Performance evaluation systems can be classified along a number of dimensions that capture variations in their structure, content, and process characteristics.
Among the most important dimensions are the following: 1.
Who/what is evaluated? Do we evaluate the individual, the workgroup, the division? 2.
Who performs (and has input into) the evaluation? Is it done by each individual's immediate supervisor? Peers, subordinates, or customers? How much input does the person being evaluated has into the evaluation and in appealing the results? 3.
Time frame: short to long.
What is the time frame over which data are collected (either formally and objectively or informally) before evaluations are rendered? 4.
Objective/formulaic versus subjective/impressionistic evaluations.
In some cases, performance is measured very objectively, using unambiguous measures of different aspects of performance.
For example, a salesperson might be scored on Euros sales, new customers developed, and increases in orders by old customers, and each of these being put on some standard scale (e.
g.
, standard deviations from the mean performance of salesmen in the organization) and then weighted 40%, 40%, and 20%, respectively.
In contrast, employees in a facility might be evaluated and rated based on the subjective overall impressions of their immediate superiors.
When objective or formulaic evaluations are used, there is the further issue of how closely tailored the formula should be to the situation of each individual.
At one extreme, every similarly situated individual in the firm (say, every salesperson) is evaluated using the same rigid formula.
The middle ground includes cases in which individuals are evaluated against their own previous performance; improvements are noted, but the same categories are used for each individual.
At the other extreme are systems in which each individual in each period has a specially tailored set of goals and objectives.
A prime example of this is management by objectives schemes, in which each individual takes part in designing his or her set of objectives.
- Relative versus absolute performance.
In some instances, employees are evaluated on an absolute scale-for example, sales volume, units produced per week, touchdowns scored, or dollar value of hours billed to clients.
In other instances, performance is evaluated on some sort of relative basis, or performance is measured on a mixture of absolute and relative performance.
Sometimes, the benchmark that is used is the performance of other individuals, either within the organization or outside, who are presumed to face the same productive environment and constraints and to possess similar capability levels.
In other cases, performance is measured relative to the individual's own previous performance.
- Forced distribution versus unspecified percentages.
When summary categories are used, a forced distribution (so many percent in category 1, so many in category 2, etc.
) may be employed, or the percentages may go unspecified.
Note that where forced distributions are used, there must be some sort of relative performance evaluation going on, even if only implicitly.
- Multi-source versus single-source evaluation.
In some systems, data are gathered entirely or largely from a single source, such as the individual's supervisor.
Other evaluation systems gather performance appraisals from many sources-customers, peers, supervisors, and so on-where each source is asked to appraise those aspects of performance that the source can reasonably be expected to know about.
- Multi-criterion versus single summary statistic.
In perhaps the majority of performance evaluation systems, all the data are ultimately massaged into a single summary rating statistic of overall performance.
Many dimensions of performance may enter into this statistic, but the final outcome is one-dimensional.
In some other systems, there is no attempt to formulate a single statistic.
In the middle are systems where there is a summary statistic that is very coarse (almost everyone is in the same category), grading many dimensions.
- Fine versus coarse performance distinctions.
Related to the previous point, a final dimension concerns how fine the distinctions are among levels of performance.
To take an academic example, at some universities, grades are reported as numbers, on a to 100 scale.
Elsewhere grades are "massaged" into broad categories like A, B, and C, in the majority of the group being lumped together into a single grade category.