and skills required.)
The suggestion in the last two tables combined is somewhat
e �
intriguing i.e., the Assessment Center appears more appropriate
in identifying the "comers" as opposed to branding the "losers".
-8-
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
Other findings of Bray and Grant further support this possibili-
ty in that CeirLer predictions of success in sales acti%ities
matched indepeldently obtained field ratings 100%. Center pre-
dictions of failure in sales matched field ratings only 10t.
Who operates the Assessment Center?
Since the Assessment Center is a set of procedure!, designed
conjointly by behavioral scientists and managers, applied by
managers in evaluating...etc., the Center clearly is operated
by the management of the organization for whom the Center is
designed. In other words, after the job-related situational
tasks are designed, experienced managers in the given organiza-
tion are selected for training as assessors (observers, raters)
in the Center. The preferred training technique is to permit
the managers to deal with the same situational tasks the future
candidates are to face. In this fashion, assessors both are
made aware of the special demands of the tasks and also assist
in "debugging" the design of the tasks selected.
The behavioral scientist continues to contribute to the
Center in three basic areas: (1) he contributes psychometric
data responsive to highly specific questions about candidates;
(2) he is available for consultation regarding unusual problems
of behavior observed or observation of behavior; (3) he main-
tains current validity data regarding Center findings and "on-
the-job" performance.
-9-
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
When is the Assessment Center used?
This is a critical question bearing on the overall utility
(cost-effectiveness) of the Center. Obviously the Center
cannot accept all employees in the organization. Some career-
development critical point should be identified e.g. the level
in the organization regarded as First Line/Level management.
Having identified this critical point, the next question is
whether inclusion in the Center processing is to be automatic
or at the individual's option.
Where conduct the Assessment Center?
This last point, while seemingly simple-minded, is hardly
so. Candidates tend to be more spontaneous and less inclined to
pursue rigid, "school solution" behaviors when they are removed
from institutional surroundings. Most importantly, managers
operating as assessors, tend to set aside assessor tasks when
"day-to-day" office concerns are pushed upon them i.e., in
institutional surroundings, they are too easily distracted from
Center activities by phone-calls, "urgent" memos, and the like.
Consequently, "isolated" and/or "protected" surroundings are
both desirable and necessary for efficient operation of the
Center.
064.4"-4'd&
BERNARD L.
Psychologist
Msy � 197h
-10-
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
APPENDIX B
Paper Entitled:
"More About Assessment Centers"
DDA EXCHANGE April 1978
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
�
_
IP
medical services
MORE ABOUT ASSESSMENT CENTERS
Bernard 1. Mooney, Ph D.. OMS
In the October issue of ODA Exchange,
reference %vas made to Asseswint Centers
and Assessment Center methodology In two
separate articles One article described Um
use of the Ascessment Center technique as
one specifically designed to study an indivi-
dual's potential for responding to the do-
minds of managerial positions. while the
other described the technique as one de-
signed 10 study an Individual's potential tor
responding to the demands of Project AIM
positions none of which %lordly involved
niniiiinerlal responsibilities 'I his mild%) may
help to ciailly for some readers these
seemingly c onti nclIcloi y desci lotions of As-
SOCSM011t C011t01 methodology.
The got minni notion of the Center. pio-
� neered by the Assessment Shill of OSS (the
�
i
sai
. -C
-.
.4
r
progenitor of today's Psychological Services
Staff-OMS) placed the primary emphasis
upon the design of simulations of real-life
tasks .. simulations reflecting the key as-
pects of performance situations that OSS
candidates might eventually be required to
handle in the held 1 hrough observation of
candidates as they faced these simulations.
Assessors-Observers attempted to generate
dynamic descriptions of the candidates in an
effort to support accurate prediction of
"most likely" Individual behaviors in future
real-life performance situations. Very early in
the development of Center methodology, it
became clear that unless there actually are
observable behaviors rather consistently as-
sociated with both successlul as well as
unsuccessful performance In real-life situa-
tions. then design of a Center is an impossi-
ble task. Granted valid behavioral criteria of
success vs 1.111WO performance (typically
defined by means of consensus among
"expoits" regal ding the performance stud-
ied), the Center methodology can be ex-
tended to emetically any moo of poi form-
(ince-work. .
;
glow.
.1
I
-
\
i
,
-
-
Approved for Release: 2024/08/05 C06158506
- Approved for Release: 2024/08/05 C06158506
�
1
�
" Milk�
Curren'ty. Centers are being used to
evaluate potential for performance In sales,
In teeline:al a ea. and In entrepreneurial
activities, the pi ogress of AB degree candi-
dates preluding the award of bona lido
undergraiunte credits for successful per-
formance in Ilw Center), the qualilications of
applicants for licensure as Psychologist, the
potential of uniformed police for advance-
ment to he pc .1ition of Detective., and so
the lest goes .cently. the American Psycho-
logical Association has funded a feasibility
Study to explore the utility of Centers in
evaluating Psychologists for award of nation-
ally accepted i.ertilication of excellence Oho
so-callec Diplornate) across 4-5 professional
specialties.
The valicaly of Centers Is a matter no
longempen to critical debate The utility of
Centel s (the overall cost effectiveness) Is.
however, a miller which must be carelully
considered by potential consumers Unfortu-
nately, most consumers seldom possess
data acequate to the task of evaluating
Center utilities While a small number of
41.
�
�
orb
%owe
consumers do have Information req.ddlno
the dollar:. and cent , I of operateui Vick
standard milli mon ....lecher) syatems, ol
most none arc. able to cite data regal ding
the predictive validity of their prewnt sys
loins (how accurately theyC an predict mu:
likely !Wive performance). the cos
to the consumer ol making a � bad ' selechoi
decision, the value to the cot...timer
making a "good 'del v.ion, and, most imam
Lanny. adequate and pragmatic definitions c
"bad" and "good' dec isions. I c � "success
and "failure" In the real-life situation
While the question of Center utility I
problematic, the mut Ii is clear . Centel
enjoy their greate.,1 utility when they ai
u;ed as a aunt* mem to the consumer
extant selection mechanisms The mo
effective Centers are those designed I
measure solely thow elements of pe
form:ince potential which are not
dressed-inewarreif by the consumer's pm
cid evaluation-elcction sy..tem Commit'
In the article in the October ismie abo.
using the Center to replace a system whit
� vei IP � - �
IT
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
��
APPENDIX C
Behavioral Definitions of
Agency Assessment Center Dimensions
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
BEHAVIORAL DEFINITIONS OF AGENCY ASSESSMENT CENTER DIMENSIONS
DIMENSION BEHAVIORAL DEFINITION
A. AWARENESS OF
DETAIL
B. CAREER AMBITION
C. CLARITY OF ORAL
COMMUNICATION
D. CLARITY OF WRITTEN
COMMUNICATION
E. CREATIVITY
F. DECISIVENESS
G. DELEGATION
in problem situations, carefully
considers all relevant facts; does
not overlook important though often
minute details of the problem.
clearly expresses desire to move to
higher job levels; demonstrates
active efforts towards self-develop-
ment and self-improvement.
oral communication is concise and
to the point"; style is characterized
by proper grammar, pronunciation and
articulation; body language emphasizes
rather than distracts from
communication.
expresses written ideas clearly;
shows a mastery of the mechanics of
English, e.g., grammar, syntax,
spelling and punctuation.
shows the capacity to generate as
well as to recognize and accept
imaginative solutions and innovative
courses of actions in approaches to
problem situations.
shows the readiness to make decisions,
to render judgments, to take action
or to commit self; is able to
recognize those situations where
decision delays will be damaging
vs. those where no urgency exists.
assigns responsibilities effectively
to subordinates; clearly understands
the levels (organizational) at which
given decisions are most effectively
made; gives adequate directions to
others and provides sufficient
guidance when delegating.
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
DIMENSION BEHAVIORAL DEFlNlTIo%
H. DEVELOPMENT OF
SUBORDINATES
I. EFFECTIVE
COMMUNICATIM
J. ENERGY
K. FACILITA1ES CROUP
PROCESSES
L. FLEXIBILlTY
M. FORESIGH"
exerts effort to maximize human
potential of subordinates through
training and development asniqnments
related to both current ac well as
future jobs.
keeps peers and subordinates and
superiors informed of plans and
activities; avoid conflicts, "snafu's"
and needless duplication of effort
by both sharing and seeking out
information.
achieves and maintains a high level
of involvement in work activities;
level of involvement is matched by
level of output on a continuing
rather than sporadic basis.
in problem-solving situations when
working with a group, deals with
others in such a way that group
efforts remain directed upon the
problem rather against each other.
in problem situations, when given
management approaches or behavioral
styles prove ineffective, is able
to modify and vary approach and/or
style in order to attain stated
goals. -
characteristically thinks several
Steps beyond present problems;
tries to anticipate impact both
of problem resolution and side-
effects of problem-solvinq techniques
to be used; tends to include the
future in addressing problems of the
present. (This Dimension later
melded with PLANNING under
Dimension X, PLANNING AND ORGANIZING
... see below.)
ii
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
DIMENSION BEHAVIORAL DEFINITION
N. IMPACT
0. INITIATIVE
P. LEADERSHIP
Q. LISTENING
ABILITY
R. MANAGEMENT
CONTROL
is able to create a good "first
impression" which endures; shows
an air of confidence through basic
interpersonal style; almost auto-
matically and apparently effort-
lessly, commands attention and
respect.
actively influences situations and
events rather than passively accepting
them; takes actions beyond those
obviously and necessarily called
for; is proactive rather than merely
reactive.
is effective in winning acceptance
for plans and ideas from individuals
and groups; is effective in guiding
and directing individuals and groups
towards efficient accomplishment of
goals; is able to stimulate others
to greater efforts and higher levels
of attainment.
is able to grasp and retain key ele-
ments of ideas presented by others;
conveys a sincere interest so that
others make special efforts to present
their ideas; on occasion, is able to
perceive new relationships or con-
cepts "buried" among ideas presented
by others.
understands the principles of control
mechanisms over tasks, processes,
products and people; institutes and
maintains effective control mechanisms;
makes provisions for follow-up of
actions decided upon.
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
DIMENSION BEHAVIORAL DEFINIlloN
S. MOTIVATION expresses strong desires to achieve
FOR WORK in the area of work respoiv,ibalities;
personal satisfactions involve
primarily accomplishments attained
in the area of work.
T. ORGANIZATIONAL
ABILITY (as a
Manager)
is knowledgeable about availability
of resources; is knowledgeable about
the capabilities of resources; brings
together optimum combination of
resources for effectively attacking
problems or accomplishing assigned
tasks. (This Dimension later
melded with Dimension X., PLANNING
AND ORGANIZING ... see below.)
U. ORGANIZATIONAL shows a clear understanding of work-
ABILITY (as an unit demands of tasks assigned; tends
Employee) to restructure tasks so as to use
personal assets most effectively;
estimates time requirements of tasks
with accuracy; persona] scheduling
of activities makes optimum use of
time. (This Dimension later melded
with Dimension X., PLANNING AND
ORGANIZING ... see below.)
V. PERCEPTIVITY
AND SENSITIVITY
W. PERSUASIVENESS
X. PLANNING AND
ORGANIZING
accurately perceives the needs which
motivate others; reactions to others
reflects awareness of and respect for
needs of others; shows understanding
of the impact oneself has on others.
is able to present own ideas and
proposals in such a manner that most
other persons react to them in a
positive, accepting fashion.
effectively identifies key elements
in problems or tasks to be accomplished;
quickly establishes meaningful
priorities among these key elements;
effectively establishes course of
action for oneself and for others;
makes reality-based assignments of
personnel and committment of resources
in accomplishing specific goals within
time constraints.
Iv
Approved for Release: 2024/08/05 C06158506
Approved for Release: 2024/08/05 C06158506
DIMENSION BEHAVIORAL DEFINITION
Y. PROBLEM ANALYSIS
AND JUDGMENT
Z. RISK TAKING
AA. SOCIAL
ADAPTABILITY
BB. STRESS
TOLERANCE
CC. TENACITY
shows skills in identifying and
defining problems; secures problem-
relevant information and logically
isolates probable problem sources;
is able to evaluate direct and
indirect effects of courses of action;
is able to generate several alter-
native approaches to problems.
shows awareness of both positive
and negative consequences of alter-
native courses of action; to
maximize gain, may take actions
where losses can be sustained but
has carefully calculated likelihood
of loss beforehand; does not require
100% guarantee of success before
taking action.
is able to maintain effectiveness
across a wide range of social situa-
tions and work-group combinations;
responds to differing social styles
by altering personal style.
whether operating under time, personal,
social or situational pressures, main-
tains a stable, effective level of
performance.
shows the capability to stay with a
problem, pursue a line of reasoning
or remain focussed on a task (which
is within the reasonable capabilities
of the person) until the matter is
settled.
Approved for Release: 2024/08/05 C06158506