Thinking, Fast and Slow

By Daniel Kahneman

Book Note by Christine Drew and Richard Erdmann

Published in 2011, this book continues to change how we understand decision-making and behavior. It is the basis for the early blogs in Learning about Leadership. Rather than attempt to review a 400 page book, we decided to provide a taste of why you should read it.

Assume that you have a jar full of blue and red marbles, 50% of each. “A random event does not lend itself to explanation, but collections of random events do behave in a highly regular fashion.” As a very patient person, you begin drawing marbles out in collections of four. A second person draws out the marbles in collections of seven. Which person is the most likely to draw a collection of the same color; the person drawing four at a time or the person drawing seven at a time? The chances of the person drawing four at a time getting four marbles of the same color are 12.5%. The chances of the person drawing seven at a time are 1/8th of that, 1.56%. The smaller the sample size, the greater the possibility of creating an extreme case – all marbles of the same color. The smaller sample size will have greater variability. Common sense – right?

Now, let’s assume that you are looking at high schools. You rank high schools based on average student performance measured in a couple of different ways – maybe graduation rates, scores on some kind of graduation test – like an SAT test, and college enrollment. Will the smaller high schools have more schools at the top of the ranking than larger high schools?

Let’s hire some experts to help us. We ask them to accumulate a list of the most successful schools in the country and to present us with some kind of explanation of success. It turns out that there are a disproportionate number of small high schools at the top. The experts suggest that the smaller the school, the higher the probability of success. In explaining why, there may be a number of explanations but one that seems to make sense is that there is more attention given to each student. It is the tender loving care quotient. You have an explanation and begin a small schools initiative. You invest $1.7 Billion dollars in the initiative.

But, we know that smaller schools are likely to have greater variability within the school itself. One outstanding student can influence the average more than that same student could in a large school. While you are more likely to draw four blue marbles, you are also more likely to draw four red marbles than the person drawing seven at a time. Let’s assume that red is the low achieving high school and blue is the high achieving high school. You have more probability of both than the person with the larger sample.

If that principle were applied to the school study, we should find that there are a disproportionate number of small schools that are low performing also. In fact, that turns out to be the case. If we look only at high achieving high schools, we will find that more are small schools that we first thought. If we were to look at low achieving high schools, we would also find that there are more small schools than we might have expected.

Over time, as the data were disaggregated, the larger schools tended to outperform the smaller schools and the belief is that the higher number of electives allowed for greater personalization in the larger schools but these were all conclusions observed after the fact, even though they were there for anyone to see. Kahneman calls this the law of small numbers.

Now, before we all start throwing rocks at the Gates Foundation and the small schools initiative, many high schools combined the best of both worlds. They preserved the ability to choose within a school by creating their small schools within larger schools. Kahneman also warns against creating “whether or not” decisions. A ‘whether or not decision’ says that we either do the small school or not. The middle ground that many found was to locate the small school within a larger school in order to preserve a wide selection of course choices while creating the tender loving care characteristic that the small school could offer. While we do not know the national averages for how these schools within a school worked out, we do have a sample. When looking for high schools to present at a conference we did on secondary school reform, all the success stories we found were small high schools within larger high schools.

Our guess, and it is a guess, is that these schools worked for two reasons. The students selected to be there – they exercised choice. The staff also chose to be there. The fact that they were small was probably not the factor that most determined their success other than through Kahenman’s law of small numbers.

Thinking, Fast and Slow is not full of educational examples but it is full of problems and decisions that most of us get wrong. It is a book that has spawned an entirely new field of economics – behavioral economics. It’s thought provoking problems have dramatically improved retirement savings in the U.S. and elsewhere. It is a slow read because every four or five pages you encounter a new idea, but it is a fascinating read and we will all see ourselves in the book, making ill-advised decisions.