You are on page 1of 10

Critical Issues in

Performance Appraisal
by David K. Banner and James M. Graber
DePaul University

Introduction
This article summarises a number of recommendations emerging from
research and practice on performance appraisal Although the work is
based upon North American experience it is suggested the principles
have wider application.
In a few areas of applied behavioural science, much research and prac-
tice seems to lead, in some crucial ways, to a "roadblock". Despite the
attention of management consultants, academic researchers, and
organisational practitioners, performance appraisal seems to be one such
area. In several key areas, there seems to be a consensus about PA:
(1) methodologically, there have been great strides forward in terms
of reliable, valid instrumentation;
(2) traditional "trait" and other similar subjective measures of per-
formance are increasingly shunned by sophisticated organisations;
and
(3) despite the obvious gains in PA technology the gap between
theory and practice remains wide.
In this article, we will explore the state of PA theory/practice and offer
some suggestions about avoiding common "pitfalls" in its application.

What is Performance Appraisal?


PA can simply be defined as "the basic, yet baffling, process of deter-
mining how an individual employee is performing"[1]. Usually this ap-
praisal is relevant to a predetermined time period of employee per-

26 | Journal of Management Development 4,1


formance. From an organisational perspective, PA typically originates
from the personnel function and requires the manager to give appraisals
of his/her subordinates; these appraisals usually form the centrepiece
of a much larger human resource programme (promotion, career plan-
ning, salary administration, EEO, management development, redundan-
cy planning). PA seems to be many things to many people. At its worst,
it is viewed as a "make-work" yearly hassle by the manager, and an in-
timidating, relatively unproductive ritual for the subordinate. At its best,
it can be a technology for communicating managerial (and organisa-
tional) expectations of subordinates, allowing for timely feedback on
performance, and forming the cornerstone of a rational, equitable human
resource management system.

Critical "Pitfall" Areas


Given our current methodological sophistication, why has the transi-
tion from theory to practice proven so difficult? We can identify several
problem areas:
• the issue of "employee effectiveness"—how is "effective" defin-
ed? Does effective depend upon specifying certain desired out-
comes? Are certain general approaches to PA (such as MBO) in-
herently more effective than others? Is there one best way to
measure performance or are different methods needed for dif-
ferent employee groups?
• the issue of "measurement" strategy—who should evaluate an
employee's performance: supervisor, peers, or subordinates?
• the issue of the appropriate PA "roles"— should PA be the foun-
dation for other human resource programmes? What role does
PA play in the overall managerial role (such as communications
with workers or work design)?
• the issue of motivation/implementation—is there a way to in-
troduce a PA programme that will get a more enthusiastic
response from supervisors and employees? What is the impact
of involving the employee being evaluated in the PA process?
Let us examine each of these potential PA issues with an eye towards
practical, implementable strategies to avoid them.

The "Effectiveness" and Measurement Issues


The "effectiveness" question has been around for a while. In organisa-
tion theory, meaningful comparative research has been thwarted because
of the multitudinous definitions (and criteria for measurement) of organ-

Critical Issues in Performance Appraisal | 27


isational effectiveness. So it is with performance appraisal. In general,
though, measures of performance effectiveness have centred around
(1) job-related behaviours (e.g. presented report clearly, followed up
promptly with client); and
(2) job-related outcomes (e.g. reduced amount of absenteeism by five
per cent, increased net sales by seven per cent).
So the first step in determining effectiveness is defining the context of
a job using these two categories; this is best done by formal job analysis
or a similar process[2]. Job performance is not a singular, unidimen-
sional concept. In some companies, both job-related behaviours and
outcomes are evaluated, even for the same job. In any event, job per-
formance measurement involves defining what criteria are to be
measured.
The PA process, then, involves how performance criteria, once defin-
ed, will actually be measured. The ideal case, of course, would be job
performance criteria that were all easily quantified (e.g. number of
widgets per month). However, it is difficult to get quantitative measures
for many jobs[3]. And, in most organisations, managerial and profes-
sional job performance is not typically measured but appraised (which,
of course, implies subjectivity)[4]. When judgmental indices are used
for PA, the measurement process becomes susceptible to problems of
rater error, bias, and reliability.
In reviewing the PA literature and from our personal experience, we can
make the following recommendations about measurement and
"effectiveness":
Recommendation 1
Job analysis should be performed using both behaviour-oriented (job-
knowledgeable employees are asked to help identify behaviours of peo-
ple effective at their job) and results-oriented (job-knowledgeable
employees are asked to describe the activities a typical employee would
perform in the job) criteria.
We suggest that behaviours are the correct evaluation criteria when end
results are out of the employee's control or when the manner of perfor-
mance is itself of importance. End results can be out of an employee's
control due to a dependence on co-workers, the state of the economy,
client demand, availability of parts and materials, or for other reasons.
By "manner of performance" we mean such things as the courtesy of
a clerk, use of proper etiquette by a waiter, or the co-operation and team-
work of an executive.
When end results are within an employee's control and important, they
are the most appropriate measurement criteria. We find that results-

28 | Journal of Management Development 4,1


oriented criteria will usually fall into one of the following four categories:
quality, quantity, cost effectiveness/level of performance, or timeliness.
Recommendation 2
These criteria should then be shared with the ratee, allowing for modifica-
tions, clarifications, and other processes consistent with mutual agree-
ment on rating criteria[5,6].
In fact, we often find it useful to have employees participate with other
"subject matter experts" (anyone knowledgeable about the job) in a
group process that first identifies the areas of the job for which an
employee should be evaluated and then develop standards for how well
each of these areas should be accomplished. Note that the setting of
standards is a management prerogative, and employee participation
should not be construed to mean that employees have an equal voice
should there be differences of opinion on what the standards should be.
Recommendation 3
The actual rating process should clearly allow for the ratees' self-ratings.
Prior to the formal appraisal meeting and/or during it, the subordinate
should be given a substantial chance of input into the results of the
appraisal [7].
We have seen this done in a number of ways. Some companies have
employees actually fill out an evaluation form on themselves, while others
ask employees to come to the appraisal review ready to discuss their
perceptions of their performance.
Recommendation 4
When appropriate, alternatives to supervisory review, a combination of
peer, supervisory and self-ratings should be considered. By this we mean
that peer review, self-review, client review, or even subordinate review
may sometimes be best.
For example, we know of a government agency where supervisors are
evaluated on their managerial practices. They receive half their points
on this element from the results of a subordinate attitude survey. Similar-
ly, it is often done to ask one's customers to participate in the evalua-
tion process. The question to ask is, "Who is in the best position to
evaluate this element of performance?" We must caution that there are
potential risks in involving parties other than the immediate supervisor,
but these are usually problems of politics rather than problems of rating
objectivity and validity.
The "Appropriate Role of PA" Issue
A number of credible researchers have questioned the usefulness of PA
as a multipurpose foundation for human resource management[8].
Despite these doubts, PA is often a building block for programmes other

Critical Issues in Performance Appraisal | 29


than performance improvement, e.g. salary administration, training, pro-
motion and demotion, etc. Furthermore, PA is often the only formal
system organisations use to communicate to the employee what his/her
job is. What is expected of PA in an organisation is as important as
the actual PA methodology used in determining how well it works; we
argue that organisation expectations of PA may be unrealistic and at
times even conflicting [9].
The purposes or uses of PA have dramatically expanded over the years.
Up until the early 1960s, PA's de jure organisational use was ad-
ministrative (firing, promotion and placement type decisions). However,
perhaps because of the lack of validity in the common "trait" or sub-
jective appraisals, there was a very loose relationship between the ap-
praisal results and resulting administration decisions, e.g. personnel deci-
sions (termination) ran counter to appraisal results, high-rated employees
were often fired, low-rated employees retained or rehired[10]. In the early
1960s, PA was called upon as a guide to employee career consulting and
development. With Drucker's[11] and McGregor's[12] landmark work
in management-by-objectives (MBO), PA became a widespread corporate
planning tool in the late 1960s and early 1970s. Finally, with the 1966
and 1970 EEOC guidelines from the federal government, legal considera-
tions created added pressure on organisations to rationalise and organise
their PA practices[13]. It became readily apparent that PA was being
called upon to do more than it was (at that time) capable of delivering.
This problem was exacerbated by two factors. One, while uses for PA
have expanded, many organisations continue to use PA methods
developed decades ago ("trait" ratings) when constraints on and uses
for PA were minimal. Two, some purposes of PA are not necessarily
compatible with other PA uses. Some authors have argued that asking
managers to rate their employees for salary/promotion decisions is in-
compatible with asking them to use PA as a counselling/development
tool for the same employees[14, 15].
These questions have led to the development of "hybrid" systems of
PA which combine behavioural systems (e.g. BARS—Behaviourally An-
chored Rating Scales, and MBO) into a powerful, multipurpose tool.

RCA Corporation and Black and Decker are just two of a growing
list of companies that have moved to a "hybrid" approach. RCA's
system uses multiple raters, emphasis on behaviour rather than traits,
and the use of training to help managers give feedback to
employees[16]. Black and Decker's MBO-oriented approach requires
goals or targets for both personal development and for work activities,
periodic review sessions (rather than annual) are required of raters,
and rater and ratee both collaborate on the goal setting and the
evaluation.

30 | Journal of Management Development 4,1


Recommendation 5
When used for multiple purposes, PA systems should be designed for
those purposes, e.g. separate objectives, measurement criteria, etc., for
career development and for the work activity. Most often, this is handl-
ed by having a separate section on the appraisal form for career develop-
ment. Impetus for career development goals may come from the
employee if he/she is anxious for advancement, or from the organisa-
tion if it is interested in improving the skill of its workforce or ensuring
the availability of suitable replacements. Often, there are different im-
plications of succeeding on work performance goals and succeeding at
career development goals. The former usually carries more weight.

Recommendation 6
Innovations like multiple raters and instruments, emphasis on behaviours
and outcomes rather than traits and PA training for raters and ratees
should be used, and periodic feedback throughout the year should be
implemented. This way the data from the process can be more nearly
tailored to multiple corporate purposes.
Many companies have special staff and resources assigned to career
development, freeing supervisors from this responsibility. Though this
is a luxury more affordable for larger companies, it is a good idea because
many supervisors either lack the necessary skills or are unwilling to get
involved in this. Training for supervisors in performance appraisal of
course is intended to result in more valid ratings, which becomes in-
creasingly important as more and more decisions are made based on
PA. Currently, training on how to conduct the appraisal interview is
widespread, and training for the development of evaluation criteria is
somewhat less prevalent. "Rater error" training is also commonplace.

The "Motivation-Implementation" Issue


One of the most common complaints heard from practitioners in the
field is that PA "doesn't work". When pinned down on what exactly
they mean by that, they say that, despite the use of well designed
sophisticated PA systems, people (raters and ratees) don't seem to be
motivated to use them correctly. A landmark paper on motivation and
PA behaviour[17] deals with what seems to motivate the behaviour in-
volved in carrying out performance appraisal in organisations. These
authors argue that "we go about our day-to-day activities privately ap-
praising the behaviour of ourselves and others" (p. 2). The motivation
behind appraising others seems to be part of a basic need to unders-
tand, predict, and control our individual worlds[18, 19]; appraisal can
also be used as a source of self-esteem through social comparison, for
data to be used for obtaining extrinsic rewards (e.g. arguing for a pay

Critical Issues in Performance Appraisal | 31


rise) and to fulfil mandated role behaviour (e.g. doing an appraisal of
a subordinate). Clearly, appraisal of others is a naturally occurring pro-
cess that serves a variety of purposes.
However, when placed in a formal organisational context, the use of
a structured appraisal system for purposes of evaluating work behaviour
becomes problematic. We make the following suggestions to improve
the enthusiasm and commitment with which PA is received.
Recommendation 7
PA systems need to be designed to be compatible with the organisa-
tional culture or context in which they are embedded. For example, we
know an advertising agency that uses open ended questions and an essay
format to evaluate its people. While an essay approach would generally
not be the method of choice, it can be defended in this case because
a "check system" where employees are given ratings from unacceptable
to excellent is abhorrent to the employees at the agency. They find such
a system dehumanising. Employee self-appraisal is a feature that will
fit better into some organisational cultures than others. Similarly, the
complexity and demands made by a PA system must not exceed the ex-
perience or the time that supervisors are willing to give to it. Again,
this varies from organisation to organisation.

Recommendation 8
PA feedback must attend to appraisee's desires and expectations. For
this reason, pay should be a topic of discussion at PA sessions[20, 21].
We have seen too many situations where the lack of connection bet-
ween appraisal and administrative decisions such as pay, promotion,
etc., has given employees and supervisors the impression that PA "doesn't
really count for anything". In fact, we can generalise this even further
by saying that any action that can underline the significance of PA should
be taken.
There are a number of things companies do to increase the perceived
relevance of PA. Some evaluate all persons with supervisory respon-
sibilities on how well they discharge these duties. Role modelling by
higher-ups in the organisation is essential. If top management doesn't
set a good example, first and second level supervisors will get the wrong
message.

Recommendation 9
Appraisers must be rewarded for giving realistic ratings of appraisees.
Obviously, appraisers are more comfortable giving high ratings than giv-
ing low ratings. This gradually leads to rating inflation. In some organisa-
tions anything less than a perfect rating is tantamount to failure.

32 | Journal of Management Development 4,1


There are several strategies: first, upper level management must sup-
port appraisers when they give a low rating, rather than tempering such
ratings because of fear of negative legal consequences. Second, appraisers
must know that they are not hurting their people by giving realistic
ratings in comparison to ratings of other supervisors. This means an
all out organisational effort to get all appraisers to rate objectively. Final-
ly, consistently "high" raters can be judged poor performers because
they fail to carry out an important part of their jobs as managers.
In summary, the improvement of PA systems is ultimately a matter of
creating a shared social definition of PA that is in congruence with the
original intent of PA in the organisation.

Summary
The management development professional can serve a number of useful
roles in bringing about such changes.
The first step is to become better educated about PA. This is a challenging
area with a great many misperceptions. One must have facts in hand
about the key ingredients of good PA to argue successfully against tradi-
tional practices that don't work.
Second, it is necessary to analyse what potential PA benefits will be
most appealing to one's organisation. Is it objective reviews free from
favouritism? The linking of performance and positive outcomes such
as rewards and recognition? The creation of the legal defensibility of
actions taken? Improved performance through clearer goals and counsell-
ing? The most attractive benefits chosen, the company should concen-
trate on designing a system to meet these needs, and selling the system
based on these merits.
Third, it must be realised that performance appraisal is not a process
that supervisors naturally do well. Unfortunately, it is often the people
who think they are doing fine that actually need the most help. Train-
ing should be mandatory for all appraisers, including upper level manage-
ment. Teaching appraisers to conduct effective reviews requires giving
them specific models to follow, not generalities, and practice through
realistic role plays.

In regard to developing appraisal criteria, the people doing the job


and supervising it must be involved. However, this is an area where
the management development professional can really help by be-
ing a facilitator. The facilitator must see that evaluation criteria
are effective behaviours or results, not traits. The facilitator can
do much to encourage clear wording, job relevance, and coverage
of important parts of the job that might be left out because they
are difficult to measure.

Critical Issues in Performance Appraisal | 33


Finally, the management development specialist can serve as cheer-leader
and journalist, offering encouragement and support while at the same
time recording successes and failures of the system. Overly optimistic
expectations must be beaten back, but people must be supported and
given feedback that they are successfully discharging their PA duties.
It may even be appropriate to survey employees to ask them how they
think the system is working.
In short, the management development professional will probably need
to be political strategist, appraisal system expert, trainer, salesperson,
and catalyst to help PA reach its real potential.
We are saying in this article to have a PA system do what it is intended
to, it must be:
(1) designed with those multiple purposes in mind;
(2) designed with the contextual elements in mind;
(3) valid and reliable through pretests and other techniques;
(4) structured with built-in motivation systems for appraisers and
appraisees;
(5) designed so that employee participation is encouraged where com-
patible with the organisational culture;
(6) oriented towards PA training to make clear both the rater and
ratee roles;
(7) designed to avoid "trait" and other subjective measures in favour
of behaviour and/or outcome measures;
(8) designed to use multiple raters when possible.

References
1. DeVries, D.L., Morrison, A.M., Shullma, S.L. and Gerlach, M.L., Performance
Appraisal on the Line, Greensboro, N.C.: Centre for Creative Leadership,
Technical Report No. 16, December 1980, p.1.
2. Ibid., p. 32.
3. Landy, F.J. and Farr, J.L., "Performance Rating", Psychological Bulletin, No.
87, 1980, pp. 72-107.
4. Patten, T.H., Jr., Paz: Employee Compensation and Incentive Plans, The Free
Press, New York, 1977.
5. Blake, R.R. and Mouton, J.S., "Power, People and Performance Reviews", Ad-
vanced Management Journal, Vol. 26 No. 3, 1961, pp. 13-17.
6. Fletcher, C.A., "Interview Style and the Effectiveness of Appraisal", Occupa-
tional Psychology, No. 47, 1973, pp. 225-30.
7. Resnick, S. and Mohrman, A.M., "An Appraisal of Performance Appraisal
Results of a Large Scale Study", Centre for Effective Organisations, University
of Southern California, 1980.
8. DeVries, et al., op. cit.
9. Ibid., p. 11.

34 | Journal of Management Development 4,1


10. Whisler, T.L., "Appraisal As A Management Tool", in Whisler, T.L. and Harper,
S.F. (Eds.), Performance Appraisal: Research and Practice, Holt, Rinehart and
Winston, New York, 1962.
11. Drucker, P.F., The Practice of Management, Harper & Row, New York, 1954.
12. McGregor, D., "An Uneasy Look at Performance Appraisal", Harvard Business
Review, No. 35, 1957, pp. 89-94.
13. DeVries, et al., op cit., p. 16.
14. Levinson, H., "Management by Whose Objectives?", Harvard Business Review,
No. 48, 1970, pp. 125-34.
15. Meyer, H.H., Kay, E. and French, J.R.P. Jr., "Split Roles in Performance Ap-
praisal", Harvard Business Review, No. 43, 1965, pp. 123-9.
16. Lazer, R.I. and Wikstrom, W.S., Appraising Managerial Performance: Current
Practices and Future Directions, Conference Board, New York, 1977.
17. Mohrman, A.M. Jr. and Lawler, E.E. III, "Motivation and Performance Ap-
praisal Behavior", Centre for Effective Organisations, University of Southern
California, 1981.
18. Kelly, G.A., The Psychology of Personal Constraints, Vol. I, WW. Norton and
Co., New York, 1955.
19. Weick, K.E., The Social Psychology of Organizing, Addison-Wesley Publishing
Co., Reading, Mass., 1979.
20. Mohrman and Lawler, op cit.
21. Prince, J.B. and Lawler, E.E. III, "The Impact of Discussing Salary Action in
the Performance Appraisal Meeting", Organisational Effectiveness Research Pro-
gress, Office of Naval Research, 1981.
Further Reading
Greller, M.M., "The Nature of Subordinate Participation in the Appraisal Interview",
Academy of Management Journal, No. 21, 1978, pp. 646-58.
Hemphill, J.K., Dimensions of Executive Decisions, Research Monograph No. 98,
Bureau of Business Research, Ohio State University, 1960.
Lawler, E.E. III, "The Multi-Trait/Multi-Rates Approach to Measuring Managerial
Job Performance", Journal of Applied Psychology, No. 51, 1967, pp. 369-80.
Mohrman, A.M. Jr., "Formal Performance Appraisal as an Intervention for the In-
terpretation of Performance and the Quality of Work Life", Centre for Effective
Organisation, University of Southern California, 1981.
Perry, L.T. and Barney, J.B., "Performance Lies are Hazardous to Organizational
Health", Organizational Dynamics, No. 5, 1981, pp. 68-80.
Primoff, E.S., Using a Job Element Study for Developing Tests, Personnel Measure-
ment Research and Development Centre, Standards Division, Bureau of Policies
and Standards, US Civil Service Commission, 1972.
Schneider, C.E., "Multiple Rates Groups and Performance Appraisal", Public Per-
sonnel Management, No. 6, 1977, pp. 13-20.
Wexley, K.N., "Performance Appraisal and Feedback", in Kerr, S. (Ed.), Organiza-
tional Behavior, Grid Publishing Co., Columbus, Ohio, 1979.

Critical Issues in Performance Appraisal | 35

You might also like