Bob C-J and Geoff Cumming
@thenewstats.bsky.social
Open science, estimation statistics, and random thoughts from Bob Calin-Jageman and Geoff Cumming. https://thenewstatistics.com/itns/
Estimation is catching on!
The esci module in @jamovi.bsky.social was installed over 3,000 times in November. The R package in CRAN is getting an additional 250 downloads a month.
Are you estimation curious? So easy to get started with esci: rcalinjageman.github.io/esci/
#stats
The esci module in @jamovi.bsky.social was installed over 3,000 times in November. The R package in CRAN is getting an additional 250 downloads a month.
Are you estimation curious? So easy to get started with esci: rcalinjageman.github.io/esci/
#stats
November 25, 2025 at 8:30 PM
Estimation is catching on!
The esci module in @jamovi.bsky.social was installed over 3,000 times in November. The R package in CRAN is getting an additional 250 downloads a month.
Are you estimation curious? So easy to get started with esci: rcalinjageman.github.io/esci/
#stats
The esci module in @jamovi.bsky.social was installed over 3,000 times in November. The R package in CRAN is getting an additional 250 downloads a month.
Are you estimation curious? So easy to get started with esci: rcalinjageman.github.io/esci/
#stats
The analysis examined correlation coefficients reported in 2022 issues of Science, Nature, and Nature Neuroscience.
Of 177 articles mentioning strength of correlation, 45% relied solely on point estimates; 53% only reported the p value to accompany these claims!
What are we even doing here?
Of 177 articles mentioning strength of correlation, 45% relied solely on point estimates; 53% only reported the p value to accompany these claims!
What are we even doing here?
November 20, 2025 at 7:27 PM
The analysis examined correlation coefficients reported in 2022 issues of Science, Nature, and Nature Neuroscience.
Of 177 articles mentioning strength of correlation, 45% relied solely on point estimates; 53% only reported the p value to accompany these claims!
What are we even doing here?
Of 177 articles mentioning strength of correlation, 45% relied solely on point estimates; 53% only reported the p value to accompany these claims!
What are we even doing here?
September 4, 2025 at 12:00 AM
Feel like an excuse to get that travel money for some sun and surf in November? Consider these two back-to-back meetings Down Under. 16-18 and 19-21 November 2025.
researchintegrityconf.com
aimos-inc.github.io/aimos.confer...
researchintegrityconf.com
aimos-inc.github.io/aimos.confer...
July 30, 2025 at 10:51 PM
Feel like an excuse to get that travel money for some sun and surf in November? Consider these two back-to-back meetings Down Under. 16-18 and 19-21 November 2025.
researchintegrityconf.com
aimos-inc.github.io/aimos.confer...
researchintegrityconf.com
aimos-inc.github.io/aimos.confer...
Look what you can find in the latest version of JASP:
**esci**
That's right, all your favorite estimation-focused analyses, strong hypothesis testing, and meta-analysis are available in the esci module for JASP 0.95.
#stats #metascience
Many thanks to the good people at @jaspstats.bsky.social
**esci**
That's right, all your favorite estimation-focused analyses, strong hypothesis testing, and meta-analysis are available in the esci module for JASP 0.95.
#stats #metascience
Many thanks to the good people at @jaspstats.bsky.social
July 29, 2025 at 2:27 AM
Look what you can find in the latest version of JASP:
**esci**
That's right, all your favorite estimation-focused analyses, strong hypothesis testing, and meta-analysis are available in the esci module for JASP 0.95.
#stats #metascience
Many thanks to the good people at @jaspstats.bsky.social
**esci**
That's right, all your favorite estimation-focused analyses, strong hypothesis testing, and meta-analysis are available in the esci module for JASP 0.95.
#stats #metascience
Many thanks to the good people at @jaspstats.bsky.social
Every day, neuroscience brings science fiction closer to reality.
Today's example: A real-time brain-to-speech interface.
Not perfect, but so exciting.
Another example of the future we're losing by destroying science funding in the US.
#neuroskyence
pubmed.ncbi.nlm.nih.gov/39837322/
Today's example: A real-time brain-to-speech interface.
Not perfect, but so exciting.
Another example of the future we're losing by destroying science funding in the US.
#neuroskyence
pubmed.ncbi.nlm.nih.gov/39837322/
May 3, 2025 at 2:30 PM
Every day, neuroscience brings science fiction closer to reality.
Today's example: A real-time brain-to-speech interface.
Not perfect, but so exciting.
Another example of the future we're losing by destroying science funding in the US.
#neuroskyence
pubmed.ncbi.nlm.nih.gov/39837322/
Today's example: A real-time brain-to-speech interface.
Not perfect, but so exciting.
Another example of the future we're losing by destroying science funding in the US.
#neuroskyence
pubmed.ncbi.nlm.nih.gov/39837322/
Sex differences are often claimed based on faulty comparison of statistical significance status rather than direct comparison.
That's the upshot of an amazing talk I got to hear this week from Donna Maney @ Emory based on this paper & others: elifesciences.org/articles/70817
#stats #neuroskyence
That's the upshot of an amazing talk I got to hear this week from Donna Maney @ Emory based on this paper & others: elifesciences.org/articles/70817
#stats #neuroskyence
May 3, 2025 at 12:21 AM
Sex differences are often claimed based on faulty comparison of statistical significance status rather than direct comparison.
That's the upshot of an amazing talk I got to hear this week from Donna Maney @ Emory based on this paper & others: elifesciences.org/articles/70817
#stats #neuroskyence
That's the upshot of an amazing talk I got to hear this week from Donna Maney @ Emory based on this paper & others: elifesciences.org/articles/70817
#stats #neuroskyence
When your job is opening the door at the Tesla showroom you have plenty of time for Snapchat! #teslatakedown
April 12, 2025 at 7:03 PM
When your job is opening the door at the Tesla showroom you have plenty of time for Snapchat! #teslatakedown
Just wishing out loud. But Orcid does allow you to define websites and social media links.
I'd settle for something offering suggested text-matches from the authors' names to bsky accounts. Just want something to maker it easy to connect with authors when discussing their work.
I'd settle for something offering suggested text-matches from the authors' names to bsky accounts. Just want something to maker it easy to connect with authors when discussing their work.
April 3, 2025 at 8:17 PM
Just wishing out loud. But Orcid does allow you to define websites and social media links.
I'd settle for something offering suggested text-matches from the authors' names to bsky accounts. Just want something to maker it easy to connect with authors when discussing their work.
I'd settle for something offering suggested text-matches from the authors' names to bsky accounts. Just want something to maker it easy to connect with authors when discussing their work.
Sweet sounds best accompanied by some beautiful green lines.
March 12, 2025 at 2:25 AM
Sweet sounds best accompanied by some beautiful green lines.
An exciting addition to my “to read” pile: Unreliable by @csabaszabo.bsky.social .
What’s the word for when you are reading about problems in science and feel angry, appalled, sad and amazed all at once?
By this time next year will there even be any science to fix in the US?
#stats #metascience
What’s the word for when you are reading about problems in science and feel angry, appalled, sad and amazed all at once?
By this time next year will there even be any science to fix in the US?
#stats #metascience
March 11, 2025 at 6:20 PM
An exciting addition to my “to read” pile: Unreliable by @csabaszabo.bsky.social .
What’s the word for when you are reading about problems in science and feel angry, appalled, sad and amazed all at once?
By this time next year will there even be any science to fix in the US?
#stats #metascience
What’s the word for when you are reading about problems in science and feel angry, appalled, sad and amazed all at once?
By this time next year will there even be any science to fix in the US?
#stats #metascience
@standupforscience.bsky.social in Chicago! Cold but inspiring. Let’s get to work!
#science #standupforscience
#science #standupforscience
March 7, 2025 at 7:13 PM
@standupforscience.bsky.social in Chicago! Cold but inspiring. Let’s get to work!
#science #standupforscience
#science #standupforscience
I still see folks questioning if PubPeer and data sleuths go too far. Anyone worried about scientific vigilantism should read Doctored. Integrity mechanisms aren't working, they're more needed than we'd like to think, and the pubpeer crowd has done way more good than harm for AD research.
February 25, 2025 at 2:41 AM
I still see folks questioning if PubPeer and data sleuths go too far. Anyone worried about scientific vigilantism should read Doctored. Integrity mechanisms aren't working, they're more needed than we'd like to think, and the pubpeer crowd has done way more good than harm for AD research.
Also, the HYEMA approach van Aert introduces, by reasonably prioritizing pre-registered strudies, estimates a negligible effect of red on women's ratings of male attractiveness, much weaker than the RE meta-analysis I helped conduct that made no distinction between studies.
February 21, 2025 at 2:56 PM
Also, the HYEMA approach van Aert introduces, by reasonably prioritizing pre-registered strudies, estimates a negligible effect of red on women's ratings of male attractiveness, much weaker than the RE meta-analysis I helped conduct that made no distinction between studies.
February 14, 2025 — R.I.P. to the greatest research system in human history. Like the willful torching of the library of Alexandria. It’s the United States of Amerika now.
February 14, 2025 at 8:00 PM
February 14, 2025 — R.I.P. to the greatest research system in human history. Like the willful torching of the library of Alexandria. It’s the United States of Amerika now.
A happy reader of Intro to the New Stats sent along this image of Statler and Waldorf lounging on the book. Guessing they're roasting p values. Read it and you can laugh along, too.
#stats
#stats
February 11, 2025 at 10:11 PM
A happy reader of Intro to the New Stats sent along this image of Statler and Waldorf lounging on the book. Guessing they're roasting p values. Read it and you can laugh along, too.
#stats
#stats
Just finished The Great Pretender
It's a devastating investigation of the Rosenhan study (we should stop teaching it) -- but more than that it is an thoughtful, incisive history and commentary on the history of psychiatry. Strong recommend.
(Thanks for this book rec, @susan-nolan.bsky.social )
It's a devastating investigation of the Rosenhan study (we should stop teaching it) -- but more than that it is an thoughtful, incisive history and commentary on the history of psychiatry. Strong recommend.
(Thanks for this book rec, @susan-nolan.bsky.social )
December 16, 2024 at 6:23 PM
Just finished The Great Pretender
It's a devastating investigation of the Rosenhan study (we should stop teaching it) -- but more than that it is an thoughtful, incisive history and commentary on the history of psychiatry. Strong recommend.
(Thanks for this book rec, @susan-nolan.bsky.social )
It's a devastating investigation of the Rosenhan study (we should stop teaching it) -- but more than that it is an thoughtful, incisive history and commentary on the history of psychiatry. Strong recommend.
(Thanks for this book rec, @susan-nolan.bsky.social )
Tired: Testing against a null of exactly 0
Wired: Testing against an interval null
An interval null is more stringent and fair. Test results are: substantive, negligible, or inconclusive.
esci tests interval nulls for many designs:
rcalinjageman.github.io/esci/
#statsky #neuroskyence
Wired: Testing against an interval null
An interval null is more stringent and fair. Test results are: substantive, negligible, or inconclusive.
esci tests interval nulls for many designs:
rcalinjageman.github.io/esci/
#statsky #neuroskyence
December 2, 2024 at 4:11 PM
Tired: Testing against a null of exactly 0
Wired: Testing against an interval null
An interval null is more stringent and fair. Test results are: substantive, negligible, or inconclusive.
esci tests interval nulls for many designs:
rcalinjageman.github.io/esci/
#statsky #neuroskyence
Wired: Testing against an interval null
An interval null is more stringent and fair. Test results are: substantive, negligible, or inconclusive.
esci tests interval nulls for many designs:
rcalinjageman.github.io/esci/
#statsky #neuroskyence
Tired: comparing means
Wired: comparing medians!
esci and statpsych let you compare medians for simple & complex designs. That's right, you can express an interaction as a difference in the difference of *medians*.
dgbonett.github.io/statpsych/
rcalinjageman.github.io/esci/
#stats #neuroskynce
Wired: comparing medians!
esci and statpsych let you compare medians for simple & complex designs. That's right, you can express an interaction as a difference in the difference of *medians*.
dgbonett.github.io/statpsych/
rcalinjageman.github.io/esci/
#stats #neuroskynce
November 13, 2024 at 6:15 PM
Tired: comparing means
Wired: comparing medians!
esci and statpsych let you compare medians for simple & complex designs. That's right, you can express an interaction as a difference in the difference of *medians*.
dgbonett.github.io/statpsych/
rcalinjageman.github.io/esci/
#stats #neuroskynce
Wired: comparing medians!
esci and statpsych let you compare medians for simple & complex designs. That's right, you can express an interaction as a difference in the difference of *medians*.
dgbonett.github.io/statpsych/
rcalinjageman.github.io/esci/
#stats #neuroskynce
This seems very reasonable to me:
July 14, 2024 at 11:23 PM
This seems very reasonable to me:
esci is now on CRAN!
Use esci to generate effect sizes and confidence intervals for lots of basic designs, beautiful figures emphasizing estimation, meta-analysis, and *interval hypothesis tests*.
install.packages("esci")
docs: rcalinjageman.github.io/esci/
#stats #rstats
Use esci to generate effect sizes and confidence intervals for lots of basic designs, beautiful figures emphasizing estimation, meta-analysis, and *interval hypothesis tests*.
install.packages("esci")
docs: rcalinjageman.github.io/esci/
#stats #rstats
March 21, 2024 at 8:43 PM
esci is now on CRAN!
Use esci to generate effect sizes and confidence intervals for lots of basic designs, beautiful figures emphasizing estimation, meta-analysis, and *interval hypothesis tests*.
install.packages("esci")
docs: rcalinjageman.github.io/esci/
#stats #rstats
Use esci to generate effect sizes and confidence intervals for lots of basic designs, beautiful figures emphasizing estimation, meta-analysis, and *interval hypothesis tests*.
install.packages("esci")
docs: rcalinjageman.github.io/esci/
#stats #rstats
2nd edition of ITNS is ready for your stats/methods classes! Estimation first, rigorous testing, with meta-analysis and Open Science integrated throughout. Get a desk copy here: tinyurl.com/yc4yn42s
Big thanks to all who helped make this possible. Feedback welcome!
#stats #OpenScience
Big thanks to all who helped make this possible. Feedback welcome!
#stats #OpenScience
March 7, 2024 at 1:41 AM
2nd edition of ITNS is ready for your stats/methods classes! Estimation first, rigorous testing, with meta-analysis and Open Science integrated throughout. Get a desk copy here: tinyurl.com/yc4yn42s
Big thanks to all who helped make this possible. Feedback welcome!
#stats #OpenScience
Big thanks to all who helped make this possible. Feedback welcome!
#stats #OpenScience
#StatsSky #rStats
Here's a fantastic tutorial by Jeffrey Spence on prediction intervals with practical examples using the predictionInterval package in R: journals.sagepub.com/doi/10.1177/...
Will definately be adding PIs to esci.
I especially like this bit:
Here's a fantastic tutorial by Jeffrey Spence on prediction intervals with practical examples using the predictionInterval package in R: journals.sagepub.com/doi/10.1177/...
Will definately be adding PIs to esci.
I especially like this bit:
February 25, 2024 at 2:33 PM
#StatsSky #rStats
Here's a fantastic tutorial by Jeffrey Spence on prediction intervals with practical examples using the predictionInterval package in R: journals.sagepub.com/doi/10.1177/...
Will definately be adding PIs to esci.
I especially like this bit:
Here's a fantastic tutorial by Jeffrey Spence on prediction intervals with practical examples using the predictionInterval package in R: journals.sagepub.com/doi/10.1177/...
Will definately be adding PIs to esci.
I especially like this bit: