Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
shakedown.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
A community for live music fans with roots in the jam scene. Shakedown Social is run by a team of volunteers (led by @clifff and @sethadam1) and funded by donations.

Administered by:

Server stats:

283
active users

shakedown.social: AboutStatusProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.7

#correlation

0 posts0 participants0 posts today
Replied in thread
Public
Eric Maugendre

@data @datadon 🧵

How to assess a statistical model?
How to choose between variables?

Pearson's #correlation is irrelevant if you suspect that the relationship is not a straight line.

If monotonic relationship:
"#Spearman’s rho is particularly useful for small samples where weak correlations are expected, as it can detect subtle monotonic trends." It is "widespread across disciplines where the measurement precision is not guaranteed".
"#Kendall’s Tau-b is less affected [than Spearman’s rho] by outliers in the data, making it a robust option for datasets with extreme values."
Ref: statisticseasily.com/kendall-t

LEARN STATISTICS EASILY · Kendall Tau-b vs Spearman: Which Correlation Coefficient Wins?Discover why Kendall Tau-b vs Spearman Correlation is crucial for your data analysis and which coefficient offers the most reliable results.
#normality#normalDistribution#modeling
Public
Austin Huang ❤️

Sure, it's good to promote healthier life #habits, but doing so via #CreditScore still seems to be a way to punish the #poor for being poor (and thus having to do grocery irregularly, esp. for those living check-to-check). Also, desire to seek more #correlation with #creditworthiness will result in more and more factors being collected, furthering various biases... Not to mention #privacy...

theconversation.com/no-credit-

The ConversationNo credit score? A grocery list could be the next best thingWhat you buy at the grocery store can predict whether you pay your credit card bills on time, new research suggests.
Public
qurlyjoe

We’ve all heard the thing where mom says, “don’t go in the water for at least an hour after eating.”

I wonder how many generations of mothers, going back how many hundreds of thousands of years, in how many human and pre-human species, have cautioned their children about this because they saw what could happen?

And scientists are finally getting around to testing whether it’s really #causation, or just #correlation.

#science
Replied in thread
Public
Ramesh #NotGoingBack

⬆️ Connecting the dots regarding #China, #Russia, #Ukraine, #Iran, #SaudiArabia, #Israel. #BFD

Meanwhile, #Russia too is distancing itself from #Trump #VP #JDVance if its state TV coverage of #Vance reported by #JuliaDavis is to be believed

All of the above are discrete data points, and we know that EVERYONE who confuses #correlation and #causation ends up dying.

Still, it is tantalizing to imagine the implications if these are indeed connected 👀

@hanse_mina @tobie1 @GreenFire @lawyerjsd

Replied in thread
Public
Ramesh #NotGoingBack

⬆️ @timkmak

>> One of interim Iranian President Mohammad #mokhber first phone calls since ascending to the position was with the #kremlin. In October 2022, Mokhaber went to #Moscow to sell #Iraniandrones and #ballisticmissiles to #Russia

Now, I know that everyone who confuses #correlation and #causation ends up dying.

Still note correlation in timing of #Putin’s #China visit and #Raisi helicopter crash.

That correlation, and ➡️ mastodon.social/@rameshgupta/1

Why meet with TOP defense & security?

Public
Dr. Robert M Flight

What do you do when you encounter a weird thing in your research? Blog about it!

Inspired by today's sleuthing about weird looking correlation histograms, here is a post actually digging into it at least a little bit (there's probably way deeper one could go, but this was good enough for me).

Weird Correlation Patterns
rmflight.github.io/posts/2024-

Original :mastodon: post was: mastodon.social/@rmflight/1122

rmflight.github.ioDeciphering Life: One Bit at a Time - Weird Correlation Patterns
#BlogPost#Correlation#RStats
Public *
Dr. Robert M Flight

#RStats #statistics #correlation help?!

I've got a bunch (3.5 x 10^6) gene-gene _Kendall-tau_ correlations, where each correlation is based on 12 samples. When I visualize the histogram of correlations, with a default of 30 bins, I see pic #1. Which looks fine, personally.

When I go to my personal favorite number of bins, 100, I get pic #2, with weird lows below the peaks.

Is this just me being paranoid as to something weird causing the alternating high and low values?

edit to add cor type

3.5 x 10^6 gene-gene correlations, in a histogram of 30 bins.
3.5 x 10^6 gene-gene correlations, in a histogram of 100 bins.
Replied in thread
Public
MidsouthMouth.OctaviaKeats

@todayilearned
Sometimes.
W.E.I.R.D.-ly.

"Even the highly cited examples of 150-people networks have been criticised as overwhelmingly skewed towards rich, educated, and industrialised societies, with non-western cultures rarely mentioned. Confirmation bias may well be a factor in the popularity and acceptance of Dunbar’s number."

oxsci.org/end-of-dunbars-numbe

The Oxford Scientist · The end of Dunbar’s number: Have our social networks changed for good? - The Oxford ScientistHelen Collins explores evidence for and against Dunbar's number, the idea that our species' social groups are limited to around 150 people.
#sociology#psychology#anthropology
ExploreLive feeds
About