Big cuts for science?

House Republicans have a proposal for cutting spending starting next month: Summary here

To emphasize, the proposal is for cutting funds already allocated for this year as well as for the coming year.

A few numbers:

NIH cuts for 2011:

• $260M, inflationary increases on non-competing grants ($260M for next year)
• $ 48.5M, Office of the director (Common Fund ) ($66M for next year)
• $ 77.3M, buildings and facilities ($0.1M for next year)
• $300M, eliminate global AIDS transfer (this program? ) ($300M for next year)
• $304M, eliminate Project BioShield Special Reserve Fund Transfer ($304M for next year)
• $639.5M, general reduction to 2008 levels ($639.5 for next year)

NSF cuts for 2011:

• $150M for research and related activities ($550M next year)
• $ 62.5M for equipment and facilities construction ($110M for next year)
• $ 70M for education and human resources ($85M for next year)

All told, about $1.5B for NIH (~5% of NIH's $31B budget) and $300M for NSF (~5% of NSF's $7B budget).

These are big cuts, but to keep things in perspective, NSF's 2011 budget is 6.7% higher than 2010, and the cuts at this point are simply proposals.

The challenge is that both NSF and NIH make multi-year commitments, which means that any cuts either have to be passed on to existing grants or taken disproportionately out of new awards. If they're smart, they'll try to spread the pain rather broadly.

[0 comments]

Debunking Discrimination

A blunt article in PNAS by Ceci and Williams provides considerable evidence that the underrepresentation of women in mathematically intensive fields is not due to systematic discrimination:

Women’s current underrepresentation in math-intensive fields is not caused by discrimination in these domains, but rather to sex differences in resources, abilities, and choices (whether free or constrained). Thus, current initiatives direct energy toward solving past problems rather than current ones. Women’s underrepresentation today results from a complex set of interrelated factors, some of which society could meaningfully address if the focus was placed squarely on them. One key to such success is moving beyond historical issues and confronting current ones.

The paper systematically examines the evidence for gender discrimination in grant and publication acceptance rates and in hiring, and finds it lacking. The story is consistent: for almost every paper discussed that provides evidence of discrimination, subsequent papers that do more rigorous analyses find no effect. (The authors do acknowledge that discrimination does occur, but that discrimination is the exception rather than the rule)

the evidence shows women fare as well as men in hiring, funding, and publishing (given comparable resources)

The key is the "given comparable resources" part - the authors note that women tend to be particularly poorly represented in research intensive positions that would give them the time and resources needed to publish more / produce more grants.

The explanation for the differing resource availability?

That women tend to occupy positions offering fewer resources is not due to women being bypassed in interviewing and hiring or being denied grants and journal publications because of their sex. It is due primarily to factors surrounding family formation and childrearing, gendered expectations, lifestyle choices, and career preferences—some originating before or during adolescence - and secondarily to sex differences at the extreme right tail of mathematics performance on tests used as gateways to graduate school admission.

Pretty much in line with Larry Summers' talk, and there are lots of footnotes. (Today's Times, by the way, has an interesting discussion of the Summers brouhaha.)

One thing that was new to me was a study of gender differences in career preferences:

adolescent girls often prefer careers focusing on people as opposed to things, and this preference accounts for their burgeoning numbers in such fields as medicine and biology, and their smaller presence in math-intensive fields such as computer science, physics, engineering, chemistry, and mathematics, even when math ability is equated. In a recent metaanalysis of >500,000 participants, the male-female effect size for preferring people vs. things overall was d > 0.90, and for engineering, 1.1, both substantial differences.

( Here's the cited paper )

The authors aren't saying that the differing numbers of women in mathematical sciences is not a problem because it is a chosen situation. They emphasize:

To the extent that women’s choices are freely made and women are satisfied with the outcomes, then we have no problem. However, to the extent that these choices are constrained by biology and/or society, and women are dissatisfied with the outcomes, or women’s talent is not actualized, then we most emphatically have a problem. With a redirection of resources, this problem might be addressed by education and outreach to young women and girls and to academic administrators. Past strategies to remediate women’s underrepresentation can be viewed as a success story; however, continuing to advocate strategies successful in the past to combat shortages of women in math-based fields today.

The paper suggests some promising strategies for addressing remaining barriers to greater participation of women in math-intensive fields. One in particular is programs like Berkeley's "Family Friendly Edge"

Definitely a paper worth reading of the issue of women in science is of interest.

[1 comment]

Science & Baseball

Back in the 1990's, the Oakland A's were a mediocre team. In baseball, the traditional way to improve a team has been to hire a bunch of super star players. The trouble with being a mediocre team, though, is that it's hard to sell tickets, which makes it tough to pay for expensive talent. To make matters worse, A's were one of the poorest team in baseball, with a total payroll of only 1/3 of the richest team, the Yankees. So the A's were stuck.

Desperate to change things, Billy Beane, the A's general manager, embarked on a new and strange strategy: he started using ideas from an obscure, mimeographed fan-zine on baseball statistics in his hiring decisions. There was tremendous push-back and resistance to the new approach, and indeed, the new hires were a strange lot. Instead of the young, athletic guys the recruiters were used to courting, now they were hiring older players, overweight players, even a pitcher with a club foot. But the new players were cheap, and Oakland started winning, and within a few years, they made it most of the way through the World Series, losing only to the Yankees.

What happened? It turns out that the amateur baseball statisticians that Beane started listening to were on to something. The traditional measures of baseball performance miss some important things. For example, batting averages, the traditional way of rating hitters, don't credit players who choose not to swing at wild pitches and instead get walked. Beane's new ways of evaluating players proved better at measuring their actual contribution to the game, and that gave him a huge advantage: he was able to identify players who were valuable in ways that nobody else could see, and he was able to hire them at a fraction of the price of similar talent that was more identifiable.

Michael Lewis tells the full story of the A's rise in his excellent book Moneyball: The Art of Winning an Unfair Game. Even if you don't care for baseball, I highly recommend the book. The reason: it's not really about baseball. It's about metrics, why they matter, and how difficult it is to get people to pay attention to the right ones.

Science funding is heavily influenced by metrics, in particular the NSF's Science and Engineering Indicators. Are we measuring the right things? Or are there things we should be doing better, like the Oakland A's, to direct resources to untapped opportunities?

I've been invited to participate in a National Academies assessment of existing science metrics, so over the next 2 years or so, we'll see.

[0 comments]

Test, test, test

Another article in the Times' excellent coverage of useful research on how we learn: To Really Learn, Quit Studying and Take a Test. Taking tests turns out to be a very effective mechanism for learning. Definitely something to consider when structuring one's courses.

The specific test mechanism looks pretty simple, and seems easy to turn into something one's students can do on their own:

The final group took a “retrieval practice” test. Without the passage in front of them, they wrote what they remembered in a free-form essay for 10 minutes. Then they reread the passage and took another retrieval practice test.

[1 comment]

"The PhD student is...

...someone who forgoes current income in order to forgo future income." Choice comment from a letter to the Economist in response to their recent article, The Disposable Academic

[0 comments]

Taxing times?

A prediction: many postdocs will face higher taxes in the years to come. Why? A recent supreme court ruling on the tax status of a similar group: medical residents.

Some postdocs are not required pay social security taxes because their fellowships are not classified as compensation (it's an issue I don't pretend to understand well, but the NPA has a useful summary of postdoc-related tax issues here ).

Having to pay social security could make a significant difference in postdocs' take home pay - it's about 7% of your income. And it could mean a hit to universities, too, since employers contribute an equivalent amount.

Something to look out for.

[0 comments]

"The mathematics of narcissism"

Fellow mathematician Jordan Ellenberg has an unusual take on the NRC's rankings: in Slate he compares the NRC's approach to ranking graduate programs to a new method psychologists are using for classifying mental illnesses.

The article is worth reading in full, but the gist of it is that there are two standard approaches to dealing with high dimensional data sets: you can cluster items into groups, and you can use statistical techniques to reduce dimensionality, typically by discarding dimensions that carry the least amount of information. The NRC uses one method, and Ellenberg thinks they might benefit from using the other.

The forthcoming Diagnostic and Statistical Manual of Mental Disorders (the DSM-V) is switching from a clustering-centric approach to a dimension reducing approach, replacing clusters like "narcissistic personality disorder" with a collection of 6 measurements ("negative emotionality, introversion, antagonism, disinhibition, compulsivity, and schizotypy"). This is apparently leading to grumblings from psychologists who find value in the clusters as opposed to the more abstract 6-dimensional vectors.

The NRC has also chosen a dimensionality reduction approach, boiling 20 program measurements down to a single quality dimension. Ellenberg suggests that a clustering approach might be more helpful, and cites a recent experiment:

The NRC, on the other hand, might have done better to toss the idea of rankings entirely, and just clustered the departments into natural groupings. The statistician Leland Wilkinson ran a quick and dirty clustering on the NRC data for math departments. He found that the departments broke up into five clusters: 10 elite departments, a big group of 59 upper-tier departments, 47 lower-tier departments, and two smaller clusters whose meaning, if any, isn't clear to me. This is much coarser information than a full ranking—but it has the advantage of not depending on politically contentious choices as to which criteria matter most.

It's an interesting idea, and I think there's some value to the approach. Indeed, the Carnegie Foundation already does something similar for universities, though probably not in a particularly statistically rigorous fashion. Having well chosen clusters would provide for saner comparisons - it doesn't really make sense to compare some kinds of programs directly, as they really cater to very different audiences with different goals.

That said, I very much doubt that the clustering approach would prove any more satisfactory than what the NRC actually did. Do you think that a prospective student or department chair would be any happier to learn that a program fell into a set of 59 "upper-tier departments" than to know that the program ranked between 16th and 27th on the NRC's quality scale?

While a clustering approach sidesteps the need to explicitly choose important criteria, there is very much a devil-in-the-details problem. Different clustering approaches can yield very different clusters. Even the simplest methods involve many choices - at the very least you have to choose a measure of similarity, and that in turn will emphasize and de-emphasize different program characteristics. You're essentially trading an explicit, principled choice about what's important for an implicit and opaque choice.

Regardless, I'd be curious to see more details of Wilkinson's approach. I imagine he just did some kind of k-means clustering - simple, but likely interesting.

[0 comments]

Professional Science

A heartening holiday article in the NY Times this week: A Master’s for Science Professionals Sweeps U.S. Schools. The Professional Science Masters is catching on big time:

The degree, which a few universities quietly pioneered in the mid-1990s, combines graduate studies in science or mathematics and business management courses. In 2008, 58 universities were offering the professional science master’s degree, or P.S.M., according to the Council of Graduate Schools in Washington. By the start of this academic year, the number had nearly doubled to 103, and is set to climb further. The number is certain to grow because the professional science master’s degree is being adopted by at least six state university systems.

The great thing about the PSM is that interaction with industry plays a big role in the degree. Students spend time in internships so they learn skills that they can't get in universities, and industry gets technology transfer through students. More importantly, to run successful programs, universities have to build relationships with local companies, which is a great way for faculty members to get clued in about what kinds of skills working scientists outside of academia really need.

Kudos to the Sloan Foundation for getting the ball rolling and to the NSF for additional funding.

A PSM + a PhD sounds like a much more effective ticket to a great industry job than a regular PhD. Given the ratio of PhDs to faculty positions, we'll need a lot more PSMs.

[1 comment]