Manasvini Singh

Please apply! This year we're especially interested in early-stage projects so slide decks are welcome :)

Ugh meta I don't want my glasses to pull up people's names, I want my glasses to pull up the relevant journal article any time someone says "I saw a study that ...."

I started yesterday. kicked things off with a good sob when Sophie evaporated into rose petals

There no logic, people like Helen just have careers backfilling justifications for xenophobia. Immigrant groups that underperform US-born are lazy morons who don’t deserve to be here, those who over perform are ruthless strivers destroying valuable White Ways of Knowing & Doing. There’s no winning

Please apply! This year we're especially interested in early-stage projects so slide decks are welcome :)

From my grandmother's story: "One by one our good decent teachers disappeared and instead Nazi teachers replaced them."

Can guest lecture on many health policy topics and/or causal inference/ applied metrics. Would love to help.

Added a new -- and permanent, I fear -- line to my paper acknowledgements section.

Going to add my own two cents to the academic convo on AI... I do think there is an AI bubble + it will burst. That doesn't mean AI will go away, but it does mean it will be less ubiquitous than many of the tech leaders say it will. Also, i think a lot of us are frustrated. As an academic... 1/7

"Because our students enter into Income Share Agreements (ISAs), in which successful graduates "pay it forward" to the next generation, the impact is in potentially unbounded."

Amazing work. (Also a nice bonus that @johanneshaushofer.com is a shining light in the academy)

an econ JMP thread in 2025? Ah, a sight for sore eyes

An incredible read.

There are more of us than there are of them.

Food $500
Rent $1500
Utilities $300
Gas $150
Bribes and fascism stuff $20 trillion

Someone who is good at the economy please help me budget this, my society is dying

My vote in the 2028 D primary for POTUS will hinge almost entirely on the % of this Administration they promise to imprison

🤯🤯🤯 Now this is what I'm talking about!

nooooooo it was one coding session, it didn't mean anything

yes absolutely ... the dangers of errant merges are dire 😭 and it feels even more dangerous when i'm just learning R and there is no simple way to inspect my merges. at least tidylog may provide a good first step!

omg perfect!!! thank you! tidylog seems to be exactly what I was looking for (I wanted dplyr's many-to-many error but for other kids of errors as well, e.g., if i tried doing m:1 but there were multiple records in the second data)

yep, basically wondering whether i have to write this every time i use join()

merge_flag = case_when(
!is.na(x) & !is.na(y) ~ "matched",
!is.na(x) & is.na(y) ~ "X_only",
is.na(x) & !is.na(y) ~ "Y_only",
TRUE ~ "error"
)

yup yup i just meant to say that diagnostics are trickier. I've tried to write my own function, but at least at this stage of learning it's not pretty easy for me :) if i use my own (clunky) function, it seems I lose the rich functionality of R's native join? Maybe soon though! 🤞

This is the only package I found aimed at that issue. But the "relationship = ..." doesn't show warnings when it should

randrescastaneda.github.io/joyn/

I finally made the switch to R from stata, and as a fierce long-time defender of stata, I can concede that R is leagues better in almost every way.

Still, I do wish there was an easier way to diagnose merges w/o complicated code. GPT hasn't helped, and joyn doesn't seem to do it right. Any advice?

dammit i did not see the first comment

Crying (complimentary)

JDM folk: please submit :)

A woman protesting outside the ICE facility in Broadview, Illinois, was brutally assaulted by goons who used mace and sting balls on her.

so many dead canaries and we just keep on mining

New The Effect materials today: introductions to basic coding and data manipulation in R, Stata, and Python. Get the wheels turning on using these languages with data with these intro pages and exercises:
nickchk.com/Coding%20and...
nickchk.com/Coding%20and...
nickchk.com/Coding%20and...