This is a critical stage in the process. There is a tendency on the part of many supervisors to think of easily measured results and then define performance in terms of those results. In fact, many of the most important results are difficult or impossible to measure. If performance management is to work, it is essential that it concentrate on relevant performance results, not merely those that are easily measured.

There are two kinds of results. The first kind occurs as the direct outcome of duty completion. The second kind is referred to in the system as "symptoms." These are measures which may be indicative of performance on a duty but which are not direct outcomes of the duty.  For example, a supervisor experiences high turnover of direct reports because they are promoted into other Rutgers positions.  This can be taken as a sign of high performance in the managerial responsibility of "Training and Development of Assigned Staff." Particular care must be taken when choosing results to be used as symptoms to make sure that the measure chosen really is a symptom of desired performance on a duty and not merely some measure that is easy to get.

To develop a list of key results, refer back to the list of key duties. Focus on how it is known a critical duty has been completed. Are there deliverables? If so, what are they? Could an incumbent manipulate the result and still do a poor position on the duty? An analyst, for example, might produce any number of reports, but they may be worthless as an input to decisions. Is the incumbent solely responsible for the result? Much work in Rutgers is based on team or group effort, and a group member might perform at very high levels but be pulled down by other members of the group. Likewise, a team member may disguise poor performance by hiding behind the efforts of the group. Finally, external forces beyond the control of the employee may distort many results.

How will the result be measured? Even if a result is an appropriate measure of performance, it may be difficult to get any concrete measure of it. If a supervisor has a staff member in charge of alumni relations, and the result wanted is "satisfied" or "happy" alumni, getting a measure of "satisfied" or "happy" will be difficult. Surveys of alumni satisfaction could provide measures, and it is possible that a measure of that aspect of performance is important enough for the supervisor to do so.

Results associated with an employee's performance generally have two characteristics. The first is quantity (quality is assumed), and the second is time. Any specification of results desired should include both. A standard result for most duties will be framed in terms of the performance cycle. That is, if performance is being planned for the next year, the result specified will be for the year. Some results may have shorter time specifications.

To summarize, when drawing up the list of results relevant to the position:

Consider the duty list developed for the position.
For each duty try to think of one or two major results which really matter. It should be difficult for the incumbent to achieve the result as defined and measured and still not complete the duty satisfactorily.

Consider duty priorities as results are developed.
There should probably be no more than 8 to 10 results listed for any position. Many positions have no readily discernable results or outcomes. In these cases, behaviors can substitute for results.

Ensure reliable measures of each result can be attained that would be difficult for the staff member (or anyone else) to distort. Consider the time framework relevant to the result.

Now you are ready to write performance standards for the position.