Andrey Kamenov, Ph.D. Probability and Statistics
A question in a survey conducted by the Centers for Disease Control and Prevention asked a respondent to assess his or her own health. Just one in every six people considered their own health “excellent;” most stuck to “very good” or even just “good.” But how do we use this data to visualize the relationship between health levels and other factors?
The problem is, the scale itself is not quantitative. It doesn’t explicitly state how much better “good” is than “fair.”
In order to measure the health levels of specific groups in the general population, we’ll use a quite common procedure. It provides values between 0 percent (meaning all respondents have poor health) and 100 percent (where everyone has excellent health). The exact values of everything in between are based on the (empirical) nationwide percentages.
Let’s see how our health index fares against two of the most obvious factors. The first chart shows the speed of peoples’ self-perceived decline in health with age: