Each summer the UVM Honors College holds a faculty seminar in mid-August. This year the honors college seminar was titled “Big Data: Engaging and Critiquing the Production of Knowledge in the Digital Age.” This sounded intriguing to me so I applied to attend and was accepted. FreshTracks has a number of investments in companies that are making use of “big data” and/or “machine learning” to solve a customer problem. On Monday August 11th at 9:00 AM I found myself back in the classroom as a student of sorts having spent Sunday August 10th cramming my head full of information from the required pre seminar readings. This last minute cramming is actually not atypical for me as I was at best a lackluster college student at least as an undergraduate. I have never confessed to my children (who are far more diligent students than I am) how poor my undergraduate GPA was. On Monday morning I found that I was one of the few non tenured or tenure track professors attending the seminar and the only “Lecturer” (UVM’s lowest level of instructor) attending the event. I was also the only one with a “day job” outside the University that immerses me in the real world. Seminar attendees came from a broad variety of departments and schools (medicine, geography, romance languages, art history, engineering, math, natural resources) and were an engaging and learned bunch.
We spent considerable time as you would expect discussing theoretical concepts and defining terms. What is big data? Is X big data while y is just data? Is there bias in data? Are some data streams more biased than others and why? Who controls data? Who should have access to data? Then we got into some more interesting territory. One of the readings (Big Data: new epistemologies and paradigm shifts by Rob Kitchin) discussed the changing landscape of scientific inquiry. In the old paradigm that many of us were taught in junior high science, there is a scientific method and scientists make educated guesses and develop hypothesis and construct a model that can be proven by experimentation. A social scientist might study the field of early childhood education for instance and develop a hypothesis that pre-kindergarten training improves students’ readiness for elementary school and that if they are ready for school then they are more engaged in their learning and they are less likely to become juvenile delinquents (I am obviously simplifying things here). The researcher would try to prove this hypothesis by a rigorous examination of available data. In the modern world however, big data can be mined with powerful computer processing power and this can reveal patterns that can prove effects and make conclusions without the need for hypothesis creation or perhaps without the need for humans at all. In that world there is no need for making educated guesses or forming hypothesis and no need for rigorous testing and perhaps no need for highly trained researchers. Simply analyze data enough and the truth will reveal itself. Machine learning can be as good as the human brain. Perhaps the modern citizen scientist will analyze Twitter streams using powerful processing technology available on their cell phone and find a pattern which proves that the theory: “pre-kindergarten training reduces juvenile delinquency” is valid or perhaps they prove the theory is bunk. Obviously I am being a bit facetious here but there are certainly people who are predicting that “citizen scientists” will drive major discovery and innovation in the not too distant future.
Needless to say it’s this last bit that worries my Honors College Seminar colleagues the most. If science can be democratized through better and better computer processing power and more availability of open public data sources then “citizen scientists” may be able to make world changing discoveries. Who needs school, who needs laboratories and who needs funded research (think of the federal tax dollars we would save) and really who needs teachers guiding the way. We can simply empower the average Joe to harness Twitter data streams and the computing power of the next generation smart phone to find the cure for cancer. If this comes to pass then the importance of university and perhaps the importance of higher education is vastly diminished.
It was with this backdrop that two days after I finished the Honors College Seminar I read with interest the recent Atlantic article (http://www.theatlantic.com/features/archive/2014/08/the-future-of-college/375071/) which takes a look at Minerva University a radically different approach to education. Minvera uses the theories of Stephen Kosslyn who taught cognitive psychology and neuroscience at Harvard and has published widely the results of his research on the science of learning. Basically Minerva has built a technology platform that allows a teacher to engage with a class of no more than 19 students in real time and the technology facilitates the teaching based on the science of learning that Kosslyn researched. This is not a massive online course or MOOC and this is not the last generation online technology platform used by most colleges including all the colleges I have taught at. The Minerva technology platform is something quite different. Minerva envisions having dozens of campuses around the world and students would spend their four years rotating among these campuses worldwide so they would get an immersive experience in an international environment but the educational content would be delivered in real time to classes of no more than 19 students via very sophisticated technology and the technology would force the teacher to teach using the best teaching methods. No more boring lectures, no more graduate assistants delivering study sessions, no more inability to measure the outcomes of teachers or determine whether or not learning takes place. Oh and Minerva claims to do all of this for $28,000 including room and board. I note that UVM’s in state tuition, fees and room and board is $29,674 for the upcoming year and its out-of-state tuition (on which it heavily relies), fees and room and board is $51,736. If technology enabled, radically restructured education can deliver a better product for less money, then as a taxpayer and as a tuition paying parent, I am all ears. As an investor this sounds to me like the disruption that has hit other industries during the past 20 years and it seems like an “investable” opportunity.
During the final day of the UVM honors college seminar the talk turned to something akin to “inside baseball” for professors, primarily tenured or tenure track professors. This talk is always fascinating to me because I am blissfully ignorant of these problems. I sign a contract, I teach classes, I get evaluated by my students at the end of the semester and I leave. If my evaluations are decent I might get asked back to teach again. I have no worries about publishing or research or serving on a faculty committee or any of that. My tenure track or tenured faculty colleagues seem to universally hate being evaluated by students at the end of the semester. They don’t agree with the way that universities (not just UVM) determine promotion and pay increases. The professors must publish, they must teach, they must conduct research they must serve on faculty committees that produce reports that the administration never acts on and they must be evaluated by 18 year olds who are ticked off that they are getting a bad grade because they partied too much most of the semester. (In a side note I spoke recently to a colleague at Green Mountain College who said that there are studies that reveal that better looking teachers get better evaluations from their students).
It was with this backdrop that I read the following article in the Chronicle of Higher Education: http://chronicle.com/article/Can-Colleges-Use-Data-to-Fix/148307/,which details a software product called Lecture Tools, which serves as a platform between students and teachers to enhance teaching but also to enhance communications between students and teachers. Theoretically this tool engages the students quickly and provides immediate feedback between student and teacher. A teacher using the tool will more quickly identify problems and be able to take corrective action. On the other hand the proper use of the tool by a teacher should allow the teacher to modify their teaching approach to have the most impact on student learning. I was intrigued until I visited the Lecture Tools website which noted that the business model is one where the teacher gets to use the software for free while the student must purchase a software license. From my perspective (as both a tuition paying parent and as a teacher) the idea of saddling a student with a license fee to help a teacher do his or her job better makes me queasy. Isn’t it the teacher’s job to become a better teacher? Isn’t it the university’s job to properly incent the teacher to become a better teacher? Why should we make the student’s pay an additional fee to help teachers and colleges do what they should already be doing. Yes I realize that we assign texts or other course materials that cost money but this seems different somehow.
Lecture Tools may simply be an interesting product in search of the right business model and as an investor I have come to learn that proper business models are incredibly important. A great product with the wrong business model is a recipe for failure. This brings me to an interesting but largely unknown experiment happening right now in Vermont. It’s called Oplerno (www.oplerno.com) which received approval this spring from the Vermont Department of Education to grant college credits for courses taught through its platform. To be clear Oplerno is not yet accredited although it intends to become accredited as soon as possible. Why do I think Oplerno is interesting? Because it has a business model that makes sense to me. Oplerno is essentially a platform that allows professors to offer courses to the world but gives the professor complete control over the business model. Professors can set the price for their courses (it could be $500 or $5,000 for instance to take a three credit course). Professors will receive up to 90% of the revenue generated from their courses but course enrollment is capped at 25 per class. Think of Oplerno as the first real multi-sided business platform to hit higher education. This business model makes complete sense to me. It seems to me that each Professor is a “brand” unto themselves. Students if given a chance should seek out gifted professors with domain expertise and with teaching chops. Yet the best professors today have harnessed themselves to a union governed system which has traded “academic freedom” and tenure and pay caps for job security although to be fair the present system also provides the “distribution channel” for a professor’s teaching product. Ironically this system seems to produce less accountability than those of us who exist in the real world have to contend with. As an entrepreneur and investor this makes zero sense to me. A good professor has something to sell. It’s a product. They should be paid well for it and more importantly they should want to be paid well for it. Heck entire college departments should decamp en masse for Oplerno and set up shop there. There is certainly a problem with the current higher education business model. Continued tuition increases and increased student and parent borrowing is not the answer.
Perhaps the answer lies in better technologies combined with better business models.
- Cairn Cross
Interesting post. I'd say competition already exists... Students vote with their choice of major and choice in electives. If you don't get enrollment, you won't keep teaching. The pier no offer of 90% is a dream! but the hyper competition might yield unintended consequences (fluff content, sensational tactics to attract students, grade inflation...). Besides, without 'Harvard' our 'uvm' next to your education, how can we judge your abilities (snark)?
ReplyDeleteThanks for the mention of Oplerno.
ReplyDelete