There's one surefire way to convince me to never buy your product

[conuly]

And that's bogus statistics.

"100% of pediatricians suggest that if possible a child would go barefoot - however they also note that it is not possible as a result of sharp objects like glass, rocks, tacks, tree roots or even toys on the floor - so a shoe which is very flexible is needed."

100%? Really, exactly, precisely 100%?

How did they find that out? Did they poll every pediatrician on the planet, and keep an updated list from med schools so that as new pediatricians are created, their opinions are also noted?

No group of people ever agrees 100% on any subject. I promise you, there's even one pediatrician out there right now who thinks that sugar water is an appropriate diet for a baby, or that children grow best on a diet of chips and ice cream. Sure, that pediatrician isn't very well qualified, but that's not the point, is it? (And this is an issue of an entirely different nature, anyway.)

Their site is also poorly edited. Never use a period when a comma is better, folks.

You know, it's not the fact that these stats are wrong, it's that they're blatantly wrong. It's insulting!

Just like I tell Ana: "If you can't lie well, you shouldn't lie at all." It's bad enough being lied to, if they'd said 99.9 I would have been taken in - but to lie badly, that's just... just... that's rude!

Comments

[rho]

I don't think that the statistics are as bad as you're making out. Every number gathered from the real world comes with a margin of error. In scientific circles, you typically state your margin of error explicitly, but in more casual circumstances, you don't. If someone tells me that they're 5'3" tall, I don't assume that they mean exactly that height. I assume that they're 5'3" plus or minus half an inch.

If they'd done a poll of, say, 100 paediatricians and they'd all said that kids would like to go without shoes, then their best guess for the overall population of paediatricians would be that 100% of them agreed. That is to say that of all the possible outcomes, 100% is the most likely one based on the sample data collected.

Now usually, if you calculate the most likely single outcome, that also turns out to be the weighted average of all the possible outcomes (weighted by likelihood). The distribution is symetric, so for every possibility higher than the most likely value there's another value that's lower by an equal amount to cancel it out.

With this case where the best guess is 100%, that obviously doesn't apply. We know that the real answer can't possibly be greater than 100%, so even if 100% is the most likely single answer, there are many possibilities which are below 100% and the combined probability of all of them together makes "somewhere below 100%" more likely than 100%. This is what you picked up on, I suspect.

Even so, quoting the highest likelihood single value is not unreasonable. The problem is that from the figure quoted there's no way of knowing what the margin of error is. If they'd asked just one paediatrician, then the most likely extrapolation from that would still be 100%, but it would be something like plus or minus 90%. If they'd polled 1000 paediatricians then the margin of error would be much, much smaller.

So in conclusion, I don't think that the statistic is necessarily false. It's just almost entirely meaningless, and potentially misleading.

[conuly]

This makes sense, though I couldn't explain it to somebody else if I tried.

[sunnydale47.livejournal.com]

I strongly doubt they did a poll. And if they did, it's highly likely that they didn't poll enough pediatricians to derive a statistically significant result. And even if they did, they probably kept trying until they got their desired result -- that's very common in advertising.

I agree with conuly -- I can't imagine 100% of pediatricians agreeing on anything except maybe whether measles is contagious. It's ludicrous to imagine that they would with a subjective issue like this.

Tthe sentence is rife with weasel words (http://en.wikipedia.org/wiki/Weasel_words#Purposes): "100% of pediatricians suggest that if possible a child would go barefoot- however they also note that it is not possible as a result of sharp objects like glass, rocks, tacks, tree roots or even toys on the floor...."

They are in no way claiming that 100% of pediatricians stated that all children prefer to barefoot at all times, but that going barefoot can't be done anywhere because it's too dangerous. Of course, that's the impression they're trying to give -- but that's not what they're saying.

What they do say is that the opinion came from 100% of n pediatricians -- where n could equal as few as two or three. They don't mention a "poll," so there's no reason to expect that they did one at all.

These pediatricians don't state, aver, or even say that children prefer to go barefoot -- they "suggest" it. "Note" in the second half is equally meaningless.

The ad also doesn't say that all children like to go barefoot -- it says "a child." The question (even if they didn't just make up the whole thing) appears to have been something like "Do you have at least one patient who always wants to go barefoot?" I'd believe that 100% of pediatricians would answer yes to that one.

But they didn't ask if children like or prefer to go barefoot. They asked if they "would" go barefoot. Well, all children would go barefoot if they didn't put shoes or socks on! And when you deconstruct the sentence, that's all it says -- that they "would" go barefoot.

"If possible" is meaningless in this context -- it's simply a red herring.

The ad doesn't specify what is "not possible." You're supposed to infer that "going barefoot" is what is not possible, but it doesn't say that. Ad writers are skilled in tricks like this in order to escape the Federal Trade Commission's truth-in-advertising regulations.

For example, an ad can't legally say that "Clean-X Detergent is better for your laundry" unless they can present proof -- but it can say "Clean-X Detergent is best for your laundry" because legally, "better" means "best" and "best" means "equal to." (http://home.olemiss.edu/~egjbp/comp/ad-claims.html)

So this ad carefully doesn't specify what is not possible. What is not possible is going barefoot 100% of the time. No one would dispute that sometimes it's necessary for even that one child who dearly loves going barefoot to wear shoes. The whole clause "it is not possible as a result of sharp objects like glass, rocks, tacks, tree roots or even toys on the floor" really says that it's not safe to go barefoot when there are sharp objects on the floor. It very carefully says nothing about times when there are not sharp objects on the floor. (It doesn't say anything about outdoors, either -- only on floors.)

And even with all the weasel words, they slipped up and used a phrase which could be legally challenged! It is possible to go barefoot no matter what's on the floor. It may not be safe, or wise, or advisable, but it's unquestionably possible. Since that's the basis of the entire claim, the whole statement is fallacious.

I know you agree that the ad is meaningless and (more than potentially, IMHO) misleading. But I don't think even their statistics are supportable. (And just as you explained the statistics because you clearly enjoy doing it, I deconstructed the ad because I enjoy doing it, not because I thought you were taking it seriously! )

[rho]

I'm very glad for that last paragraph of yours there; for a minute or so I was worried that my abilities to express myself were even lower than I worried that they might be.

You're absolutely right, of course, and I agree with pretty much everything (not meant as a weasel term myself there, just that I was too lazy to go back and read through in fine detail to check I did actually agree with everything) that you've written. It's a fine critique, and the better/best is something I didn't know before.

The point I was trying to make is that the statistics aren't necessarily bogus, in and of themselves. I'm fairly sure that they are bogus, but it's only by considering the rest of the ad (as you've done) that I think you can say that for (almost) sure.