- By Ian Bradley
- 0 Comments
In my executive coaching practice in Montreal, many of the business issues that I discuss with my clients concern prediction. These prediction problems are rarely conceptualized as such, instead they are couched in jargon of the particular business using words such as “forecasting,” “planning” or “analysis”. However, the underlying psychological process is prediction.
For example, in the retail clothing business, senior executives and their teams make predictions at least four times a year about what fashions will be in style six months down the line. Credit teams in financial institutions make routine predictions about a potential lender’s ability to repay his or her financial obligation. Marketing executives attempt to determine the public’s response to their newly created campaigns.
Predictions are not only routine in most executive jobs, but they are also extremely important since they can be the main drivers of the company’s success or the source of an employee’s bonus.
Now here’s the sobering news from psychology. We, humans, are not very good at this type of cognitive activity. My profession of Clinical Psychologists went through this epiphany over forty years ago when a generation of dissertations tackled the following question:
Who’s better at predicting things about patients: “ expert clinicians or machines?
The machines were represented by statistical algorithms that used historical data to make predictions about such things as;
- which patients will improve with psychotherapy
- who is will likely get discharged earlier from the hospital
In short, which was better: clinical or statistical prediction?
Well, after numerous studies, the verdict was resounding- statistical prediction based upon tested algorithms were better than the trained experts. It was a hard pill for us psychologists to swallow but we survived and moved on.
What are the implications for business?
Firstly, recognize that prediction is tough. Research in many fields has shown that people making the predictions are notoriously over optimistic about their abilities. In fact, the more of an expert they are – the greater their certainty, but not necessarily their accuracy.
Secondly, require that business people making predictions be explicit about the decision and what is underlying their prediction. What factors did they consider? How did they weigh these factors? What things did they not consider?
At one point in my career, I was running a large department of clinical psychologists who routinely gathered at weekly Rounds to present and discuss cases. The meetings were interesting with a lively debate about the likely outcome of various therapeutic interventions. However, the meetings also reflected the old joke that two psychologists in a room yields three different opinions. In the clinical meeting, all the opinions were offered with justifiable reasons, some even citing relevant research. But, almost invariably, each proffered statement was countered by an opposing equally elegant point of view. It was like a merry-go-round, great fun, but when the music stopped everyone was exactly in the same position that they began but further convinced in the rectitude of their own view.
I decided to change things by introducing a bit of reality feedback. Henceforth, at the end of the meeting, each psychologist was required to document his or her prediction in a one-page sheet that recorded the prediction and the underlying reasons, listed in terms of importance. Then, six months later the patients were presented again, and the predictions evaluated. The results of my changes were interesting. People began to be more cautious in their judgments, they began to consider more information and listen more to the ideas of others. I didn’t stay in the position long enough to see if feedback had any long-term effect on accuracy, but I suspect it did.
Here’s the bottom-line for business.
- Recognize when you are making predictions.
- Then ensure that people making the predictions are explicit not only about the outcome but the underlying decision-making process.
- Finally, collect data and feedback the results in a way that employees can learn from their successes and failures.