Jan Vanhove :: Blog - Collinearity isn’t a disease that needs curing
#statstab #37 Collinearity isn’t a disease that needs curing
Thoughts: This shifted my thinking from "assumptions are nuisances to check" to understanding what they actually do and mean.
#MachineLearning #regression #stats #colinearity
janhove.github.io/posts/2019-0...
27.02.2024 16:38 — 👍 15 🔁 5 💬 1 📌 1
Why then are correlation coefficients so attractive?
Only bad reasons seem to come to mind.
Worst of all, probably, is the absence of any need
to think about units for either variable. Given two
perfectly meaningless variables, one is reminded
of their meaninglessness when a regression coefficient
is given, since one wonders how to interpret
its value. A correlation coefficient is less likely to
bring up the unpleasant truth—we think we know
what r = -.7 means. Do we? How often?
Sweeping things under the rug is the enemy of
good data analysis. Often, using the correlation
coefficient is "sweeping under the rug" with a
vengeance. Being so disinterested in our variables
that we do not care about their units can hardly
be desirable.
How have I never read this scathing paragraph about correlation coefficients from Tukey (1969) before? 🧪 #statistics
02.03.2024 16:37 — 👍 75 🔁 25 💬 0 📌 0
ggplot2 3.5.0: Legends - Tidyverse
The 3.5.0 version of ggplot2 comes with an overhaul of the guide system. Read here what is new for legends.
One of the big improvements in the new version of #ggplot2 are related to legends.
Teun van den Brand lays out all the detail in this blog post. Get ready for the legend extension extravaganza #stats
26.02.2024 08:23 — 👍 30 🔁 13 💬 0 📌 3
Secrets in the canopy: scientists discover 8 striking new bee species in the Pacific
By lifting their gaze to the treetops rather than poking around on the ground, researchers discovered eight new species of masked bees.
New species out! Exciting to participate in this paper describing new yellow-faced bees from the South Pacific. Like many of the Hawaiian Hylaeus, they tend to stay up in the forest canopy, which kept them hidden for a long time! Makes you wonder how many more are out there…
🧪
26.02.2024 06:40 — 👍 39 🔁 13 💬 2 📌 0
Estimating Bee Communities is a Tall Task, Analysis Shows
A new meta-analysis of bee-sampling studies reveals best practices—and the breadth and scope necessary—for thoroughly documenting a local bee community.
Studying bee communities is critical to assessing their role in ecosystems and agricultural systems. A new meta-analysis of 72 bee-sampling studies reveals best practices—and the surprising breadth and scope necessary—for thoroughly documenting a local bee community.
14.02.2024 14:13 — 👍 14 🔁 8 💬 1 📌 0
Great resource for researchers who want to use #RStats for robust causal analysis. Helped me understand the logic behind DAGs (directed acyclic graphs): www.r-causal.org Thanks @malcolmbarrett.malco.io @lucystats.bsky.social & @travisgerke.bsky.social 🙏
12.02.2024 21:17 — 👍 19 🔁 4 💬 0 📌 0
#statstab #22 Plot model coefficients with {ggcoef_model()}
Thoughts: Very useful for quick plotting while drafting a write-up. Helper packages are good, but we can easily fall back into the cookbook method of stats.
#stats #dataviz #plots #rstats
larmarange.github.io/ggstats/arti...
06.02.2024 17:50 — 👍 4 🔁 3 💬 0 📌 0
