Sign in

Mark Kelson

@markkelson.bsky.social
460 followers 1.8K following 40 posts

Data botherer and parkrun bore Currently reading: Lessons

PostsRepliesMedia
Mark Kelson @markkelson.bsky.social · 03/06/2025
It’s blurry photos time! (plus a nice photo of my cat - something for everyone) Great talks at the AI for a Healthier Future event. Xujiong Ye Mike Allen Gareth Hawkes Yanda Meng Beverley Shields Dr. Cédric Mesnage Jonathan Phillips James Davies Steph Selway Helen Chapman Aline Villavicencio
A seminar roomA seminar roomA seminar roomA seminar room
130
Mark Kelson @markkelson.bsky.social · 25/04/2025
Sign up here for a UKRN organised symposium on checks for reproducibility April 30th 13:00-14:30. Would Iove to get feedback on what we’ve done recently at Exeter. I promise i will start my talk with an anecdote of questionable veracity! www.ukrn.org/event/ukrn-w...
ukrn.org
UKRN Webinar: Symposium on Checks for Reproducibility | UK Reproducibility Network
030
Mark Kelson @markkelson.bsky.social · 08/04/2025
Hey everyone! The Institute for Data Science and AI at Exeter is now on BlueSky! @idsaiexeter.bsky.social
011
Mark Kelson @markkelson.bsky.social · 08/04/2025
My talk on our recent Reproducibility Evaluation at the GW4 Open Research Week is now available! www.youtube.com/watch?v=DroH... @drgbuckingham.bsky.social
youtube.com
GW4 ORW: Exploring the reproducibility of research publications: A Reproducibility Evaluation Pilot
YouTube video by GW4 Alliance
030
Reposted by Mark Kelson
UK Reproducibility Network @ukrepro.bsky.social · 06/03/2025
Great to see so many of our Institutional Leads and Local Network Leads together in Cardiff yesterday. Lots of discussion, ideas, plans and bright Welsh sunshine. Huge thanks to Cardiff School of Journalism, Media and Culture for hosting us. #collaboration #openscience #reproducibility
Group photo
033
Reposted by Mark Kelson
UK Reproducibility Network @ukrepro.bsky.social · 04/03/2025
Congratulations shortlisted nominees for the 2025 Dorothy Bishop Prize, recognising work of early career researchers promoting robust, rigorous and reproducible UK research: - Andre Maia Chagas - Annayah Prosser - Lewis Jones - Michael Biddle - Sacha Beniamine - Zahra Rahemtulla #OpenResearch
ukrn.org
DB Prize 2025 shortlisted nominees
0145
Reposted by Mark Kelson
Daryl O’Connor @daryloconnor.bsky.social · 05/03/2025
What a terrific, engaging and fun UK Reproducibility Network (UKRN) Institutional Leads Retreat in sunny Cardiff. Lots of exciting sector-leading work going on & lots more to come!
021
Mark Kelson @markkelson.bsky.social · 05/03/2025
A really great institutional leads away day for the UKRN in Cardiff. Lovely to meet so many e-colleagues in person. Great discussions and exciting initiatives planned! Thanks to Cardiff and JOMEC for hosting!
Room sign saying UKRN institutional leads meetingUKRN leads milling in a roomA view of the sea from a train windowPoynter's top 40 tips for journalists
010
Mark Kelson @markkelson.bsky.social · 12/02/2025
You only have until Friday (14th) to get your proposals in for the Metascience conference 2025 metascience.info/call-for-pro...
metascience.info
Call for Proposals - Metascience
The call for proposals for Metascience 2025 is open with a deadline of 7 February 2025.
010
Reposted by Mark Kelson
UK Reproducibility Network @ukrepro.bsky.social · 23/01/2025
The UKRN Dorothy Bishop Prize is open for nominations until midnight of Sunday 26 January 2025! www.ukrn.org/2024/12/12/t...
A lego minifig of Dorothy Bishop with a laptop, standing in front of the UK Reproducibility Network logo.
03224
Mark Kelson @markkelson.bsky.social · 23/01/2025
UKRN is on here now everyone! Get following!
000
Mark Kelson @markkelson.bsky.social · 13/01/2025
Had fun meeting our new Statistical Data Science MSc students today. Really excited about this programme which is a collaboration between Computer Science, Mathematics & Statistics and the Humanities. Looking forward to seeing what they can do! #SDS #UniversityOfExeter
A regular lecture room with some bright eyed data science students who have yet to become jaded with the world.
010
Reposted by Mark Kelson
Dr Matt Lloyd Jones (he/him) @befriendabacterium.bsky.social · 07/01/2025
Ah nothing like sitting down after a hard day looking through badly designed studies and relaxing with a bit of telly. What's this? Eat Smart: Secrets if the Glucose Goddess. Sounds fu-
031
Reposted by Mark Kelson
Alex Prichard @profalexp.bsky.social · 02/01/2025
Floella Benjamin introducing Rik Mayall is very funny, and very touching, and his five mantras for a full life sound pretty anarchistic to me…
0134
Reposted by Mark Kelson
Maarten van Smeden @maartenvsmeden.bsky.social · 23/12/2024
Let us start 2025 in a positive mood: here are 10 methods things researchers can worry *less* about in 2025
media.tenor.com
a countdown clock with the number 10 in the center
ALT: a countdown clock with the number 10 in the center
15259119
Reposted by Mark Kelson
Rude Law Dog @esghound.com · 01/01/2025
My new year's resolution is simple, automated and doable
Screenshot of Bsky app requiring alt text before postin
03010
Mark Kelson @markkelson.bsky.social · 27/12/2024
The image this tweet conjures never fails to amuse me. Also, it's my current mood. @mattywin.bsky.social
A tweet response to a tweet asking to say something about your dad. It says "He once opened a party popper directly into the bin to save on cleaning"
041
Mark Kelson @markkelson.bsky.social · 12/12/2024
A great day today. Great atmosphere on campus for graduation, a departmental social with lovely colleagues, and I saw an uprooted tree!
Daytime graduation millingAin't no party like a maths & stats departmental social in a uniquely ill-suited spaceAn uprooted tree
010
Mark Kelson @markkelson.bsky.social · 12/12/2024
As ever, graduation was a joy. Congrats all Apploed Data Science and Statistics students! I'm so proud! #UniversityofExeter #ADSS
Selfie of me in a gownExeter forum filled with people celebratingGroup photo of all the graduates outside the great hall
010
Mark Kelson @markkelson.bsky.social · 12/12/2024
Storm Darragh did a number on this tree on the way into Streatham campus
A fallen over tree with roots exposedFallen down tree
000
Reposted by Mark Kelson
Elizabeth Gadd @lizziegadd.bsky.social · 11/12/2024
I’m sighing a lot today.
Automating the assessment of impact- good bad or ugly’ slide
041
Reposted by Mark Kelson
Alfredo Sánchez-Tójar @asanchez-tojar.bsky.social · 10/12/2024
In 2020, we showed that studies published in ecological journals WITH a code-sharing policy showed a 21% reproducibility potential at max (Culina et al 2020, @plosbiology.bsky.social) We've now investigated journals WITHOUT a code-sharing policy. The results are worrying. 📰 doi.org/10.32942/X21...
doi.org
Code-sharing policies are associated with increased reproducibility potential of ecological findings
48635
Reposted by Mark Kelson
Noah Haber @whaleactually.com · 10/12/2024
In the later stages of getting to do a dream: a fully replicable manuscript. Every single number, stat, figure, etc. in the manuscript has a complete, open and pushbutton full pipeline all the way from data collection to seeing the number on the page.
6392
Mark Kelson @markkelson.bsky.social · 10/12/2024
*adjusts Ronnie Barker glasses* Police have described it as a mash and grab
000
Reposted by Mark Kelson
Dax Kellie @daxkellie.bsky.social · 09/12/2024
If you missed my talk but still want to learn how to make your R code more reproducible, my slides are here 🙂: daxkellie.quarto.pub/will-your-co... All the links to packages and resources I mentioned are there, so hopefully this can be a nice reference, too! #ESAus2024 #rstats #quartopub 🧪🌏
Title slide: Will your code run again? Tips for making code reproducible in R
15347107
Reposted by Mark Kelson
Daniel Padfield @padpadpadpad.bsky.social · 09/12/2024
Two days left to apply for this! Please share.
01221
Reposted by Mark Kelson
Khoa @khoavuumn.bsky.social · 08/12/2024
Real footage of a synthetic control model
14642118
Reposted by Mark Kelson
Dax Kellie @daxkellie.bsky.social · 08/12/2024
It’s true, if you’re feeling worried about your R code, soothe your stress like a chamomile tea & come learn some tips about reproducibil(tea) with me tomorrow at 4:30pm #ESAus2024 😀☕️ (sorry, the pun was there & it would’ve been rude not to take it)
1368
Mark Kelson @markkelson.bsky.social · 06/12/2024
I'm workshopping a new questionable research practice acronymn: HOSPITALing Highlighting Only Significant P-values In The Abstract Like-the-last-30-years-of-science-reform-never-happened What do you think? #QRP #MetaScience
120
Reposted by Mark Kelson
Jamie Cummins @jamiecummins.bsky.social · 23/07/2024
Introducing RegCheck: a tool which uses Large Language Models to automatically compare preregistered protocols with their corresponding published papers and highlights deviations. @malte.the100.ci @ianhussey.bsky.social @ruben.the100.ci @bjoernhommel.bsky.social regcheck.app
regcheck.app
RegCheck.app
RegCheck is an AI tool to compare preregistrations with papers instantly.
510250
Reposted by Mark Kelson
Layal Hakim @drlayalhakim.bsky.social · 04/12/2024
Incredibly proud to see the work-in-progress talks from our 4th year MMath students at the seminar series. It's great to see the depth of their research, a true reflection of how much they've developed since I first taught them Mathematical Methods in 1st year🥹 Looking forward to the next seminar! 🙂
042
Mark Kelson @markkelson.bsky.social · 03/12/2024
Fascinating collection of talks & interests at the Maths & Stats, Computer Science and Business school workshop for research collaboration right now. Ranging from fully applied work to pure maths in search of application. Loads of super important work happening. Lots of connections being made!
A full room of nerdsEdwin Ip presenting Sonia Oreffice presenting on some work of huge societal importance
021
Reposted by Mark Kelson
Sébastien Fauconnet @daryo.me · 01/12/2024
La fameuse fenêtre d’Overton
4155521373
Mark Kelson @markkelson.bsky.social · 27/11/2024
Could these *be* any nuttier?
A packet of walnuts called "Chandler Walnuts". Geddit?
020
Mark Kelson @markkelson.bsky.social · 26/11/2024
Ok, this is great.
000
Reposted by Mark Kelson
Andrew Heiss @andrew.heiss.phd · 26/11/2024
Lots of the students in my dataviz class speak other languages, so I wrote up a little guide about how to change the locale in #rstats to get localized month/day names with {lubridate} and other R functions datavizf24.classes.andrewheiss.com/news/2024-11...
You can actually change R’s localization settings to get output in different languages!

If you want to see what your computer is currently set to use, run Sys.getLocale():

> Sys.getlocale()
## [1] "en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8"

There’s a bunch of output there—the first part (en_US.UTF-8) is the most important and tells you the language code. The code here follows a pattern and has three parts:

- A language: en. This is the langauge, and typically uses a two-character abbreviation following the ISO 639 standard
- A territory: US. This is the country or region for that language, used mainly to specify the currency. If it’s set to en_US, it’ll use US conventions (like “$” and “color”); if it’s set to en_GB it’ll use British conventions (like “£” and “colour”). It uses a two-character abbreviation following the ISO 3166 standard.
- An encoding: UTF-8. This is how the text is actually represented and stored on the computer. This defaults to Unicode (UTF-8) here. You don’t generally need to worry about this.For macOS and Linux (i.e. Posit Cloud), setting locale details is pretty straightforward and predictable because they both follow this pattern consistently:

- en_GB: British English
- fr_FR: French in France
- fr_CH: French in Switzerland
- de_CH: German in Switzerland
- de_DE: German in Germany

If you run locale -a in your terminal (not in your R console) on macOS or in Posit Cloud, you’ll get a list of all the different locales your computer can use. Here’s what I have on my computer:

 [1] "af_ZA" "am_ET" "be_BY" "bg_BG" "C"     "ca_ES" "cs_CZ" "da_DK" "de_AT" "de_CH" "de_DE" "el_GR" "en_AU" "en_CA"
[15] "en_GB" "en_IE" "en_NZ" "en_US" "es_ES" "et_EE" "eu_ES" "fi_FI" "fr_BE" "fr_CA" "fr_CH" "fr_FR" "he_IL" "hi_IN"
[29] "hr_HR" "hu_HU" "hy_AM" "is_IS" "it_CH" "it_IT" "ja_JP" "kk_KZ" "ko_KR" "lt_LT" "nl_BE" "nl_NL" "no_NO" "pl_PL"
[43] "POSIX" "pt_BR" "pt_PT" "ro_RO" "ru_RU" "sk_SK" "sl_SI" "sr_YU" "sv_SE" "tr_TR" "uk_UA" "zh_CN" "zh_HK" "zh_TW"

For whatever reason, Windows doesn’t use this naming convention. It uses dashes or full words instead, like en-US or american or en-CA or canadian. You can see a list here, or google Windows language country strings (that’s actually RStudio’s official recommendation for finding Windows language codes)Those functions have a locale argument, though, so it’s really easy to switch between languages:

> wday(x, label = TRUE, abbr = FALSE, locale = "en_US")
## [1] Friday
## Levels: Sunday < Monday < Tuesday < Wednesday < Thursday < Friday < Saturday

> wday(x, label = TRUE, abbr = FALSE, locale = "fr_FR")
## [1] Vendredi
## Levels: Dimanche < Lundi < Mardi < Mercredi < Jeudi < Vendredi < Samedi

> wday(x, label = TRUE, abbr = FALSE, locale = "fr_BE")
## [1] Vendredi
## Levels: Dimanche < Lundi < Mardi < Mercredi < Jeudi < Vendredi < Samedi

> wday(x, label = TRUE, abbr = FALSE, locale = "it_IT")
## [1] Venerdì
## Levels: Domenica < Lunedì < Martedì < Mercoledì < Giovedì < Venerdì < Sabato

> wday(x, label = TRUE, abbr = FALSE, locale = "zh_CN")
## [1] 星期五
## Levels: 星期日 < 星期一 < 星期二 < 星期三 < 星期四 < 星期五 < 星期六You can also set the locale for your entire R session like this:

> Sys.setlocale(locale = "de_DE")
## [1] "de_DE/de_DE/de_DE/C/de_DE/en_US.UTF-8"

Now month() and wday() will use German by default without needing to set the locale argument:

> month(x, label = TRUE, abbr = FALSE)
## [1] Juli
## 12 Levels: Januar < Februar < März < April < Mai < Juni < Juli < August < September < Oktober < ... < Dezember
> wday(x, label = TRUE, abbr = FALSE)
## [1] Freitag
## Levels: Sonntag < Montag < Dienstag < Mittwoch < Donnerstag < Freitag < Samstag
48717
Reposted by Mark Kelson
Alberto Cairo @albertocairo.com · 26/11/2024
Fun #dataViz!
0145
Reposted by Mark Kelson
Evan Peck @peck.phd · 20/11/2024
Trying something new: A 🧵 on a topic I find many students struggle with: "why do their 📊 look more professional than my 📊?" It's *lots* of tiny decisions that aren't the defaults in many libraries, so let's break down 1 simple graph by @jburnmurdoch.bsky.social 🔗 www.ft.com/content/73a1...
921577458
Mark Kelson @markkelson.bsky.social · 23/11/2024
Inauspicious sign at parkrun this morning
A Torbay Velopark sign half ripped down in the wind
110
Reposted by Mark Kelson
Rohan Alexander @rohanalexander.bsky.social · 22/11/2024
A student did their applied stats research project on The Strokes and it's taking all my academic integrity to not just give it 100% regardless of any flaws.
1121
Reposted by Mark Kelson
Jan Broder Engler @jbengler.de · 22/11/2024
The brand new `legendry` package by @teunbrand.bsky.social takes ggplots2 guides to the next level 🤩 teunbrand.github.io/teunbrand_bl...
1623355
Reposted by Mark Kelson
Layal Hakim @drlayalhakim.bsky.social · 21/11/2024
Snowy day! Postponing lectures to another day is a good idea. Staff and students might be struggling to come in, schools are closed, and let’s not forget, some students are parents too. Stay safe and cozy! ⛄ Here are some photos of our snowy campus @uniofexeter.bsky.social
051
Reposted by Mark Kelson
Daniel Padfield @padpadpadpad.bsky.social · 03/10/2024
PLEASE SHARE: PhD opportunity to start September 2025! Come work on climate change and antibiotic resistance with me, Ed Feil, Angus Buckling, and others. We will use lab experiments, modelling and sequencing to look how plasmid transfer, selection, and costs of resistance change with temperature.
21224
Mark Kelson @markkelson.bsky.social · 20/09/2024
Ran an induction for the latest Applied Data Science and Statistics MSc cohort today. Looking forward to seeing how you all do! Former ADSS grads: let me know what you're up to!
020
Mark Kelson @markkelson.bsky.social · 10/09/2024
In March 2021, just as we emerged from lockdown, I was approached to be the independent statistician for a trial (AFRI-c: check out the logo to see a typically heroic trial acronym) exploring whether air filters were effective at preventing respiratory infections (including Covid-19) in care homes.
AFRI-c study
Air
Filters to prevent
Respiratory
Infection including
Covid in Care Homes
141
Reposted by Mark Kelson
Alex Peden @alexpeden.bsky.social · 20/08/2024
I've had a go at starting a starter pack for #ResearchCulture, I'm pretty new here so please don't be offended if I haven't found you yet. Happy to add people to this, let me know! #Research #ResearchDevelopment #OpenResearch #ResearcherDevelopment #ResponsibleMetrics go.bsky.app/FAzhuGh
185331
Reposted by Mark Kelson
Laura Helmuth @laurahelmuth.bsky.social · 16/08/2024
What is 1 – 1 + 1 – 1 + 1 – 1 + ...? The question puzzled some of the greatest mathematicians of the 18th century and led one to claim it explained how God created the universe. www.scientificamerican.com/article/1-th... 🧪
scientificamerican.com
The Paradox of 1 – 1 + 1 – 1 + 1 – 1 + …
Why a mathematician thought this infinite series explained how God created the universe
64412
Reposted by Mark Kelson
Gordon Hodson @gordonhodsonphd.bsky.social · 15/08/2024
Analysis of Russell Group UK universities finds: ✴️ Women take much longer to get academic promotions. Think of the cumulative effects of such disparities. www.tandfonline.com/doi/full/10....
21411
Reposted by Mark Kelson
Peder M Isager @isager.bsky.social · 13/08/2024
New blog post: Seven basic rules for causal inference: pedermisager.org/blog/seven_b..., in which I state the most basic ways that causal relationships can produce correlation, and provide R code to demonstrate how each rule plays out in data.
pedermisager.org
Seven basic rules for causal inference | Peder M. Isager
Personal website of Dr. Peder M. Isager
075
Mark Kelson @markkelson.bsky.social · 15/08/2024
Just used Sky Bridge to port my twitter followers over to this platform. Worked really well. Recommend! Details here. medium.com/@jamestplunk....
medium.com
How to leave Twitter
Be part of the crumbling
100