Previous month:
March 2007
Next month:
May 2007

April 2007


On Friday I attended the inauguration of Lehigh University's 13th President, Dr. Alice Gast. (I am a faculty member at Lehigh.) Dr. Gast actually became President on August 1, 2006 but the enthusiasm for her appointment has not abated - faculty members attended her inauguration in droves with many showing up without even RSVPing, to the professed dismay but inner satisfaction of the person in charge of logistics. (The issue with logistics was that the faculty sit in front of the stage in a specially designated area, and because quite a few professors did not RSVP there was some concern that we would not have enough seats. But in the end it all worked out.) Everybody on campus speaks very highly of Dr. Gast. Before coming to Lehigh she served as Vice-President of Research at MIT for 5 years and was a Professor of Chemical Engineering at Stanford University until 2001. (Another disclosure: her time at MIT and mine overlapped, but we never met then.)

Of course Lehigh recruited her to help bring the institution to the next level. Outsiders do not always realize the level of commitment and dedication of Lehigh alumni to their alma mater - no love-hate relationship here upon graduation! The well-roundedness of the undergraduate students, at least the ones I teach, has not ceased to astonish me since I arrived on campus. (They are a bit spoilt, but very good, and better than they think.) Lehigh also benefits from some superb strategic planning: I do not know who had the idea of developing global studies in the undergraduate curriculum, for instance through the Global Citizenship Program, but that person struck gold. This is absolutely a time where US graduates need a deeper understanding of the world and I expect global studies to play a big part in increasing Lehigh's visibility outside the Northeast.

The main question mark is on graduate research. Many institutions want to raise their profile by increasing the amount of research they do (grants, number of doctoral students in each department) but the quality of grad students becomes very unequal outside the very top institutions such as MIT, Stanford and Berkeley. We do get some excellent graduate students, in particular from Turkey (thank you Bilkent University) and some diamonds in the rough but the truth is, there is too little supply (worthy grad applicants) and too much demand (slots in US programs). Ultimately institutions which strive for excellence will have to rethink their assumptions on how to reach the top. Top student research at the graduate level might come to be viewed as icing on the cake rather than the cake itself. Parents send their kids to college to give them a great foundation and Lehigh is very well positioned to do just that. Dr. Gast emphasized in her inaugural remarks Lehigh's commitment to education through the university's flagship programs in engineering but also business and the arts and humanities. She has already struck me as a very poised and thoughtful person who is skilled at personal relationships (the fact that she stood by herself in the hall minutes before her own inauguration as the faculty marched by to shake every professor's hand and thank us for coming - our attendance wasn't mandatory - certainly endeared her further to the faculty.) While much has been made of her being a woman, and of MIT and Harvard both picking women presidents, I am looking forward to her presidency not because she advances the cause of women, although that does not hurt, but because she has the skills and ability to do a fantastic job, and I am glad she is going to do it at Lehigh. 

Downwards, graduates!

If you work in academia like me, you have been knowing for several months what the New York Times announced on April 4: this has been an exceptionally good year for college admissions. The last two years had already seen growing numbers of applicants and increasing test scores but many universities had, I believe, made an attempt to accommodate the children of the baby-boomers. A third year of overcrowding the dorms simply wasn't possible for most, and many talented kids are getting "bumped down" to safety schools they had never imagined they would need. Administrators rejoice: lower admission rates and higher SATs will look good in the US News rankings. But the sentence that really caught my attention was in parenthesis towards the bottom of the article: "The statistics project that the number of high school graduates will peak in 2008."

There has been little discussion of what will come after that. Parents with shattered dreams of Ivy League have been busy convincing themselves that their talented children will be as happy at lesser known colleges, and administrators with an eye on the annual rankings have been busy admiring their school's newly found selectivity. The fact that the number of graduates is set to decrease, though, does not bode well for the classes of 2007 and 2008. That is because employers do pay attention to the university a student graduates from - companies do not have the resources to recruit on every single campus and prefer interviewing candidates at schools they have previously hired from. Even if 2012 witnesses an increase in job openings large enough to accommodate all the college graduates, the peak in 2008 suggests that companies will not have the time to change their recruiting habits before things get back to normal. So the student who would have been admitted to Ivy League schools in a heartbeat three years ago might well drag her bad luck with her a little while longer. Unless...

I looked up the projected high school graduation rates up until 2018 and, dear future graduates of the class of 2008, let me assure you that things aren't as bleak for you as that sentence buried in the NYT seemed to indicate. Because the peak the article was referring to is not a peak at all, but rather a plateau until 2010, followed by a tiny movement downward in 2011-2012. And then, just when your little brother who makes so much fun of you now is getting ready to graduate, the numbers spike up faster than a roller-coaster at Disney Land - the grand total is expected to cross the 3-million threshold in 2016. In other words: universities, high school students, parents and recruiters are in this for the long run, and everyone will have to adjust. Colleges that used to play second-fiddle to the big leagues will get their share of the limelight thanks to the outstanding high school seniors falling in their lap, courtesy of the capacity constraints at the top places. Everything else being equal, I'd suspect the brilliant kids who were denied admission at all the Ivy Leagues to be more motivated than ever to shine and prove the admission committees wrong. Reputations are about to be made anew; exciting times await us...

Double underlines

Yesterday I noticed something odd as I was browsing the web. If you look up trading, or hedge fund, or Value-at-Risk on, you will see some words with a double underline. On webpages, words with a single underline indicate a link to another site, so you would be forgiven for thinking that this is a link to something relevant and useful. You would be wrong. Instead, you are staring at the latest installment in the great saga of Internet advertising - in this specific case, having your mouse over any of the words with double underlines will open a text advertisement for MSN Money. Now let me ask you this: once you have figured out that double underlines meant junk ads, why in the world would you get anywhere near them again?
One of these days we're going to need a TiVo for the pages we browse online.

A band-aid on illiteracy

As I mentioned in my last post, a few days ago I stumbled upon the 2005 assessment results of the Nation's Report Card for eighth-graders in US public schools. In the public schools of half the states (not including D.C.), at least one eighth-grader out of four scores below basic on the reading test. Below basic! The extremes are Hawaii at 42% and North Dakota and Massachusetts, both at 17%. And some states are supposedly watering that test down to help schools avoid sanctions? Oh well, that's reassuring.

All the talk about the merits and flaws of testing takes the focus off the dirty truth, which is that so many kids can't read. The issue, I believe (and I mentioned something along those lines in an old post on competitive engineering), is that bringing literacy to everyone in America isn't a sexy topic - people get their competitive juices fired by attempting a task that has never been done before and can be clearly documented through a single observation; I guess you could call it qualitative innovation. Quantitative innovation, i.e., bringing something that is already available to a large segment of the public, is more of an operational problem. (Ever noticed that people prefer to think of themselves as leaders rather than managers? It is the same idea. Managers deal with the nitty-gritty stuff that must happen for the project to be successful; leaders provide vision and strategy. Who doesn't want to do the latter rather than the former?)

A honest discussion of ways to enable all schoolchildren to function normally (we're not talking about reading Shakespeare, just the menu at Denny's or the driver's license manual) would be time better spent than the usual arguments on the perverse effects of the No Child Left Behind Act. Because the only other option I can think of - besides doing nothing and hoping the problem goes away - is to sell the portable scanning devices currently being developed for the visually impaired to people with reading deficiencies. (Again, breaking new ground appeals a lot more to the community than making the playing field level for everyone; on the plus side, the now gigantic size of the market would certainly help bring the price tag of the device down.) Shame on the American educational system when that happens.   

No Child Left Behind

Much of the talk surrounding the reauthorization of the No Child Left Behind Act has focused on whether the tests have been truly measuring reading and math proficiencies and whether they have had any perverse effects on K-12 education, for instance forcing teachers to focus on brute-force test preparation and enciting lawmakers to lower standards. (Complaints have been simmering for years; Sam Dillon at the New York Times wrote an article on the topic as early as May 2003.) I had been wondering whether there was any obvious bias in the statewide test score averages so I looked up the 2003 results for eight-graders and, being a data-cruncher at heart (ah, those engineers), tried to explain them by running a linear regression on several variables: the 2004 statewide unemployment rates for 16- to 19-year-olds, the average high school graduation rate for freshmen entering high school in 2000, two-year average (2003-04) of statewide median income, and percentage of the state population below the poverty rate in 2003. (Ideally I would have used 2005-06 unemployment and graduation data for students who took the test in 2003 but it was not available. Also, New York and Wisconsin were removed from the set due to data unavailability.)  My goal was to use these variables to explain the test score and study how the scores predicted by the regression model differ from the true ones.

So here are the results of my regression:

  • In case you are wondering, the graduation rates for Class of 2000 high school freshmen explain only 57% of the Grade 8 total scores, i.e., math plus reading (R-square=0.57 - if you have forgotten everything about regression, the goal is to find the input variables that make R-square as close to 100% as possible).
  • The statewide median income has no predictive power at all on the test scores (0 is in the 95% confidence interval for that coefficient, so you can't even be 95% certain whether increasing the median income would increase or decrease the test scores).
  • If you include the unemployment rates of the 16- to 19-year-olds in the regression, R-square barely increases to 61% (my idea in considering this factor was that students drop out if they struggle in class and/or there are jobs available for them).
  • If you add the percentage of the state population below the poverty level to the freshman graduation rates, R-square jumps to 70%. My guess is that the percentage of the population below the poverty rate captures the average quality of the environment outside school.
  • Note that R-squares change if you try to explain the freshman graduation rates using the test scores and the other variables. For instance, only 21% (gasp!) of the freshman graduation rates is explained by the test scores of eighth-graders, although you would assume that students drop out because they are not doing well in school.
  • 57% of the variability in high school graduation rates is explained when you consider both test scores and percentage of state population below poverty level.

I was still trying to decide whether the tests are useless (shouldn't they predict graduation rates a little better?) when I stumbled upon the 2005 assessment results of the Nation's Report Card for eighth-graders in US public schools. But that will be for my next post.

The right kind of entrepreneurship?

The World Economic Forum ranks the US at the sixth place for competitiveness (behind Switzerland, Finland, Sweden, Denmark and Singapore), and at the seventh place in technological readiness (behind Denmark, Sweden, Singapore, Finland, Switzerland and the Netherlands) in its 2006-2007 reports, down from the first place in both categories. The size of the US economy compared to those other countries takes a bit away from the impact of the news, but it should be a wake-up call nonetheless - while the US have the best universities and research facilities in the world, many countries are doing a better job when it comes to bringing technology and innovation to the general public.
This made me think about the state of entrepreneurship in the US. After all, if the country that invented the Internet (thank you DARPA) is falling behind in promoting access to its inventions (according to Internet World Stats the Internet penetration rate in the US is only 69.6% as of January 2007), no other achievement is safe. It is well-known that the culture in the US is highly favorable to entrepreneurship - there is little red tape to start a business, and much less stigma for failing than in other countries.  According to the Small Business Administration, there were 671,800 small business starts in 2005 compared to 544,800 closures, and between 1995 and 2004, an average of 0.3 percent of adults per month became primarily self-employed. The number of nonemployer firms (companies that do not have employees) reached a high of 19.86 million, and in 2004 10.2 percent of the workforce chose self-employment. I could keep on going but you get the idea: seven years after the dotcom crash, entrepreneurship is doing very well. And yet, this begs the question: is the right kind of entrepreneurship flourishing? Americans love to be their own boss, they say, but during the heyday of the dotcom bubble few had any reluctance in teaming with others to give instant fortune a shot. So self-employment might well be a response to longer and longer workweeks in corporate America: people might be dropping out of the rat race to have more control over their schedule, as a way to enjoy their work (and life!) more. While there is nothing wrong with that, a striking feature of the most important scientific and technological advances in recent memory is that they all involved some level of teamwork. That is obvious in the fight against cancer but also in the rise of open-source software and the development of high-tech consumer goods, the idea being that if only one person can do it in a reasonable amount of time, it can't be that complicated, and hence that life-changing for the rest of the population. So is the growing number of self-employed and nonemployer firms a testimony to the strength of entrepreneurship in the US, or the sign that entrepreneurship, like technological readiness and competitiveness, is under threat because evolving in the wrong direction (in part because of the trauma of the dotcom crash), while other countries can promote small, nimble companies with a critical mass of employees and faster response times?