Lightnews — Scholar-powered news

Reposted by Michael Friendly

Aidan Wright @aidangcw.bsky.social · 1d

A new take on the limitations of "psychometric networks" now out in Nature Human Behavior. You don't want to put too much confidence in individual edges. Something we cautioned against in 2017.

1/2

www.nature.com/articles/s41...

Statistical evidence in psychological networks - Nature Human Behaviour

Psychometric network models have become increasingly popular in psychology and the social sciences. Huth et al. show that a large proportion of reported network findings are based on weak or inconclusive evidence inviting caution when interpreting results.

www.nature.com

4 19 56

Michael Friendly @datavisfriendly.bsky.social · 13h

Welcome back!

How To Read This Chart @howtoreadthisch.art · 1d

The first edition of the new newsletter is out! Check it out IF YOU DARE (halloween)
www.howtoreadthisch.art/the-glorious...

The glorious return of How To Read This Chart

Chapter 1 So, uh, what have you been up to? The last time anyone received an issue of "How To Read This Chart," President Trump hadn't yet deployed the military to D.C., the Mets were only a game-and...

www.howtoreadthisch.art

2

Michael Friendly @datavisfriendly.bsky.social · 13h

Just to be clear, this thing with Penguin glyphs is an experiment. The first part was, can I actually do this-- draw parametric glyphs to represent Penguin data? Yes, I now can.
Second part: Can you see anything from this?

Michael Friendly @datavisfriendly.bsky.social · 21h

#rstats #dataviz
A "line-up" test has been proposed as a human significance test: Can an observer spot a difference that rejects a null hypothesis?

Here's glyphs for 20 penguins, representing the main variables with visual features.

There are THREE multivariate outliers here. CAN YOU FIND THEM?

1

Michael Friendly @datavisfriendly.bsky.social · 14h

#TodayinHistory #dataviz #OTD 📊
💀Oct 12, 1926 Edwin A. Abbott died in London, England 🇬🇧󠁧󠁢󠁥󠁮󠁧󠁿

1884: In _Flatland_, he gave literary description of life in a 2D world for people living in a 3D world
Celebrate FLAT EARTH Day! 🗺️🗺️
💡 But it gives an idea of how we 3D people can understand a 4+ D world.

Cover of Edwin Abbott's _Flatland_, subtitle: A Romance of Many Dimensions, By: A Square. It shows a pentagon image of a Flatlander's home

Image of a 3D sphere rising through the 2D plane of Flatland, showing that it would appear as a circle changing in size over time

Color portrait of Edwin Abbott, wearing clerical robes

2

Michael Friendly @datavisfriendly.bsky.social · 21h

The answers will be given to anyone who follows me, likes ❤️this post, or can wait 48 hours or less.

You will be surprised!

Michael Friendly @datavisfriendly.bsky.social · 21h

#rstats #dataviz
A "line-up" test has been proposed as a human significance test: Can an observer spot a difference that rejects a null hypothesis?

Here's glyphs for 20 penguins, representing the main variables with visual features.

There are THREE multivariate outliers here. CAN YOU FIND THEM?

1 2

Michael Friendly @datavisfriendly.bsky.social · 21h

#rstats #dataviz
A "line-up" test has been proposed as a human significance test: Can an observer spot a difference that rejects a null hypothesis?

Here's glyphs for 20 penguins, representing the main variables with visual features.

There are THREE multivariate outliers here. CAN YOU FIND THEM?

5 3 6

Michael Friendly @datavisfriendly.bsky.social · 2d

#TodayinHistory #dataviz #OTD 📊
💀Oct 10, 1687 Johann Heinrich Schulze died in Germany 🇩🇪

1727: Experiments paving the way to the development of photography: 🎞️Images obtained by action of light on a mixture of chalk, nitric acid, and silver salts

4

Michael Friendly @datavisfriendly.bsky.social · 2d

#rstats #dataviz
Everyone loves Penguins, right?

What about visualizing them using ~ schematic Chernoff faces? Here's a 1st proof-of-concept example Claude helped me with.
Bill sizes are way too small. Sex is mapped to eye shape, but can't easily be seen. Anyone want to help with this?

Glyphs representing a sample of 20 observations from the penguins data set, mapping the variables body_mass, flipper_len, bill_len, bill_dep to visual features, colored by species

2 2 7

Michael Friendly @datavisfriendly.bsky.social · 3d

Interesting field work in the AI jungle.
Here's an earlier small scale study I did with Claude

nightingaledvs.com/what-does-ai...

1 6

Michael Friendly @datavisfriendly.bsky.social · 3d

Stop SEED HACKING.
Reproducible research now demands set.seed(42) everywhere

Paul Julian @swampthingpaul.bsky.social · 4d

While digging through some code from a manuscript I recently read ... yes, that rabbit hole I came across this line and I think I just found my new favorite set.seed(...) 🤣

set.seed(i+42) # Don’t Panic. “What is the meaning of life, the universe, and everything?”

#Rstats

6

Reposted by Michael Friendly

Andrew Gelman et al. @statmodeling.bsky.social · 3d

The worst research papers I’ve ever published
statmodeling.stat.columbia.edu/2025/10/09/t...

The worst research papers I’ve ever published | Statistical Modeling, Causal Inference, and Social Science

statmodeling.stat.columbia.edu

2 7

Michael Friendly @datavisfriendly.bsky.social · 3d

# STOP DOING THIS: Dynamite plots
# STOP DOING THIS: Dynamite plots

2

Reposted by Michael Friendly

Gretchen McCulloch @gretchenmcc.bsky.social · 3d

Higgledy piggledy
Timothée Chalamet
Has a name meriting
Endless design

Much like his forerunner
Benedict Cumberbatch:
Hexasyllabically,
Easy to rhyme

tweet from childish branzino @absflora:
timothée chalamet is the new benedict cumberbatch in the sense that you can say ANYTHING and we know who you mean. tiffany chevrolet. timpanogos charlemagne. symphony cabernet. jiminy castaway.

52 630 2.1K

Michael Friendly @datavisfriendly.bsky.social · 3d

if(TRUE) {UGH!}

Where is this described or illustrated?
Do I have to look through all the graphs in my book?

1 2

Reposted by Michael Friendly

Everyday ggplot2 extension 📊🛠 @ginareynolds.bsky.social · 4d

Join us to hear the {ggiraph} story from @davidgohel.bsky.social at the #ggplot2 extenders meetup next Wedsnesday! #rstats 📊

extenders: bit.ly/ggplot2exten...
sign up: bit.ly/join-gg-exte...

ggplot2 extenders meetup questionnaire

bit.ly

4 6

Michael Friendly @datavisfriendly.bsky.social · 4d

#TodayinHistory #dataviz #OTD 📊
🎂Oct 8, 1873 Ejnar Hertzsprung born in Copenhagen, Denmark 🇩🇰
1911-1913: The Hertzsprung-Russell diagram, a log-log plot of luminosity as a function of temperature for stars
Not so beautiful as a plot, but those red giants catch the eye.
Revolutionized stellar physics

The original Hertzsprung-Russell diagram, a log-log plot of luminosity as a function of temperature. A cluster of points is circled in the upper right which violated the otherwise linear relationship of the main sequence ; these proved to be the Red Giants

Modern rendition of the H-R diagram, with named clusters identified

4

Michael Friendly @datavisfriendly.bsky.social · 5d

Great post by @andrew.heiss.phd on animating #rstats {dplyr} operations

www.andrewheiss.com/blog/2024/04...

Visualizing {dplyr}’s mutate(), summarize(), group_by(), and ungroup() with animations | Andrew Heiss

Visually explore how {dplyr}’s more complex core functions work together to wrangle data

www.andrewheiss.com

1 8 54

Michael Friendly @datavisfriendly.bsky.social · 5d

Cool. I didn't know about the axis dodge trick

scale_x_discrete(guide = guide_axis(n.dodge = 2))

IBD Doctor @ibddoctor.bsky.social · 6d

Every once in a while I need to fiddle with long category labels in #rstats ggplot2.
It always helps me to refer back to this excellent blog post by #AndrewHeiss @cararthompson.bsky.social
www.andrewheiss.com/blog/2022/06...

Quick and easy ways to deal with long labels in ggplot2 | Andrew Heiss

Explore different manual and automatic ways to rotate, dodge, recode, break up, and otherwise deal with long axis labels with ggplot2

www.andrewheiss.com

1 6

Reposted by Michael Friendly

Jan Broder Engler @jbengler.de · 6d

Being free and open-source, this is how you can support #tidyplots ❤️

tidyplots.org
github.com/jbengler/tid...

#rstats #dataviz #phd

2 13

Michael Friendly @datavisfriendly.bsky.social · 5d

But worse-- the width of the stairs to your basement. Years ago, I had to disassemble a dryer to get it down there 😬

Jordan Nafa @jordannafa.bsky.social · 5d

Pro-tip: Measure the width of your front door before you buy appliances. Learned this lesson the hard way 😮‍💨

1

Reposted by Michael Friendly

Nick Brown @steamtraen.eu · 5d

😅

1 13 56

Reposted by Michael Friendly

Jan Broder Engler @jbengler.de · 6d

This is how you can add normal data ellipses in #tidyplots 🐣

#rstats #dataviz #phd

library(tidyplots)

pca |>
tidyplot(x = pc1, y = pc2, color = group) |>
add_data_points() |>
add_ellipse()

2 26

Reposted by Michael Friendly

Melanie D’Arrigo @darrigomelanie.bsky.social · 5d

I feel like this photo of masked, armed men pepper spraying a pastor protecting his community is going to be a defining picture of this moment in America for a long, long time.

1.7K 21K 56K

Michael Friendly @datavisfriendly.bsky.social · 5d

😍
But maybe precision = invariance (inverse var)

Matthew Kay @mjskay.com · 5d

missed opportunity to rename all dispersion measures following the same scheme

OG variance = squariance
SD = stairiance (also known as rootsquariance)
Gini = fairiance
IQR = iquariance (pronounced with a schwa, like "aquarius")
MAD = marriance
precision = prairiance
etc

1 1 6

Michael Friendly @datavisfriendly.bsky.social · 6d

This is SO AMAZING! Data Vandals rock!

Jason Forrest @jasonforrest.bsky.social · 6d

We are simply delighted to have the opportunity to share with the commuters of the LA region. There's really a lot to discuss in the exhibition, and we hope you can check it out for yourself! More info here: art.metro.net/artworks/exh...

1 2