Traditional Interviews Deliver Just Scraps

Suppose you didn't have the luxury of a grocery store in your neighborhood. You have to catch your own game if you want to eat. But game is scarce where you live and distances to food sources are far. You decide you need a good dog to scour the wilderness and bring you breakfast, lunch and dinner every day. So, you buy a hunting dog from a friend. After a brief get-acquainted period, you notice that 50 percent of the time the dog brings you a hearty meal, but the rest of the time he fetches back pieces, scraps and trash. He can't tell the difference though; it all looks the same to him.

Imagine our "owner" is an organization and the "hunting dog" is a recruiter whose job it is to find and deliver badly needed talent. Interviews are like the hunting dog that fetches above-average results 50 percent of the time and below-average results the other 50 percent. Yet they continue to be one of the most widely used hiring tools - despite the fact that they are no better than chance at predicting success or failure on the job. Recent research has shown that interviews are good tools to get to know the applicant. Yet even the most accomplished interviewer is unable to consistently measure the two most important traits that research studies have associated with job performance- conscientiousness and emotional stability.

Conscientiousness is reflected by how much the person cares about producing a quality product. Emotional stability is reflected by how much calmness and predictability people have toward coworkers, customers and problems in general. In study after study, both of these traits consistently have been associated with successful job performance.

But this is not rocket science. Common sense dictates that good employees are generally calm, easy to get along with and care about the quality of their work. You just can't measure this during a job interview because people "fib".

I don't believe that anyone knowingly hires "pieces, scraps and trash" but interview technology is as blind as the hunting dog - it does not discern the difference between the good, the bad and the ugly. Measuring this difference begins with an understanding of the job. Every job in an organization can be reduced to four general requirements. They are cognitive ability, organizing ability, interpersonal skills, and the right set of attitudes, interests and motivations - all areas that a face-to-face question and answer session just cannot measure with any accuracy.

Accurate hiring means that recruiters first must define how much of each requirement is necessary for successful job performance, then they must measure the results. Interviewers can use realistic simulations to measure interpersonal skills; tests, cases and exercises to measure organizational and cognitive ability; and special tests to measure attitudes, interests and motivations (AIMS). The actual choice and implementation of each hiring tool should be supported with a job study and validation study.

Does this sound like a lot of work? I agree. Traditional interviews are fine tools to use if you are picking people for the company softball team. They give you a chance to get to know the players who just want to have fun. But if your organization expects players to look and act like a world championship Olympic team, you need to toss out interviews and start using some measurement tools that work.

What are the tools that work best? Well, start by throwing out silly tests that measure communications style, leadership style or any other "style" taught in a training workshop, unless you can produce solid proof that test scores are strongly correlated with job performance. Not only will style tests serve as lawsuit bait, but they also won't get you the high producers you expect. By the way, "solid proof" is not someone's personal opinion or a casual observation of "what we need around here." Proof requires rating a large number of jobholders, comparing test scores to ratings, examining the results for adverse impact, and statistically correlating scores with job performance.

One-on-one simulations give you the most accurate results when measuring interpersonal skills for jobs like management, sales, customer service, or self-directed teams. They have about a 70 percent association with performance.

Jobs that require continual learning or serious problem-solving, like professional or managerial jobs, need case studies or mental alertness tests to determine if the applicant has the right stuff to deliver good decisions. The same is true for planning and organizing ability - cases and exercises. Both these measures have about 25 percent association with job performance.

Finally, you need a test that measures job attitudes, interests and motivations. Scores on this test provide insight into the hidden factors that affect a person's willingness to perform a job. This is a test designed for selection, not training. And it needs to be validated in a study similar to that described above. It has about a 5 percent association with job performance.

So, here are the options faced by the hunting dog's owner. Use interviews and continue getting pieces, scraps and trash for lunch. Or, assemble a comprehensive battery of simulations, cases, exercises, tests, and AIMs that deliver hearty meals about 90 percent of the time. It's your choice.