Using Competency Models to Assess Employment Candidates and Employees

Validation of Selection Criteria
When determining the validity of competencies for use in either candidate selection or employee assessment, there are two distinct kinds of validity that should be considered. These are:
Content Validity
Empirical Validity
Content Validity
Content validity refers to the knowledge, skills, and abilities needed for successful job performance by virtue of the work itself. A good question to ask when establishing whether a given competency is content-valid is, "What is the degree of impact this competency has on job performance? How critical is it?" Another factor to consider is, "How often is this competency used in performing the work"?
When establishing selection criteria, from a legal defensibility standpoint, it is best to focus on "skills and abilities" rather than "knowledge." The reason is, courts will question whether or not the employer provides (or could have provided) short-term training to develop the necessary knowledge. As an employer, this argument can be tough to defend.
In order to be considered valid, job-content based criteria must be "verifiable." There needs to be some statistical evidence showing the evaluation criteria being used is "essential" or "critical" to successful job performance. In order for criteria to be considered "verifiable," the employer needs to show there has been an independent validation of the data by multiple raters (usually 3 or more) who are job knowledgeable.
Job knowledgeable persons, sometimes referred to as an "expert panel," are persons who have actually performed the work, but can also include others who supervise the work and are intimately familiar with the requirements of the job.
Process for Establishing "Content Validity"
The recommended process for establishing content validity is as follows:
Create an expert panel, 3 or more persons who are "job knowledgeable"
Choose persons who have actually performed the work as well as
those who have supervised it
Panel should always include two or more current (or past) job incumbents
Review job description to determine key challenges, responsibilities, and expected results.
Use group process to define a "preliminary list" of competencies felt to be critical to job performance. (HR Technologies, Inc. provides a "competency lexicon" and software to facilitate this process.)
Have each panel member independently (without discussion) assign a weighting to the listed competencies using a 5-point weighting scale, thereby reflecting the "relative" importance of each competency to successful job performance.
Use group process to fine-tune these weightings and arrive at a final, consensus weighting for each competency.
Finalize the competency list, with the group agreeing to eliminate lower-weighted criteria and arriving at a final competency set (i.e., the "assessment model").
Focus the group primarily on selecting competencies that are "skills and abilities," eliminating those that are knowledge-based (especially those competencies easily addressed through short-term training).
Use the group to establish "behavioral anchors" (i.e., definitions or examples of performance levels) for each competency contained in the selection model, thus providing an objective, uniform scale for assessing employees and determining their competency "level."
Empirical Validity
Empirical validity requires that evaluation criteria be "statistically verifiable". This means the employer must demonstrate there is a high degree of mathematical correlation between a given competency and successful job performance.
If, using performance evaluation data, you can demonstrate that a high percentage of top performers possess a certain competency and can further demonstrate this competency is relevant to their high performance level, you can make a pretty strong case that the competency in question is "empirically valid".
In the case of sales positions, for example, if you can show there is a high degree of correlation between sales volume and certain competencies, those competencies are considered to be empirically valid. They are statistically relevant to sales performance. In such cases, you need to be able to show that those who posses these competencies consistently achieve high sales volume (i.e., they consistently beat sales quota). On the other hand, those who lack these same competencies consistently deliver lower sales volume (i.e., they are consistently below quota).
Put simply, to test for the validity of the competency selection model:
Employees who fit the model consistently beat sales quotas
Those who don’t fit the model consistently fall short of quota
Competencies that have been empirically validated are highly defensible from a legal standpoint and can easily be demonstrated to be valid criteria for either candidate selection or employee assessment.
Where possible, therefore, empirically validated competencies need to be identified and added to those that are job-content based to arrive at a final competency model.
Employee Assessment & De-selection Process
When using the competency model for de-selection purposes, it is important employee evaluation be a "multi-rater" process. Preferably there should be 3 or more raters, each having first-hand knowledge of the employees to be rated. This should obviously include the employee’s immediate supervisor and other management personnel familiar with the quality of the employee’s work.
Additionally, where available, the last two performance evaluations for each employee should be provided to the members of the evaluation team, as well as any other performance-related documentation contained in the employee’s human resources file. Also, when using empirically validated competencies, you should have the last two years of measurable performance data available for use in the assessment process. For example, if a sales employee, empirical data might include:
Sales volumes for last 2 years
Sales volume increases for last 2 years (dollars and percentages)
Established sale quotas (if any)
Performance against quota (i.e., percent of quota achieved) for last 2 years
The following process is then recommended for use by the evaluation team in the actual evaluation of employees:
Evaluation team reaches consensus on a uniform employee evaluation scale (usually 1 to 4) based on employee competency level.
4 = Superior level of competency (has few, if any, peers)
3 = Above average level of competency (competency is higher
than most)
2 = Average (acceptable) level of competency
1 = Below average (unacceptable) level of competency
Where they exist, pre-established "behavioral anchors" (i.e., pre-determined performance levels for each competency, along with their respective definitions) should be used in the place of the above general, numerical rating scale.
Team members independently (without discussion) evaluate each employee against the competency model and record their ratings
Independent rating scores are then averaged for each competency, and the employee then receives a final score by multiplying the average evaluator rating for each competency times its pre-assigned weighting.
A "cut-off score" is then established so that it yields the number of employees needed to reach the program’s objective.
The cut-off score is applied and employees whose scores are below the cut-off are identified and placed on a "target list."
Adverse Impact Analysis
Once employees are identified for de-selection purposes, an analysis needs to be performed to determine whether there has been adverse impact on any particular "protected class" employee grouping (i.e., race, creed, color, age, sex, national origin, or disability). Where the percentage of a certain protected c