Shinichi Nakagawa
@itchyshin.bsky.social
Behavioural ecologist, meta-analyst & manga-lover
Around 3 years ago, we started writing papers during our weekly lab meetings. This is the first one from our new lab at
@ualberta.bsky.social
Location-Scale Models for Ecologists!
preprint: ecoevorxiv.org/repository/v...
tutorial: ayumi-495.github.io/Eco_location...
Enjoy! #glmmTMB #brms
@ualberta.bsky.social
Location-Scale Models for Ecologists!
preprint: ecoevorxiv.org/repository/v...
tutorial: ayumi-495.github.io/Eco_location...
Enjoy! #glmmTMB #brms
July 25, 2025 at 12:52 PM
Around 3 years ago, we started writing papers during our weekly lab meetings. This is the first one from our new lab at
@ualberta.bsky.social
Location-Scale Models for Ecologists!
preprint: ecoevorxiv.org/repository/v...
tutorial: ayumi-495.github.io/Eco_location...
Enjoy! #glmmTMB #brms
@ualberta.bsky.social
Location-Scale Models for Ecologists!
preprint: ecoevorxiv.org/repository/v...
tutorial: ayumi-495.github.io/Eco_location...
Enjoy! #glmmTMB #brms
The biggest "meta-analysis" ever? - 17 million effect sizes - elifesciences.org/articles/95857
With wonderful @coreytcallaghan.bsky.social and Will Cornwell
We question the abundance-occupancy relationships using ebird data - we find ~ zero correlation
With wonderful @coreytcallaghan.bsky.social and Will Cornwell
We question the abundance-occupancy relationships using ebird data - we find ~ zero correlation
May 29, 2025 at 5:23 PM
The biggest "meta-analysis" ever? - 17 million effect sizes - elifesciences.org/articles/95857
With wonderful @coreytcallaghan.bsky.social and Will Cornwell
We question the abundance-occupancy relationships using ebird data - we find ~ zero correlation
With wonderful @coreytcallaghan.bsky.social and Will Cornwell
We question the abundance-occupancy relationships using ebird data - we find ~ zero correlation
A new paper from our lab
onlinelibrary.wiley.com/doi/10.1111/...
All meta-analysts can use this to get extra insights
onlinelibrary.wiley.com/doi/10.1111/...
All meta-analysts can use this to get extra insights
May 4, 2025 at 2:06 PM
A new paper from our lab
onlinelibrary.wiley.com/doi/10.1111/...
All meta-analysts can use this to get extra insights
onlinelibrary.wiley.com/doi/10.1111/...
All meta-analysts can use this to get extra insights
We have a fully funded MSc position at UofA, Canada
"Ecological time machines: bridging paleontology and community ecology"
www.i-deel.org/opportunitie...
co-supervised by Christine Sosiak (scholar.google.com/citations?us...)
Please repost!
"Ecological time machines: bridging paleontology and community ecology"
www.i-deel.org/opportunitie...
co-supervised by Christine Sosiak (scholar.google.com/citations?us...)
Please repost!
April 17, 2025 at 5:45 PM
We have a fully funded MSc position at UofA, Canada
"Ecological time machines: bridging paleontology and community ecology"
www.i-deel.org/opportunitie...
co-supervised by Christine Sosiak (scholar.google.com/citations?us...)
Please repost!
"Ecological time machines: bridging paleontology and community ecology"
www.i-deel.org/opportunitie...
co-supervised by Christine Sosiak (scholar.google.com/citations?us...)
Please repost!
April 8, 2025 at 12:23 PM
April 8, 2025 at 12:11 PM
Also - there is caterpillars plots too in the same package - orchaRd
April 8, 2025 at 12:07 PM
Also - there is caterpillars plots too in the same package - orchaRd
3 key versions of effect sizes.... (beautiful figs by Ayumi) - Joel says this is confusing but any ideas for improvement - welcome
April 6, 2025 at 2:42 AM
3 key versions of effect sizes.... (beautiful figs by Ayumi) - Joel says this is confusing but any ideas for improvement - welcome
Read the MS and find this fig - literately nested random effects and crossed random effects explained
April 6, 2025 at 2:42 AM
Read the MS and find this fig - literately nested random effects and crossed random effects explained
A new review on different types of repeatability (ICC) and how you may get it
ecoevorxiv.org/repository/v...
With my good #squid friends (Holger and my 10th variance-related paper).
ecoevorxiv.org/repository/v...
With my good #squid friends (Holger and my 10th variance-related paper).
April 6, 2025 at 2:42 AM
A new review on different types of repeatability (ICC) and how you may get it
ecoevorxiv.org/repository/v...
With my good #squid friends (Holger and my 10th variance-related paper).
ecoevorxiv.org/repository/v...
With my good #squid friends (Holger and my 10th variance-related paper).
The model conceptual visualization is here (by @szymekdr.bsky.social )
March 25, 2025 at 2:44 AM
The model conceptual visualization is here (by @szymekdr.bsky.social )
March 25, 2025 at 2:33 AM
With Beautiful artwork by Szymek
March 25, 2025 at 2:33 AM
With Beautiful artwork by Szymek
A new paper with @szymekdr.bsky.social and others
ecoevorxiv.org/repository/v...
We can now test adaptive radiation, pleiotropy or phenotypic integration beyond mean, co-divergence and "contra-divergence" of traits (see next)
Probably my favorite contribution so far to Evolutionary Biology
ecoevorxiv.org/repository/v...
We can now test adaptive radiation, pleiotropy or phenotypic integration beyond mean, co-divergence and "contra-divergence" of traits (see next)
Probably my favorite contribution so far to Evolutionary Biology
March 25, 2025 at 2:33 AM
A new paper with @szymekdr.bsky.social and others
ecoevorxiv.org/repository/v...
We can now test adaptive radiation, pleiotropy or phenotypic integration beyond mean, co-divergence and "contra-divergence" of traits (see next)
Probably my favorite contribution so far to Evolutionary Biology
ecoevorxiv.org/repository/v...
We can now test adaptive radiation, pleiotropy or phenotypic integration beyond mean, co-divergence and "contra-divergence" of traits (see next)
Probably my favorite contribution so far to Evolutionary Biology
March 19, 2025 at 2:21 PM
This is something useful for all meta-analysts (or systematic reviewers) - especially I like the analysis of alternative metrics - led by amazing Yefeng!
www.cambridge.org/core/journal...
www.cambridge.org/core/journal...
March 19, 2025 at 2:18 PM
This is something useful for all meta-analysts (or systematic reviewers) - especially I like the analysis of alternative metrics - led by amazing Yefeng!
www.cambridge.org/core/journal...
www.cambridge.org/core/journal...
A new paper led by @mlagisz.bsky.social - how to know how good your search string is
March 7, 2025 at 4:29 PM
A new paper led by @mlagisz.bsky.social - how to know how good your search string is
Do you use survival (time-to-event) analysis and interested in calculating repeatability (ICC - intra-class correlation) - here is how (it is just a simple formula)
authors.elsevier.com/sd/article/S...
authors.elsevier.com/sd/article/S...
February 20, 2025 at 8:58 PM
Do you use survival (time-to-event) analysis and interested in calculating repeatability (ICC - intra-class correlation) - here is how (it is just a simple formula)
authors.elsevier.com/sd/article/S...
authors.elsevier.com/sd/article/S...
Also, we can model 4 types of publication biases: 1) small study effect, 2) decline effect, 3) small study divergence and 4) Proteus effect.. Sometimes, not modeling the scale part can create the small study effect or decline effect when there are none (this is pretty cool -we think)
February 12, 2025 at 5:15 AM
Also, we can model 4 types of publication biases: 1) small study effect, 2) decline effect, 3) small study divergence and 4) Proteus effect.. Sometimes, not modeling the scale part can create the small study effect or decline effect when there are none (this is pretty cool -we think)
Although we cannot get I2 (heterogeneity) in both parts - we can get CV in both parts (and these CVs can be comparable - on the same scale)
February 12, 2025 at 5:15 AM
Although we cannot get I2 (heterogeneity) in both parts - we can get CV in both parts (and these CVs can be comparable - on the same scale)
We can also model between-study variances in both location (mean part) and scale (variance part) and look at their correlation like below - each mountain is a study - some study has more variation
February 12, 2025 at 5:15 AM
We can also model between-study variances in both location (mean part) and scale (variance part) and look at their correlation like below - each mountain is a study - some study has more variation
There are often variations in variability among effect sizes like this. We can model such differences in variances (heteroskedasticity) using location-scale meta-regression models
February 12, 2025 at 5:15 AM
There are often variations in variability among effect sizes like this. We can model such differences in variances (heteroskedasticity) using location-scale meta-regression models
We may always start #MetaAnalysis with a multilevel location-scale meta-analytic model in #Ecology and #Evolution. We can probably reanalyse many previously published meta-analytic data (getting new insights - like 2 findings with the price of 1)
ecoevorxiv.org/repository/v...
ecoevorxiv.org/repository/v...
February 12, 2025 at 5:15 AM
We may always start #MetaAnalysis with a multilevel location-scale meta-analytic model in #Ecology and #Evolution. We can probably reanalyse many previously published meta-analytic data (getting new insights - like 2 findings with the price of 1)
ecoevorxiv.org/repository/v...
ecoevorxiv.org/repository/v...
led by @mlagisz.bsky.social - an important message to all learnt society #EcoEvo
royalsocietypublishing.org/doi/full/10....
royalsocietypublishing.org/doi/full/10....
February 9, 2025 at 8:27 PM
led by @mlagisz.bsky.social - an important message to all learnt society #EcoEvo
royalsocietypublishing.org/doi/full/10....
royalsocietypublishing.org/doi/full/10....