Hi new friends. see this thread for how this account works, how it's somewhat different from
@GraphCrimes
and how it's somewhat different from bad take twitter
The question wording is "do you agree with the phrase 'it's ok to be white'"
This is a phrase that has been co-oped by racists as a dog whistle for their white nationalism.
This is the quality of chart you would expect to be published on a dot info domain. Someone paid someone else money for this and thought it was a perfectly fine way to visualize data. I mean, where to begin...
@elonmusk
Question for you: did you fire the team that removed tons of bot accounts daily? Because if you did then....uhh you are inflating exactly how you accused twitter of doing.
The question asks if you agree with the "phrase". Reading the question my thought was to answer no because I disagree with the phrase how it is usually presented. It's not innocuous. It has socio-contextual meaning that is being washed away by rassmussen and Matt Walsh
There may actually be some interesting things in the dark deep crevices of the report. But calling the US military weak based on vibes is just bad data
My followers should know this already, but the science we have points towards to Huanan Market as the origin of the Pandemic. Every time they bring up an alternative they use bad data, mislead, or lie to do it.
Also pls stop saying this is how women view the height changes in men. The more famous version of this graph is literally the version with a skirt. Why are you weird
From a data perspective it's absolutely a red flag that
@jessesingal
says that the data is not collectable when we already have some data around the exact problem he is talking about.
If I was admitting that my career was based on something backed by zero evidence I would simply not accuse someone else of fomenting a moral panic in the very next tweet.
I'm going to explain it here. But basically you used an unnecessarily wide Y axis to make it seem like the temperature didn't increase. Why have a -10 and +30 range? It's WAY too wide
Apparently they have it fixed on the website
don't know the precise origin (was this taken on the TV? thats a question for
@DrGBuckingham
) but we see this a lot. Design/production ppl take a chart and go without checking with the data folks first.
This sucks so much
So much of the fun of Twitter exists because of these little bots that ppl made. So many cool little tools that let you analyze tweets or take screenshots.
This will be the end of them.
‼️REALLY excited to report my talk "Building a ggplot2 rollercoaster: Creating amazing 3D data visualizations in R" was accepted to rstudio::conf(2022)! Can't wait to see everyone in person and show you how to use
#RStats
to make incredible 3D
#dataviz
!
#rayshader
#rayrender
Can't have red in both Russia AND the countries that have imposed sanctions. Gotta use like green or something to differentiate or don't use the Russian flag
Good thread talking about the gas stove issue. At this point it seems likely that gas stoves are increasing asthma risk in children. But there is a good bit of uncertainty around how much. Lots of factors to consider
This new "gas stoves cause 12.7% of childhood asthma" thing is a great example of how we often aren't very good at communicating uncertainty
I don't really agree with these headlines 1/n
@MattZeitlin
One of the highlights of my career was finding an error in McKinsey data that they tried to bury after they pettily criticized my analysis
@Noahpinion
Did you find any data suggesting that increasing police training actually reduces propensity for violence? Did you check to see if police departments that increase training see reductions in use of force?