How to progress from judgment to understanding

Kai-Alexander Schlevogt | Wed, 06/18/2008 10:45 AM | Management

Imagine a fairy promises you to fulfill one wish. What would you ask from her? If you are smart, you will request perfect causal understanding!

Discerning the true causes of events and behavior among the "fog of war" is a necessary condition for leadership success. However, many executives are not good at causal reasoning. For example, when judging others, they tend to underestimate situational factors.

As a consequence of erroneous inferences, leaders are prone to overlook powerful levers of change. Imagine you ascribe voiced opinions to attitudes instead of circumstances. From the few negative comments of your employee, you may jump to the conclusion that his general attitude is bad. Then it is more likely that you will fire him than that you will implement measures aimed at changing the task environment, for example. You thus might miss the opportunity to achieve breakthrough performance improvements, which might be possible by simply modifying the working conditions a little bit.

Here are five recommendations on how to improve your causal reasoning which will transform you into a "scientific leader". Through an iterative process of deduction and induction, you will achieve both analytical breadth and depth. Among other things, the proposed steps can be used to lead yourself, explain events and behavior, as well as to evaluate the causal reasoning of others.

1. Consider both personal and situational forces at work

Many leaders switch to "autopilot mode" when they search for the driving forces of events and behavior. By default, they opt for one favorite type of explanation that makes sense for them or simply is convenient, possibly because it fits with prejudices. Yet what holds true in evolution, applies in the context of leadership, too: Diversity increases the chances of survival!

Therefore, whenever you want to truly understand something, open your mind to a broad array of explanations. You might discover multiple forces at work, which may even pull into opposite directions. The explanation for the observed result then would hinge on the balance of these forces.

At the initial stage of causal reasoning, you should emulate scientists and engage in what I call "centrifugal thinking", following different trails instead of quickly zooming in on one solution.

One way of diversifying your explanatory options is to always ensure that you consider both personal and situational factors. You might even set up the following "red alert" system: Whenever you spot yourself inferring dispositions from observations, you should immediately use this insight as a trigger for generating a list of contextual driving factors. A good approach to avoid jumping from observations to personal blame is to think like an impartial judge and ask: What are the mitigating circumstances here?

Conversely, the red lights should go on when you find yourself or others blaming external circumstances, such as unfavorable winds in a game of tennis, for personal failure.

2. Turn assertions on their head

It is particularly important to think out of the box. What I call "counter-thinking" is the process of turning assertions on their head. The most famous example from science is the claim that the earth rotates around the sun, contradicting the long-held belief that our planet is the center of the universe.

The first method of counter-thinking entails exploring the opposite of your current explanation without changing the causal ordering. Imagine your salesman did not close a deal with a customer. First, you have to decide how you interpret this observation itself. Did he lose business or did he consciously turn down a customer? Let us assume you framed this observation as a failure. Then you are likely to ascribe it to lack of ability or effort, for example. If you perceive it as a conscious rejection by the salesman, your explanation might be that he hates the company. He may engage in sabotage, deliberately turning down customers. Then imagine that the opposite is true: The employee chose to forego the business because he actually loves the company and thus is eager to act in its best interest.

A variant of such counter-thinking is to leverage the insight that correlation is never a proof of causation, and reverse the causal order. For example, instead of arguing that share prices went down because of unstable political leadership, you could argue that the government is shaky because of market turmoil.

3. Develop a logical case for your hypothesis

As soon as a leader has generated a hypothesis, he needs to check whether he can make a convincing case for it, which is supported by facts. As regards the salesman, you might reason as follows: Your employee understands that many companies lose significant amounts of money especially at the front, because salespeople lack "margin discipline". They often offer a customer generous discounts and, under the radar screen of corporate controllers, regale him with gifts to increase sales at the expense of profit. In our case, the salesman might have thought that the price demanded by the customer would have led to an unacceptably low contribution margin. This may have prompted him to turn down the business for reasons of prudency.

An alternative way of reasoning, starting from the hypothesis that the business was lost rather than voluntarily surrendered, could take into account the context. One logical argument would be that the salesman lacked adequate resources to reach the corporate objectives. Or the competition is ferocious and the product lacks a compelling value proposition. You could also argue that one of the key global accounts switched suppliers for reasons beyond the control of the salesman.

4. Exploit the power of contrast to choose the correct explanation

To determine which of the hypotheses and supporting reasoning are correct, you have to engage in empirical analysis, which focuses on the facts. One approach is to use my "Five P-framework for effective causal reasoning", consisting of past, project and peers to find supporting and disconfirming evidence. Through various contrasts, you will be able to ascribe observations to either person or periphery, or both.

Here is an example of how you can use the framework to understand the salesman's behavior: First, you need to check whether in the past, he was successful. If this is the case, the recent performance drop might be due to unusual events beyond his control. As regards the "project" dimension, check whether he is experiencing problems only with one specific product or across your entire range. If his sales are low in all categories, it might be he who is responsible for the failure. Besides, you need to examine how peers are faring. Has the performance of all your salesmen declined? If this is the case, the loss of business might stem from systemic factors.

Once you have identified whether the problems are due to personal or situational factors, or a combination of both, you need to use the 5-P framework to drill deeper, identify those factors that you can control and prioritize these levers in terms of impact. For example, if you have zoomed in on personal factors, you need to collect facts to discover the specific dispositional root causes of poor performance, such as his personality, incompetence or insufficient efforts. If you chose to focus on ability, you have to pinpoint which specific competencies need to be improved and which measures would be most effective to do so.

As regards the periphery category, you must examine situational factors, such as the work environment. You even have to check whether your own leadership weakness has caused or at least aggravated the problems! You also should broaden your perspective and determine whether industry structure and broad trends, such as increased environmental awareness, could have caused the bad performance.

5. Use experiments to validate results

An excellent method for validating your choice of explanation and thus gaining correct causal understanding is to conduct controlled experiments. Like a scientist, you should try to change one variable while holding other factors constant and examine its impact. For example, if you lack data on the "project" dimension of my framework, just give your employee a broader but still manageable range of products to sell and see whether he fares better in other categories. In case you cannot conduct controlled experiments, you might be able to benefit from natural experiments, observing the outcome caused by factors that changed without your intervention.

If you find that, after your comprehensive and granular analysis, there is more than one plausible explanation for what you have observed, I advise you to be positive and optimistic, choosing the rationalization that offers the strongest upside potential. For example, if you cannot empirically disconfirm that situational factors caused the performance problems, you might well give your employee the benefit of doubt. Your strong belief in his personal qualities could become a self-fulfilling prophecy, since it will probably increase his self-confidence and ultimately his performance, too.

Alas, you rarely encounter a fairy who endows you with magic gifts. But if you heed the advice on how to improve your causal reasoning, you will not need a supernatural intervention to become a "scientific leader", who does not only judge but also understands!

"Prof. Kai on Strategic Leadership" Column Number 4. Kai-Alexander Schlevogt (D.Phil. Oxford) is a professor of strategy and leadership at the National University of Singapore (NUS) Business School and author of The Art of Chinese Management (Oxford University Press). He can be reached at schlevogt@schlevogt.com

Correction

In column No.3 by Kai Alexander Schlevogt titled The Leadership quest for plausible explanations, we failed to mention the writer's name in the byline. We apologize for the error. --Supplement editor

Comments (0) | Post comment

A | A | A |