Lightnews — Scholar-powered news

I think the bottom line is I need one app for things other people want me to do (calendar/email) and one app for things I want to do (reminders/todo).

Roger D. Peng @rdpeng.org · Nov 7

I've found putting things that I *want* to do in my calendar has an opposite effect and I end up not doing it. So for those kinds of things I use Todo/Reminders apps that are pinned to specific dates/time so they pop up like a calendar (but not on my calendar). 🤷🏻‍♂️

1 1

Roger D. Peng @rdpeng.org · Nov 7

I wouldn't include it in the "Talks" section but might maybe in the "Teaching" section if it could reasonably be considered a class of some sort (or part of one?)

1 2

Roger D. Peng @rdpeng.org · Oct 20

The best part of this job is working with great people!

Lucy D’Agostino McGowan @lucystats.bsky.social · Oct 20

What a delightful week working with @rdpeng.org & @stephaniehicks.bsky.social on Analytic Design Theory — stay tuned, exciting things happening here!

2 8

Roger D. Peng @rdpeng.org · Sep 25

I like these too.

Roger D. Peng @rdpeng.org · Sep 22

It has a punch card as a bookmark!

Roger D. Peng @rdpeng.org · Sep 22

Got to see @rafalab.bsky.social this week while visiting Boston and he gave me this amazing present!

2 5

Reposted by Roger D. Peng

Karl Rohe @karlrohe.bsky.social · Sep 19

What are the biggest **empirical** successes of “multivariate statistics”?

Should be fields/industries that a built around them. Eg:

pageRank in web search
factor modeling in finance
phylogenetic reconstruction in evolution

These are currently my best examples. What else?

1 2 1

Roger D. Peng @rdpeng.org · Sep 11

I'm experimenting with data from the World Cube Association (www.worldcubeassociation.org/export/results). Not sure about causal inference but maybe prediction. Data is very clean and requires merging across tables.