"In the past a leader was a boss. Today's leaders must be partners with their people..."

--Ken Blanchard

Home >> Motivation Initiatives >> Performance Management >> Stage 3

The Rating Scale

No matter what the game, one of the basic tenets of good sportsmanship is to let the players know how the game is scored. The same idea applies to performance management. General Statute 126-7 requires employees' overall performance to be rated on a scale consisting of five levels, with the middle level representing "meets expectations." The Office of State Personnel developed a rating scale, with labels and definitions for each point on the scale, which has been used for a number of years and continues to be used.

Therefore, the rating scale is the way employee performance is "scored." Supervisors are advised to share this rating scale with their employees during the performance planning discussion so that employees know well in advance the "rules" under which their performance will be graded. Similarly, it is important at the end of the performance cycle for supervisors to review the rating scale so that they can apply it faithfully as they complete their employees' performance appraisals.

Some agencies, however, have adopted a modified rating scale. Be sure to check your agency's policy for the definition of the rating scale and for the recommended procedures for deriving overall performance ratings for employees.

Rating Each Expectation

Here is a very practical approach to rating results and behavioral expectations. Imagine checking the temperature of different food items with a probe thermometer. The "neutral" reading of the thermometer is whatever the room temperature is. As soon as you insert the probe into a food item, the reading will go up or down depending on whether the item is heated or refrigerated.

Now, apply the same logic to rating performance. You start out with the "rating meter" at "meets expectations" (the midpoint of the rating scale). Looking at the information for the employee's first expectation, the meter will go up or down depending on how actual performance compares with expectations. Performance that meets expectations does not move the needle; it remains at "meets expectations." Performance that goes beyond expectations drives the meter up; performance that fails to meet expectations drives it down.

This is a highly effective approach to rating. As you move from one expectation to the next, reset your "rating meter" to the neutral ("meets expectations") position and let the information concerning performance drive the rating for that expectation.

Arriving at an Overall Rating

North Carolina State Government requires that employees receive an overall performance rating annually. How do you go from a set of ratings for each expectation to a single rating that represents overall performance?

Different agencies may provide specific instructions for arriving at an overall rating. Check your agency's procedures. The basic logic is to assign an overall rating that best balances the mix of ratings given for the individual expectations. Consider this example using the official state rating scale labels: If the majority of expectations are rated "Very Good" along with one "Good" and one "Outstanding," the overall rating should clearly be "Very Good."

However, different expectations carry with them different priorities. Greater weight should be given the ratings assigned to high priority expectations. Furthermore, some agencies' procedures require an intermediate step in which you first assign a summary rating for all results expectations and a summary rating for all behavioral expectations, and then combine those two summary ratings to arrive at the overall performance rating. Yet another procedural step required in some agencies is to give more weight to results expectations than to behavioral expectations - usually twice as much weight.

The main point is to arrive at an overall rating that best captures the overall performance of the employee and that is clearly supported by the evidence.

Using Weights

To try to achieve consistency in assigning overall ratings, some organizations convert the ratings to numbers and the priorities to numeric weights and then compute the overall ratings. This is mathematically precise, yielding overall weighted ratings that can be carried out to two or three decimal points. This approach is best used when results expectations are expressed as metrics. When used with more subjectively defined expectations, such as responsibilities or behavioral expectations, weighting has the unfortunate byproduct of lending a false air of precision to the task of assigning overall ratings.

If this mathematical approach is used, the computed rating must be converted back to the non-numeric North Carolina rating scale. Care should be taken in making this conversion and agencies should provide a key for interpreting the computed averages. For example, if a 1-to-5 numeric scale is used, it should be clarified that a weighted overall average below a 3.0 is "Below Good" (in other words, you would not round a 2.75 up to a "Good," even though you would mathematically round a 2.75 up to a 3). In addition, it would not be realistic to require a 1.00 average for an "Unsatisfactory" rating, or a 5.00 average for an "Outstanding" rating. Under these rules, for example, an employee would have to have "Outstandings" on every single expectation. Instead, a range should be defined for "Outstanding" (such as, "greater than 4.50").

If done properly, the weighting approach can help to ensure that all employees' overall ratings are consistently assigned, based on the ratings of their multiple expectations.