A foundation funded by the nation’s sleep industry today released its latest survey finding that – ta daa! – we’re not getting enough sleep.
Too bad we can't just snooze through it. Because as survey research, this one’s a recurring nightmare.
It’s an annual event for the National Sleep Foundation, an outfit funded by sleeping-pill and mattress manufacturers, among others. Always good with the headlines, the NSF this year says: "One-Third of Americans Lose Sleep Over Economy."
Listing the problems is as simple as counting sheep. The actual number behind the headline is 27 percent; it morphs conveniently to “almost one-third” in the report’s highlights and on to “one-third” in the press release. Far worse, NSF produced this number by giving its respondents nine possible reasons for losing sleep – four of which are economic. It then toted up those four to hoist "the economy" into the top slot. As an attention-grabber it works. As neutral, balanced measurement, maybe not.
The other five possible reasons for sleep loss get just one shot each. Some are sensible (problems with health or with personal relationships), but others are more far-fetched (global warming or the wars in Iraq and Afghanistan). And there are plenty of good reasons for losing sleep that aren’t offered – trouble at work, say. The neighbor’s dog. Or, for me, manipulated polls.
While the subject sounds timely, by the way, this study was completed four months ago.
Even beyond the headline, the NSF news release, as in years past, is an exercise in cherry-picked data. It notes alarmingly that 54 percent of adults “have driven when drowsy at least once in the past year” – “a major public safety problem.” Less prominently reported is how many actually experienced an accident or near-accident due to drowsiness – 1 percent. You’ll find that on p. 42 of the full report.
Or consider, more fundamentally, the NSF’s main drift – that people are getting less sleep. It's not a bad message for pill and mattress manufacturers, even if the NSF maintains that its surveys are editorially independent. It repeats its mantra again this year – "sleep problems are on the rise."
But are they? The NSF release reports a reduction in average hours of self-reported sleep from 2001, but skips over its more recent 2004 data – from which it does not show a significant change in weekday sleep hours. It shows a slight decline on weekends, three-tenths of an hour.
Even that’s well worthy of debate. The federal government's American Time Use Survey, conducted by the U.S. Census Bureau for the Bureau of Labor Statistics, comes up with estimates that are strikingly different from the NSF's. The latest government data, based on 12,500 interviews across the year 2007, show an average of just under 8.3 hours of sleep on weekdays and 9.2 hours on weekends and holidays. NSF, by contrast (with 1,000 interviews last October and a different approach to asking the question) gets an average of 6.7 hours on weekdays, 7.1 hours on weekends.
The different approach may be critical. Two time-use experts at the University of Maryland, Profs. John P. Robinson and Steven P. Martin, suggest the federal estimates are likely to be more accurate because they’re based on a specific item-by-item listing of the previous day’s activities, rather than responses to the NSF's more general question asking people to estimate how many hours they sleep a night.
The professors suggest that the hectic nature of daily life may make people feel like they're getting less sleep than they really are. The reality, they say, is different; a report they published in 2007 evaluating 40 years of federal time-use surveys, from 1965 to 2005, found no decrease in sleep time over that period – instead, a slight increase. Its conclusion: "These latest time-diary data suggest that American adults of all ages and races and both genders continue to get no less sleep than 40 years ago."