Friday, March 26, 2010
Regardless of whether the organization adopts a custom built or readymade competency model, at some point you will develop a tentative or preliminary design. The organization may wish to have a design that is testable on a small group of employees, a specific unit, product or geographic location. A preliminary model provides the organization with a working tool, which is still under development yet one that, will permit testing and updating based on small-scale outcomes. To test the preliminary design there are several qualitative and quantitative methods available, including subject matter expert focus groups, surveys, as well as various statistical techniques. The degree and sophistication of the testing will depend, on large part on the organization’s unique needs, its resources, and comfort level with statistical analysis.
Using the example of the long distance truck driver, we may have already used questionnaires, surveys, and focus groups to capture performance related information from drivers, managers, mechanics, and others. With the performance information collected from these subject matter experts, we can develop job descriptions, interview questions, performance statements, and training programs permitting us to manage top performers. We also have a means to take the “meets” employee to the next step and reach an outstanding level of performance through the training and development tools built into the model. Finally, we have a tool that can differentiate the poor performer from others and work towards “re-potting” or releasing those individuals.
Now that we have a preliminary model, we can take that design, return to these same subject matter experts, and determine if they believe we have captured the attributes of the drivers’ performance. On a qualitative level, we could ask subject matter experts to indicate that our preliminary model of knowledge, skills, abilities, and behaviors does or does not point to outstanding, average or poor performance. Using a quantitative process, we could take those same individuals and ask them to “rate” each factor of our model on a scale of 1 to 3, with 1 indicating the factor measures the performance, 2 somewhat measures the factor, and 3 does not measure the factor. We could expand this scale to a 5 step rating factor of from 1=Completely Measure to 5 Does Not Completely Measure. Several statistical tests exist to measure the degree of fit between the model and the ratings givens by the subject matter experts.
Over time, the deployed model, if valid, should track the employee’s individual job performance. Using the design to select a new driver or promote a forklift operator to driver should increase the likelihood they will be successful as a driver. Rejecting an applicant to be a driver using the competency model should also confirm the individual would be less likely to have been successful in the organization as a driver. The model should also provide a road map of what it would take to develop an average driver into an outstanding driver. It then becomes an organizational decision whether the investment in the effort would yield the desired results.
Wtether we use qualitative or quantitative methods to test the design, at some point we will learn that our model will require updateing to bring it into alignment with the “perceptions” of the subject matter experts in order to achieve the degree of desired results, increased performance. This could occur during our pre-implementation testing or after the model has been in use for some time. During the early days of a new model’s implementation, the organization should watch for a signs the design fails to live up to its expectations. Should the organization have gotten the model wrong, continued use without corrections could harm the creditability of the organization and the designers. Remember, the goal of a competency model is to assist the organization in the selection, development, management, reward and recognition of those knowledge, skills, abilities, and behaviors that will lead to success for both the business and the individual.
