The urge to infer
A news item on cbc.ca this week had the following headline: "Need 8 hours of shuteye? Even 6 can cut diabetes danger, study suggests". The first line of the article informs us that: People who sleep less than six hours a night are nearly five times more likely than longer sleepers to develop a blood-sugar condition that could lead to diabetes, new U.S. research suggests. A longer Reuters article fills in some details: Using data from a large, six-year study, they identified 91 people whose blood sugar rose during the study period and compared them to 273 people whose glucose levels remained in the normal range. They found the short sleepers were far more likely to develop impaired fasting glucose -- a condition that can lead to type 2 diabetes -- during the study period than those who slept six to eight hours. One might ask any number of questions—for example, how did they measure the number of hours slept—but I want to focus on the interpretation. The Reuters article accur...