Are you making a difference? Managers must tell correlation from causation

Does eating chocolate make you smarter? In a study published in The New England Journal of Medicine in 2012, Franz Messerli demonstrated a correlation between the amount of chocolate which a nation eats and the number of Nobel prizes its citizens win. Unsurprisingly, Messerli’s findings were deliberately playful: he was not really attempting to assert that eating chocolate leads to prize winning research - he was illustrating the difference between correlation and causation. Chocolate consumption tends to correlate with national wealth, which also correlates with those factors which create an environment suited to Nobel prize winning research.

We might think that we already understand the difference between correlation (factors which co-exist and which may change together, but which have no causal link) and causation (one factor which drives another through a causal mechanism), and would never fall into the trap of mixing them up. But I think that if you are a manager you are at risk of mixing these concepts up all the time - I know that I am.

Consider your response to a crisis. Imagine that your project is running late, or your product quality is too low, or that one of your teams is consistently underperforming. What do you do as a manager? For many of us, our first instincts are to use all of the management tools which require our personal attention: we organise daily meetings, we run workshops and brainstorming sessions, we assign our best troubleshooters, and we ask for more reporting and risk management.

And sometimes, but not always, the problem is resolved. The project gets back on track, quality improves, the team performance picks up. We congratulate ourselves on a job well done, and make a mental note of the interventions which we believe worked. But I don’t think we ask ourselves often enough whether we really made a difference, or whether the resolution of the problem was correlated with rather than caused by our intervention.

Like all humans we are subject to confirmation bias and a desire for validation. We want to believe that we made a difference, so we look for confirmation that we made a difference. And, if we are the boss, and we start from the presumption that our direct intervention was needed, then we will usually find the confirmation that we are looking for.

Mixing up correlation and causation as managers may lead us into a habit of micro-management. We become convinced that things only get better when we directly intervene, and find ourselves intervening more and more. And we find out just how exhausting micro-management can be for teams and for managers. Team members wonder why they aren’t trusted to do their jobs, and why they spend all their time reporting and attending meetings. Managers wonder why they need to do everything themselves, and why their teams don’t seem to be able to get anything done on their own. This is the opposite of the world of autonomous, empowered teams responsible for their own performance which we think we want.

Is there a remedy for this unfortunate habit? I think that it is an easy habit to acquire, that it is driven by universal human cognitive biases and that we are all guilty of it from time to time. However, I hope that we can address it in a similar way to the way that science attempts to remedy cognitive biases and test hypotheses: we can acknowledge our cognitive biases and look for evidence which challenges our beliefs rather than confirms it.

How do we do this in the business world when we typically don’t have the capacity to run experiments (although we should create this capacity - the subject of a future blog post), and where we don’t have a controlled environment with a control group? I think that we should do what managers should do anyway, and seek feedback from our teams. We should be in the habit of running post-implementation reviews, post-project reviews and post-incident reviews. If we are doing them well, we make them blameless so that we can learn as much as possible. What if we also took the opportunity to ask whether our management interventions made a difference? What would our teams say if we asked them whether our daily meetings, additional reporting and extra scrutiny made a difference? And what would they say if we asked them what we should have done differently to help them?

Management requires humility. It requires the humility to be servant leaders to our teams, and to ask ourselves the question - did I really make a difference?

(Views in this article are my own.)