I wonder if our IAT results will show that Ignatieff is as disliked (or, at least, if people’s implicit attitudes about him, in comparison with the Liberal Party itself) as this Slate piece suggests.  I really can’t wait until our panel survey is complete and we can get to analyzing some sweet, sweet implicit attitudes about Canadian parties and leaders.

I’m going to start showing this article to anyone who questions the usefulness of social science, it’s status as an actual science, or what use my research (especially the implicit stuff) might be.  For those who aren’t interested in clicking the link, some social scientists are arguing that they can get a handle on a player’s future success in the NFL by doing textual analysis on college press conferences, etc.  These responses can then be used to map intangible psychological characteristics. Basically, they argue that it can help measure things like “leadership”, “winning attitude,” etc.

Honestly, I think it makes a lot of sense.  I’ve done some research with implicit attitudes, and I like implicit research, textual analysis, etc. because it’s a quieter way of measuring attitudes and behavior.  People construct answers to explicit things like survey questions, etc.  Methods like this textual analysis (and the IAT, and the Misattribution Test, Go-No Go, etc.) are sneaky and don’t illicit the same sort of interviewer, framing, question wording, comparability, etc. problems of standard surveys.  And apparently NFL Combine tests.

That said, there are obviously issues.  John Kerry, to the best of my knowledge, is not a terrorist and yet their analysis identified him as one.  Well, that’s not really a problem, but just an aspect of dealing in probabilities not certainties, but yeah, kind of a big mistake there.  And, the 4-quadrant graph that accompanied the bit about Aaron Rogers (who is awesome, by the way) and Aleks Smith is a little messy.  The relationships aren’t immediately clear (plus, on a visualization note, it tries to show too much.  I’d have done 80+ and >80 on QB ratings).

However, I think these issues have more to do with needing to refine the measures and theorize about which psychological aspects are most important to future success than any underlying problem with the concept.  I mean, just from a pro-sports perspective, once you get to the level of the NFL (or the NBA, MLB, NHL, etc.) you are so far to the right of the bell-curve on pure physical ability that distinguishing, physically, between Vince Young and Aaron Rogers (or Drew Brees, etc.) is probably next to impossible and likely useless if trying to predict future NFL success.  Analysis like that described in the link though, that can help you (probabilistically mind you) get inside the players’ heads and help managers/owners make more informed decision.  It may still be something of a crap shoot, but at least you’ll be playing with weighted dice.

Just recently read a brief blog post on Freakonomics (see here) about the rather poor quality of expert judgments.  Tetlock, a social scientist I otherwise strongly dislike because of his, rather anti-science, stance against the use of implicit attitudes (see here for a good smackdown of his objections), has also published an extensive and incredibly ambitious project on the fallibility of expert political judgments, primarily from a foreign policy standpoint (see here).  Basically, Tetlock finds that no one predicted the end of the Cold War and pundits are only slightly better and often worse at prediction than dart throwing monkeys.  Dubner is basically arguing that coaches and GM’s aren’t too good at predicting pro-success.  I’m not sure though how bad we (as social scientists, etc) really are at prediction or whether we allow non-relevant considerations (ideology, jersey sales, Tim Tebow) to cloud that judgment.  I mean, we’re pretty good at predicting vote choice, turnout, and a lot of other social phenomenon.  Then again, most social scientists aren’t pundits or consultants (BDM and the Freakonomics guys notwithstanding).  Also, it makes you wonder about the “wisdom of markets” and the “wisdom of crowds” if experts are so bad at prediction.  Provides a whole new spin on the economic crisis if even hedge fund managers, presumably experts, could be just as bad as chance at predicting what will go up and down.  Anyway, just an interesting area of research and an interesting aspect of social science to think about, especially considering how important we seem to think political knowledge is (Zaller 1992, 1996, Converse 1964, etc).

My and Aleks’ Canadian Election survey, complete with super cool implicit association tests (IAT) about the political parties, is up and running on Amazon’s M-Turk system.  It’s been up for about 12 hours now and, so far, we have exactly 1 respondent.  This is less than ideal, but we did post it fairly late last night, so hopefully we’ll get some more respondents today.  It took the one guy who’s completed the task only about 18 minutes to do it, which is less than we thought, so that’s a plus.  At least we’re learning about using Inquisit online and how to utilize the M-Turk system on Amazon, even if our response rate is somewhat limited so far.  The survey will stay in the field for the next week, so hopefully we’ll get a decent enough sample over that time frame.  The sample is restricted to Canada after all, maybe they don’t have wi-fi access up there in the great white north.  Anyway, so far, so good, at least functionally. I’m going to try and restrain myself from check the Amazon website every 30 minutes today to see who has completed the survey.

You might be a grad student if…

you have a catastrophic plumbing malfunction and you spend the entire time that the (very nice and competent) guy is working thinking about the strange principal-agent relationship involving you, your landlord, and the plumber.  Does this situation resemble a multiple principal relationship, a la Moe 1985 (& Moe 1990, and Lyne & Tierney 2003)?  My landlord is Congress, I’m the President, and the plumper in the bureaucracy?  Or maybe the landlord is President and I’m Congress?  I do have a singular overarching goal of being able to flush the toilet, similar to Congressional concern for reelection.  Might my landlord and I’s goals be competitive rather than cooperative, creating problems of control for us and discretionary opportunity for the plumber? I mean, I’m the only one who can directly evaluate the performance of the agent by having a functioning toilet and shower.  However, my lease is up in June and I’ll be leaving for Connecticut, so I have no incentive to monitor (or even ask for) a stable, long-term plumbing policy solution, while my landlord is likely interested in a cheap and effective long-term solution, doubly so since she’ll be moving-in in June.  Also, I don’t have to pay the guy (even to get reimbursed) so as long as the toilet flushes when he leaves (and it did, guy seemed good at his job), I have no interest in confirming if the job was hard/easy/accomplished/cheap/expensive/etc.  I guess that really does make me Congress, since as long as my toilet works (reelected), I’m less concerned with actual outcomes.  Which makes our cat chair-kitten of the Subcommittee on Trying to Escape the Whole Damn Time the Guy was Here.

See here.  Nice guest post from David (another Rice grad student) on arguably the most prominent blog in the discipline. Personally, I would have name dropped someone else in the first paragraph when discussing anarchy (never really liked the style over substance approach of the Sex Pistols, I was/am a Clash & the Damned kinda guy), but I guess you have to write for your audience.  As a general rule, political scientists have never struck me as being particularly conversant in 1st wave English punk bands.  Other than that, the points about studying caretaker governments are interesting and quite good.

Thanks to fellow Rice political science grad student Daniel Zaccariello for the link.  While facetious, the post “Stuff Political Scientists Like” has inspired me to go ahead and gather some original data for my dissertation.  Bring on the interviews and copious citations!  Maybe I’ll do some of that while summering in Connecticut this year.

All three appear in this report on some of Dietland Stolle’s work first reported on the Monkey Cage.  Apparently Aleks and I aren’t the only ones interested in the upcoming Canadian election (going on May 2nd, 2011), but I wish I’d thought of polling daycare kids.  That’s an easy write-up in USA Today or NY Times.

But a good quote from John Sides over at The Monkey Cage, “The problem isn’t that Americans can’t reason their way to a balanced budget. The problem is that policymakers don’t like the way many Americans would do it.”  See full post here.

First, let me make this abundantly clear.  I’m actually surprised and pleased that some journalist somewhere was covering the MPSA (see here). Also, I picked up the link from The Monkey Cage; you should read that blog.  3rd, I’m upset that the journalist didn’t cover either of my panels (which, granted, obviously could have used snappier titles), but that’s not the reason I’m going to critique this guy.

I’m going to critique this guy because none of the panels he mentions seemingly have anything to do with mainstream political science.  Two caveats: first, voter turnout (the fried chicken story).  We know that economic incentives boost turnout; we know that mobilization boosts turnout.  That Prof from Utah is getting that article buried in a third tier journal (from what I understand having read the Atlantic article, I didn’t see/read that panel).  Secondly, sure Theda Skocpol was the chair of the Obama panel, but seriously.  I’m pretty sure that we, as a profession, put in these types of panels so that someone other than us will be interested. It’s not really the sort of theory-driven sort of hypothesis testing that drives most political science research.  But, maybe journalists will show up if we put “In the Age of Obama…” in front of a few panel titles.

Again, I’m mostly reacting to the coverage in the Atlantic and didn’t actually see any of the panels mentioned.  But that, I think, is mostly due to the fact that I was attending panels on topics of interest to me (state-level political parties, bio-politics, political psychology, implicit attitudes, etc.).  These tend to be topics of little interest to anyone else, and I think I can make a strong inference on that point given my experience at multiple dinner parties over the last 3 years.  The sample isn’t random, but it is extensive.  On a side note, how did Warren (the journalist) miss the bio-politics stuff?  Some of it was from one of the Rice profs about ideology and smell (see here).  What better way to make the profession look like a bunch of Ivory Tower, out-of-touch crazies?

Anyway, my experience at MPSA 2011 (my first time attending) was that the average panel was outstanding.  I presented two pieces (both coauthored, see here & here if you are interested & much love to my coauthors).  The discussants were excellent on both panels, the panels were well-organized in that all the papers actually fit together, the attendance was good and the overall discussions were extremely helpful.  But I’m sure if Warren (or honestly any covering journalist) had shown up to the Parties and Partisanship panel, he/she would have noted how a few jokes were cracked at the expense of the South and the conservative nature of Southern politics (not to mention their horrible record keeping, an issue of far more importance to most scholars of American politics than the well-documented conservatism).

All in all, the profession gets a shout-out in a major outlet, and its sort of just cherry-picks things that make us look like out-of-touch, uber-liberal weirdos.  Which maybe we are, but it’s not really the foundation of our work.  Given that the profession is a good scapegoat for budget problems lately, we might want our public image to look a little better.