’science’s dirtiest secret: The “scientific method” of testing hypotheses by statistical ......

Wattsupwiththat.com ^ | March 20 , 2010 | Anthony Watts

[Lonesome in Massachussets]

It’s called cherry picking.

From what I saw of the CRU emails, they didn’t only cherry pick the data they tortured it until it confessed.

The absolute first rule in the effective use of statistics is the need to accept the results, to resist the all too human urge to fudge when the we don’t get the answers we want, or expect or “know” to be true.

I sort of take issue with the premise of the article. For instance, when Eddington first measured the bending of light by an eclipse the data were not overwhelmingly clear, the “signal” was faint, but the use of statistical technique allowed him to infer, with a certain level of confidence, that the observations confirmed the General Theory of Relativity.

[Citizen Tom Paine]

Dear reader remember that you can’t PROVE anything with statistics.

[wolfcreek]

My Wife works in R & D. In her first year, a senior engineers told her that any test she performed could be manipulated to have a *desired* outcome.

It all comes down to integrity and honor.

[colorado tanker]

This just in. The IPCC has declared there is a consensus that this article is wrong, within a 90% confidence factor . . . .

[SunkenCiv]

/bingo