New in
@PNASNews
with
@ylelkes
and
@eunjikim_media
:
Measuring dynamic media bias
Examining cable news from 2010-2021, we find bias varies over time within and across:
- channels
- program times (afternoon/primetime)
- shows
Some personal news!
My wife and I will be moving to Salt Lake City this summer where I'll be joining the Department of Political Science at the University of Utah as an Assistant Professor!
I'm so lucky to be joining such a wonderful group of scholars in such a great city.
Excited to share my new project:
Congressional Data in R
This is a comprehensive guide to congressional data, how to work with it, clean it, merge it, visualize it, and model it
A lot of hard-earned experience went into writing this.
this
@Andrew___Baker
et al paper on staggered treatment is so good and readable for folks familiar with, but not steeped in, the econometrics literature on this
@kyleichan
Fwiw, which I mentioned in the interview, it didn't work great! Which is encouraging. I was using this to determine a base rate of the laziest possible uses of chat gpt
Some personal news now that things are official:
I'll be a post doc next year at Michigan State, working with
@MattGrossmann
in the
@IPPSR
!
I'm incredibly excited to work with Matt and continue doing research while getting more involved in the state politics community
New, from
@VincentAB
and myself:
Software Citations in Political Science
We document the under citation of open source software, propose citing it in supplemental info as a partial solution, and create an R package to do it!
for folks learning programming for the first time -- I frequently have to go back to tutorials *I wrote* to look something up. that's how this works! constantly having to look stuff up
CDC DATA WOES: Some
@CDCgov
staff were mortified when a Seattle teenager managed to compile coronavirus data faster than the agency itself. “If a high schooler can do it, someone at
@CDCgov
should be able to do it.”
Now online!
Software Citations in Political Science
@VincentAB
and I document the under citation of software in political science, suggesting it disincentivizes the creation and maintenance of public goods.
@dylanmatt
It's like reading a book you don't love but you get attached to the characters and then the author gets better at writing in the second and third book
important advice for new poli sci grad students:
take the past 10 years of APSR (at least), ideally in their print version. physically printed is ok too.
shred them finely, dump them in a tub, then soak in the tub for >= 1 hr to absorb the literature
Post/ABC poll on Trump’s approval among Republicans:
July 87%
now 74%
18% of Republicans say Congrese should impeach Trump and remove him from the presidency
Excited to share my new project:
Congressional Data in R
This is a comprehensive guide to congressional data, how to work with it, clean it, merge it, visualize it, and model it
A lot of hard-earned experience went into writing this.
Maybe being obese doesn't actually raise your risk of death
"Body mass index may not increase mortality independently of other risk factors in adults, according to a new study published this week" in
@PLOSONE
.
folks: what's your favorite book for an intro methods sequence in the social science?
what I've currently got:
Mostly Harmless
Mixtape
Quantitative Social Science (Imai)
Regression and Other Stories
New paper with
@MattGrossmann
and
@MartyPJordan
:
The Correlates of State Policy and the Structure of State Panel Data ()
We introduce the Correlates of State Policy Project data, describe its structure and advantages, and use it in an applied setting 1/n
Day 2 of new
#rstats
tutorials: making gorgeous city maps in R with ggplot and the osmdata package.
I also show how to:
* highlight individual streets by name
* overlay specific locations from Google Maps (like breweries!)
* use custom fonts
I hereby claim, for purposes of securing a tenure track job, that I now have three new papers, all forthcoming in the American Political Science Review
Some exciting news!
The
@JawsPoliSci
team has been awarded an
@APSAtweets
grant to host an in-person pre-conference ahead of the 2023
@MPSAnet
annual meeting!
Join our email list () and follow our twitter account for more info and the call for papers!
i probably would have saved ~50 hours of instruction/office hours time last year if I used RStudio cloud instead of having each student install on their own machine
We're really excited about this package. Our aim was to make it easy to use a massive, comprehensive dataset on state politics and policy. Very little prior R knowledge is needed to get this up and running.
See Matt's thread for some examples!
New R package & web app:
@IPPSR
beta release of the Correlates of State Policy R package by
@joshmccrain
&
@calebjlucas
2,000+ state-year variables:
- quick search by keyword/text
- load & merge in panel & network data
- output citations
- make maps
If you don't pay legislators you get already-wealthy legislators. If you don't give legislators staff and resources, you get more influence of lobbyists in the legislative process
Now online!
Lobbyists into Government
w/
@BCEgerod
at
@qjps_editors
We find big effects of lobbyists becoming govt officials on lobbying firm revenue ($320k per year)
ungated:
We especially need to talk about salary in academia. I make more money now as a brand new UXR compared to most full professors. This isn't right.
Academia severely underpays and overworks its PhD workforce. No wonder so many tenured faculty are leaving for industry. 7/
New
#rstats
tutorial:
purrr: Introduction and Application
* learn basic functions such as `map`, `pluck` and `nest`
* use these in real applications
* see how to create multiple models and visualizations with purrr
in my department we have very little money for speakers, however if you fly out you can stay with me and i can probably set you up with a talk and then we can go skiing so think about that
Please ask for feedback and frequently present your work at lower stakes seminars and lab meetings. I know it probably feels awkward and difficult but please do your best.
New
#rstats
tutorial! A crash course in maps and shapefiles with ggplot.
I show how to:
* load in non-standard geographies (DMAs, in this case)
* join your own data in and use to make a Chloropleth
* work with states and congressional districts
a little peek into some of the cool data i'm working with:
where do House staff go when their boss loses an election?
they mostly leave the Hill, and very very few end up as lobbyists
@SeanTrende
a low-effort behavior that we all know very well at this point is extremely beneficial and saves lives doesn't seem to be something we should be proud of avoiding
New paper with
@BCEgerod
: Lobbyists into Government
We find:
* when lobbyists become congressional staff or bureaucrats, their previous firms see a 36% ($320k) increase in yearly revenue
* larger increases for congressional staff over bureaucrats
i wish i could tell my high school AP calc teacher (i did terribly in that class) that I am now writing calculus and algebra problem sets for PhD students
While looking at Gay's important 2001 paper, I became skeptical of the whole research approach.
I am not a political scientist, and it is possible that I am mistaken. If I am not mistaken, it is such an obviously flawed method, that I don't understand why the paper could pass
Since we're talking about local news...
Happy to share this newly forthcoming paper in
@PSRMJournal
with Erik Peterson:
"Local Elections Do Not Increase Local News Demand"
pre-print:
People who drank moderate amounts of coffee, 1.5 to 3.5 cups per day, were up to 30% less likely to die during a multiple year study period than those who didn’t drink coffee, new research found.
somebody should really write a how-to guide on MRP with both raw survey data and the post-stratification demographic data, it's really not obvious how to do this!