Monday, April 8, 2013

Feeling Sleepy?

Many times over the years I've pointed out surveys that, coincidentally, have results that fit an organization's interests. This Monday morning, grab another cup of coffee because you could get sleepy after this latest example.

Here's the (too long) lede:
An online survey conducted by the Sleeping Disorders Centre has found that almost three quarters of respondents were at a high risk of suffering from sleep apnoea. Half of the 52 respondents were classified as obese (with a BMI over 28), a key indicator of risk, 69.2% said that they snored loudly, and 92.2% complained of feeling tired during the daytime. 55.8% of all respondents to the survey were men, who traditionally have a higher risk level than women, although recent evidence suggests this gap is closing.
Let's try and set aside the crappy writing for a moment.  Yeah, my beginning newswriting students could do better, but as I've emphasized time and time again on this blog, this is one of those examples of getting a survey result that just so happens to nicely fit your interests.

A sleep center.  Who's likely to use their online survey?  People with sleep problems, that's who.  Here's a bit of "methodology" of the survey:
Visitors to the Sleeping Disorders Centre website were asked to complete the STOP-BANG questionnaire, a set of eight questions regarded as the most accurate means of indicating a person’s level of risk of suffering from sleep apnoea from a self-conducted survey. 
Still awake?  C'mon, the STOP-BANG thing, that's gotta get you at least a little bit interested.  What the hell is a STOP-BANG, other than something I probably did in college?  It's a real questionnaire. Here's one version, so despite the silly name it's legitimate.

My real issue, of course, is this qualifies as a really bad SLOP, a self-selected opinion poll.  In other words, non-scientific BS.

The point of all this?  Beware poll results that just coincidentally support an organization's aims and goals -- even if you like that organization.

No comments: