The problem with ‘evidence’ in HR

I admit it: I like the evidence based approach to HR. I like it because it appeals to my sense of trying to do the right thing in the right way. And I like the idea of using other people’s findings and applying them to my own work.

The evidence based approach essentially tries to apply scientific standards of causality and evidence to demonstrate how intangible human capital can be observed and shown to add tangible business results. It questions common beliefs and demands that the practitioner use methods which is proven to work. Bingo. It is an approach taken from Evidence Based Medicine (EBM), which has been around for more than 20 years.

Evidence in HR

The problem I have found with a lot of the “evidence” in HR – especially in emerging areas (such as HR analytics) – is that it is produced by people who have a vested interest in the results.

Imagine for example that you were reading a study, which concludes that a particular drug is really effective. The article highlights all the benefits (and a couple of irrelevant disadvantages) and shows how this is a new important step in curing this particular illness. It compares  the product with similar products and shows that the effect is much better. Patients report higher satisfaction levels.  Everybody is happy.

Would it make a difference if the study had been made by the producers of the drug?  Of course it would. You would question the findings. You wouldn’t necessarily think that they were lying or making up stuff. You just wouldn’t trust it completely. And so you shouldn’t. The conflict of interest is obvious.

This is a major problem in HR. Take talent management. I have seen many studies showing high ROI on TM, correlation studies which show that companies with great TM processes have higher profit, lower employee turnover and higher engagement levels. I have seen studies showing TM leads to higher market share etc. The problem is that all of these studies are made by people and companies who have a clear interest in the results; consultants, software vendors, authors/publishers etc. I have only seen very few really good research (double-blinded, longitudinal etc.) about Talent Management.

Why is that? I believe there are three primary reasons;

  1. It takes time (and resources) to make great research. A great piece of double-blinded, longitudinal research takes many years to produce. Unfortunately many don’t have the patience for this.
  2. There is too little. Because clients like data and research and there is too little available from neutral researchers, people and companies with vested interest will have to produce it themselves.
  3. The conclusions are ‘better’ if produced by consultants. When consultants conclude something it goes something like this “we can conclude that better talent processes lead to the creation of significant shareholder value”. When a university professor concludes something, it goes something like this: “Our research point to several processes which may benefit the talent as well as the company but the effect depends on the context and other unknown factors”. Which one sells?

The fact that much of the evidence is produced by people with a vested interest does not make the studies or conclusions wrong. They may be as correct as they come. The problem is that as long as there is too little ‘proper’ evidence it is difficult to really trust it.

This is a problem because if we want people outside of this profession to take us serious – such as CEOs and CFOs – we must be able to show them data which cannot be questioned. So lets make those surveys and that data which can do that.

5 thoughts on “The problem with ‘evidence’ in HR

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s