Rating is used by the majority of organizations with performance management or appraisal systems and there are good arguments for doing so. But there are also persuasive arguments against rating. The pros and cons are discussed below.
Arguments for rating
The arguments for rating are that:
It satisfi es a natural wish from people to know where they stand. But this is only desir-
•
able if the manager’s opinion is honest, justifi ed and fair, and the numbers or letters convey what is really felt and are meaningful.
It provides a convenient means of summing up judgements so that high or low per-
•
formances can easily be identifi ed (as long as the judgements are consistent and fair).
It motivates people by giving them something to strive for in the shape of higher ratings
•
(as long as they know what they have to do to get a better assessment).
It is not possible to have performance-related pay without an overall rating (assuming
•
performance-related pay is wanted or needed).
It can provide a basis for identifying high fl yers for a talent management programme
•
or for generally predicting potential. But past performance is only a predictor of future performance when there is a connecting link, ie there are elements of the present job that are also important in a higher-level job.
Arguments against
Ratings are largely subjective and it is diffi cult to achieve consistency between the ratings given by different managers (ways of achieving consistent judgements are discussed below). Because the notion of ‘performance’ is often unclear, subjectivity can increase. Even if objectivity is achieved, to sum up the total performance of a person with a single rating is a gross over-sim- plifi cation of what may be a complex set of factors infl uencing that performance – to do this after a detailed discussion of strengths and weaknesses suggests that the rating will be a super- fi cial and arbitrary judgement. To label people as ‘average’ or ‘below average’, or whatever equivalent terms are used, is both demeaning and demotivating. The whole performance review meeting may be dominated by the fact that it will end with a rating, thus severely limit- ing the forward-looking and developmental focus of the meeting, which is all-important. This is particularly the case if the rating governs performance or contribution pay increases.
Analysing and Assessing Performance 157 Furnham (2004) raised a number of questions about the rating process including the issue of what should be observed and recorded, the availability of reliable performance standards, the evaluative and judgemental nature of the process
There are also many well-known rating errors. Grote (1996) lists nine as follows:
Contrast effect. The tendency of a rater to evaluate people in comparison with other
•
individuals rather than against the standards for the job.
First impression error. The tendency of a manager to make an initial positive or nega-
•
tive judgement of an employee and allow that fi rst impression to colour or distort later information.
Halo or horns effect. Inappropriate generalizations from one aspect of an individual’s
•
performance to all areas of that person’s performance.
Similar-to-me effect. The tendency of individuals to rate people who resemble them-
•
selves more highly than they rate others.
Central tendency. The inclination to rate people in the middle of the scale even when
•
their performance clearly warrants a substantially higher or lower rating.
Negative and positive skew. The opposite of central tendency: the rating of all individ-
•
uals as higher or lower than their performance actually warrants.
Attribution bias. The tendency to attribute performance failings to factors under the
•
control of the individual and performance successes to external causes.
Recency effect. The tendency of minor events that have happened recently to have
•
more infl uence on the rating than major events of many months ago.
Stereotyping. The tendency to generalize across groups and ignore individual
•
differences.
Powerful attacks on rating were made by Coens and Jenkins (2002) and Lee (2005) as set out overleaf.
Attacks on rating
Coens and Jenkins
Ratings are not a good idea because of the ‘unintended consequences – the insidious, destructive and counterproductive effects of giving people ratings about their work performance. Whether accurate or not, people are psychologically affected by ratings.
And except for people rated at the highest end of the scale, the impact is usually negative… Our ability to fairly measure the performance level of an individual is severely hampered by the unknowable effects of systems and random variations.’
Lee (2005)
The rating process is actually a by-product of the attempt to measure
•
performance outcomes. An excessive emphasis on measurement can be misguided. The desired end that is lost in measuring performance is not measurement at all, but rather description.
Poor ratings can stigmatize performance and cause unnecessary resistance to
•
the acceptance of feedback.
The goal is to have the employee assist us in describing, interpreting and
•
redirecting performance feedback, not reacting to the ratings. Feedback can accomplish the same positive goal as a rating without the negative side effects.
If the goal is performance improvement, then feedback – not labelling past
•
efforts – is the preferred tool.
Although ratings can be positive they can also be punitive and focus attention
•
on the negative rather than the possible. The only message the employee gets from a poor rating is: ‘Stop doing what you have been punished for doing.’ This kind of rating may not even be an adequate description, since many ratings are a summary of a number of activities collected over time. It does not focus attention on what to do to get better.
Ratings are feedback but feedback of the worst kind.
•
Conclusions on rating
There are strong arguments both for and against rating. But the majority of organizations favour rating for three main reasons: 1) it informs performance pay decisions; 2) it identifi es high fl yers for talent management purposes or poor performers for remedial action or dis- missal; and 3) it tells employees where they stand. Some either ignore the cons or are unaware of them. But many are concerned with the real problems of inaccuracy and inconsistency and
Analysing and Assessing Performance 159 ways of tackling these are discussed below. Some organizations have adopted alternative approaches, as described in the next section of this chapter.
Achieving accuracy in ratings
Murphy and Cleveland (1995) suggested that rating accuracy is improved when:
Good and poor performance are clearly defi ned.
•
The principle of distinguishing among workers in terms of their levels of performance
•
is widely accepted.
There is a high degree of trust in the system.
•
Low ratings do not automatically result in the loss of valued rewards.
•
Valued rewards are clearly linked to accuracy in performance appraisal.
•
Achieving consistency in ratings
The following methods are available for increasing consistency.
Training
Training can take place in the form of ‘consistency’ workshops for managers who discuss how ratings can be objectively justifi ed and test rating decisions on simulated performance review data. This can build a level of common understanding about rating levels. This is sometimes called ‘frame of reference training’ (Bernardin et al, 2000). The purpose of frame of reference training is to calibrate trainers so that they agree on: 1) how to match the specifi c behaviours of the behaviour of those they are rating to the appropriate performance; 2) the effectiveness levels of alternative behaviours; and 3) the rules for combining individual judgements into a summary evaluation for each performance dimension.
Peer reviews
Groups of managers meet to review the pattern of each other’s ratings and challenge unusual deci- sions or distributions. This process of moderation or calibration is time consuming but is possibly the best way to achieve a reasonable degree of consistency, especially when the group members share some knowledge of the performances of each other’s staff as internal customers.
Monitoring
The distribution of ratings is monitored by a central department, usually HR, which chal- lenges any unusual patterns and identifi es and questions what appear to be unwarrantable dif- ferences between departments’ ratings.
Consistency at a price can also be achieved by forced distribution or ranking as described earlier.