By David Creelman

We learn the most about evidence-based management when we look at real cases. Here is nice one:

An organization routinely hired large numbers of people for an important, but low level, job. They wanted to have some kind of automated screening that would weed out poor candidates, leading to better quality of hire and saving managers a lot of time. They engaged a leading assessment vendor to create a scientifically-validated screening tool.

The tool seemed to work. The managers saved time because they did not have to interview as many people. The people they did interview (only the ones that passed the screening) were strong candidates. The program looked like a success. There was only one thing that bothered the recruiters: the assessment took a long time for a candidate to complete, they worried that good candidates would dislike the effort required. In fact, a certain percentage of people who started the test did not bother to finish. Yes, the assessment was screening out poor people, but maybe they were also losing the cream of the crop.

What would you do?
Too often organizations resolve this kind of question with an argument. Some people love the new assessment and the time it saves. Other people are uncomfortable at removing the human element and fear the process is driving away the best candidates. There are good arguments and strong emotions on both sides of the issue. Normally, the decision is made on the basis of who is the most senior or who is most tenacious. That is not what happened here.

This organization had an evidence-based mindset. Some recruiters had forwarded a reasonable hypothesis: that the burden of the assessment was causing the best candidates to drop out. Rather than argue about this, they decided to test it. The recruiters tracked those people who dropped out, and followed up with a telephone interview. It turned out that these people were dropping out because they were a poor fit, not because they did not like the screening process. The assessment was working.

It is easy to walk away from this story thinking “assessments are a good thing”, however I want you to concentrate on the deeper lesson that “evidence is a good thing”. The company might have found out that the people dropping out were indeed high performers, in which case they would have changed how they used the assessment. Organizations need to move away from arguments based on opinion, to the evidence based mindset of “good question, let’s find out.”

Another example comes from Google. They liked the idea of doing a lot of interviews to make sure they got absolutely the right people. This could get out of hand with applicants going through 10 or more interviews. Some people in HR questioned whether this was too hard on applicants and a waste of the company’s time. The evidence-based response to this issue is: “Good question, let’s find out.” So instead of debating whether this intensive interviewing was integral to their culture, or whether getting the very best people was crucial, they simply checked to see at what point more interviews ceased to add value—and they found that there was rarely any point in exceeding four interviews.

The idea of gathering data is hardly revolutionary; the point to keep in mind is that organizations often fail to do so. Frequently, decisions that could be made based on fact are made based on opinion.  We need to cultivate a culture that asks for data and that encourages experimentation.  Instead of taking and defending positions people need to form hypotheses and, if they seem reasonable, say “Let’s get some data to find out if this hypothesis is correct.”

David Creelman is CEO of Creelman Research, providing writing, research and speaking on human-capital management. He works with a variety of academics, think tanks, consultancies and HR vendors in Canada, the U.S., Japan, Europe and China. Mr. Creelman can be reached at dcreelman@creelmanresearch.com.<