Lies, Damned Lies, and ...

Cecil’s column Man Wrecks Planet: Are the popularly published environmental statistics accurate? - The Straight Dope brings up a good point: far too many statistics are simply credulously repeated without really checking them out. One researcher counts 48 chromosomes in the human cell and no one checks it out for more than 40 years. Scientists, the UN, and everybody else accepted that the number of 48 chromosomes was an unquestioned, established fact.

The same thing happens in the media. A police chief known to be corrupt reports that Ryan Lochte and friends trashed a restroom and assaulted security guards and every media outlet in the world accepts it as established fact – until one team from USA Today actually goes to the gas station in question and views the security camera footage and finds out that Lochte was basically telling the truth.

It was established fact that radiation leaking from the Hanford complex was causing thousands of cancer deaths per year among the Native American tribes living along the Columbia River – until someone noticed that the purported number was greater than the total cancer deaths in the area by a factor of over 100.

People credulously accept that one in five children go to bed hungry in the US because it is a common statistic that is constantly repeated. It is absolutely untrue. The same goes for the number of children who are contacted by child molesters on the Internet.

Given the wide disparity of common statistics and easily observable facts, it is no wonder that people question the t-shirt rant about destruction of the earth. Sure, all the scientists and all the UN believe it – but if you track it down these numbers were created by activists who all had an axe to grind and whose scientific integrity is less than sterling.

This creates a problem. If you really want people to be concerned about the environment, why not encourage them to check the facts themselves? Better yet, encourage more kids to go into research. Have them actually look to see if the results of the original research can be verified. Otherwise, no one is ever going to trust the environmentalists. The public has been lied to far too often – and lot of these statistics are being used as rationalization to push socialist economic agendas and giant government programs that on their face seem to have little to do with actually solving any of these problems. Is it any wonder, then, that people are more than a little skeptical?

4 out of 5 SDMB posters agree !!!

Doesn’t everyone know that 75.893558% of statistics are just made up?

The lottery is a tax on the mathematically challenged … it funds schools so don’t knock it.

Oh, people can come up with statistics to prove anything. Forfty percent of all people know that!

Can we have cites for, well, any of those claims at all?

Well…that depends on whether you use probability or odds for your calculations. The probability is extremely low but the odds are 50/50.

Since when does the odds for something equal 50/50 if there is a tiny chance that it will happen and a huge chance that it won’t? I can’t find any source saying that’s a typical use of the term. I find various sources saying that it isn’t:

And it’s a shuck by the anti-tax brigade, since what usually happens is that direct school funding is cut in proportion to expected lottery revenues - which may or may not reach those expectations. The idea that lottery money is *added *to school funding is complete nonsense in most if not all states.

Whoosh?

Either it will or it won’t–50/50!

Obscurely academic tidbit: This is actually called the Principle of Indifference and has a learned history of debate and discussion in probability and philosophy…

Six out of Seven Dwarfs are not Happy.

You can probably find a cite for anything. That’s the problem.

But here is an example regarding the oft-quoted (and false) statistic that one in five children in America goes to bed hungry:

OK, there we go, that’s a cite for two of the eight claims in your OP. Care to support any of the other six?

The chromosome issue is corroborated in the book Adam’s Curse by Oxford geneticist Bryan Sykes. OP is only off on the duration of the error.

The original miscount, by Theophilus S. Painter, took place in 1923. He had a decent excuse in that at the time it was impossible to view chromosones except as a jumbled mass. Painter spent several months trying, but failing, to get an accurate count. A greatly refined preparation technique was discovered by T.C. Hsu in 1952, but incredibly he and others were so convinced there were 48 chromosones that they somehow counted 48 even though 46 and only 46 were right there in plain sight. It remained for Albert Levin to repeat the experiment without preconception, and to convincingly report his conclusive finding of 46 human chromosones in 1956, 33 years after Painter’s error.

In this vein Cecil wrote a column about a claim making the rounds which reported that US grade school children of the 1950s had vocabularies of something like 30,000 words.

Now, Shakespeare’s vocabulary was “only” about 26,000 words, so the claim being made for the grade school children was obviously ridiculous. Cecil was able to track down one or two original investigators who informed him that the total vocabulary of all the grade school children studied was 30,000, and certainly no one child had a vocabulary of anywhere near 30,000.

Similarly, it was for years widely reported that 8 cups/64 fl oz of water per day were essential for good health. Then someone tracked down the original study, and that study did recommend 64oz. water per day, BUT the water in food definitely counted, and since food is ~70-80% water, the non-food fluid water requirements are nowhere near 64oz. So drink only when you are thirsty, and (I might be a bit overboard with this last part, but only a bit) your fluid water requirements may actually be close to zero if you are physically inactive and eat a lot of stuff like watermelon.

It’s a statistical fact that most Americans have more than the average number of eyes.

It’s true. Virtually nobody has a third eye. But there are plenty of people who have lost an eye. If only one person out of a 100,000 has lost an eye, then the average number of eyes per person is 1.99999. And that means everyone with two eyes has more than the average number.

Thank you for the post which so adeptly combines information, wit, and insight.

I don’t know where SDMB would be without such treasures!

No. I am too lazy. Blame it on my RH- blood type if you wish.

Oh, all right – I recommend the book The Half-Life of Facts: Why Everything We Know Has An Expiration Date by Samuel Arbesman. Fascinating read. He uses the chromosome thing as Exhibit A.

The USA Today investigation into Ryan Lochte can be found by searching their web site. The Wall Street Journal makes a frequent point of the fact that reporters rarely do any independent investigation, but instead either just copy each other or uncritically accept any news release that comes along. A guy who took frequent advantage of that was Alan Abel. His “Society for Indecency to Naked Animals” was a classic.

The Hanford incident was reported in both the Seattle Times and the Post-Intelligencer many years ago. It may not exist on the Internet (frankly, I figure you can use Google as well as I can if you are really interested), but it was commonly used in the 1970s by college statistics professors (including mine) as an example of bad statistics. The problem arises from the fact that no researcher is genuinely impartial. If a researcher does not find what he or she is looking for, then they assume that the object of research, such as a disease, is under-reported. Or, as in the case of autism, they simply expand the definition until they come up with the numbers they want.