In response to UCU's Marking and Assessment Boycott (MAB), my employer @RoyalHolloway has decided to implement emergency regulations which in my view seriously call into question the rigour of degrees awarded (1/7) The regulations include allowing marks for a module to be scaled proportionately (you did 50% of the coursework; that counts for 100%) (2/7)

Everyone knows the most fun way to watch the World Cup is to support the more democratic nation in each game. So here, thanks to @vdeminstitute data, is your group-by-group rundown! (1/n) We start in Group A, where the Netherlands is clearly in pole position, and Qatar clearly in last place (2/n)

This use of MRP makes me incredibly anxious. There are considerable technical difficulties here. Begin nerdgasm... (1/15)

https://twitter.com/jamesjohnson252/status/1548291779522244608

MRP works by modelling responses as a function of different demographic and political characteristics, and then making predictions for different voter types (2/15)

Towns are sexy in UK polisci, aren't they? Let's talk a little bit about the super-sexy Towns Fund gov.uk/government/pub… (1/n) As @estwebber and @georgegrylls reported this morning, the Perm. Sec at MHCLG gave evidence to the Public Accounts Committee this week thetimes.co.uk/article/robert… (2/n)

Strong, strong political reasons to support Osaka in this match #USOpen Azarenka and Lukashenko have a "friendly" relationship nytimes.com/2020/09/09/spo…

I have some reservations about the claim that Covid-19 divisions now run deeper than Brexit (1/n)

https://twitter.com/pollymackenzie/status/1304301641810706433

The basis for the claim is that the proportion of mask-wearers who hate, resent or think badly of non-mask wearers (58%) is greater than the proportion of Remainers who think badly of Leavers (33%) (p. 11 of report) (2/n)

Tonight I'll be presenting my #APSA2020 paper, "The voting power of demographic groups". You can find the paper at drive.google.com/file/d/1G-SHiy… and a video presentation at (1/n) The idea behind the paper is simple: the power of a voter group is the number of seats where the result would have been different had that voter group not voted—or alternately, the number of seats in which that voter group was pivotal (2/n)

Seven levels of recursion on politics Twitter (0/7) Step 1: Oh, the government has said it will break the law. That's bad, I should tweet about that.

https://twitter.com/BBCPolitics/status/1303311625643008000

(1/7)

I'd be very interested to know if this implies that non-democratic regimes are better at tackling climate change

https://twitter.com/UKandEU/status/1303294073319624704

Difficult to think how you would test the claim: democracies are generally richer, and richer countries have different climate preferences and capabilities. Rich non-democracies also tend to be natural resource rich, which further shifts preferences

Does anyone have experience of combining quantities of interest after multiple imputation *that aren't coefficients*? I ask because I'm combining average marginal fx which are very not normally distributed, & I find transforming them before applying Rubin's rules gives "more sensible" results than leaving them untransformed. But it's all got a bit messy and Taylor expansions frighten me, so...

Anyone I know who's used a multinomial or conditional logit model on British electoral data and lived to tell a tale of Hausman McFadden tests? Update: a mixed logit model with random alternative-specific intercepts seems to "work" (give better fit to the data), seems functionally equivalent to MNP, and fits in a little under 6 hours. Now to repeat for 4 other multiply-imputed data-sets and four model specifications...

Okay, time to dig into the underlying research (1/n)

https://twitter.com/thhamilton/status/1273164297116569603

The study itself can be found at doi.org/10.1080/088240… ; the data is at osf.io/n2m85/?view_on… (2/n)

Has anyone who follows me used 2015 candidate spending data either from @ElectoralCommUK or @BESResearch ? There's an odd entry for Plymouth Sutton & Devonport (1/n) As you may(!) recall, Plymouth S&D was one of the constituencies where the Conservatives exceeded campaign spending limits by rebadging local spending as national spend web.archive.org/web/2017042306… (2/n)

I just saw a breakdown of a university estate by room size. A lot of the rooms had a normal operating capacity of 15 to 20 (1/n) That's close to optimal for UK universities, where most teaching is done is seminar groups of 12-15. Obviously different estates will have different modal sizes, and historic estates might conform less to this pattern (2/n)

One way of investigating racist behaviour is to conduct audit studies, where you present people with the same information (job application, letter requesting help) but change the name of the sender to strongly signal a particular race en.wikipedia.org/wiki/Audit_stu… (1/n) Last year Sarah Birch and @philiphabel published an audit study on UK MPs'* responses to emails asking for help doi.org/10.1111/lsq.12… (2/n)

* = well, MPs or their staff

Unpopular opinion: BBC journalists shouldn't editorialize & the Editorial Guidelines are good since they construct an elaborate system for managing conflicts of which the BBC (rather than politicians) is the ultimate arbiter For the longer version of this argument, buy my academic hardback at only GBP88: amazon.co.uk/Broadcasting-P…

Did Australian legislators who were out of step w/ their constituents on the issue of same-sex marriage suffer electorally in the 2019 election? A new WP by me and @jillesheppard says "no, not to any substantively meaningful degree" drive.google.com/file/d/1X6m0Us… #teamprecisenulls Timeline:
2017: Postal Survey: en.wikipedia.org/wiki/Australia…
2018: votes on same-sex marriage
2019: federal election

Found an odd pattern today whilst simulating party membership - yr thoughts welcome (1/n) Here are two plausible party-joining functions, illustrated for a party on the left.
LHS/"proximal": join the closer you are to the party.
RHS/"distal": join the more you are to the left/that party's "side" (2/n)

Niche questions about UK parties' policy positions in 2017 follow: a thread (1/n) Does anyone know what the Conservative and Liberal Democrat position on hate speech was in 2017, ideally with contemporaneous quotes from spokespeople? (2/n)

Today my book, A Court of Specialists: Judicial Behavior on the UK Supreme Court, comes out in the UK. You can buy/request for your library via @OUPLaw at global.oup.com/academic/produ… (1/n) (I say this: I haven't yet seen a physical copy of the book, and I'm slightly worried that thanks to The Present Situation it's trapped in an office I have no access to) (2/n)

I had my first Zoom academic conference experience y'day. The conference was great, but Zoom the app is shady as f**k I was never asked for my superuser password on my Ubuntu install, and yet somehow the app is in /usr/bin - how?