The Rational and Irrational

The last two weeks I've been in the midst of a move, soon to take on the mantle of Dean [edit: I'll be Dean of Academic Support services, including responsibility for the library, IT, IR, faculty professional development, and service learning] at a small urban HBCU. I've gotten to enjoy the three things I enjoy least: moving, painting, and plumbing. The biggest moving challenge was around 200 boxes of books, half taken from the lovely library (pic) I built only a few years ago. The contractor hasn't been returning my calls, so it may be a while before the new one is built.

I've tried twice now to get a new driver's license, but the DMVs were full to bursting, and I'll have to get there first thing in the morning or else spend the day waiting apparently. I always keep a book handy for such occasions. It used to be A History of Western Philosophy by Bertrand Russell--the only such history I've enjoyed reading. Now I keep The Prince by Nicolo Machiavelli in the car. Both can be browsed easily. Lately I've been toting around a book I found while unpacking, called The Evolution of Cooperation by Robert Axelrod (wiki). I had bought it years ago when I did a session for high school counselors during a seminar on ethics. Most of the other sessions were touchy-feely. Mine addressed cold war strategies and TIT FOR TAT, a game theory strategy that Axelrod makes much of. I recommend the book, and while you're at it, The Selfish Gene (wiki) by Richard Dawkins, which addresses game theory in the context of biological evolution.

The argument is made that in relationships with others, the most robust (or at least one very robust) strategy is to have a short memory about past interactions, and to reciprocate immediately both cooperation and betrayal (called defection in the book). In terms of the workplace, this could be something as simple as being honest with co-workers. [aside: I always put the hyphen in co-worker, because otherwise it might be interpretted as cow-orker. I'm not sure what that is, but I wouldn't want to be one!] If initial trust is reciprocated, then the relationship works to the advantage of both parties. On the other hand, if someone is less than honest, TIT FOR TAT would have you reciprocate. In this case, probably not by returning the same behavior, as that would ultimately be self-defeating, but by exposing the behavior or otherwise addressing it head-on. One of Axlrod's main points is that reciprocity of negative acts can 'echo' in destructive cycles. Think of acts of ethnic-based violence, for example. He gives insightful examples of life in the trenches of WW I, where opposing units would often come to implicit terms of temporary truce. Violations of the truce were dealt with immediately and harshly.

One of the most interesting results from game theory is that in short-term relationships, the rational strategy is to betray (or defect). Only in circumstances where you will interact for an indefinite number of times with others does cooperation make rational sense. In human societies, this is ameliorated by the fact that our reputations follow us. But witness the behavior of people who are anonymous (as on Internet message boards) to see evidence of the unfettered behavior emerging.

Another book, this one on my Amazon wish list, is Predictably Irrational by Dan Ariely. I found an outline of it here that summarizes the main points. There are several bits here that are of interest to marketing higher education. These have to do with price and perception. The first is the power of comparison. An example is given of a bread-maker that wouldn't sell until the manufacturer came out with a 'deluxe' model, which made the basic model seem like a bargain. At the same time, higher prices are associated (arbitrarily, it seems) with more value. One institution I consulted for raised its tuition dramatically over three years just to be in the same league as the ones it aspired to be. That is, the increases were not financially motivated, but for marketing purposes.

The second chapter--on anchoring--has some powerful lessons for marketing higher ed, I think. Again, the theme is perception becomes reality. Black pearls are worthless until they're stuck in a pricey display window with a ridiculously high number on the tag. Since perceptions of prices in higher education are already high, the strategy would be driven by competition locally or within a niche. The lesson of Starbucks (given as an example) is to create something that's the same but looks new--to create new expecatations of price and experience. If there's a recurring theme, it's that perception drives experienced reality.

Finally (for this post), there's the idea that 'free' is often overpriced. That is, people often value something that's putatively free disproportionately over merely cheap. Amazon.com's free shipping is an example of a success story. Free online applications to college (waiving the normal fee) could be another. At my new university, students get a 'free' laptop--never mind that they pay a hefty technology fee. It makes me wonder what else we could package that way...

Challenges with First Generation Students

We've run an attitude and behavior survey for the last three years, which distinguishes on one item whether or not a student's parents attended college. Testing means on the other items using this as the key yields some interesting characteristics. Sample size is around 350 of each type.

First-generation students are less likely to be happy with campus conditions, quality of instruction, use the library, like the general education curriculum, read for fun, have a mentor, or say that tuition was worth it. They have a lower assessment of their thinking and communications skills. They are more likely to cite detrimental family influences on studying, say that money problems might cause them to drop out, and have chosen the college as their first choice. The picture of first-generation students is one of generally unprepared for the financial, culture and intellectual requirements of the academy.

Plans to deal with this division are well under development. Interestingly, I got an email today about a similar plan already implemented at Manchester College, a private school in Indiana. I found the article online here. The elements are the same as the one we are developing: assure parents and students of a successful transition to and through college into the workplace (or graduate school). Their students are probably more prepared than ours, judging by their graduation rate. We have the additional challenge of bridging the cultural capital gap for the first-generation students.

Elements of a successful plan probably include addressing marketing (internal and external), adjusting financial aid policies to include continuous review, academic support services, adjustment of the curriculum (to rebrand the liberal arts, in our case), and career development. In our case, it's particularly hard to market a liberal arts product to first-generation students who have no idea what those words mean. A major part of the plan is repackaging those ideas in a more understandable way. It also requires an honest re-think on our part, asking if we are really delivering what the implied promises are.

Content Management and Portal Intergration

One of the biggest headaches for any reporting unit like institutional research, is maintaining web sites that push out information to constituents. On a bigger scale, the whole institution has a similar problem. Nowadays a content management system (CMS) is pretty much required to maintain sanity. These packages, like drupal or joomla (both open source), will help separate design from content--a big step in the right direction. But that still leaves the problem of generating reports (often dynamically) and getting them on web pages in a suitable format.

In the past I've relied on custom code for this sort of thing, but that is not maintainable past a certain scale. So I was interested to read an article in Campus Technology about a product called bluenog that integrates a CMS with reporting functions (they call it business intelligence, but it looks more like reporting to me) and a portal designer to boot, all integrated. The demo videos look interesting, but don't give a lot of detail. I found a review here (which led me off to read about other interesting web 2.o worksavers like this). The Bluenog company has an interesting business model--they sell you the software, but then give you the source code. I'm downloading it to give it a spin tonight. It's over 700mb for the full (trial) version.

Update: After installing the trial version of their ICE (integrated system), and playing with it for a while, I think it's more complicated than what I'm looking for. I can see using this in a very large installation, but my head is swimming at the development complexity of the thing. Maybe I've misjudged it, however, and will continue to play around with it for a few days.

Gloom and Doom for Traditional Universities?

There's an interesting article at WebWire that I found linked to U. Bus. predicting the end of the traditional model of universities. This is predicated on the existence of disruptive technologies, that the higher ed industry is mature and incapable of much innovation, and that it is currently overpriced. The following quote attributed to Peter Druckeris is cited from Forbes:

Thirty years from now the big university campuses will be relics. Universities won’t survive. It’s [Internet technology] as large a change as when we first got the printed book…The cost of education has risen as fast as the cost of health care…Such totally uncontrollable expenditures, without any visible improvement in either the content or quality of education, means that the system is rapidly becoming untenable. Higher education is in deep crisis.

As an example of things to come, the author gives us Andrew Jackson University's 'sponsored tuition' program. It's a fascinating idea--pay for the marginal cost of instruction for those students through commercial sponsorship. Here's a quote from AJU President Don Kassner, taken from Wikipedia:

Most universities spend a tremendous amount of money to recruit students. Many spend as much as thirty-five percent of their revenue on marketing and advertising. They have to keep their tuition high to recover these costs. We eliminated these costs by structuring relationships with strategic partners that refer potential students to us. Therefore, we can operate a quality, degree granting institution without the escalating tuition and excessive fees deemed necessary by many schools.

That's a glimpse of the disruptive technologies part. What about inflexibility of the current system?

Anyone who's ever served on a committee will not find it hard to believe that higher education as an industry is not going to change overnight. A study from The National Center for Public Policy and Higher Education and Public Agenda called The Iron Triangle (pdf), which seeks to illuminate the gearworks of change in higher education. The authors quickly get to a central problem: "The various stakeholders must agree on the definition of the problem." Moreover, principles (according to the article) are locked into a mentality of thinking of cost, access, and quality as forming what mathematicians would call a partition--improving one variable necessarily has adverse effects for the other two. It's a zero-sum game, in other words. The stakeholders do not necessarily hold the same view--the industry is seen as bloated an unresponsive. Within the report are found comments by presidents on the subject of cost, access, and quality. These are very interesting and highly recommended for reading in detail.

Not all presidents bought the premis of the iron triangle. Here's a quote from one president, which resonates in light of current events in the auto industry:

Years ago I heard a speaker from the auto industry who asked, ‘What
happened to the auto industry in the 1970s? It wasn’t bad design. It wasn’t
planned obsolescence. It wasn’t unions. Fundamentally, it was hubris. It was
a belief that the American automobile industry had always built the best
vehicles, always would, and that the public would buy whatever we built.
We saw our problem not as a product problem, but as a marketing problem.

Maybe. I think it's a little more complicated than that. Here's one idea to consider.

What if the market for the best students has become more liquid, driving up their price? That would mean that for the same cohort of talented students, universities generate less revenue from them because they're bidding against each other. Who makes up the short-fall? Costs can get pushed down to the need-based crowd, which is what's been happening (see practically any issue of Opportunity for such an argument). This has been enabled by cheap money. So the combination of aid leveraging for talent and easy money could be part of the answer. It would be very interesting to see the financial aid matrices for various institutions over the last ten years, to see how they've evolved.

Education Feeds

I had some more fun with yahoo pipes mashed up with a google docs spreadsheet, to take the top two RSS feeds from five higher ed news sources, and embed them below. It's not beautiful, but it's live (will update when the feeds do). I'll work on it some more to improve the asthetics, when I get a chance. Sources: insidehighered.com, New York Times, The Chronicle of Higher Ed, Wired, and University Business.



I got started using this blog.

Virtual Loans

I mentioned an idea in the last post about institutional loans. A better word might be virtual loans because the money doesn't really exist until it comes due. That is, it's a loan with no capital at risk. You'd think this would get the attention of a bank or two, but in these times who knows.

Direct loans from the government are limited in amount, and subject to other kinds of restrictions. Loans in general are likely to be harder to get for students in the current climate. I've argued that the easy money of the last decade has allowed tuition to rise in its own bubble along with housing, and that there will be a return to Earth in short order. Virtual loans are possibly a way to ease this pain.

Here's how it works. Most institutions give out institutional aid that is a simple discount from tuition. We don't ask anything in return for this discount, and it's used to get students with the kind of ability we want to attend, or (to a lesser extent) help students who can't afford the cost. For the latter group particularly, it's probably a good idea to tie the aid to student work. This helps alleviate the need for part-time workers, and gives the students a point of engagement with the institution that should help with their retention. But I digress.

Instead of a no-strings grant from the institution, why not ask for something? Work study is one idea, but there are lots of others. Future participation in alumni activities, recruiting, etc. are some ideas that could be the strings attached. Another one is repayment for a part of the total aid. That is, the institution "loans" the student money to make up part of the tuition instead of granting it directly. Since this money never really exists, the terms of what I'm calling a virtual loan could be quite generous. How about a zero-percent loan, deferrable two years after graduation, and paid out over 10 years? Even if many of these default, this program only has to pay for its own administration in order to break even, since there's no real capital at risk. The goal would be to provide a future revenue stream with no present capital outlay.

Even better would be to involve a bank in the processing of the virtual loan. Banks do loans all the time (well, they used to, and this will give them something to do). They have the regulatory infrastructure in place, and are set up to receive payment with no additional bureaucracy needed. This should lower the administration cost. If you know a banker, ask him or her on what terms would he or she loan money at no risk.

There may be some legal reason why this all can't happen. I should have opportunity in the near future to explore this idea with people who know more about the regulations than I, and I'll find out if this idea is feasible. Maybe someone is already doing it. The virtual loans wouldn't do anything to alleviate the current financial woes, but would start to create a new long-term revenue stream that would start to pay off in a few years. With judicious planning, an institution could be audacious enough to eventually offer these kinds of loans to the extent that they replace traditional loans, and thus secure a favored place in the market. That is, if the virtual loans were managed so that they push out more traditional loans, it would be very attractive to students because of the much lower interest rates. This could only happen by institutions weening themselves off of the interest-bearing loans over time, but in principle there's no reason for that to be impossible.

The fact that virtual loans are not ubiquitous seems to me to be an inefficiency in the marketplace. But it may be that the conditions for such loans are not favorable enough to warrant the administrative costs. If the default rate is near 100%, for example, it might create more problems with alumni (who often voluntarily give money) than it is worth.

Financial Aid Woes

The front page of InsideHigherEd this morning has an article about Syracuse University, and the impact of the financial climate on student appeals for financial aid.

Syracuse University recently sent an appeal to alumni warning that approximately 400 students will be unable to return for the spring semester unless the institution can raise an additional $2 million in scholarship support by the end of January 2009.

They are putting out the call to donors for additional aid to bridge the gap. I tried to find a figure for the university's discount rate, and stumbled on an old self study that puts the 1989-90 rate at 10.5%, and describes the financial aid leveraging process they're putting in place.

An innovative awarding and measuring tool, based on econometric methods, factored in student academic qualifications and family financial need. This model enabled the prediction of enrollment outcomes and guided the University to a considerably more competitive pricing position. In the course of its implementation, the undergraduate tuition discount has increased from 10.5% in 1989-90 to 37.8% in 1996-97. While this approach has accounted for a considerable shift in net available revenue, it is also credited with helping to stabilize the quantity and quality of undergraduate enrollment. The undergraduate tuition discount rate is expected to level at about 38% in 1998-99 and remain there for the immediate future.

That puts the rate at 37.8% ten years ago, after a dramatic-sounding reformulation of aid policies. The InsideHigherEd article quotes $16,737 as the average award, but I wasn't certain from the context if this was all institutional. Apparently it is, because the IPEDS report tool puts last year's average institutional award at $17,136. That would be a discount rate of a whopping 52%, far in excess of the 38% target. This assumes, however, that they are not including other fees and such when they calculated the discount previously.

What to make of the plea to alumni for additional aid? This looks like a short-term emergency response, and as such might be quite reasonable. I would also argue, however, that they should also be taking a hard look at their leveraging model. All of us who give institutional aid in order to shape incoming classes will have to do the same. The basic calculation, is: what do you gain by letting these students walk out the door? With an average net revenue (tuition-institutional aid) of $15,443, the price tag would be almost $6.2M if all 400 left. Assuming that the requested additional $2M in aid is the right number to keep these students, it seems fairly obvious that the university is better off creating the aid out of thin air if it can't get the donations, rather than letting students drop out for financial reasons. Adding 2M to a 155M aid budget would be a small price if that's the total impact of this recession.

The article also mentions loans and the inability of students to get them as an impediment to attending. I tested my idea about institution-based loans with our Business VP yesterday. It seems like an idea worth pursuing, and I'll follow up here when I get a chance.

GRE Trend

There's an article in insidehighered.com that says "The number of students taking the Graduate Record Examination will decline in 2008, the first time ever that the GRE has seen a fall in test-taking during an economic downturn."

I wondered what google trends would show about searches for GRE. See below.

Interestingly, the number of searches has been averaging less and less over the years, despite the article's assertion that this was the first year they'd seen a drop in actual tests. I wondered if the results have some bias, and tested that by asking for statistics on searches for toys.
Nope--this seems as regular as a heartbeat, with a spike every holiday season. Searches for 'graduate school' have trended down, while searchs for 'jobs' have trended up. Steak seems to be trending down a bit, while potatoes are perhaps on the rise. Try 'wedding' and 'divorce'--Atlanta seems to be hard on marriages.

Google Trends

Would you like to know if your students or applicants are more concerned about tuition, aid, or loans this year? One global measure is how many times certain terms were Googled. Amazingly, this information is yours for the asking at google trends. You can restrict to geographic locations and time periods, and ask for several terms at once. An example is shown below, where I requested "tuition" and "financial aid" searches in the USA.
Note how regular these patterns are. Other trends are apparent too. There are more searches for tuition than financial aid, and the latter tails off in the winter. Also, searches for both terms have increased over the last years, but not dramatically. This could be due to any number of things, and one can get an idea by searching for "best college", which also shows an increase. It's probably a good sign that the charts don't change dramatically for the current cycle. If you look at "unemployment" on the other hand, it's a different story. A report at the bottom shows which locales most frequently googled the term. Fascinating stuff, and potentially important for your marketing.

A previous post mentioned a debate about SAT. Its chart is fascinating. Whereas the search trend is to decline slightly on average, the news volume (bottom graph) is clearly trending up.
The same thing (even more dramatically) is happening with ACT. Of course, "act" is another word that might be googled frequently, so that's more ambiguous.

As a bonus you can link to the results and export as a CSV file. From there, we could link to google documents or yahoo pipes. You could, for example, create a dashboard-type feature to show how many times your school's name was googled in your state (or was in the news), and graph it on an administrative portal beside applications received and awards granted.

The Wealth of Majors

I've argued before that average graduate salaries would be interesting to use in program and institutional assessment. I've thought that the government ought to report out these average salaries for graduates of an institution to use as an outcomes measure. They have all that data anyway--why not use it for something? I've even made this pitch to people who work in the Department of Education, but I invariably get a stony response.

There's a report shown in the Wall Street Journal showing data collected by a private company to accomplish the same thing. Take a look and see if this isn't useful information. I wish that for comparison's sake, they'd included people who never finished college, or never finished high school.

Disambiguating cause and effect is difficult, of course. Impossible with the data shown. We don't know if MIT graduates earn more than SIU graduates because the school taught them better or the students were better to begin with, or if perhaps they settled in locations nearer their schools, which have different economies. It may say more about what kinds of people attend what kinds of institutions and major in what kinds of subjects than it does about the effectiveness of a program of study in the eventual production of wealth. Still, this kind of information is useful for parents and students considering their options. The fact that median salaries for philosophy majors is slightly more than that of marketing majors makes me happy.

If the government really were to do such a study, using tax records, financial aid data, and clearinghouse data, some clever person might be able to find a control group in order to untangle causes and effects. A good salary shouldn't be the sole end of education (see this book on that topic), but it's an important factor, and undoubtably useful.

Yammer

I heard about Yammer the other day on NPR's All Tech Considered segment. It's like twitter, except for the workplace. Some people swear by twitter, but I haven't been able to see the use of it myself, given the time investment seemingly required. (I know--I probably just don't get it.) Anyway, yammer seems like a solution to a problem.

What problem? Well, spam in short. Not the external kind that clogs up the Internet in general, but internal spam. At a large enough institution, it's an example of a tragedy of the commons. Way more than half of my emails are stuff I don't particularly want to read. If it were tagged with a minimal amount of meta-data then it would be much more useful to me. For example if a message came with standard tags I could instantly decide whether or not I'm interested in reading it. Better yet, if I could just let it blow by without even entering my In-Box, so much the better. That's essentially what yammer does. It's like sending out a chat on a group channel, but with meta-data possibilities. It's simple to embed a tag: just use a pound sign. For example:

Trying to learn the yammer API #web

would tag the message with the web keyword. Users could elect to follow that thread, or even create an RSS feed from it. There are also groups. I'm not sure what the relative advantages are of the two options, other than that groups can be made private.

There is a desktop application that's very pretty. It runs on Adobe's AIR platform. A picture is shown below.
It's very pretty, as you can see.

Is this really a company-wide spam killer? That remains to be seen. I'll first test this in my areas of responsibility and see how it works. With any luck, others will become interested too.

Yammer is free for basic functionality. If you want more administrative functions, you have to pay.

Edit: I can already tell this isn't really a tool to eliminate much internal spam. It generates more traffic if anything. It's for a different purpose, and I will explore this (you can actually see what the creators envision on a short video on their site). I'm thinking that perhaps posting mass mailings to a portal message board is the solution to the spam problem.

Cost of College

There's an article in the New York Times about the decades-long rapid increase in the cost of higher education. Another one is on cnn's website. Both refer Measuring Up 2008, The National Report Card on Higher Education (pdf) from the National Center for Public Policy and Higher Education. In the Times article, the Center's President is quoted saying:

The report, “Measuring Up 2008,” is one of the few to compare net college costs — that is, a year’s tuition, fees, room and board, minus financial aid — against median family income. Those findings are stark. Last year, the net cost at a four-year public university amounted to 28 percent of the median family income, while a four-year private university cost 76 percent of the median family income.

I found median family incomes by state here, and looked up our state to find that it's about $53,000. According to the statement above, a private university would have a net cost of around $40,000. We're nowhere near that. In fact, family contributions are less than a quarter of that figure. But we have also not raised tuition much in the last decade--certainly nothing like the graph shown in the article (reproduced from the original report below).

In 1995 our tuition was $10,488. Now it's $18,792, but discount rates have increased too. Leaving that detail aside, it would be an increase of around 79%, or 4.6% per year average. The graph shows an increase of about 293% in the same period, or 8.6% per year. The point of this is that the averages conceal considerable variety.

Loans increased by over 100% in the last decade, according to the report (Stafford Loan borrowers increased by 50%). This leads naturally to the speculation that tuition prices have been inflated by the credit bubble along with housing. With loans harder to get, the grim calculus isn't hard to compute: some combination of fewer students enrolled and lower net tuition paid. This will reduce net revenue to many institutions.

What to do? Undoubtedly, there are efficiencies to be found at any institution. The luxury of NCAA division II (scholarship athletes) will be under pressure. Division III nominally has no athletic awards, which adds up to a lot of money quickly. Other ideas include:

• Instead of direct institutional aid, require some work-study for the money. This saves on part-time employees.
• Instead of direct institutional aid, make institutional loans. I don't know of anyone who does this, but it worked well for GM for a long time. Small institutions could form a consortium for this. The way it would work is to loan students money with repayment starting at some point after graduation. The money loaned isn't real--it's just discounted tuition, so no real money changes hands until repayment. There are probably legal issues here I'm not aware of, but it would alleviate the loan crunch and also let colleges eventually generate interest from these loans. Maybe later on they can divide them up into tranches and sell credit swaps (just kidding).
• Add or increase after-work classes for working adults. The infrastructure for a college or university is large and expensive. Using classrooms at night and hiring instructors is a small additional cost for the additional revenue generated. Flexibility in tuition structure can make it easy to meet demand with supply.
• Online classes are trickier because of the competition. This may or may not be a good investment for the institution, but even traditional or evening classes can be made more efficient by using a hybrid approach. If some of the class meetings are online, there are additional types of learning opportunities, as well as the potential to reduce some costs and make it more convenient for commuters.

In any event, I predict a growth industry in education consultants to come and tell you how to make your operation more efficient. This give the administration someone to blame when they eliminate positions.

Dehne Visit Notes

We had a visit from consultant George Dehne of GDA Integrated Services today. A lot of what he said resonated with our retention research that led to the Plan 9 project. Specifically, he said that first-second year attrition is generally a sales problem, and after that it's programs. A small rural liberal arts college has three strikes against it, according to Dehne, because the so-called Millenial student is looking for an urban experience (lots to do), doesn't care much about size, and doesn't understand what liberal arts is (maybe it involves crayons?). To underscore the second point, he said that their research tells them that 65% of applicants don't know how big the institution is that they're applying to. But generally, preferences for size have grown from an ideal of 2500 ten years ago to about 6500 now.

One of their specialties is finding a distinctive element (big idea) around which to rally marketing and delivery of the educational product. The big idea is the project of brainstorming, involvement with the institution, and market research. I think we're capable of doing most of the first two parts, but aren't in a position to do our own market research. Part of the idea there is to cast a large net and find out what part of the college-bound pool isn't applying to our institution. Maybe we're missing part of our ideal market. The other part is to test out these
"big ideas" before adopting one, to make sure it really works.

One interesting idea that worked at another school was re-branding the core requirements as a major--every student has two majors (they value majors, and don't value 'general education', being the hypothesis). Another interesting idea was having first year seminars on topics like The culture and history of football.