The Knewton Blog

Our monthly newsletter features edtech and product updates, with a healthy dose of fun Knerd news.

The Limits of Lecture Courses (or: Evolve, Damn It!)

Posted in Knerds on March 8, 2010 by

Kalyan is a top teacher at Knewton, where he guides students through their LSAT prep.

Lecture based classes are common largely because they are easier to implement from an organizational standpoint, thus helping keep tuition costs lower than they would otherwise be. But there is a tradeoff – they are also far less optimal from a pedagogical standpoint than smaller, more interactive classes because the single-paced nature of the lecture just isn’t appropriate for the needs of a large number of students.

I encountered this problem rather frequently at the undergraduate level. For the better part of three hours every Tuesday evening, I would fight to avoid falling asleep in a packed auditorium, while the professor would lecture and write scarcely legible chemical reactions on a chalkboard that happened to be much too small for the room. The conundrum I found myself in was that I could learn the material covered in three hours of class if I spent 45 minutes alone with my textbook. I didn’t need to be an economist to figure out that spending three hours in that suffocating and dimly lit auditorium was an incredibly inefficient allocation of my time.

Ten years later, I am finishing up my third Masters degree (in Economics of all things) and am confronted with yet another iteration of the same problem, except I now have additional commuting time to deal with. It now takes me a total of 4.5 hours to assimilate the information I could in 60 minutes on my own. Attendance is compulsory, as a result of which I estimate that I will waste 42 hours in class this semester.

But I am not the only type of student whose needs aren’t being met. Consider the predicament of a student who pays attention for 3 hours in class and for one reason or another, still doesn’t quite get it. This student might be better served by spending 5 hours on the same material. This brings me to my central point- given the technology available, it seems absurd to put the two of us in the same lecture and settle on a mean duration of 3 hours which would neither address my needs nor hers/his.

We typically assume that “student ability” is normally distributed i.e. the student described above and I would fall on the periphery of this distribution with most students clustered around the mean. While that might appear to mitigate the severity of the problem, I have yet to see any concrete measurement of “student ability” or, more pertinently, evidence of its application on a group by group basis. I am, of course, talking about learning, and not grading, which tends to be more accommodating, if dangerously so.

The detrimental effects of compulsory attendance requirements coupled with ineffective teaching are understandably more amplified for students with non-traditional learning styles and/or other constraints.

My security guard, for example, works full-time to fund his college degree, and is also taking a full load of classes. It is inevitable that he will not be able to attend every session, and it is quite likely that his grades will suffer even if he does a stellar job on all the assigned exams and homework material. This is a consequence of the fact that in far too many cases, in addition to a student’s academic performance, attendance is treated as a mandatory requirement. This is most frequently manifested in the form of “If you miss more than x number of classes, you will drop one (half/full) letter grade”.

There is something fundamentally wrong with this premise. Learning styles can be quite diverse and it is dangerous to assume that learning is well correlated with attendance, particularly in large classes. In an ideal world, this would be the case, but in a system riddled with inefficiencies– one in which it is the norm to judge professors on the basis of their research prowess instead of their merits as teachers, this connection may well prove tenuous.

Such a grading scheme also doesn’t account for how long an individual spends on the material outside class, which could, at least for some students, be a better determinant of performance. In the case of the aforementioned security guard, most of his learning happens while he is sitting at the guard’s desk, in between pushing buttons to let people into the building. Should he be penalized for where he chooses to study?

His case is made stronger by the fact that unlike days past, teachers aren’t the only resource that students have. The advent of search engines, online journals, online encyclopedias etc. have meant that the proactive student will never suffer a dearth of resources.

It is also worth considering the evolution of the intellectual toolbox that educational institutions aim to impart to students. This toolbox certainly shouldn’t be the same as it used to be. The collection of tools and skills required to thrive today isn’t the same as it was in the 1960s or even the 1980s. For example, with ready access to search engines, recalling arcane details is no longer as vital. What is more important, however, is being able to make sense of the endless reservoir of information that is the internet, and being able to separate reliable, well-founded information from inadequately supported, misleading and/or downright false information. This necessarily means that students need to receive more training in elementary logic and critical reasoning in order to avoid these pitfalls of unsupervised learning. Existing curricula need to place a far greater emphasis on this fundamental skill than they currently do.

This will also need to be accompanied by the evolution of the role of a teacher from “sole information provider” to “learning facilitator”. Persisting with the “sole information provider” strategy in this day and age may be dangerous because it engenders an approach to learning that revolves solely around attending class and not putting in much effort outside of it, which can be a serious handicap to overcome once students are outside the academic establishment.

From the perspective of a graduate student, there may not be a better substitute to face-to-face interaction with exceptionally accomplished researchers. Research institutions are a place where experts aggregate and great minds often collide. The benefit to society from these activities certainly cannot be overstated. Graduate students are typically better off than their undergraduate brethren because class sizes in graduate programs tend to be much smaller, with students being more rigorously selected, thereby allowing for an environment in which every individual’s needs can be addressed (most of the time).

Undergraduate students, however, are vastly different beasts and are not normally given the privilege of the intimate class setting (the rare individuals that are pay dearly for it). The typical introductory undergraduate course has at least 40 students and consequently takes the form of passive learning from lectures designed to meet the needs of the amorphous “average student.”

Compounding these problems is the fact that the tenure system, despite its benefits, has made it possible for lazy researchers to impart little or no useful knowledge to their students in the classroom, particularly in schools that cater to disadvantaged demographics. Considered a bad influence on graduate students, such teachers are often relegated to teaching- you guessed it- undergraduate courses, where they are free to be as lazy and pointlessly tyrannical as they please.

From a pedagogical standpoint, any system that aims to provide a single solution for a diverse variety of learning needs is necessarily bound to be inefficient. The tools provided by technology offer solutions, however.

If evidence were required that an online learning platform can be as effective as a physical one, the success of Knewton’s model provides an excellent example. With our team-teaching approach we are able to utilize the benefits of a lecture environment while simultaneously being able to address individual student concerns and questions that may have otherwise prevented them from fully understanding the material that followed.

Importantly, we offer a variety of algorithm based learning tools. With a web-based approach, it is possible to identify individual student weaknesses on the basis of their performance in well characterized areas of a test. Further, we are able to give students the opportunity to hone those skills in a focused manner (e.g. with Create a Quiz on the LSAT course).

This list, however, is merely a humble beginning. We have only recently begun to really delve into the wealth of student performance data we generate on a daily basis and are working furiously to produce practical tools that will help students generate a customized self-study program which will be complemented by the live lessons and archives that are currently available.

While our product offerings are currently limited to standardized tests that we have developed curricula for, there is no reason why this model cannot be applied to high-school and college-level courses. Since online learning platforms are massively scalable, a greater number of students could realistically be taught by the best instructors in the field. Perhaps even more importantly, they will have a wealth of learning tools as well as virtual resources to help them diagnose the specific concepts they are having trouble with and to develop custom learning programs that cater to their individual strengths and weaknesses rather than abstractions like “mean student ability.”

Next time: How standardized testing can (contrary to public perception) help level the playing field for everyone.