The semester is dead! Long live the semester!

So it’s not quite dead yet. I still have a final Saturday afternoon and a paper due next week. But classes are over. Hurray! In a matter of weeks I will be 12 credits closer to the magical 144 credits required to get me out of here with my doctorates in hand.

Philosophy of Education Society

I spent last weekend at the Philosophy of Education Society Annual Conference in Cambridge, Massachusetts. Highlights: meeting another grad student who is working in my area (perhaps the only other grad student working in my area), having the top senior scholar in my field attend my works-in-progress session, and finding three terrific books on philosophy and social science at Raven Books.

Other observations: Harvard’s campus is glum. You’d think with a $40 billion endowment you can afford some sort of ornamental ground-covering. Cambridge is nice though, especially the Charles River.

On a more professional note, there seems to be a significant and positive relationship between a philosopher’s reputation and her/his confidence that her/his sagacity extends beyond say, the writings of Levinas, to all other matters, including: education policy, public health, the death penalty, the Iraq War, Sudan, whether Obama will beat Hillary, the aesthetic merits of Charles Dickens compared to Queen Latifah.

And to believe we philosophers are footnotes to Plato, whose character Socrates’ singular claim to wisdom is that he knows that he knows nothing.

Triathlon, my new favorite blog, and prom for grown-ups

I've already written about my mixed feelings about multisport. Namely, I worry that the barriers to entry make it such that most participants in the United States are white and wealthy, and that part of what attracts people to the sport is its ability to signal to other white/wealthy people that one is of their caste. My new favorite blog has a post which makes the point with more humor and poignancy.

In related news, I’m skipping Collegiate Nationals this weekend to spend more time with my family and books, in that order. I will have a little leisure; Owen has their “Capitalist Ball” Friday night, which is essentially prom for MBA students. Owen students are, in general, very different from doctoral students, whether it be their conceptions of a meaningful life or their estimation of the value of knowledge. I learn a good deal about myself by being around them.

Praise for Eastside Cycles

There’s nothing like having an honest bike mechanic in the neighborhood. Yesterday,
I took my beloved 2003 Fuji Roubaix in for a yearly tune-up at Eastside Cycles. I bought the bike used last year from a person who took pretty good care of it. I’ve kept it clean and lubed, but figured a little preventive maintenance might be a good idea.

The guy at Eastside looked my bike over and asked me if I had noticed anything wrong with it. After I told him that I hadn’t, he suggested that I not bother with the full $65 tune-up. I just needed to have the rear wheel trued—which they did in five minutes and below the listed cost.

They could have easily talked me into wasting money, but instead found the cheapest way for me to get back on the road. I went ahead and purchased a flat kit from them and would have bought new tires, but they didn’t have the ones I was looking for. I’ll make sure to go to them first with my future cycling needs.

I went to the wrong school

Now in my fifth year of graduate school, midway through two doctoral programs, I realize that it has been time utterly wasted. I could have just taken a 10-week course.

This company makes promises that no other serious philosopher, dead or alive, has ever made. I saw them advertised in the subway in New York.

Follow up on Ed School Rankings

A classmate of mine who runs a much more popular blog suggests I may be mistaken about points three and four in the previous post about school rankings. He argues first that multicollinearity matters only in regression equations, not indices. Second, he suggests that “Mean GRE” scores and “Acceptance Rate” may not be muticollinear anyhow. Third, he argues that "Expenditures" and "Expenditures Per Faculty" are not collinear because different institutions have widely differing numbers of faculty.

I’m checking on an answer to one. If multicollinearity is not an issue, answers to his second two points will be beside the point.

Regarding his second point: Assume that the distribution of GRE scores across a school’s pool of applicants follows a normal distribution. Imagine that a given school in year one receives 100 applicants and accepts 10 students out of 100 applicants. Imagine that the same school receives 150 applicants in year two, but admits the same number of students (10). If the school continues to give the GRE the same weight when making admissions decisions, we would expect the GRE scores of the incoming students in year two to be higher than the incoming students in year one. The acceptance rate has also dropped from 10% to 6.67%. There is thus strong reason to suspect that mean GRE and Acceptance Rate are highly correlated.

Regarding his third point: My classmate's point is valid, but only for institutions that have widely differing numbers of faculty. For schools that have roughly the same number of faculty, those with large research expenditures get double credit over their peers in the rankings; first, for having more research money; second, for having more per faculty. Further, it may be the case that smaller schools tend to have less research expenditures per faculty (perhaps there are economies of scale). These schools get doubly punished. This should cause some suspicion that these two variables may be highly correlated. I’m betting dollars to donuts they are. Now just to see whether this matters...

Peabody Makes Number Two

The U.S. News and World Report Rankings of Graduate Schools were released a couple of weeks ago. I’ve long been suspicious of rankings as such—I’m just not sure how one can meaningfully represent university quality with statistical methods. I know that there are better and worse schools and that we may usefully (if imprecisely) categorize institutions in broad tiers. But the latest reports have the pretense of surgical precision.

Time is scarce, so I will focus just on the methodology, rather than the deeper philosophical blunders. Here are a few of the problems:

1. The response rate for the survey of deans is 47%. The response rate for superintendents is 24%. No attempt was made to control for non-response bias. The reader has no idea whether respondents differed in non-random ways from non-respondents.
2. Superintendents were allowed to respond ‘don’t know’ if they were unfamiliar with a school. Such responses did not count against the school. This could mean that smaller schools may have fewer raters, thus making their score more sensitive to any particular rater. Also, it is possible that people tend to ‘not know’ a school when its research is of low quality. The only people who actually know these schools are its alumni, who may give biased ratings. Thus, mediocre schools may end up with artificially inflated scores.
3. “Mean GRE” scores and “Acceptance Rate” are likely to be collinear. As more students apply to a school, the school is able to limit its acceptances to those with the highest GRE scores.
4. “Total Research Expenditures” and “Average Expenditures per Faculty Member” are collinear. This is perhaps the most inexcusable statistical blunder in the report.
5. The specialty rankings are made by education school deans. Deans are unlikely to be experts in more than one specialty. The rankings would be more meaningful if they were by faculty in each specialty, or at the minimum, by department chairs.
6. Given all of these (and other problems), we would expect confidence intervals to accompany the rankings. For instance, they could say that Peabody is somewhere between the 1st and 5th highest ranked university. My suspicion is that the p-values were such that the confidence intervals would expose the meaninglessness of most year-to-year changes in rank.

I gave a conference paper last year on Great Britain’s now defunct Research Assessment Exercise. The RAE ranks universities according to peer review. I was disturbed by the idea of government rankings, but I’m now equally worried about private rankings. Rankings have come to wield enormous influence. There is no escaping this. There ought to be an alternate ranking system which is open, subject to peer review, and driven by a desire to distribute information rather than sell magazines.

Grandmothers of the world, unite!

Matlock, Season 1, comes to DVD this Tuesday.

The Answer to Last Post's Question

I'm taking the jobs and giving them my absolute best. I'll just have to work smarter.

The Perils of Summer Funding

The situation:

1. You want to graduate next year, which requires you to focus on writing during the summer. But you don’t get paid to do your own research, and thus would need to take out loans.
2. Your advisor offers you interesting and well-paid summer work which is in no way related to your research interests, dissertations, or graduation in general.

Do you:

1. Try to do both, hoping that the paid work won’t result in you publishing less, delaying graduation, etc. You might succeed, but the odds are uncertain.
2. Turn down the well-paid work and stay true to the graduate-student vow of poverty.
3. Accept the well-paid work and give up on the accelerated-graduation pipe-dream.