Thursday, 19 March 2009

How not to do data analysis

io9 presents a chart that purports to show that shark-jumping has an effect on television ratings. I’ll freely concede that Battlestar Galactica has had its, er, weaker moments, but the chart doesn’t actually show that creatively weak episodes had any effect whatsoever on the ratings that can be distinguished from the underlying, secular downward trend in ratings.

Since I had about 300 more important things to do, I decided to analyze the data myself. First, I reentered the ratings data from here into an spreadsheet and then identified the “shark-jump” episodes with a dummy variable, with the help of IMDB. I then created two new variables: a simple ratings difference variable for each episode, and a dummy variable to indicate whether or not an episode immediately followed an identified shark-jump.

I then converted to a CSV file, opened R, and estimated a linear regression: Delta = a + b(FollowShark). While the effect of an episode following a shark-jump was negative (about 0.025 ratings points), the effect was not statistically significant (p ≈ 0.736, two-tailed). Throwing out “Razor” and “The Passage,” to focus on episodes io9 says showed ratings losses improves the coefficient to about -0.042 ratings points, but it is still not significant (p ≈ 0.613, two-tailed).

So, the moral of the story: the episodes identified may have been “shark jumps,” but they didn’t seem to have a discernible effect on the ratings of subsequent episodes. And, besides, any analysis that doesn’t identify the crapfest known as “Black Market” as a shark-jumping incident isn’t worth taking seriously to begin with.

Bear in mind that TV ratings themselves leave something to be desired; variations of several tenths of a ratings point are within the expected margin of measurement error.

Wednesday, 18 March 2009

Memo to colleagues: I will be out sick for a week in early 2010

Mass Effect 2 has officially been announced for the PC and XBox 360. Hopefully this time Bioware will be spared the fake controversy that arose last time with the original Mass Effect, but then again maybe the fake controversy will help sell games.

Saturday, 14 March 2009

I am now Scalzi-complete (I think)

I finally got around to ordering John Scalzi’s Agent to the Stars and read it last night at the hotel in Dallas. Agent is an eminently enjoyable read, and it does for Hollywood what The Android’s Dream did for government bureaucracy (which is to say, skewers it mercilessly). Definitely strongly recommended.

Thursday, 12 March 2009

Must… resist… snark

von at Obsidian Wings writes:

I’m fairly confident that the US would respond with overwhelming military force if the de facto government of Mexico was randomly firing rockets at Laredo and McAllen, TX.

People in Laredo have problems dealing with such basic things as rain and temperatures below 80 degrees; I can’t imagine how they’d react to mortars and rockets.

Wednesday, 11 March 2009

Quickie Poll Fail

I would like to thank CNN for giving me the ultimate example of the Worst Web Poll Ever for my semi-annual classroom rant about self-selected Internet surveys:

Vote on this!

Huzzah and kudos to Burt Monroe (via Facebook) for the link. If the EPOVB section of APSA ever makes a T-shirt, this should be the picture on it.

Tuesday, 10 March 2009

Rushing to Judgment

Cassandra at Villainous Company points at some real data that indicate that a large chunk of Americans really, really dislike Rush Limbaugh. While perhaps this is the result of false consciousness or just uninformed judgment, my gut feeling is that it stems from Rush being a self-important, loud-mouthed blowhard who gets far too much credit among the party his continued existence on the face of this earth (or at least in the media) is increasingly a liability for due to his marginal role in rallying Republican support in the 1994 midterms that led to the GOP takeover of Congress.

Fault in ourselves and not the stars?

Prompted by the Wisconsin Lutheran story, PTJ at The Duck of Minerva objects to the current state of scholarship in political science:

Once people are hired, they also have to figure out what to assign to their students; for that purpose, they need books and articles. Naturally, people want to assign the current, contemporary research in their field if they can, but not only does that not say much about civic engagement or the future of the political landscape, but it doesn’t even say what it does say in a way that is particularly accessible to undergraduate students. “The Role of Parties’ Past Behavior in Coalition Formation,” to pick just one of the articles from the most recent issue of the American Political Science Review, doesn’t exactly sound like a page-turner. And yes, I know full well that other disciplines also have a dichotomy between their contemporary research and the kinds of things that one assigns to undergraduates, but the gulf is particularly pronounced in contemporary Political Science. (At least Anthropology and Sociology have classics that can be profitably read by undergraduates; once one gets outside of the social sciences, the humanities have works of art and literature, and the natural sciences have textbooks and laboratories.) I remember serving as a TA for an American politics class while in grad school; the professor told a lot of stories about how actual politics worked, but the reading material talked about such scintillating topics as fire-alarms versus trip-wires in governmental oversight regimes. So the students, not surprisingly, ignored the reading and listened to the stories.

I think the students were on the right track. If one wants to actually do much serious thinking about civic life and one’s individual responsibility within it, one would be well-advised to stay as far away from the last several decades of Political Science scholarship as possible. Undergraduate education in politics shouldn’t be about learning how to solve extensive-form games; it should be about learning how government works. But contemporary Political Science isn’t much help to that task. This implies that if we want students to come to articulate their own sense of civic engagement, we ought not send them to the Political Science department, but could achieve the same effect by sending them elsewhere. And to make matters worse, people trained in Political Science probably aren’t likely to know how to facilitate this for undergraduates, which further undermines the need for a Political Science department in a liberal arts college.

Now, I’m not saying that every liberal arts college ought to go around eliminating its Political Science department. (In fact, Political Science departments at most liberal arts colleges I know are actually quite far removed from the mainstream of the discipline; I don’t think this is an accident.) But I am saying that the decision makes a certain amount of sense, since the discipline of Political Science is so far away from the goals of a liberal arts undergraduate education. And that’s too bad—bad for Political Science, not bad for the liberal arts.

Laura at 11D voices her concurrence to this line of thinking. And while it is tempting to agree, I think this is a symptom of a larger problem with political science as a discipline. Nobody, for example, would suggest that a physicist studying muon decay in a particle collider should somehow make his or her research directly relevant to people building nuclear reactors, yet somehow we (or some of we) expect cutting edge political science research to be directly relevant to the average citizen, and if it isn’t then it’s the fault of the research agenda.

I think this results from fundamental confusion about the endeavor of political science: political science is not civics, just as psychology is not Dr. Phil. Yet in “Intro to Psychology” they don’t teach Dr. Phil—they explain the basic findings of psychological research. So why do we teach civics in our introductory courses rather than teaching political science? (And the answer isn’t just “the Texas legislature says we have to,” because in point of fact we could make up 6 hours of anything related to U.S. and Texas government to fulfill the Texas gen ed requirement, yet we all end up doing the same old civics-based crap for the most part.)

To the specific points made by PTJ: yes, there are accessible, “classic” works in political science that can be fruitfully used with undergraduates. I’ve used works such as Key’s Southern Politics, portions of The American Voter, and excerpts from other scholarship from the 1940s-1960s successfully in a number of undergraduate courses. Yes, undergraduates do like “stories” instead of theory… but they prefer that in virtually any discipline. Math students would much rather listen to stories about Newton fleeing Cambridge during a plague rather than learning the calculus he co-derived,* yet I don’t think any serious person would think that the students are right in that circumstance.

Moreover, animating the theory of fire alarms versus police patrols in congressional oversight is part of the responsibility of a professor, and could be easily be tied to real contemporary political concerns (think of the treatment of detainees in the War on Terror, where Congress reacted to dramatic media coverage much more readily than to, say, John Yoo’s memoranda on the topic), just as explicating the whole business going on with the cave is the job of a professor who’s trying to explain Plato. And I dare say the APSR‘s normative theory articles are no more penetrable to the average undergrad than the game-theoretic stuff, so this isn’t an issue of the “teched up” research agenda oft-lamented in the discipline. And, no, I don’t think undergraduates need to “solve extensive-form games” but I think it’s reasonable to expose them (at least at a very simple level) to theories like the median voter model or spatial models of realignment that do explain how “actual politics” works. And, no thanks to the rewards structure of our discipline, there are quite a few good books out there for students that do these things for students at the junior and senior levels.

On the broader question of whether political scientists should be inculcating a sense of civic engagement in students rather than publishing research, I am probably one of those odd ducks who doesn’t think it’s our job to push students to engage in political activity without some evidence that the student is actually interested in doing so. (I think it’s a perfectly legitimate position for a citizen in a liberal democratic republic to not want to have anything to do with politics.) But I realize I’m an outlier in this regard, and there is certainly no lack of research and writing by prominent scholars on youth participation and civic engagement (both Russ Dalton and Martin Wattenberg have books on the topic, and they’re both in the same R-1 department, to say nothing of the whole Putnam coterie) and other “practical politics” concerns, even if it’s not appearing in the contemporary APSR.

* Although modern calculus is more derived from Liebnitz’s approach that Newton’s; Newton’s notation is better reflected in differential equations.


I enjoy Nate Silver and Andrew Gelman’s writing most of the time… so why I am I so bored with’s content these days—to the point of having dropped my feed subscription—even when I agree with the substance of the content, even if not the aggressively Obama-cheerleading tone? (Or, to be fair, a link to a Nate Silver post.)

Maybe I’m in the minority—hell, I probably am—but I guess I subscribe to the philosophy of “dance with whomever brung you.” Silver at least has a keen analytical mind that probably would be better spent on his comparative advantage of data sifting and presentation rather than armchair political analysis from a perspective that’s available from, and done more thoroughly and thoughtfully, by folks like Kevin Drum.

Thursday, 5 March 2009

The model of a modern attorney general?

Orin Kerr and Eugene Volokh are talking up the likely run for state attorney general by Ted Cruz, the state’s former solicitor general. While I can’t say I’m thrilled about all of Cruz’s political positions, particularly on the social conservative dimension where Cruz makes much of his advocacy for Ten Commandments tomfoolery and takes pride in undermining foreign relations, he does at least seem to be eminently qualified for the post.

As a semi-related aside (perhaps brought on by my learning-more-about-while-teaching Texas government this semester), while in general I’d favor taking a rather large scythe to the number of statewide elected offices in Texas in favor of more gubernatorial appointees in line with the federal model, I’d probably favor keeping the attorney general’s office a separately-elected post, mostly to better promote checks and balances on executive power in a more transparent way.


Laredo managed to wedge its way up to the trough to get $31.5 million in stimulus money (along with $57.2 million in other funds) to build the Cuatro Vientos project in south Laredo, which is basically a bypass for U.S. 83 (Zapata Highway). Hopefully TxDOT can move relatively quickly on this project, since I know based on the MPO long-range planning workshop I went to last month that folks on the south side have been looking for some traffic relief.

Death of a discipline

Inside Higher Ed reports (as does the Milwaukee Journal Sentinel) that Wisconsin Lutheran College has decided to eliminate its political science department and, with it, two apparently-tenured faculty members to better focus on its “liberal arts mission.” I find myself in agreement with the thoughts of Michael Brintnall, executive director of the APSA:

“It would be thought to be a central component of a liberal arts education,” [Brintnall] said. “The subject matter is too central to civic life and understanding where we are going in the world to not offer the content.”

There is an argument to be made that the political dimensions of life can be explored in other social science and humanities disciplines—principally, through history, economics, sociology, anthropology, philosophy, and psychology—but somehow I doubt Wisconsin Lutheran will be devoting the attention those dimensions deserve in a well-rounded education.

Then again, Wisconsin Lutheran may have made the right decision in its current circumstances: according to the Journal Sentinel article, the abolition of political science only affects 5 majors directly. Considering that we had political science majors beating down the doors at Millsaps, which isn’t much bigger than Wisconsin Lutheran, I’m not sure what is going on with that.

þ: John Sides and Steven Taylor.

Tuesday, 3 March 2009

DC Talk

As you might expect, I am in complete agreement with my occasional co-bloggers James Joyner and Steven Taylor that while in principle the residents of the District of Columbia ought to have the right to vote for representatives (and, for that matter, senators)*, the proposed way of doing so—by passing a piece of ordinary legislation, allegedly pursuant to Congress’ plenary police powers over the District—is so blatantly unconstitutional I’m almost surprised members of Congress can vote for it while maintaining a straight face.

While I’m not convinced that we actually need a separate capital district in this day and age—other federal republics, such as Canada and Germany, seem to function perfectly well without a distinct federal district—as long as we have one we really ought to follow the rules. And, as I’ve noted before, the rules themselves are not so onerous as to justify bypassing them—if they were, the Constitution never would have been amended to give D.C. residents the right to choose presidential electors in the first place, during an era when the idea of granting voting rights for the district’s largely African-American population was much more politically contested than it is today.

* Note that this is a subtly different statement than “D.C. should have the same representation as a state,” a dubious proposition at best; there is no reason why D.C. residents couldn’t be treated as residents of Maryland (or for that matter Virginia or Wyoming or California) for the purpose of reapportionment and voting in federal elections but not otherwise be subject to state law.