I groaned. but I approve.

instant publish

it's currently 1 hat every few months. clear trajectory towards a few hats every month.

terrified of the day it becomes a multi hat day.

I bought my first hat---in about 10 years---approximately two years ago. It's been a quadratic function since.

I will not give up my em-dashes but I've pivoted to explicitly using 3 dashes instead of the proper em-dash

very jealous.

Boston Public Market is a good food hall near Fanieul. Pretty good overall and variety. But may be a bit far from seaport/convention center

Taiwan Cafe and most things in Chinatown. Shy Bird. Bardo's Bar Pie is about as far north (sort of...) as you'll find south shore bar pizza (a regional specialty). Most things in Seaport are good but can be $$$$ (Row 34, Trillium, etc...)

a lazy hack I've used when I really want access to something hiding or not exported is :::

reprex.lazyload:::myfunc()

but admittedly I don't think this is the norm and probably only for emergency use (or laziness)

I'm at the airport and the airline forced me to do that at the kiosk -- which makes sense

It printed a form I had to fill out by hand with emergency contacts

No one wants to take it. I keep trying to give it to airline staff and they say "You dont need that. We just have to ask"

Cool.

We should chat (you know how to find me)! We've been working on this a bit

yes actually! I don't know if an abstract/write up is available but if of interest, please let me know and I'll dig this up

scholar.google.ca/citations?vi...

if Doug Flutie and Arnie *weren't actually there*, then I choose my false memory that they were.

The seats were incredible regardless. Basically high five distance from from Mick Jagger as he danced around

agree I'd prefer CIs. but I lean more comprehensive reporting of everything because "p < .05" could be .0499 or .0101

put it all out there. it's all useful!

software as a replacement of look up tables: totally agree.

but the exact p-values themselves are not always reported. so we need those other bits of information to help (and verify)

I'd rather see it all reported/displayed (especially N or df) than reporting of just one thing or another.

it took me a minute but I remembered the suite of R packages developed across companies to replace other tools with R:

pharmaverse.org

After IBM acquired SPSS, they also started offering direct integration with (or even pass through to?) R

A big, perhaps qualitative, sign of movement is in bio/pharma industry and how they're now making/publishing R packages to support this work (and move away from SAS and others)