Next month:
April 2007

March 2007

Should teaching be a part-time job?

It is no news that the baby-boomers are set to retire. It is no news either that this is going to change the face of employment. A few months ago the Economist (which is, as you might have guessed by now, my favorite reading of all time) published a special report on manpower in its issue of January 4th, 2007; it suggested that some of the baby-boomers would consider part-time work rather than plain retirement, for instance to maintain their standards of living. We have all seen those customer service employees at Wal-Mart or at airports who appear to be well into their seventies - the first wave of part-retirees might well have already begun. Of course some people need two (or three!) jobs to make ends meet, so part-time for the company and part-time for the employee are two very different things.

A friend of mine taught a course in electrical engineering at a university in the San Francisco area last year - she has a well-paying industry job, but she wanted to teach. Of course the emergence of life coaching and the proliferation of yoga studios throughout the country (in a manner not unlike Starbucks before the market saturated) suggest that more and more individuals today want to live a life that matters. This urge is now taking new forms, far beyond the usual "coach my son's Little League team" or "donate 10% of my earnings to charities". Many educators try to encourage college students to become schoolteachers, but it is not a choice many make - it is a tough job, even in schools with modern facilities and good students (briefly stated, you don't always make the impact you signed up for). Science and engineering majors in particular can easily find far better-paying jobs and, quite frankly, pay does measure how much society in general values certain groups of individuals. A one-way street, in my opinion, no matter how much lip service people pay to their children's teachers - which might not even be much, given the growth of the homeschooling movement. Part-time teaching would allow people to give back to their community and neighborhood without losing their primary professional identity. It could become some kind of "paid volunteer work" - you would not be doing it for the money. Maybe K-12 teaching will become the  next global warming: the issue failed to gain traction for years and suddenly we are all watching "An Inconvenient Truth". Twenty years from now we might all spend a few hours a week in front of a blackboard.


The Customized Book

There is an article in this week's Economist (dated March 24th) about the future of books, and in particular about how people will read now that so many books are being digitized (thank you, Google). The article states: "Change is least likely in the physical medium of books. [...] The biggest changes are likely to be seen in what becomes a book in the first place." While there is some truth to the second part of the claim, I very much hope future developments will prove the first part wrong. I'd like to see print-on-demand become a more widespread practice in the publishing industry. Of course these days it is very similar to printing a PDF file: you don't get any choice. But getting a Word file instead opens a world of possibilities. One could imagine readers changing the font type and size, or the margins, or the binding, of the books they want to purchase before sending the job to the printer. (For instance, I prefer Garamond to Times New Roman, and I think large margins are a waste of paper.) The printed matter is much more than the succession of ink patterns on the page, and with everyone questioning the future of the physical book, it is high time this relic from the Gutenberg era started differentiating itself from its digital counterpart. The difficulty lies in production: having different paper types requested for each book makes it necessary for a real person to change the paper in the feeder annoyingly often. For now this is obviously not commercially viable (although it would allow printers to move to a high-end segment of the market and increase their prospects of long-term financial viability). But in the same way as people pick options for their car, I hope that soon they will be able to pick options for their books.   


More on grade inflation

GPAs haven't stopped their upward trend since my post yesterday, so here comes the second batch of valuation strategies in higher ed.  The widespread explanation for the trend is that students take a consumerist view of education and - as customers - must be satisfied by the product they bought when they enrolled. But if you take the analogy one step further, companies want to know about the product they buy when they hire students straight out of college. If universities aren't able to provide that information through GPAs, then a response to grade inflation would be for companies to take the lead in assessing students who apply for their jobs. And since companies are not necessarily able to get this extra work done in-house, this opens plenty of opportunities for human resource companies in the US (new opportunities overseas, in particular to address the manpower shortage in China and India, are described in the January 4, 2007 issue of the Economist.)

So what can be done to help companies measure students' performance if GPAs no longer seem to do the job? Students interested in finance positions in New York might be invited to take the GMAT, for instance; others will tackle with the GRE Subject Tests, which, as GRE's parent company ETS explains, "gauge undergraduate achievement in [eight disciplines: biochemistry/cell/molecular biology, biology, chemistry, computer science, literature in English, mathematics, physics, psychology.] Each Subject Test is intended for students who have majored in or have extensive background in that specific area." While the subject tests' primary purpose is now to evaluate applicants to graduate school, one can imagine additional uses that would open much more sizable revenue opportunities for ETS. After all, the tests are already there. Another avenue would be for companies to partner with testing services to develop evaluation processes tailored to their needs. (Perfect fit comes with a comfortable premium.) Consulting firms already rely on case studies to judge applicants, and other businesses might enjoy knowing that a student knows the basics of the function he is hired to perform. A drawback of this approach is that it favors students who already know at the expense of those who learn fast. (The testing companies would reply that this is not a problem, they can devise a test for that too.)

But this leaves us with the question: does grade inflation matter? A company which interviews only students whose GPA exceeds a certain threshold can easily keep track of the number of job applicants on each campus. When the recruiters notice a substantial increase of eligible students, they will raise the threshold. Universities with lots of students achieving Dean's List status will raise the threshold as well, and will start keeping track of GPAs with, say, four decimals or out of 100. A 4-point scale to measure GPAs will soon seem as outdated as floppy disks.  If Stojstaczer's predictions are correct and GPAs keep increasing by 0.146 per decade, the average GPA at Harvard will be 3.78 by 2024 (from 3.42 in 1999) but its tuition, which has been rising by 4.5% a year on average, will be over $70,000 a year. I'd like to conclude grade inflation will be the least of college kids' (and their parents') problems 20 years from now; however, the sad truth is, we will probably have gotten used to $70,000-a-year tuition by then.   


Valuation Strategies in Higher Ed

It is no secret by now that grade inflation has been pushing all grades towards As. According to Where all grades are above average, an op-ed piece by Stuart Rojstaczer in the Washington Post dated January 28, 2003, and its supporting documentation, the average GPA in American colleges and universities has increased from 2.94 to 3.09 between 1991-1992 and 2001-2002; these numbers are typically accompanied with gloomy predictions of As showered on the whole student body by 2050. But one item is missing from this picture - GPA is just a proxy of how well a student learnt what she was taught during her four years in the higher education system. Quantifying a student's proficiency matters not only to the kid in question and her proud parents, but to the companies which are interviewing her. When the means no longer relates to the end goal, there is not much to do besides picking another performance metric.

The fact that performance is not much captured by letter grades (as happens if almost everyone gets an A) does not mean that professors are unable to compare students. For instance, professors can assign a numerical grade to papers and convert that into a letter grade at the end of the semester. If GPAs stop fulfilling their purpose, it becomes necessary for recruiters to ask the academics directly what they think of this or that student. Faculty members do not have any incentive to distort their evaluation; depicting mediocre applicants as super-stars will reflect poorly on the university once the student starts working, and companies unhappy with the recommendations they receive can threaten to stop recruiting in that department. Because graduates' placement record is often tracked by the department (with various degrees of care) and much touted to prospective majors, this is a potent threat that encourages professors to be honest. A positive side effect of putting faculty members on the spot is that it encourages them to know the students enough to provide thoughtful comments to recruiters (which does not mean much, but putting names on everyone's face in a section would be a good start). In essence, this is the "graduate school application" model, where each student must provide 3 or 4 letters of recommendation from faculty members.

Companies, however, might want a fuller picture of an applicant's potential without having to contact 15 different faculty members. Ranking students precisely is fraught with fairness issues (when do you assign the same rank to two students?) A possible solution becomes to supply the Registrar's Office with numerical rather than letter grades, have it compute an average percentile for the student's performance and summarize this information to (for instance) honor students in the top 10% and top 25%. (Note that the numerical grades can also be used to compute letters and then GPAs.) In other words, you want to recognize that most students perform similarly while rewarding students who are above the pack.

The inconvenient is of course that companies would depend on universities to implement change, and such solutions offer not much more than a tweak of an old system. I'll discuss company-centric ways to measure students' value in my next post.


Creative Engineering

Creative engineering sounds a lot like creative accounting - taking liberties with well-established rules does not quite seem the best way to build a bridge, or a house, or a car. And yet the creativity of engineers is much-touted in the recent rebranding of the profession. (See for instance A call for K-16 engineering education.) How can that be? The most glamorous part of engineering is certainly not to replicate other people's work or follow someone else's blueprint. The engineering process shares many similarities with mystery novels: set a end goal (say, create a product with some specific properties),  run an investigation (select possible materials), question the prime suspect (perform tests), review the clues (analyze the test results), question the other suspects (perform other tests), arrest the culprit (determine the material to be used).

But as anybody who has ever watched CSI will tell you, you can't make the most of the information you have unless you know that, for instance, these three elements combined represent that product. At an even more basic level, you need to know how to run the tests to actually gather information. In other words, science and engineering require extensive training before you get to the "cool stuff", in contrast with English or business (you can write a book without being an English major, you can start a company based on a sudden idea). And training, needless to say, is not creative at all. Yet engineering is creative, and not just for students who go on to graduate school - not in the process, which is a good thing, but in the steps, the observations to be taken, the choices to be made.

Sadly the creativity factor might not be enough to motivate high school students to join the science and engineering ranks. In a country which cherishes self-employment, science and engineering suffer from two major drawbacks: 1) they are not portable (a lab in the garage doesn't necessarily sound like a good idea, and equipment is expensive) 2) most of the ground-breaking work is done in teams (just look at the number of co-authors on papers published in Science or Nature.) And it will not have escaped the attention of the non-self-employed part of the population that research groups have a very flat structure (the scientist, his colleagues, his postdocs, his graduate students), which makes the engineering world unappealing to the hierarchy-crazed as well. So if you want to do science or engineering, you need to find someone to give you money and/or equipment, you need other people to help you do the job, and you just have one layer of hierarchy underneath you. This does suggest that most engineers are not too high on the totem pole of large companies.

But that is because for years engineers have focused on producing new toys, and marketing people have focused on finding use for them. The lukewarm interest of US students for engineering and science majors compared to their European and Asian counterparts (Educating Engineers for 2020 and Beyond) has motivated a shift in the traditional training of the engineer, in particular to integrate management courses in the core curriculum. Engineers are no longer just, say, computer geeks but accompany the products they invent longer in the development pipeline; they are better able to articulate their inventions' purposes in front of the puzzled marketing people, and they can also think big rather than marginally improve an existing object. Engineers have become mutants, or caterpillars - they take new form after a few years, and move on to management. Not least creative in the engineering path is the turns it takes along the way.


Competitive Engineering

Everybody remembers Sputnik, even people who weren't there. More seriously, everybody remembers Sputnik, even people who know nothing about engineering. Putting a man on the moon was a direct reaction to the Soviet challenge - a matter of wounded pride, of superiority to reclaim. (The Manhattan Project provides another example of breakthrough engineering via crisis management.) There has been some effort in recent years to foster a similar attitude towards global warming, but the public has been slow to internalize the sense of urgency, at least until Al Gore's "Inconvenient Truth".

If the US need to feel threatened in their engineering supremacy to be jolted into action, where will the next threat come from? The result should be clearly observable (there is a satellite in the sky), it should be meaningful for a large part of the population, and the success of the enterprise should be beyond doubt. A South Korea scientist briefly epitomized such a challenge when he announced he had cloned human embryonic stem cells in 2004-5, but since some of the data had been manipulated, the most recent defining moment in engineering remains Game 6 of Deep Blue vs Kasparov (Deep Blue won). One could imagine politicians up in arms, falling all over themselves to throw money at the National Science Foundation, if a super-computer powered by, say, Indian or Chinese algorithms were to beat an American super-computer at a chess game. More realistically, my bet is on nanocomputers, health care or emergency services. Nanocomputers - the technology works or doesn't work. Health care - maybe gene therapy. (A cure for cancer would be wonderful, but any cancer survivor will tell you that true success is difficult to define.) Emergency services - a global epidemic or a natural disaster that other countries would reply faster to.

In the meantime, DARPA is promoting some competitive spirit of its own, and is offering a $2 million dollar prize for the development of new autonomous-vehicle technology in its Urban Challenge. Remarkably, DARPA does not impose any citizenship requirements on the team participants, with the exception of the team leader (who must be a US citizen). The X Prize Foundation, which counts Google's Larry Page as one of its trustees, is offering $1 million prizes on topics ranging from spaceships to genomics, in what the Economist called "one of the most intriguing trends in philanthropy: promoting change by offering prizes" (issue dated March 1st, 2007). It is unclear whether teams always know of each other's existence before the prize is awarded, though - the teams involved in the genomics contest do not appear to meet at a single, specific location, for instance, and the adrenaline factor will not kick in as much if they don't experience any outside pressure. As the Sputnik episode demonstrated 50 years ago, some of the most extraordinary technological advances are achieved when, and only when, the leader feels the breath of the runner-up on its neck.         


More on providing access

In my previous post I discussed how magazines provide their readers with access to experts through conferences (not out of the goodness of their heart, but to increase their revenue). Many of them also provide another type of access: they supply advertisers with detailed knowledge of their readers. I am referring to the trend of offering vanity memberships into "preferred readers' councils" (with the promise of special events, advance information, etc) if the readers agree to fill surveys and receive samples of advertisers' products. I was reminded of this when I saw an ad in the April 2007 issue (page 136) of Yoga Journal (of all places): "Join our free Reader Council and become part of an exclusive group of Yoga Journal readers. From time to time, we'll invite you to participate in surveys, send you product samples, and enter you in Yoga Journal sweepstakes. All of this, just for sharing your opinions."
The idea is not new; again, the New Yorker has long been implementing a similar revenue management technique. When I gave the New Yorker an old (then still valid) email address about two years ago, it promptly sent me all kinds of advertisements, including an invitation to join a select group of readers given favored treatment if I filled a survey. (The keywords, for the councils as for the conferences, are becoming familiar by now: access, exclusive, preferred, select...) When things seem to good to be true, they often are, but out of curiosity I did follow the link to the survey. It consisted in page after page after page of questions about the products I use (shampoo, toothpaste, etc) - a gold mine for advertisers, who will rightly pay an exorbitant sum of money for such precise information on individuals. I closed my web browser after two pages and with the end nowhere in sight, but many people will give away personal information if the rewards are big enough. Maybe magazines will soon propose trades: I'll give you free access to subscribers-only content on my website if every week you tell me what you think about this or that product, and I'll discontinue your access if you don't. Testing the goods of the New York Times' advertisers and answering a battery of questions might seem a small inconvenience if we get to read the newspaper's columns too.


Magazines' new business model: providing access

A few weeks ago I noticed that the New Yorker's newsstand price had increased from $3.99 to $4.50 - I guess the price of paper is to blame. It seems that the magazine benefits from highly successful revenue management practices, though; the New Yorker festival was sold out last October. (The festival brings together authors and journalists for paid readings and discussions. Most events are in the $20-$30 price range.) Apparently that was not enough; the New Yorker is now advertising a Spring New Yorker conference, which caters to a whole different segment of the population - the kind willing to pay $1,200 for one day and two nights with the cognoscenti (the subtitle of the conference is: "2012: Stories from the near future" - basically the idea is for the New Yorker to provide access to tomorrow's brilliant minds and advance knowledge of impending innovations.)

Of course the New Yorker is not the first magazine to use its name to attract conference participants: MIT's Technology Review holds a conference on emerging technologies every Fall, for instance. (Price tag for the three-day conference and workshop: $1,245. Participants can register for just one day, though. There is an interesting pricing problem hidden in there: how do you price access to information and people?) This points to a nascent trend in publishing: a number of magazines are establishing a foothold in the knowledge and relationship management business. There is much talk about the impact of the Internet on newspapers, and of online advertising revenues possibly replacing dwindling subscription numbers (see for instance The Economist's report on the newspaper industry in its issue dated August 26th, 2006) but that discussion assumes that the first and foremost purpose of magazines and newspapers in the twenty-first century is still to disseminate news. It is possible, however, that the purpose itself of newspaper companies has changed; the need for differentiation puts a premium on magazines that can "be more" to their readers. The New Yorker, in providing access and analysis, is betting that conferences will provide a strong revenue stream - maybe its main source of profits a few years from now. At $1,200 per registration, they will certainly help defray the production costs of the print magazine quite handsomely.