Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Locked thread
vodkat
Jun 30, 2012



cannot legally be sold as vodka

paging trig for expert analysis

Adbot
ADBOT LOVES YOU

vodkat
Jun 30, 2012



cannot legally be sold as vodka

LP0 ON FIRE posted:

https://www.youtube.com/watch?v=E1a231faW-k

*owns dad with instagram presence; sips chocolate syrup water*

jezus that was hard to watch, despite her mentally disabled dad she seemed cool, i wish her the best in all her non-computer related endeavours :3:

vodkat
Jun 30, 2012



cannot legally be sold as vodka
https://www.youtube.com/watch?v=eR68wr2pTe8

vodkat
Jun 30, 2012



cannot legally be sold as vodka

i think my brain might be broken forever

vodkat
Jun 30, 2012



cannot legally be sold as vodka

Trig Discipline posted:

actually there's a strong argument to be made (and it is currently gaining ground) that parameter estimates and confidence intervals are a much more appropriate way to do most of what people currently do with frequentist stats, so you kinda know the good bit

i haven't read this book about it but i've heard it highly recommended: https://www.amazon.com/Understanding-New-Statistics-Meta-Analysis-Multivariate/dp/041587968X

basically the idea is that null hypothesis rejection is a pretty poo poo way to do statistics, and even within the frequentist paradigm we can do way better.

big scary monsters posted:

obviously a lot has been written since, but there are two classic papers on the subject by jacob cohen (who was a psychologist, so he really knew bad stats) that are easy to read and explain the problems of null hypothesis testing really well

Things I Have Learned (So Far), 1990 is an informal discussion of problems with common statistical methods, while The Earth is Round (p < .05), 1994 is a very nice and clear criticism of NHST with examples, and should be required reading for anyone doing science imo




Loving the stats chat, with the lads :cheers:

Adbot
ADBOT LOVES YOU

vodkat
Jun 30, 2012



cannot legally be sold as vodka

Sagebrush posted:

i don't really play video games any more because now i'm good enough at 3d modeling and embedded programming and other creative technological stuff that it just feels like a game. so instead of building a computer out of logic gates in minecraft, i go build one for real. as a side effect it allows me to feel extraordinarily smug

please stop reading https://www.reddit.com/r/iamverysmart/

  • Locked thread