We build structured hiring processes for our corporate clients. Some of the processes we see, prior to the revision, are long-drawn out systems of multiple steps and the collection of data that one would be hard pressed to review, much less show that the step was job related. The issue of valid selection criteria is emerging as a area of interest to employment lawyers. Today, you need to be able to show the validity and business necessity of any selection criteria you use. You must be able to document that the standards you use for selection are in fact essential to the job and/or predictive of success on the job, while not causing disparate impact on protected classes. You must be able to show through statistical analysis of your workforce data that the selection criteria is in fact related to factors demonstrated by your own workforce at particular established levels.
Criteria Validation: There is a lot of new information out there to be concerned with in devising your selection criteria. If you can answer these questions with a “yes” you are probably on the right track:
- Did you do Job Analysis on the job with several individuals (high, medium and low performers) to determine the essential functions of the position, the context and working conditions of the job, the technical skills or knowledge needed to do the job, the ADA standards required of workers?
- Do you have a compliant job description which is up-to-date and incorporates essential functions of the job based on the information you gathered in Job Analysis?
- When you established your job posting for the position, did you ensure that you are using criteria that is evident in the current work force?
- Do you have multiple steps in the process that produce job-related information on the candidate upon which you base your decision?
- Do you rely too heavily on a single phase of the process for making your final decision?
- Do you handle each applicant the same way and does every qualified candidate go through every step?
- Is your selection process free from steps that cause disparate impact within protected groups?
- Do you know whether your selection criteria is valid? And by this, I don’t mean that you pulled a focus group together to look at the new criteria and everyone agreed it was fine.
Establishing Valid Selection Criteria: In these times, employers need to be particularly careful in managing the selection process because people are so desparate for work and many candidates are willing to assume significant risk in challenging an employment decision which negatively impacts them.
Here is a recent posting on the Department of Labor website regarding the Uniform Guidelines for Selection:
The degree of relationship between selection procedure scores and criterion measures should be examined and computed, using professionally acceptable statistical procedures. Generally, a selection procedure is considered related to the criterion, for the purposes of these guidelines, when the relationship between performance on the procedure and performance on the criterion measure is statistically significant at the 0.05 level of significance, which means that it is sufficiently high as to have a probability of no more than one (1) in twenty (20) to have occurred by chance. Absence of a statistically significant relationship between a selection procedure and job performance should not necessarily discourage other investigations of the validity of that selection procedure.
While you might wonder how you are going to accomplish this, it might not be as difficult as you think. What this particular guideline is requiring is that you can show the relationship between the criteria or steps you use in the selection process and actual performance of the job. For example, if you do an analysis of your workforce and find that only 15% of the workforce has a 4-year degree and you are requiring the degree in your job posting or in the job description, you have a problem. However, if you say that a degree is desirable and will accept relevant work experience instead (providing this is true of your analysis of the existing workforce), your criteria should pass.
If you say you require a particular level of math skills, such as would be demonstrated through the administration and scoring of a test of basic math, but the job does not require the use of basic math, here again, you have a problem.
Begin by asking what particular skills are required in the job. Does the person need to write memos? You can find out by doing some job analysis with people in the job in your own company. If you find that this is an essential part of the job, then you might assess writing skills. But be sure that the content of the test is job specific and that the scores you use for establishing the pass rate are statistically significant at the .05 level with the scores demonstrated by existing people in the job. Unless you are up on statistics and test construction, I wouldn’t advise that you attempt this yourself. There are plenty of good consulting firms you can hire at a reasonble cost to do the work for you.
Personality Tests: How about behavioral measures, such as personality skills assessments? Again, avoid any measures that incorporate non-job related content. If you can show the relationship through statistical analysis between the results of the assessment and its ability to predict success on the job at the required level, then you are on a far better footing than if you simply use the test provider’s documentation of validity.
Many test providers give you an extensive validity report that shows that there is no disparate impact caused by the instrument or that the instrument is valid and reliable in measuring what it purports to measure. The problem is that the OFCCP or the EEO won’t allow you to rely solely on the documentation for the instrument, unless you do the diligence of validating your selection standards within in your own workplace. And of course, if there are holes in the test documentation, they will use that in making their case against your invalid selection criteria.
There are a few exceptions to this on-site study guideline, such as you have too few people in a role to do a study, performing a study is virtually impossible since everyone with the job does considerably different work, the role might be new to your organization, AND the job you have is similar to one in which the validation study work was originally completed. However, you still will have the obligation to monitor the impact of the assessment regularly, to show you did some Job Analysis after a period of time to understand the work context, and that you have a good and up-to-date job description. If you are a good sized organization (over 50, although many regulations are applied at the 15 or above level), you need to look into the establishment of valid selection criteria and show how the criteria is statistically related to workers or the work actually being performed.
Tomorrow, I will discuss validation of interview questions, so tune back in.
If you would like additional information on validating selection criteria, visit the Department of Labor at either the www.eeoc.gov/policy/docs/factemployment_procedures.html or www.ofccp.gov/policy/docs and you will find some very useful documents there to enable you to understand the issues more fully.
Look for our upcoming Webinar on Establishing the Validity of Selection Criteria. The sign-up sheet will be posted on our blog.
