What a debate about p-values shows about science

Re:What a debate about p-values shows about science by cwyers on Monday October 02, 2017 @09:41PM (#15389435)

> Ideally, Lakens says, the level of statistical significance needed to prove a hypothesis depends on how outlandish the hypothesis is.

> Yes, you’d want a very low p-value in a study that claims mental telepathy is possible. But do you need such stringent criteria for a well-worn idea? The high standards could impede young PhDs with low budgets.

Watching someone try to be Bayesian without using Bayesian techniques is painful.

> There are also new, advanced statistical techniques — like Bayesian analysis — that, in some ways, more directly evaluate a study’s outcome.

Almost as painful as watching Bayesian analysis described as "new."

[ Parent | on HN ]

Re:What a debate about p-values shows about science by sideshowb on Tuesday October 03, 2017 @09:24AM (#15392199)

Give them a break. It's a pop science article. Such articles can describe 1.5 concepts at most, so having used that budget on p-values, it's not available for Bayes, so explaining the concept from a frequentist perspective is reasonable.

Also Bayesian analysis (as a mainstream stats technique) is relatively new as it has only become mainstream with the advent of cheap computing power.

[ Parent | on HN ]

Re:What a debate about p-values shows about science by cwyers on Tuesday October 03, 2017 @04:13PM (#15395737)

Can I criticize the preprint from Lakens and his collaborators, which says the same thing as this "pop science" article?

https://psyarxiv.com/9s3y6

[ Parent | on HN ]

Re:What a debate about p-values shows about science by sideshowb on Tuesday October 03, 2017 @06:12PM (#15396661)

Oh alright go on then ;)

[ Parent | on HN ]

Re:What a debate about p-values shows about science by vinchuco on Tuesday October 03, 2017 @04:54AM (#15391077)

Naive question: Is there a good technique for a bot that detects factually wrong statements?

[ Parent | on HN ]

Re:What a debate about p-values shows about science by 09094920394314 on Tuesday October 03, 2017 @05:01AM (#15391106)

Sure, just give me definition of factually wrong and I'build it for you:)

Less snarky: not really

[ Parent | on HN ]

[dead]

Re:What a debate about p-values shows about science by sideshowb on Tuesday October 03, 2017 @09:21AM (#15392183)

Ahem. Liberal here working on such a system (in concept anyway).

[ Parent | on HN ]

Re:What a debate about p-values shows about science by 201709User on Tuesday October 03, 2017 @10:36AM (#15392673)

Then machine learning it is.

[ Parent | on HN ]