Promises, Promises: How to Identify a Bad Hiring Test (Part II of II)
In Part One I explained why a test user has to be exceptionally careful about trusting a vendor’s claim their test is suitable for hiring. Without due diligence on the part of the test user, junk tests lead to hiring too many wrong people and turning away too many right ones, amounting to an estimated 20% and 50% of yearly payroll. In this section, I’ll continue explaining how to identify good hiring tests.
Personality Versus Skill
One would think a score on a personality test would predict a skill (i.e., do high analytical-trait scores actually mean high analytical skills?). Unfortunately for people who rely on personality traits to make hiring decisions, no.
There are several theories about personality. Some people would argue that personal traits are organic brain structures; others argue they represent a description of how we want people to perceive us; and, still others argue they define how we perceive ourselves. I tend to believe people are genetically predetermined to think and act in certain ways which they continuously modify depending on life experiences. No matter what your theory, a personality score is no more or no less than a self-reported opinion that has extremely low correlations with specific job-skills.
Birds of a Feather
One thing self-reported tests do well is identify people who tend to think and behave alike (remember we are not talking about being job-skilled). If you hire people who score high in bleeding-heart tree-hugging, you will probably have a workforce that is so tree-focused it cannot be productive. Likewise, if you only hire people who score high on dominance, you will probably get cold-hearted autocrats who share a survival-of-the-fittest mentality. Hiring people based on personality scores can lead to a workforce that suffers from “group think.” Groups that think and act alike are often unable or unwilling to consider all sides of an issue. The inaccuracy of interviews produces wide performance variances. Adding more-accurate tests without knowing the consequences will narrow your variance between employees, but not in ways you might expect.
Legality, EEOC, and Discrimination
When a test vendor claims his or her test has been approved by the EEOC, you can be sure they do not know what they are talking about. The only time the EEOC “approves” a test is when a user gets audited; and, after thorough investigation, proved they followed Uniform Guidelines and the Standards for testing. The EEOC is not a certifying test body. Vendors who claim their test is EEOC-compliant should be required by law to wear big floppy shoes, rubber noses, and grease paint when they make sales calls.
When a test vendor claims his or her test has no adverse impact, you can be sure the test is probably not doing its job. Let’s use an example to explain how this works. Assume the National Forest Service has discovered people with big feet are 173% more efficient stamping-out forest fires than people with small feet. Incensed at this insult to their heritage, powerful lobbies of small-footed Munchkins converge on Washington claiming the Forest Service treats Munchkins unfairly. They demand legal protection.
Fearful Congressmen, anxious to get the stature-challenged lobbyists off the lawn and back home, pass a law declaring Munchkins a protected group. Is the Forest Service now required to hire all Munchkins regardless of their shoe size? No; they just have to show that big feet are a job requirement and business necessity for firefighters. Big-footed Munchkins will always make the cut, but at the Munchkin group level, there will always be disproportionately fewer Munchkins than other groups. Does the Forest Service have adverse impact? Yes. Are they justified? Yes. Are they guilty of violating any laws? No.
Now suppose a test vendor approaches the Forest Service. They claim their test is a great tool for selecting firefighters because it does not discriminate against Munchkins. Remembering the Forest Service has already shown big feet are 173% more effective at stomping out fires, do you think the test would actually identify the most physically qualified firefighters?
Biology Trumps Logic
A long time ago people lived in small bands. When someone met a stranger it was vital to quickly assess whether they were friend or foe. As you might imagine, people with poor judgment did not live long enough to pass on their clueless-gene. This genetic legacy lives on today every time a recruiter or hiring manager has an exceptionally strong drive to “get to know” a candidate. You often read articles describe it only takes 30 seconds to form a long-lasting opinion or hear managers claim they “know ‘em when they see ‘em.” These are deep-seated biological drives doing their survival job.
Unfortunately, getting to know someone is not the same as learning whether he or she is job-skilled. Just look at the employees who are hired. Don’t you wonder why only 20% of the salespeople generally produce 80% of the sales? Why good producers almost always make bad managers? Why top employees out-produce bottom ones by at least two-to-one? Or, why people are so adamant about using any other hiring tool other than face-to-face interviews? It’s biology.
What are some of the symptoms of a bad hiring system? A shortage of promotable candidates; excessively long training; attempting to fix broken employees by sending them to workshops; not knowing the details of the Guidelines for selection or the Standards for testing; an 80/20 workforce; failing to recognize interviews are weak hiring tools; neglecting biological drivers, or, not having the patience to work through about six candidates to find one who is qualified.
So what if a vendor offers you indemnification from lawsuits? Big deal. Lawsuits are often the least of our worries. Only a handful of hiring challenges ever get to court. Most of them are against big companies and most are settled. Hiring the wrong people and rejecting the right ones is much more expensive than lawyers. Organizations need to worry more about job performance. The cost estimates of poor performance ranges from 10% to 50% of payroll every year.
Here is the problem and the solution as succinctly as I can describe it. You have to be your own watchdog. There are literally hundreds of test vendors promising to solve your hiring problems. You can often screen out the blatantly unqualified ones by giving the vendor a call and asking, “Is your test specifically designed to predict future job performance; and, can you please send me documentation describing the development process and any validation studies you have conducted?”
HR and corporate attorneys, above all, must know the basics of the Guidelines and the Standards. In my experience, corporate attorneys know a great deal about what happens after the fact, but they cannot be expected to know how to use the Guidelines and Standards to minimize the potential for litigation. Organizations that follow the Guidelines and Standards maximize the potential for mounting an effective defense, minimize the potential for an EEOC challenge, and save the organization 20% to 50% of base payroll every year. That’s the highest ROI of any imaginable organizational program. The Guidelines don’t require you to follow them, but they are best practice.
Any system, device, interview, source, or process intended to separate qualified candidates from unqualified ones is a test. You don’t need to worry about lawsuits as much as you do productivity, turnover, and training expense. Be patient. It takes about six candidates before you find one who is qualified. An effective hiring system does not discriminate against skilled people. Skilled people come in every color, ethnicity, gender, age, and level of physical ability.
If you are unfortunate enough to be challenged, your management will probably demand to know you why you did not follow best practices. They might have resisted every time HR insisted on following the Guidelines and Standards, but don’t expect management to forgive you when the EEOC determines your organization did not follow them.
Original post created by: ERE Articles