Most competency models include knowledge areas, skills, abilities, and other personal attributes (referred to as “KSAOs”). The structure of competency models can vary, but they usually have competency titles and specific behavioral descriptors which define the competencies. Some competency models define the competencies in terms of levels, such as core, intermediate, and advanced.

Traditional job analysis has been around for more than 100 years, whereas competency modeling is relatively new. However, they both have essentially the same objective – to understand the requirements to perform in a particular job or role. But there are differences:

• Job analysis is heavily focused on tasks, whereas competency modeling is focused primarily on the KSAOs required to perform those tasks.
• Job analysis usually involves collecting a lot of statistical data from incumbents and their managers, whereas competency modeling uses more qualitative techniques.
• Job analysis is focused on understanding the current situation, whereas competency modeling is more future oriented (e.g., what trends are emerging that will impact skill requirements tomorrow?).
• Job analysis tries to cover all the details, whereas competency modeling intentionally deals with those elements that are most important to success.
• Job analysis describes the average performer, whereas competency modeling differentiates performance excellence from the average.

That depends on what you want to do with it. At one extreme, if the competency model will form the basis of a performance management system, then it should be as simple and streamlined as possible. At the other extreme, if the competency model will provide the framework for developing job specific training programs, then it should be very detailed. What many companies do is develop a fairly detailed competency model, with many competencies and behavioral descriptors, and then extract from that competency model the content needed for particular applications.

The answer here follows along the same lines as the answer to the previous question – it depends on what will be done with the competency models. Performance management systems typically cover a large number of jobs, whereas training programs are often focused on only one job or a few jobs.

The term “validated” is most appropriately applied to some instrument or process (e.g., a selection test) where the purpose is to predict some future event or outcome (e.g., job performance, retention). In this case, validity is expressed as a statistic, such as a correlation coefficient, which indicates how well it predicts. Competency models, on the other hand, are “valid” in the sense that they (a) include the most important knowledge and skills for successful job performance, and (b) don’t exclude key areas. So, validity is “built into” the competency model during the process of developing it. Rather than trying to establish statistical proof, Censeo’s approach is to ensure that senior managers, who understand the job and where it’s going, go through a simple process of reviewing the competency model, making changes, and then endorsing it.

Some consulting firms offer competency modeling tools where you select from an existing pool of items those which are applicable to a particular job. The main advantage is that it’s easy and fast. The main disadvantage is that the resulting competency models may not capture what’s uniquely important in the particular situation.