When you become aware of a statistic do you question it, or just accept it as fact?

I think you’d enjoy Bad Science by Ben Goldacre. Great book which addresses lots of things, in addition the misleading presentation of statistical results.

More specifically…it depends on the source and how consistent the new information is with more verified aspects of my worldview.

Question it. Nothing is absolute.

After taking several statistics courses in college, the most important thing I’ve learned is to never trust statistics.

Studies show that 100% of people are heavily influenced by statistics. And that half of all statistics are made up. :P

These things aren’t even statistics, they’re just factoids. I like to see the real numbers behind these things, otherwise I take them with a grain of salt. Many times the wrong numbers are reported, for example there was a story saying the rate of workplace suicides had doubled. Turns out it meant there were 6 instead of three in the previous year. (I exaggerate because I don’t recall the exact numbers). But saying the rate of something doubled if the original rate was very low isn’t all that meaningful.

I question everything. I had a critical thinking course along with a lot of science courses and I’ve learned to check sources and methods before buying a statistic.

Secondly, statistics can be easily manipulated, so really, they’re kind of meaningless for the most part.

Haha! GA @wildpotato

I am not sure if you can say I “question” a statistic, but I try to be critical and understand what it actually is based on and what does it actually say.

60% of the time, it works every time!

The mean number of testicles per person on Earth is 1.0 (...to 1 decimal place).

Always question it.

@virtualist I’ve never been more happy to be an anomaly!

@JONESGH Logic axioms and postulates are absolute and true !

Question everything.

@virtualist as far as we know. for hundreds of years the world was absolutely flat. doesn’t mean anything.

@JLeslie Eggs are of no danger for your cholesterol levels.

Not at all. If you know statistics, you know there are so many ways to distort the outcome. You also need to know the margin of error when reading statistics and how/what data was obtained.

@jeffgoldblumsprivatefacilities On the average every person on earth is an anomaly.

The math behind it is very important. Not all results can be called statistically significant. So it’s really about trust in the organization publishing statistics. Not all of them employ professional mathematicians and statisticians.

@oratio what made you say that? By the way if I cut out yolks and sweets that are made with eggs like cookies and cake my cholesterol goes down 50 points wihout cutting anything else within a month. I have done it many times in the last 20 years, and it always works. Last time I “worked” on my cholesterol I cut my animal intake in half and it went down another 10 points after doing that for 10 days. Gotta wonder if I went vegan if it would actually be normal. My sister is vegan and she is the only person in my family with normal levels of cholesterol, and hers were high like mine when she was younger and not a vegetarian. What is interesting now is the science associated with vitamin k2, which is found in yolks and helps keep calcium from settling in the wrong place like arteries and other soft tissues.

It depends on the source, although I’ve had enough statistics to realize that if you manipulate the date, you can show whatever you want.

Question it. If you’ve ever taken a stats or a research methods class you know you need to question it.

The smell test.

@virtualist wrote: The mean number of testicles per person on Earth is 1.0 (...to 1 decimal place).

(I know it’s a joke, but I’m using it as an example).

So clearly, if there were an equal number of men and women in the world, then you could have one testicle per person. But are there an equal number of men and women in the world? If we researched that, we’d probably find that there are a significantly larger number of women. So I begin to suspect something is wrong her.

Then I think that there are a number of men without testicles. They may have been removed on purpose or lost by accident or to illness. So that’s two strikes against it. Enough to doubt the statistic and to ask for more information on how the number was derived.

How to Lie with Statistics is a quick and easy read. 100% correct and available at your public library.

@daloon – Thanks for putting it so well, I had the same exact thought but have had insomnia for the past week and my brain ain’t working right.

Good question. I question every thing. It’s like those ads that say “2 out of 3 dentists use X toothbrush!” Right, and exactly how many dentists were questioned? Three? And how were they questioned? Were they lead into making some comment about X toothbrush that they could take out of context?
Always question everything.

I never accept them as infallible facts.
If it’s a stat I dislike, I’ll do my little research.
If it’s one I like, I’ll wait for opponents to do the research.

I am a doubting Thomas by nature… especially when it comes to statistics. I do also consider the source… especially if the stats are being used to try to sell me something.

That’s why I really hate the media and their political polls… Report facts about the issues and what the candidates’ records are relative to those issues. Don’t tell me that 41% of people polled like candidate A, while 52% prefer candidate B and 7% are undecided. Whom did they poll? Most people whose opinion I respect don’t take calls from pollsters. Besides, I like to make my own decisions regardless of what the masses think.

@daloon @ 0.95 to 1.04, I’m covered. They all round off to 1.0 (purposefully stated that way <g>) .I’m only screwed in my joke if the asymmetries you find are different by more than +/- 4–5%

@virtualist LOL! Um. I think you’re screwed in your joke! At least from where I sit!

I agree with the above comments that it’s important to question statistics, but I also want to point out the other side: don’t become such a skeptic that you disbelieve everything. Scientists continue to use statistics because they are the best tool we have, but we are aware of the limitations of statistics. We can never really say we’ve proved anything—all statistics can do is show us that one hypothesis is more likely to be true than another hypothesis.

I think it’s important for people to be aware that statistics can be presented in a way to justify opposite sides of the same issue. For this reason, a person should look into the specifics of a study and see if it wasn’t skewed to one side.

As for the oatmeal thing, it is based on the following: oatmeal has fiber, and apparently fiber can clean out the build-up of cholesterol in one’s arteries. So increasing your fiber intake counteracts the cholesterol deposited in your arteries. If you decrease your cholesterol as well, the fiber will work even better. So unless Wheatina actually has a significant amount of fiber, it probably will make a difference if you switch to oatmeal instead.

@drdoombot I just want to see that study. The study that says Oatmeal is better than other grains everything else being equal. Personally I don’t even feel convinced that fiber helps clear out arteries. I think eating fiber, which is found in non-animal products most abundantly, will lower your cholesterol because you are not eating the animal products, not that the fiber is really doing anything. I do think the fiber does some positive things don’t get me wrong, I just really question the cholesterol connection. I tried to google how the studies were performed, diets before and during the study, and I cannot find those details.

I also always questioned the idea that HRT would reduce the risk of heart attack. in women Seems now medical science has come around to agree with me. I don’t think it is the female hormones that are the biggest influence over women and heart attacks. I think it was an assumption made based on correlations and then became the umbrella belief. But thats just me, no one has to agree with me.

@JLeslie Exactly. How much of this stuff is completely, totally made up just to sell products!

@virtualist Well, there you go! A rounding error! You can round something off and then it says one thing, while the data really show something else. That’s the kind of thing that gives statistics a bad name—dishonest analysts, displaying something that is true, but in a way that puts the actual message on its head. That’s statistics with an agenda.

And another thing. Pride. So often pride gets in the way of good science. When someone comes along with either new data or a reanalysis of the old data that shows a different reality that current thinking, the people who created the current thinking often take it personally. Like you’re attacking the person instead of describing reality more precisely. Scientists take things personally, and get defensive, and it makes it much harder to see the truth. No longer does the data talk. Only the Oracle (scientist) is speaking and describing a reality that only the Oracle can see.

@virtualist I bet you had no idea you were going to become such an object lesson! ;-)

@Val123 oh, and there is that, good point. But, also the doctors who take on these beliefs as fact bother me, maybe bother me more than the Quaker Oats where their intention is obvious.

Benjamin Disraeli once said: “There are three kinds of lies: lies, damned lies, and statistics.”

That was supposed to be Mark Twain who said that!

@daloon Actually, Marc Twain made it popular in the states. Desraeli said it first.

@daloon ..... ” @virtualist wrote: The mean number of testicles per person on Earth is 1.0 (...to 1 decimal place).”

@daloon wrote ”(I know it’s a joke, but I’m using it as an example).”

I think you’ve about run my joke into the ground.
You’re singing maniacally to the choir. <g>

@gussnarp Just for you. Factoid: noun: something resembling a fact; unverified (often invented) information that is given credibility because it appeared in print.

As a skeptic, I distrust every statistic, even those that sound plausible at first. I never accept anything at face value, unless it has been proven to me, and always more than once.

Definitely question it. Skepticism is valuable.

@Psychedelic_Zebra Bravo!

I just heard on tv that people who take statins are 50% less likely to die from the flu, and researchers think it should be tested. Come on!

@gussnarp What or whose statements can you trust? It would appear none; but maybe only statements that agree with some statement you have made. You seem to have tied yourself to a life of infinite frustration and paranoia. Nothing can be proven absolutely to your satisfaction except formal statements of logic progressions. There are statistics and then there are Statistics; they are everywhere. If one spent even 5% of one’s lilfe analyzing each one for satisfactory personal belief it become a life of Hell. When the Singularity arrives one of her first tasks should be to free us of this frustration! All statements of fact +/- some statement of precision and inaccuracy would the gospel truth , statistically. That would become another Law of Robotics as initiated by Asimov!

Anybody can come up with made up statistics in order to win an argument or prove a point… 85% of people already knew that!