[Ads-l] p-hacking

[Dan Goncharoff]

In the way that a successful crime is positive for the criminal but
negative for society.

On Mon, Oct 8, 2018, 11:06 AM Dave Hause <dwhause at cablemo.net> wrote:

> But positive from the viewpoint of the hacker.
>
> -----Original Message-----
> From: Dan Goncharoff
> Sent: Monday, October 8, 2018 8:12 AM
> To: ADS-L at LISTSERV.UGA.EDU
> Subject: Re: p-hacking
>
> I would put this closer to the older meaning of computer hack, a way of
> getting around a system to obtain something.
>
> In this case, one gets around mathematical methods of determining data
> significance.
>
> This is a negative hack, not a positive hack.
>
> On Mon, Oct 8, 2018, 12:57 AM Dave Hause <dwhause at cablemo.net> wrote:
>
> > Might this version of hack be in the sense of a clever way to do
> > something,
> > similar to NOAD, 2nd, n.: 2. . . .a piece of computer code that performs
> > some function, especially an unofficial alternative or addition to a
> > commercial program.
> > Dave Hause
> >
> > -----Original Message-----
> > From: George Thompson
> > Sent: Sunday, October 7, 2018 1:15 PM
> > To: ADS-L at LISTSERV.UGA.EDU
> > Subject: p-hacking
> >
> > A recent blog post used a term quite unfamiliar to me -- but, then, it
> > would be.  But it's not in the OED either.  It appeared in an
> > installment of a blog originally issued by NPR under the title
> > "Cornell Food Researcher's Downfall Raises Larger Questions For
> > Science"; the blog is called FOOD FOR THOUGHT.
> > September 26, 20183:07 PM ET
> >
> >
> https://www.npr.org/sections/thesalt/2018/09/26/651849441/cornell-food-researchers-downfall-raises-larger-questions-for-science
> >
> > "There's nothing wrong with having a lot of data and looking at it
> > carefully," Althouse says. "The problem is p-hacking."
> >
> > To understand p-hacking, you need to understand p-values. P-values
> > tell you how likely it is that a result from an experiment is due to
> > chance or natural variations in the experiment. For example, if you go
> > on a diet, the p-value can estimate the chance that the weight you
> > lost was the result of natural background fluctuations in your weight
> > as opposed to because of your new diet.
> >
> > P-hacking is when researchers play with data to arrive at results that
> > look like they're scientifically significant. For instance, they can
> > cherry pick data points, re-analyze the data in multiple ways or stop
> > an experiment early.
> >
> > I suppose the expression is connected with the OED's
> > hack, v.3, section 1.  transitive. To make a hack of; to use in an
> > indiscriminate way; to make common, stale, or trite by such treatment.
> > At least, the phrase in the definition "use in an indiscriminate way"
> > seemed appropriate.
> >
> > GAT
> >
> > --
> > George A. Thompson
> > The Guy Who Still Looks Stuff Up in Books.
> > Author of A Documentary History of "The African Theatre", Northwestern
> > Univ. Pr., 1998.
> >
> > But when aroused at the Trump of Doom / Ye shall start, bold kings,
> > from your lowly tomb. . .
> > L. H. Sigourney, "Burial of Mazeen", Poems.  Boston, 1827, p. 112
> >
> > The Trump of Doom -- also known as The Dunghill Toadstool.  (Here's a
> > picture of his great-grandfather.)
> >
> >
> http://www.parliament.uk/worksofart/artwork/james-gillray/an-excrescence---a-fungus-alias-a-toadstool-upon-a-dunghill/3851
> >
> > ------------------------------------------------------------
> > The American Dialect Society - http://www.americandialect.org
> >
> > ------------------------------------------------------------
> > The American Dialect Society - http://www.americandialect.org
> >
>
> ------------------------------------------------------------
> The American Dialect Society - http://www.americandialect.org
>
> ------------------------------------------------------------
> The American Dialect Society - http://www.americandialect.org
>

------------------------------------------------------------
The American Dialect Society - http://www.americandialect.org