The NYTimes and other news orgs may have just fallen into another correlation/causation trap. Yesterday they published an article which implies that you will live longer if you read books:
Reading books is tied to a longer life, according to a new report.
Researchers used data on 3,635 people over 50 participating in a larger health study who had answered questions about reading.
The scientists divided the sample into three groups: those who read no books, those who read books up to three and a half hours a week, and those who read books more than three and a half hours.
The study, in Social Science & Medicine, found that book readers tended to be female, college-educated and in higher income groups. So researchers controlled for those factors as well as age, race, self-reported health, depression, employment and marital status.
Compared with those who did not read books, those who read for up to three and a half hours a week were 17 percent less likely to die over 12 years of follow-up, and those who read more than that were 23 percent less likely to die. Book readers lived an average of almost two years longer than those who did not read at all.
The NYTimes made it sound like reading more lead to living longer (see the headline)
when in fact there’s no evidence showing causation which is exactly what the researchers concluded.
Edit: I followed up by email, and one of the paper’s authors told me that:
The NYT article did a good job summarizing our key findings: reading books confers a survival advantage, after adjustment for a whole host of covariates (sex, age, wealth, health, education, depression, etc). For this reason, we believe there is a strong connection between reading and longevity.
What are the chances they are right?
I for one don’t believe them. But then again they are the ones with the PhDs.
Do you think there’s a real connection, or is this a spurious correlation?
What I think we have here is a study which looked at population data and found a correlation between life expectancy and reading. And without data to explain the connection, that correlation doesn’t mean anything. Without evidence of a real connection, I see it as it’s as much a spurious correlation as the relationship between suicides and US spending on science, tech, and space.
image by ZapTheDingbat