#p-hacking

[system-of-a-feather]

Wanted to share this very good video on the Replication Crisis as someone who took a whole college course on this taught by one of the researchers dedicating her career to this topic ^^

[cuttlecology]

sorry we p-hacked your boyfriend. yeah, it looks like he's statistically significant but he's actually not. your boyfriend is part of the reproducibility crisis now.

[spiritualdirections]

As I talk to MIT science types off the record, they admit that there's a lot of bad science out there, and for much of the reason in this comic strip: producing useless science is easier than establishing causality. Especially in any discipline that studies people in uncontrolled environments (like real life)

This just reinforces the point that I've made before, which is that science is a human activity, and therefore subject to human error, sin, pride, and other foibles.

That's not anti-science, but it's anti-scientism, the idea that 'what scientists say' = truth & rationality. (Scientists who are atheists tend to adhere to scientism, despite all evidence it is false.)

Stefan Gasic on Twitter

“@nntaleb @sapinker That’s the problem; taking hokum seriously 😑 #SocialSciences”

[thesociologicalcinema]

I'm not a big fan of the commonly seen *, **, *** system for denoting p-values on graphs. Let's embrace 21st century symbology and do better: p < 0.1 :) p < 0.05 :) :) p < 0.01 :) :) :) p < 0.001 :) :) :) :) p < 0.000000001 :) :) :) :) :) ~ @nelson_lab

[bookanimeart]

A researcher is having trouble P-hacking their study, so they call up some friends: a programmer and a urologist.

[adventuresinsocialpolicy]

Sanjay Srivastava

We can talk about bias from experimental effects without impugning someone’s character, so can we try to discuss bias from data analysis choices in the same way? I.e. less getting side-tracked by how the honest researcher wouldn’t ever p-hack, more systematic consideration of safeguards.

[joshisanonymous]

Interpreting uninterpretable P-values.

Lately, I’ve been trying to learn more about open science and how it relates to research I’ve done, research I’d like to do, and how it relates to sociolinguistics in general. One topic that comes up regularly when talking about open science is pre-registration. For those who aren’t familiar with this process, pre-registration refers to publishing a detailed, time-stamped description of your…

View On WordPress

[ushf]

The great statistics debate rumbles on

The general idea is that P-values are useful enough if taken in the right context and extremely misleading if viewed in isolation: https://www.nature.com/articles/d41586-017-07522-z http://andrewgelman.com/2017/09/26/abandon-statistical-significance/