McDonald's and BMI

[Trolldor]

Well, on the bright side I don't have to pay a monthly membership to do what I can do myself at home.

Why get a running machine when I can just go running?

[Coito ergo sum]

Well, on the bright side I don't have to pay a monthly membership to do what I can do myself at home.

Why get a running machine when I can just go running?

Very true. And, why go to a "medical professional" when you can check your BMI, see that you need to lose about 20 pounds, give or take, and then go out and run.

[Trolldor]

Because you aren't qualified to assess your own physical health and take in to account a number of factors, you are qualified to run.

[GreyICE]

It doesn't matter. It's a ratio whereby, if a global application returns abnormal results then a population may have a problem. In the same way that humans have no ideal height per se, yet if fifty percent of a population is in the ninety fifth percentile then something is wrong.

If 50% of the population is in the 95th percentile, then you really need to relearn statistics. Because, percentiles aren't averages. I'm very, very very certain that in all populations everywhere 5% of the population is at or above the 95th percentile.

It's based, by the way, on the simple volume-mass law (as a 3d object scales in dimension, mass increases with the cube of a single dimension).

The problem should now be apparent. Yes, it does blow chunks as soon as people vary at all from the "normalized" height, why?

[Warren Dew]

Yes, individual cases. The 'sample' is not enough for it to be statistically applicable to a given population.

Actually, the current data provide two statistically significant results. First, the people answering the poll are drawn from a population significantly less likely to be overweight than the overall U.S. adult population with greater than 99.9% certainty (assuming they're not lying about their BMI). Second, in that population, people who are overweight are less likely than people who are not overweight to have an opinion on the McDonald's law with greater than 95% certainty.

Unfortunately, because of the lack of overweight people with an opinion of the law, we don't yet have enough data for statistical significance regarding the hypothesis that overweight people are more likely to support the law than people who aren't overweight.

[Trolldor]

Actually, the only thing it tells you is that the people inclined to look at this thread are inclined to vote on polls.

Maybe. There's no control, no way to varify results etc. etc. etc.

Don't try to pass this off as valid of anything.

[Meekychuppet]

I'm off to bang my head against a wall to which cracks first.