- 1 Appraisal Instruments
Inherent in the deWnition of job performance is the development of an instrument
for assessing it. Hence, the appraisal instrument is more than a form to be
completed ‘after the fact’ at the request of the HR department. It is a diagnostic
tool to be used on an a priori basis to facilitate self-coaching and the coaching of
others as to what a person needs to start, stop, or continue doing on the job. In
short, it speciWes for both the manager and the employee what should be observed
for attaining organizationally desired outcomes.
In practice, many organizations focus on the outcomes themselves such as
measures of goal attainment. The dangers in doing so are at leastWvefold. First,
outcome measures are often aVected by factors beyond a person’s control (e.g. the
economy). Consequently, a person receives a high or low evaluation undeservedly.
Second, outcome measures are often deWcient in that they do not take into account
factors for which a manager should be held accountable (e.g. sharing information
with other divisions; team playing). Third, they can encourage a ‘results at all costs’
mentality where ‘winning the game is perceived to be more important for one’s
career than how one plays the game.’ Fourth, a focus on outcomes is counter to the
coaching process inherent in performance management. A focus on outcomes
keeps people informed of the score, but it does not tell them by what means and
to what extent they have inXuenced their rate of promotion, salary increase, or
their team’s production. Fifth, outcome measures increase the probability of
hindsight bias, a decision-making error discussed subsequently. Behavioral instru-
ments, on the other hand, should be designed to specify how an organization’s
strategy is to be implemented in order to attain its goals.
Behaviorally anchored rating scales or BARS (Smith and Kendall 1963 ) provide
the rater with behavioral illustrations of the diVerent points on the rating scale (i.e.
‘behavioral anchors’ for deWning a 1 , 5 ,ora 7 rating of an employee). Behavioral
observation scales or BOS (Latham and Wexley 1977 ) require appraisers to rate the
frequency with which they have observed an employee demonstrate speciWc
behaviors. In contrast to these two instruments are trait-based scales which require
the rater to make inferences from behavioral observations to personality disposi-
tions (e.g. integrity, dependability). This inferential leap makes it diYcult to
calibrate speciWc ratings to actual employee behaviors. Hence inter-observer reli-
ability between two or more appraisers is usually low. Giving an employee con-
structive feedback without deWning each trait behaviorally can cause confusion:
‘What did she mean when she said I need to be more assertive with peers?’ Thus, it
is not surprising that the North American courts take a dim view of them (Latham
and Wexley 1994 ).
BARS are based on the assumption by Smith and Kendall ( 1963 ) that it is
helpful to give raters a frame of reference in the form of behavioral illustrations of
what constitutes unacceptable, acceptable, and highly acceptable performance.
performance management 367