Michael Phelps is a swimmer. He has a great body. Ian Thorpe is a swimmer. He has a great body. Missy Franklin is a swimmer. She has a great body.
If you look at enough swimmers, you might conclude that swimming produces great bodies. If you want to develop a great body, you might decide to take up swimming. After all, great swimmers develop great bodies.
Swimming might help you tone up and trim down. But you would also be committing a logical fallacy. Known as the swimmer’s body fallacy, it confuses selection criteria with results.
We may think that swimming produces great bodies. But, in fact, it’s more likely that great bodies produce top swimmers. People with great bodies for swimming – like Ian Thorpe’s size 17 feet – are selected for competitive swimming programs. Once again, we’re confusing cause and effect. (Click here for a good background article on swimmer’s body fallacy).
Here’s another way to look at it. We all know that basketball players are tall. But would you accept the proposition that playing basketball makes you tall? Probably not. Height is not malleable. People grow to a given height because of genetics and diet, not because of the sports they play.
When we discuss height and basketball, the relationship is obvious. Tallness is a selection criterion for entering basketball. It’s not the result of playing basketball. But in other areas, it’s more difficult to disentangle selection factors from results. Take business school, for instance.
In fact, let’s take Harvard Business School or HBS. We know that graduates of HBS are often highly successful in the worlds of business, commerce, and politics. Is that success due to selection criteria or to the added value of HBS’s educational program?
HBS is well known for pioneering the case study method of business education. Students look at successful (and unsuccessful) businesses and try to ferret out the causes. Yet we know that, in evidence-based medicine, case studies are considered to be very weak evidence.
According to medical researchers, a case study is Level 3 evidence on a scale of 1 to 4, where 4 is the weakest. Why is it so weak? Partially because it’s a sample of one.
It’s also because of the survivorship bias. Let’s say that Company A has implemented processes X, Y, and Z and been wildly successful. We might infer that practices X, Y, and Z caused the success. Yet there are probably dozens of other companies that also implemented processes X, Y, and Z and weren’t so successful. Those companies, however, didn’t “survive” the process of being selected for a B-school case study. We don’t account for them in our reasoning.
(The survivorship bias is sometimes known as the LeBron James fallacy. Just because you train like LeBron James doesn’t mean that you’ll play like him).
So we have some reasons to suspect the logical underpinnings of a case-base education method. So, let’s revisit the question: Is the success of HBS graduates due to selection criteria or to the results of the HBS educational program? HBS is filled with brilliant professors who conduct great research and write insightful papers and books. They should have some impact on students, even if they use weak evidence in their curriculum. Shouldn’t they? Being a teacher, I certainly hope so. If so, then the success of HBS graduates is at least partially a result of the educational program, not just the selection criteria.
But I wonder …
How do you evaluate whether a given company (or product) will be successful or not? Clearly, you’ll want to study the business plan, interview management, talk to customers and competitors, and get an unbiased evaluation of the product. But is that enough? Unfortunately, it’s not — you’re only looking at those things that you can look at. You’re looking at the known knowns but omitting the known unknowns.
So how do you see what you can’t see? Instead of just looking at the individual case, look also at the category that the case derives from. If you’re evaluating a start-up company, look at all the details of the company. But also look at the category called “start-up companies”. What’s the success rate there? Studying the category won’t always tell you why companies succeed but it will tell you how often they succeed. That can provide you a “base rate” with which to compare your individual case. For a balanced view, always look at the case and the category.
Learn more in the video.