Adapting Analysis and Visualization to Our Fast and Slow Brains

The SIGMA employee book club is reading Thinking, Fast and Slow by Daniel Kahneman. If you haven’t read it yet, it’s a fascinating book on psychology and decision-making, and highly relevant to those of us who analyze and share data for the purposes of extracting insights and developing marketing actions.

As analysts we put a lot of time and effort into producing analyses to support decision-making efforts, but often don’t put in enough effort to synthesize and summarize the output. Unfortunately, all that hard work is lost if our audiences don’t get it or jump to the wrong conclusions.

And that brings me back to Kahneman’s book. He refers a lot to the two sides of our brain, System 1 and System 2 – terms originated by Keith Stanovich and Richard West.

  • System 1 is fast; it’s our autopilot and intuitive side that’s ready to make snap judgments
  • System 2 is slow; it’s our logical and thoughtful side that we use when we need to focus on something

According to Khaneman, when faced with a problem or decision, our brains will tend to take the easy route. While people considered to be more rational and logical apply System 2 thinking more readily, others do not. We should keep these systems in mind as we develop and share our analyses, since the presentation can influence the decisions made.

Three of my favorite insights follow:

Recommended for YouWebcast: 5 Growth Hacking Techniques to Increase Your Revenue in 30 Days or Less

Insight 1: The Law of Small Numbers

All too often we are asked to produce insights that explain why a certain sales team or individual produces such outstanding results. In many cases, this may simply be luck. Remember, if you toss a coin 1,000 times and you toss a string of 10 heads in a row in the middle of the exercise, it’s not significant, it’s just the luck of small numbers within a larger population.

Insight 2: Balancing Representativeness with Population Sizes

In one study, subjects are asked to guess which college major an individual is enrolled in. The description of the individual represents majors in which there are very few students enrolled. In most cases, the subjects give far greater weight to the description than to the population size of each major.

I have seen a similar occurrence with penetration rates versus total population. From the chart below, you might assume that this business has a very large number of customers in the Legal Services segment.

When we provide additional information about the underlying population, it shows that Legal Services has the smallest number of customers. This might lead us in an entirely different direction regarding strategy and actions for this business.

Insight 3: Regression to the Mean

Kahneman tells of air force flight instructors who praised pilots for great performance and yelled at them for poor performance. Those who were praised did more poorly the next time; those who were yelled at improved their performance. The instructors concluded that praise didn’t work but punishment did.

Regression to the mean tells a different story. It is natural that the follow-up to an excellent performance would be less so, and that the follow-up to a poor performance would be better, simply due to the fact that these two observations were likely outliers on a bell curve. When we are asked to apply a forecast to a dataset, we should take this into consideration. Rather than evenly applying a forecast percent, Kahneman suggests that we consider adjusting up for segments that previously underperformed, and adjusting down for segments that overperformed.

These are but a few of the insights found in Thinking, Fast and Slow. Read the book and you’ll find plenty more that will get you thinking about ways to improve your analysis, visualization and communication skills to support decision-making.

  Discuss This Article

Comments: 0

Add a New Comment

Thank you for adding to the conversation!

Our comments are moderated. Your comment may not appear immediately.