iiinova — where your digital self keeps evolving

Ed Kroc @edkroc.bsky.social · 02/12/2025

Definitely the reason I prefer the quasi-likelihood approach for binomial regression.

Ed Kroc @edkroc.bsky.social · 02/12/2025

This follows from first principles. Mean of a Bernoulli is p and var is p(1-p). In the log-reg setting, your estimating the mean via your (generalized) linear combo of predictors and their SEs are functions of the (residual) var. Hence, if the var is off, so too is the mean and vice versa.

120

Reposted by Ed Kroc

Common Tern Project @commonternproject.bsky.social · 12/11/2025

There’s one month left to apply for the PhD position on interactions between contaminants and parasites in common terns, co-supervised by @alicecarravieri.bsky.social, Bernd Sures from @unidue.bsky.social and myself. You can do so here: euraxess.ec.europa.eu/jobs/380142

02232

Reposted by Ed Kroc

Víctor Martín-Vélez @victormartinvel.bsky.social · 08/11/2025

Interested in spatial ecology & One Health in urban gulls? 🧪 Join us at @imares-group.bsky.social @icmcsic.bsky.social as a #JdC postdoc in the #ALLIES project! 📅 Apply by Dec 10 📩 Contact: joan@icm.csic.es 🔗 See the call: acortar.link/yhYLRt #Postdoc #AppliedEcology #UrbanEcology #OneHealth

2128

Reposted by Ed Kroc

Charles Margossian @charlesm993.bsky.social · 07/11/2025

Applications for the PhD and MSc programs in statistics at UBC are now open! 📆 Deadline for PhD program is December 1st 📆 Deadline for MSc program is January 5th The department covers all areas of statistics and we have a lot of momentum in Bayesian computation!

2105

Ed Kroc @edkroc.bsky.social · 05/11/2025

That gorgeous black to white contrast! Reminds me of L. marinus.

010

Ed Kroc @edkroc.bsky.social · 08/08/2025

I have one but it's much more stats oriented than eco (a little of that too though). Also, I can get my hands on others but need to be discreet. Can I email you?

100

Ed Kroc @edkroc.bsky.social · 07/08/2025

To be fair, it's "lazy" because it's a theorem that requires a proof but often people just take it as definition. Casella and Berger were famously unhappy with the phrase LOTUS, taking a shot at Ross - who introduced the term - in their Statistical Inference text.

110

Ed Kroc @edkroc.bsky.social · 21/07/2025

Would be interesting to formally quantify that though.

020

Ed Kroc @edkroc.bsky.social · 21/07/2025

Yes, Spearman is not immune. It is quite literally a Pearson correlation on the rank data. But the deleterious effects are undoubtedly greatly mitigated because transforming data into ranks makes everything look more uniform, and uniform marginals will span the full theoretical -1 to +1 range.

120

Ed Kroc @edkroc.bsky.social · 16/07/2025

It's been 10 years for me and I still can't bring myself to chuck them. Someday!

110

Ed Kroc @edkroc.bsky.social · 14/07/2025

Lots of little gull floofs at the Tsawwassen ferry terminal this morning.

000

Ed Kroc @edkroc.bsky.social · 10/07/2025

Haha, yes, that happens to me too! 😆 I think serious statisticians can make good arguments for most paradigms. But when it comes to non-statistician scientists appropriately applying them, I think the Fisherian approach is the most foolproof.

000

Ed Kroc @edkroc.bsky.social · 10/07/2025

Science, unlike politics, is not about making binary decisions. It's about weighing available evidence and updating plausible beliefs. Those are continuous actions, not discrete.

100

Ed Kroc @edkroc.bsky.social · 10/07/2025

That's a good post. I still think the Neyman-Pearson framework completely misses how science is typically done. Rarely do we believe a hypothesis by default. We just have one that aligns best with the evidence or absence thereof.

100

Ed Kroc @edkroc.bsky.social · 10/07/2025

Building height continues to be the strongest predictor of nesting propensity for GWGUs.

010

Ed Kroc @edkroc.bsky.social · 10/07/2025

Hello, Sechelt. There are two rooftop GW Gull nests in this picture. Can you guess where? Answer below.

100

Reposted by Ed Kroc

Michelle Wille @duckswabber.bsky.social · 10/07/2025

In the UK, 2021-2023, 103,497 carcasses of 155 species tested for HPAI. Gannets, Black-headed Gulls, Barnacle Goose most affected. 7500 +ve for HPAI across 80 species. Mortality data to estimate impact. 👉 www.tandfonline.com/doi/full/10....

011

Reposted by Ed Kroc

Michelle Wille @duckswabber.bsky.social · 10/07/2025

Across 13 European countries, only 3 dead Little Terns found in 2022-2023, amongst tens of thousands of breeding birds. Surprising given how other tern species have been impacted. 👉 www.tandfonline.com/doi/full/10....

031

Ed Kroc @edkroc.bsky.social · 10/07/2025

Welcome! Looking forward to some gull content. 🌊🐥

120

Reposted by Ed Kroc

British Ornithologists’ Union (BOU) @bou.org.uk · 04/07/2025

Partial Niche Partitioning in Three Sympatric Gull Species Through Foraging Areas and Habitat Selection | doi.org/10.1002/ece3... | Ecology and Evolution | #ornithology #seabirds 🪶

03310

Reposted by Ed Kroc

Maudeux.creations 🎨🐦 @maudeuxcreations.bsky.social · 21/06/2025

the ivory gull, one of the most fascinating birds of the Far North! It stays in the Arctic all winter and follows polar bears to feed on the leftovers of their kills — a brilliant and icy survival strategy ❄️🐻‍❄️ #birdart #wildlifeart #sketchbook #seabirds #natureillustration 🌿 #SciArt 🐡

1315

Reposted by Ed Kroc

Maudeux.creations 🎨🐦 @maudeuxcreations.bsky.social · 14/06/2025

More Arctic gulls! The Sabine’s gull is a beautiful bird I dream to see one day! This watercolor is part of my “Arctic Larids” poster, available as a print in Maudeux Redbubble shop. #birdart #wildlifeart #sketchbook #sabinesgull #natureillustration #scientificillustration 🌿 #SciArt 🐡

1213

Reposted by Ed Kroc

Maudeux.creations 🎨🐦 @maudeuxcreations.bsky.social · 11/06/2025

Sketches of the Mediterranean gull... I quite like the one sunbathing 😎 These humorous drawings are based on real biology! #birdart #characterdesign #natureillustration #sketchbook #wildlifeart #MediterraneanGull #funnybirds #nonbreedingplumage 🌿 #SciArt 🐡 #FieldworkSketches

Sketches of mediterranean gull characters. They are enjoying the sun with sunglasses and cocktail. The idea is to illustrate its wintering grounds—far from the cold of Northern Europe ❄️.
So the bird is shown in its non-breeding plumage (yes, the black head is just for summer!)

Sketches of mediterranean gull characters. One is flying, an other one on the water and an other one is taking off its sunglasses.

Sketches of mediterranean gull characters. One has a breeding plumage and try to seduce a partner and the other one is looking at it's tags on its legs...

0122

Ed Kroc @edkroc.bsky.social · 14/06/2025

I'm glad to see you and your family are all safe.

120

Reposted by Ed Kroc

Common Tern Project @commonternproject.bsky.social · 08/06/2025

Weekend loveliness: retrapping a gull-billed tern carrying a GPS and getting to know a few of the newest colony members.

0385

Ed Kroc @edkroc.bsky.social · 08/06/2025

I always liked Maugham.

000

Ed Kroc @edkroc.bsky.social · 07/06/2025

Haha, yes, that is a great observation! 😂

010

Ed Kroc @edkroc.bsky.social · 06/06/2025

Haha, no worries! Yes, home base is Vancouver, but I'm always on the gulls up and down the N. American Pacific coast. Just starting to expand work into the Baja Peninsula!

010

Reposted by Ed Kroc

The Waterbird Society @waterbirdsociety.bsky.social · 06/06/2025

#Waterbirds in the news! @axios.com just featured Cimino et al.'s new study, "The First GPS Observation of a Western Gull Riding in a Long-Haul Garbage Transfer Truck" bioone.org/journals/wat... ttps://www.axios.com/local/san-francisco/2025/06/04/hitchhiking-seagulls-san-francisco-climate-change

axios.com

This seagull took an 80-mile truck ride twice to find food

Researchers say it's the first recorded case of a seagull using a vehicle for transport while foraging.

1115

Ed Kroc @edkroc.bsky.social · 29/05/2025

Hmmm, I think any kind of automatic/unthinking model selection is my biggest pet peeve. Thoughtful model selection should largely address both overparameterization and sample size constraints, at least if done reasonably!

100

Ed Kroc @edkroc.bsky.social · 28/05/2025

Lol! Fair warning I guess to not suggest me as a reviewer for a paper that uses it. But I'll be a nice reviewer for other work!

110

Ed Kroc @edkroc.bsky.social · 28/05/2025

I endorse this post 100%. "Automatic inference" is mostly an abomination. I have zero tolerance for it in ecology.

100

Reposted by Ed Kroc

UKCEHseabirds @ukcehseabirds.bsky.social · 22/05/2025

Finally the first Kittiwake egg! Three days later than last year and ten days later than average. Only five later dates. Due to birds being in poor condition or the dry conditions making nest building difficult? Maybe time will tell

0283

Ed Kroc @edkroc.bsky.social · 07/05/2025

So how does the meta-analyst define a coherent X? Classic options all have flaws, but RVVMs can handle this no problem. 3/3

000

Ed Kroc @edkroc.bsky.social · 07/05/2025

Say X is a continuous moderator of interest (e.g. age, exposure). Some studies report X exactly, others discretize it, reporting only a range. But different studies discretize in different ways, incongruously. 2/3

100

Ed Kroc @edkroc.bsky.social · 07/05/2025

New RVVM blog up. Here I discuss a new paper that, among other things, uses RVVMs to fix a common meta-analysis problem: incongruent categorization of a moderator across studies. 1/3 edkroc.wordpress.com/2025/05/06/r...

edkroc.wordpress.com

RVVM Blog 5: What’s My Age Again?

Blink-182 is one of those punk rock bands that: (1) everyone knows, (2) has some great songs, and yet, (3) overall, isn’t very good. Like NOFX or The Clash. Nevertheless, like those bands, th…

100

Ed Kroc @edkroc.bsky.social · 03/05/2025

Hem! No County for Old Men The Odfather Mulholland Dive And the thrilling legal trilogy: Lien Liens Lien 3

110

Reposted by Ed Kroc

Wetland Bird Survey (WeBS) | Goose & Swan Monitoring (GSMP) @webs-gsmp.bsky.social · 24/04/2025

The UK's oldest known Oystercatcher was discovered on The Wash, at 43 years old, a new British record! Read the article by @rspbscience.bsky.social below to find out more about the importance of ringing and conserving these vital habitats⬇️ www.rspb.org.uk/whats-happen... #Ornithology #UKBirding

rspb.org.uk

UK’s oldest known Oystercatchers discovered on England’s east coast wetlands

04412

Reposted by Ed Kroc

British Antarctic Survey 🐧 @bas.ac.uk · 22/04/2025

Why is Antarctica so important? Antarctica might feel very remote - until you look at the planet from a different perspective. It's actually central to our world's ocean currents. 🎁 Check the alt text to dig into this diagram 🌊 This is from a new book by BAS' @oceanandice.bsky.social (et al.)

The image shows a map of the world's currents, but looking at a projection of the world from the bottom of the globe, with Antarctica at its centre. The diagram shows all of the world's currents cycling through major oceans, all linked together by a central circular current that goes around the outside of the Antarctic continent.

"The Southern Ocean around Antarctica is central to global ocean circulation. It enables connectivity across the whole planet by linking the Atlantic, Pacific and Indian Ocean basins, and is the key site where waters are returned to the surface from depth and then converted into new water masses. The clockwise circulation of waters around Antarctica is the Antarctic Circumpolar Current, the strongest ocean current system in the world."

(Figure 1.1 from "Antarctica and the Earth System", edited By Michael P. Meredith, Jess Melbourne-Thomas, Alberto C. Naveira Garabato, Marilyn Raphael)

212741

Ed Kroc @edkroc.bsky.social · 20/04/2025

Let me know if you pursue anything like this approach and need or want any help. I'm just starting to expand the approach to other species south down the Pacific coast and am particularly interested in how robustly predictions can be validated across different species complexes.

110

Ed Kroc @edkroc.bsky.social · 19/04/2025

This approach requires access to useable building level data (open datasets, govt sources?). I am sure the exact features that are most predictive are different in the UK vs Salish Sea, but wouldn't be too hard to figure this out.

110

Ed Kroc @edkroc.bsky.social · 19/04/2025

We have had success in Vancouver using spatial and structural features of the urban environment to predict population size and trends. Expanding this work to the entire Salish Sea now and prelim results look good. See: www.sciencedirect.com/science/arti...

121

Reposted by Ed Kroc

Louise Blight @louiseblight.bsky.social · 06/04/2025

Waiting for the fish market to open. Glaucous-winged Gull pair. #SuperSeabirdSunday #birds #gulls

0133