Probability > Base Rates and the Base Rate Fallacy.
What are Base Rates?
- In probability and statistics, “base rate” usually means the same thing as prior probabilities (used in Bayesian Probabilities). For example, let’s say 10% of people have visited an Orlando theme park in the last ten years and 90% had not. The base rate for the number of people who have visited an Orlando theme park in the last ten years: 10%.
- In epidemiology, a base rate is more specific. It is the prevalence of a disease, symptom, or characteristic in a specified population. For example, the prevalence of autism in the U.S. (as of 2014): 1 in 68 children.
- In banking, a base rate (also called the discount rate) means something entirely different. It’s the interest rate at which a central bank lends money and the level below which domestic banks cannot lend.
The Base Rate Fallacy / Bias
When you ignore (or don’t understand) general statistical data and make a judgment based on specific data, you’re falling prey to the base rate fallacy. This happens all the time; People not well-versed in the technical rules of prior probability usually don’t take the prior statistical data into account, as it doesn’t seem relevant.
If you can’t wrap you’re head around Bayesian probabilities, you aren’t alone. When given data about a women’s chance of having breast cancer, given a positive mammogram result, an alarming 80% of physicians got the probabilities wrong. See: Even Physicians Don’t Understand Statistics.