Why does statistics seem like sorcery?

[Faid]

It's based on the same principle.

In sorcery, you take a little piece of someone, nail clippings, hair, whatever and do something to it, and that same thing happens to the whole person.

Same with statistics.

You take a little piece of a population, a sample, and find something happening in it, and then that same thing happens in the whole population.

And always make sure you meet the Great Old Ones' expectations. See that the p-value is 0.05 at most, or things with tentacles start coming out of the woodwork.

ETA: When p=0.1, they start getting really aggressive.

[Pingu]

I think some people conflate "implies" and "indicates".

True, but it's mainly because of ambiguity over the word "correlation".

"Correlational studies" usually means "observational" i.e. non-experimental studies - no randomly allocated manipulated variable.

But "Correlation" can also be used in the technical sense to mean the linear relationship between two continuous variables (e.g. Pearson's correlation coefficient).

And the fact is that you can do correlational studies with discrete variables and you can do experimental studies with bivariate continuous variables.

The thing people are supposed to remember (that correlation doesn't equal causation) references the first meaning, not the second.

So you find people thinking they can infer a causal direction from an observational study as long as the have a categorical predictor (even if they had to do a median split to get it) and thinking they can't infer a causal direction from an experimental study, just because their test was a bivariate Pearson's, even though one of the variables was randomly generated by a computer, and cannot possibly have been caused by the participant's response!

[gandalf]

^s0rceress

[Pingu]

[Old Dan]

[F X]

It's funny, because Twain didn't come up with the phrase, but now he is considered the author.

[Old Dan]

It's funny, because Twain didn't come up with the phrase, but now he is considered the author.

"now"

[F X]

From Chapters from My Autobiography

Figures often beguile me, particularly when I have the arranging of them myself; in which case the remark attributed to Disraeli would often apply with justice and force: 'There are three kinds of lies: lies, damned lies, and statistics.'

[Steviepinhead]

Twain most likely got the attribution wrong, too:

Mark Twain popularized the saying in Chapters from My Autobiography, published in the North American Review in 1906. "Figures often beguile me," he wrote, "particularly when I have the arranging of them myself; in which case the remark attributed to Disraeli would often apply with justice and force: 'There are three kinds of lies: lies, damned lies, and statistics.'"[2]

Alternative attributions include, among many others (for example Walter Bagehot and Arthur James Balfour) the radical English journalist and politician Henry Du Pré Labouchère (1831–1912), Jervoise Athelstane Baines,[3] and British politician and man of letters Leonard H. Courtney, who used the phrase in 1895 and two years later became president of the Royal Statistical Society. Courtney is quoted by Baines (1896) as attributing the phrase to a "wise statesman",[4] but he may have been referring to future statesman rather than a past one.[5]

The earliest instance of the phrase found in print dates to a letter written in the British newspaper National Observer on June 8, 1891, published June 13, 1891, p. 93(-94): NATIONAL PENSIONS [To the Editor of The National Observer] London, 8 June 1891 "Sir, —It has been wittily remarked that there are three kinds of falsehood: the first is a 'fib,' the second is a downright lie, and the third and most aggravated is statistics. It is on statistics and on the absence of statistics that the advocate of national pensions relies…" Later, in October 1891, as a query in Notes and Queries, the pseudonymous questioner, signing as "St Swithin", asked for the originator of the phrase, indicating common usage even at that date.[5] The pseudonym has been attributed to Eliza Gutch.[6]

The American Dialect Society list archives include numerous posts by Stephen Goranson that cite research into uses soon after the above. They include:
Sir Charles Wentworth Dilke (1843–1911) is reported twice in October 1891 to have used the phrase, without attributing it to others:
"Sir Charles Dilke [1843-1911] was saying the other day that false statements might be arranged according to their degree under three heads, fibs, lies, and statistics." The Bristol Mercury and Daily Post, Monday, October 19, 1891The Derby Mercury (Derby, England), October 21, 1891; Issue 9223 "Sir Charles Dilke and the Bishops" "A mass meeting of the slate quarry-men of Festiniog Ffestiniog, Wales] was held Wednesday night [Oct. 14] to protest against certain dismissals from one of the quarries...." He [Dilke] observed that the speeches of the Bishops on the disestablishment question reminded him that there were three degrees of untruth--a fib, a lie, and statistics (Laughter)"The phrase, as noted by Robert Giffen in 1892, was a variation on a phrase about three types of unreliable witnesses, a liar, a damned liar, and an expert (Economic Journal 2 (6) (1892), 209-238, first paragraph; the paper was previously read at a meeting of the Australasian Association for the Advancement of Science at Hobart in January 1892). 1892 Jan talk, June pub Robert Giffen (1837–1910, Walter Bagehot's assistant editor at The Economist 1868ff; 1882-4 President of the Statistical Society): "An old jest runs to the effect that there are three degrees of comparison among liars. There are liars, there are outrageous liars, and there are scientific experts. This has lately been adapted to throw dirt upon statistics. There are three degrees of comparison, it is said, in lying. There are lies, there are outrageous lies, and there are statistics."
That phrase can be found in Nature, page 74 November 26, 1885: :"A well-known lawyer, now a judge, once grouped witnesses into three classes: simple liars, damned liars, and experts. He did not mean that the expert uttered things which he knew to be untrue, but that by the emphasis which he laid on certain statements, and by what has been defined as a highly cultivated faculty of evasion, the effect was actually worse than if he had." [7]
A minute of the X Club meeting held on 5 December 1885, recorded by Thomas Henry Huxley, noted "Talked politics, scandal, and the three classes of witnesses—liars, d—d liars, and experts." Quoted in 1900 in Leonard Huxley's The Life and Letters of Thomas Henry Huxley.[8][9]

From the wiki article on the phrase.

So basically it's a variation on an older lawyer joke.

[BardoXV]

Many years ago I took a statistics course and at the beginning of the course the professor told us "Give a good statistician the raw data, and he will prove anything you want."

Something along those lines, "Figures don't lie, but Liars can sure figure."

[BardoXV]

That's interesting, since Mark Twain wasn't his real name.

[Testycalibrated]

Many years ago I took a statistics course and at the beginning of the course the professor told us "Give a good statistician the raw data, and he will prove anything you want."

Something along those lines, "Figures don't lie, but Liars can sure figure."

My stats teacher started class with a similar warning. "Statistics are usually nothing more than the shortest path from an unwarranted assumption to a foregone conclusion." He wanted us to avoid that trap. That was a long time ago though.

[Steviepinhead]

There's no trap like an old trap.

[Testycalibrated]

Depending on the euphemism.

[Pingu]

Guys, I've been working on this:

http://www.nottingham.ac.uk/toolkits/play_14143

Any feedback would be gratefully received.

(I know that the .png tables are crap, but for some reason the gui won't let me specify the table spacing and they took up far too much screen real-estate as tables - will fix)

ETA the one on Operationalisation should come first, really, but I didn't realise they didn't get it until I'd done the rest.

There are more to come.

ETA2: they are an attempt to extend this:

http://www.nottingham.ac.uk/toolkits/play_244

Which is brilliant, but I don't have the coding skills to do, or the brilliance, basically.

[VoxRat]

Guys, I've been working on this:

http://www.nottingham.ac.uk/toolkits/play_14143

Any feedback would be gratefully received.

(I know that the .png tables are crap, but for some reason the gui won't let me specify the table spacing and they took up far too much screen real-estate as tables - will fix)

ETA the one on Operationalisation should come first, really, but I didn't realise they didn't get it until I'd done the rest.

There are more to come.

ETA2: they are an attempt to extend this:

http://www.nottingham.ac.uk/toolkits/play_244

Which is brilliant, but I don't have the coding skills to do, or the brilliance, basically.

Hey, well, you know there's only one person I know of at TR who's brilliant - by his own assessment.

Have you asked him?

[Pingu]

no

[Jet Black]

Depending on the euphemism.

Testycalibrated