Endogeneity bias
28th January 2017 | Professor Nick Lee
For the past few columns, I introduced critical issues in designing your own research, with a special focus on how to design experiments. For the next few columns, I’m going to build on those pieces to some extent, to explore a set of issues that I see as perhaps even more important – and especially so in today’s world.
Specifically, I’m going to introduce a set of key issues you need to be aware of when you draw conclusions from research that’s already been done, that you might read about in newspapers, white papers, management books, magazines, and the like (and, yes, even academic papers). Many of these issues are vitally important to what we can say and believe about the results of research, but few people are ever aware of them. Indeed, in some cases, lack of consideration of them can render the results of research completely meaningless.
The first issue I am going to introduce is called endogeneity bias. If you have training in economics, and especially with a focus on econometrics, this is something you probably already know about. But, in my view, it’s a bit of a scandal that outside that environment very few people are ever taught about it. That’s a real problem, because as you will see soon, endogeneity bias can render research on really important topics (eg social policy, as well as business-related issues) essentially useless. I’m going to explain this concept without the use of complex equations, but I will use a diagram and a few symbols to make my points clearer.
PLEASE NOTE: Subscriber-only content – To read the full article, please login or purchase a subscription. Subscription Options Login
