SAS graphics

Okay, it is barely possible to make a decent looking SAS graphic with a half page of code, painfully specified to remove the abjectly painful default look. So imagine my susrpise when R and STATA do good looking graphics with a one line command. Sure, it might occasionally be a long line but still . . . Even EXCEL does this a lot better.

Why is SAS different?

It might seem like a minor point but there is a fair bit of truth to the idea that (easy to use) graphical represenations of data are extremely helpful.

The longer I work with SAS (since 1997) the more I wonder about this . . .

Weight of evidence

One of the annoying things in bio-medicine is that we often cannot replicate findings in a quick or efficient way. Unlike, for example, many areas of physics, an unusual finding can't be looked at quickly. So what do you do with a marginal finding?

Say for example, p=0.045 and conservative analytic approaches (such as BMA or LASSO regression) exclude this parameter?

If you publish it then you risk it being oversold and creating confusion.

If you don't publish it then you might join a long list of other people who ignore an important association.

Both alternatives seem unsatisfactory in the absence of a "hypothesis generation" tag for papers. I like the clinical pharmacology literature for these papers as the best of the alternatives but wish that it'd be cleaner.

A chicken in every pot, a couple more for your HMO

Sue Lowden, the candidate who is currently on track to become the next senator from Nevada recently said about health care:

"And I would have suggested, and I think that bartering is really good. Those doctors who you pay cash, you can barter, and that would get prices down in a hurry. And I would say go out, go ahead out and pay cash for whatever your medical needs are, and go ahead and barter with your doctor."

My first thought was that she meant 'barter' in the figurative sense, that patients should try to cut a deal with their doctors, not that patients should literally give doctors goods and services in lieu of cash.

I was mistaken:

The campaign of Senate candidate Sue Lowden (R-NV) is continuing to stand by Lowden's call for the use of the barter system as a means to bring down health care costs.

On Monday, Lowden doubled down on the barter idea: "You know, before we all started having health care, in the olden days, our grandparents, they would bring a chicken to the doctor. They would say I'll paint your house."

[TPM] asked Lowden spokesperson Crystal Feldman how this could ever be a workable policy, in an era of costly procedures, tests, pharmaceuticals and provider networks? "Americans are struggling to pay for their health care, and in order to afford coverage we must explore all options available to drive costs down," Feldman told TPMDC in an e-mail.

Feldman continued: "Bartering with your doctor is not a new concept. There have been numerous reports as to how negotiating with your doctor is an option and doctors have gone on the record verifying this. Unfortunately, Harry Reid's failed leadership forces us to take drastic measures. The fact remains that instead of producing a health care solution Americans support, Harry Reid spends his time focusing on attacking his biggest threat to another six years in Washington, Sue Lowden."

Aside from comic potential here (there are a lot of services you can barter for in Nevada), this suggests an interesting thought experiment:

Assuming that medical costs were driven by individual doctors and not by hospitals and drug companies (which really can't be bartered with), what would the introduction of barter do to the economics of health care? I would think that the introduction of wide-scale bartering would make the market less efficient and would produce more maldistribution of resources. Is this always true or is this another case where the strange economics of health care produce counter-intuitive results? And what would the other consequences be?

Given that there are approximately eight gazillion economics blogs out there, is there any chance that someone who knows what he or she is talking about could answer this one for us?

Financial reporting done right

With relevant facts, important context, smart commentary and lots of obscenity (some of which is not that effectively bleeped out so you might want to wait until you get home to watch this).

The Daily Show With Jon Stewart	Mon - Thurs 11p / 10c
These F@#king Guys - Goldman Sachs
www.thedailyshow.com
Daily Show Full Episodes Political Humor Tea Party

Journal Impact Factors

Candid Engineer has a post on impact factors. I find the variability of impact factors by field quite frustrating as it often neglects the field specific differences in publication style. Take, for example, the field of statistics. I have yet to find a statistician who would not be proud of an article in the Journal of the American Statistical Society (JASA). It has an impact factor of 2.3; this does not compare favorably with (for example) Epidemiology (IF=5.4) which also publishes methodological papers. But the impact factor approach understates the effort required to beuild on or develop methods -- more sophisiticated methods take a lot longer to work their way into the literature.

In any case, comparing JASA to the New England Journal of Medicine (IF=50!!) makes it clear that the field has a lot to do with the impact of a journal.

So what do you do when you are on the borders of a field?

It's a tricky problem . . .

Tragedy of the grading commons

Unless things have changed radically over the past few years, there's a simple way of determining which teachers are grading the hardest and/or assigning the most homework: just go to parent/teacher night and look for the longest line of concerned or angry parents.

This is not to say that there is no positive feedback; you will have parents who will thank you for keeping standards high and will offer their sincere support, but they will be the minority and more importantly, their per capita impact is less because of the nature of administrators.

As I mentioned in an earlier post, there are very few promotions available to teachers. Administrator is one of the few exceptions, but once you've made the jump into administration the situation changes radically.

Here is some context from the Bureau of Labor Statistics:

Education administrators advance through promotion to higher level administrative positions or by transferring to comparable positions at larger schools or systems. They also may become superintendents of school systems or presidents of educational institutions.
...

In May 2008, elementary and secondary school administrators had median annual wages of $83,880. The middle 50 percent earned between $68,360 and $102,830. The lowest 10 percent earned less than $55,580 and the highest 10 percent earned more than $124,250.

In May 2008, postsecondary school administrators had median annual wages of $80,670. The middle 50 percent earned between $58,940 and $113,860. The lowest 10 percent earned less than $45,050 and the highest 10 percent earned more than $160,500.

When I first decided to go into teaching I asked a retired superintendent I knew for advice. The first thing he told me was, "Never trust a superintendent; they'll lie to your face." I think he was being just a bit harsh but I understand his position. Administrators live in an intensely political world where the right move can double their incomes and the wrong one can get them demoted or fired. It tends to test character.

Under these circumstances, you can understand how frightening an angry parent can be. When a mother and father storm in and demand to talk to the principal, they bring considerable pressure to bear. Enough pressure that even a dedicated educator (and most administrators fall into that category) has to be tempted to cave.

These tense conferences are often the result of an A student receiving a B, and they, in turn, often result in an equally tense principal-teacher conference.

I have never heard of an angry conference caused by an A.

Like a well-maintained commons, high educational standard are in everyone's best interests. Unfortunately, just as it is in the best interest of the individual farmer to overuse common land, so it is in the best interest of individual parents to see their children's grades inflated.

I haven't seen any recent reform proposals that will address that problem; I have seen quite a few that will make it worse.

Who could have seen this one coming?

Apparently the competition for students willing to pay huge tuitions might occasionally lead to a drop in academic standards.

From the New York Times:

Over the last 50 years, college grade-point averages have risen about 0.1 points per decade, with private schools fueling the most grade inflation, a recent study finds.

The study, by Stuart Rojstaczer and Christopher Healy, uses historical data from 80 four-year colleges and universities. It finds that G.P.A.’s have risen from a national average of 2.52 in the 1950s to about 3.11 by the middle of the last decade.

For the first half of the 20th century, grading at private schools and public schools rose more or less in tandem. But starting in the 1950s, grading at public and private schools began to diverge. Students at private schools started receiving significantly higher grades than those received by their equally-qualified peers — based on SAT scores and other measures — at public schools.

Career paths, educational reform and unintended consequences

This post by Joseph about career paths for researchers reminded me of some disturbing trends in career paths on the other side of academia, teaching.

In primary and secondary education, you can have a completely successful career by any conceivable standard and never get a single promotion. You can easily spend your entire professional life doing the same job with the same title. That might even be the ideal.

For a field requiring a degree, additional coursework and certification, this is an extraordinarily limited career path. To make up for that we have traditionally offered the following:

1. Reliable income that increases at an agreed-upon rate annually.

2. A high level of job security after a certain number of years (though this is somewhat offset by low job security before reaching tenure).

As well as creating a career path that didn't depend on promotion (and therefore largely avoided the Peter Principle), this emphasis on deferred, but steady compensation meant schools could minimize their investment in new, untested personnel (most really disastrous teachers do not make it to the tenure mark).

Two of the central tenets of the current move for education reform are elimination of tenure and replacing raises based on experience and education with merit pay. This leaves us with a career path that offers no real chance of promotion, no job security and wildly variable pay based on metrics that are largely out of the teacher's control and can easily be gamed by a biased administrator (see here).

Research Scientists

In the comments on a post on the unintended consequences of increasing post-doctoral salaries, Drug Monkey comments:

I also want to reiterate that at least 70% of the arguments going on here are complaining about the postdoc as a short training stint morphing into a substantial part of, if not the very pinnacle of, a science *career*. If this is the root of your problem, whinging for small raises ain't solving the issue. I've made myself quite clear that I think the lack of a career scientist job track is a horrible situation that bears fixing..and would save the NIH money to boot.

I think that this is really insightful. If we want to have a strong body of directed semi-independent researchers (currently the ground held by post-docs) then it might make sense to formalize the career path. My only thought is whether the new "research scientist" would end up looking a lot like a traditional professor?

On the other hand, so long as the grant system is in the present form, it's rather hard to argue for reform as the financial incentives are pretty well aligned towards making researchers who repeatedly win grants very valuable to their institutions.

And you thought you didn't want to know how they made sausage...

The New York Times has an interesting article on the world's most expensive coffee.

A discussion you ought to be following

Bruce Bartlett and Mark Thoma continue their analyses of the relationship between think tanks, the media and economics.

Here's Bartlett:

Conservatives also realized that putting out a study saying the exact opposite of a liberal study was sufficient to muddy the water and prevent a reporter from drawing a clear conclusion from the liberal study. It didn't matter that the liberal study was done by a preeminent scholar in the field and the conservative study was done by a glorified intern. All that mattered is that they came to opposite conclusions, thus leading to on-the-one-hand/on-the-other-hand stories that everyone hates but the media won't stop writing.

Conservatives understand--better than liberals, I think--that most stories are lucky to last one news cycle. If the reporter later decides that the liberal study was really worthwhile and the conservative one was worthless, he isn't going to go back and do another article on the subject. It's water over the dam.

Parenthetically, I would add that the talking head approach to policy debate on the cable news channels reinforces all the negative aspects of this development. Once upon a time, I used to do a lot of cable interviews. At first, I was often paired with people I knew at other think tanks who were slightly more liberal than I am. But because we both shared common facts and knew the limits of what could be demonstrated through serious academic research, we naturally tended to agree with each quite a bit.

Having two guests who agree with each other is the last thing cable channels want; they want their guests to be 180 degree polar opposites. So gradually I noticed that I was no longer being paired with peers from liberal think tanks, but people I had never heard of who were identified as "Democratic consultant" or something like that. Such people clearly knew virtually nothing about the subject we were discussing and were just there to endlessly repeat talking points that someone gave them.

And here's Thoma's reply:

Some of the people on these shows aren't qualified to speak as economists, but they get called again and again because they can spout talking points in an entertaining fashion. Paul Krugman paired against journalist Robert Samuelson* in a CNN debate on the deficit a week or so ago is an example of this (though from what I saw it's not clear that Samuelson satisfies the "in an entertaining fashion" constraint). Menzie Chinn was certainly frustrated by the pairing:

Nonetheless, overarching all this is a simple question. Why do we ascribe any credibility to a person with an undergraduate degree in political science (what is called Government at Harvard) in the area of economics (let alone accounting)? (The question is inspired by watching the debate between Professor Krugman and Mr. Samuelson on Fareed Zakaria GPS yesterday...)

I'm not sure what the answer is. The elevation of entertainment over facts isn't going to change as that is the most profitable strategy for the networks, and economists are unlikely to become more entertaining.

Though it won't solve the problem, one force helping to counteract the elevation of entertainment over facts is a pair of the most entertaining shows on television, the Daily Show and the Colbert Report. Not only do these shows routinely savage cable news channels for sensationalism and lack of content, they often provide more factual context in their satirical coverage than you'll see on CNN et al.

* I wonder if anyone at CNN has confused Robert Samuelson with Paul Samuelson?

Another reason to abolish tenure

You may have seen this story from Inside Higher Ed (via Andrew):

The biology professor at Louisiana State University at Baton Rouge gives brief quizzes at the beginning of every class, to assure attendance and to make sure students are doing the reading. On her tests, she doesn't use a curve, as she believes that students must achieve mastery of the subject matter, not just achieve more mastery than the worst students in the course. For multiple choice questions, she gives 10 possible answers, not the expected 4, as she doesn't want students to get very far with guessing.

Students in introductory biology don't need to worry about meeting her standards anymore. LSU removed her from teaching, mid-semester, and raised the grades of students in the class. In so doing, the university's administration has set off a debate about grade inflation, due process and a professor's right to set standards in her own course.

Anyone who has spent time behind a podium can give you examples of parents and administrators pressuring teachers to raise a students' grades. The most egregious case I know of took place at a highly prestigious and very pricey prep school, but even in public schools in the days before school choice, a call from a parent to a principal about a grade would often result in an unpleasant employer-employee chat and the loss of a badly needed hour of grading and lesson planning.

I remember those meetings, and I remember thinking how much easier it would have been to stand up to that pressure if only the administrator had had more authority to fire teachers. Thank God for education reform.

Bartlett on the intellectual decline and political rise of think tanks

Over at Forbes, Bruce Bartlett provides an insightful insider's view of the development of the modern think tank. Mark Thoma adds his comments here.

Quote of the day on financial innovation

Felix Salmon is blunt:

In the world of credit, innovation generally consists of taking risky stuff, waving some kind of diversification and/or overcollateralization magic wand, and ending up with something which is (a) meant to be safer, and (b) much more difficult to analyze on a fundamental basis: you end up having to use models instead. And models have a tendency to break.

While on the subject of evolution...

I can't miss a chance to recommend Ian Stewart's "Through the Evolvoscope," a clever and elegant discussion of fitness landscapes. I believe this originally appeared in Stewart's Scientific American column, but you can find it in Another Fine Math You've Gotten Me Into.

Way cool.