This semester I have been doing a lot more administrative work for the EMAC degree. As the program is growing, both in terms of faculty and students we have to have more conversations about sequencing the courses, course content, and overall examining the pedagogical experience of our students. As one can imagine this rather easily leads to a discussion about how many students we can put in our classes, how we can cover more sections of the same class, faculty resources, etc. So partly because of this, and partly because of the recent discussion about MOOCs I started wondering, what is the ideal student to instructor ratio for a semester. What is the Dunbar number for teaching?
The Dunbar number (rather famously used in internet and network studies) is the sorta of limit of the number of people one can maintain social ties with, the number of people your brain can hold and still remain close with. Often sociologist reference this number at 150 (and yes I am aware that this is a contested idea, but we can save that debate for another time). So you can have roughly 150 close ties with people but as you start to meet more people and develop more close ties some of those other 150 start to fade away. A human brain can only handle the cognitive load of maintaining tight social bonds with so many people. Its worth mentioning that there is probably variance within a population, some individuals having higher Dunbar numbers than others. And its probably also worth noting that one reason this comes up in social media research is because people often ask if social media can extend a Dunbar number, making it possible to maintain a higher cognitive social load.
But this got me to wondering what is the Dunbar number for instruction. That is at any one time what is the number of students I can interact with, what is the cognitive limit of the number of students I can maintain and have a consistent interaction with. I think we can all agree that the fewer the students the more time I have to spend with each, and versa visa, the more students the less time I can spend with each. But I am hypothesizing about something slightly different here. What is the educational cognitive load? What is the Dunbar number for instruction? At what point do the edges start to fray and do I lose track of the students I am interacting with over the course of a semester?
So here’s my guess, and this is just a starting point. I think my Dunbar number for instruction is about 45-60. That is over the course of a semester I can handle about 45-60 in classes and still interact with them, become invested in them, and try to treat each (as much as possible) as an individual who is having an educational experience in which I play a role. Keep that number under 60 and I pretty much “know” all of my students. If someone walks into my office and says “How is Steven doing this semester?” I know about his semester. How he is doing. How he is engaging the material, and probably some additional material about where he is at with in the program and what interests him. Keep that number under 60 and I think that is true of all of my students. But as that number rises over 60 I start to lose the ability to “keep all the students in my head.” That is I start to lose track of some, they disappear, fade into the background. I can keep tabs on most but I end up focusing on the ones who need challenging, or the ones who are struggling, and some in the middle get lost under the weight of too many students.
I think that this is somewhat independent of class size. That is I could handle one 50-60 person class, two 25-30 person classes, or three 15-20 (or heck even 5-6 classes of ten each). Yes ideally the smaller classes would be better, but I think if I were teaching just one large class I could use some of the non class time to spend extra time looking at student work, engaging with them outside of class etc. . . I also think this is somewhat independent of graduate students’ whose committees I am working on. I think that since I already have an established relation with them they are probably already in my real Dunbar number. Also I think the load is different depending on what types of students I have that semester. If I have a lot of first years that number trends down closer to 45, and if I have a chunk of students I have already had in prior classes I can push that number to the higher end. Note I am not directly talking about time commitments here. Obviously doing my own research, handling committee assignments, and a personal life all “cost” time, I am really just talking here about “cognitive load.” What is the “cognitive load” of students I can handle in any given semester?
Hypothesizing about this, I just put this question on Twitter last night, and really to my surprise most people agreed, placing their number at the 50 range. Sure this is a biased sample, and people are probably influenced by past teaching experiences. But its a rough guess and a good starting point for a conversation.
Why This Matters
There has been a great deal of discussion about MOOCs lately, expanding the number of students in a classroom, or in a related discussion classroom size, or faculty workload. If we distill down to the core what it is I do as an instructor/teacher I can only handle so many students doing that job. Sure I can lecture to hundreds if not thousands a week. Heck the tech would make it possible for me to lecture to millions (just record a lecture), but that really isn’t what I see as teaching, that’s just broadcasting content, the same as writing a book and letting someone read it. So, it seems to me that MOOCs or blended learning, or whatever, are precisely the wrong way to go. Increasing the cognitive load of teachers isn’t the answer. Our value add isn’t in broadcasting content, but rather in working closely with students.
The institutions we work for have an interest in efficiency how can you maximize the number of students who learn with a given professor. But if there is a cognitive load max there is a point at which there are diminishing returns, where by adding in more students the institution is undermining the very value it seeks to add. And by big fear here, we end up creating a tiered education system whereby a certain class of people get the in classroom, close work with professors, and another get the discount $10,000 for four years model that doesn’t involve being closely valued and attended to by an instructor.
Okay, so here is my question is there any research on this? What is the cognitive load max for you for a given semester? All else being equal what is your Dunbar number for instruction? Leave a comment, interested in hearing from folks.
I used to think 60 was my limit too. Then I took on a class of 800+. And found I had time for everyone. I didn’t have to turn anyone away for three years (except once when I had just that minute cornered the Chairperson for a budget discussion after weeks of trying).
But the reason things went well was that I had found this article before I took up the assignment – Don’t think small group. Think Mexican wave.
Google’s link is ugly so why not Google “teaching big classes making lemonade”. The .pdf comes up in position 1.
I teach K-12. I have 60-70 students. Over the course of the year, I will have the entire middle school cycle through my classes, so I’ll have 150 or so that I “know”. I have a class of 2 and a class of 6. I really know those students. I think 12 is my ideal class size. And I could see having 5 classes of 12 and be able to keep up. So I think grouping matters. In classes that approach 20, I’m aware that I’m not keeping up with everyone. The same number of total students spread out over more classes would work better for me.
It depends on the nature of the instruction. I work for an honors program that heavily invests in the minds and lives of students. I find in this intense mentoring environment with a good chunk of high maintenance students that my limit is around 30. The more I know and need to know to do this work well, the lower that number gets (and it is further depressed the more children under 5 I have at home – another form of fairly intense cognitive and social load).
I like the article. I used to teach English language as a second language for adults and children. There were afternoon classes every day and morning classes during the weekends.English is not my mother tongue, I live in Macedonia and have BA in English language and literature, just to clarify the level of education and the background. What I have been facing is overloading of information about the students, personal type of info which amazingly I was remembering. During the courses I was remembering their names, but when the courses were over, the names were fading away. and the personal info stayed in my memory. Even today , when I see someone I do not remember his or her name, I first ask if he or she was my student
…The number of students I was remembering was not more than 60 at a semester…
Kind regards.
Sixty sounds about right. When I was teaching 4 classes I would struggle to keep linking names to faces if the total reached 80 or 90.
This is an important point to the extent that online courses try to replicate the kind of student-instructor relationships found in face-to-face, classrooms, but I wonder if it is still relevant in connectivist MOOCs where the instructor’s role (to the extent that there is an instructor) is more to set activities, act as an admin/facilitator, and then participate as co-learner.
The issue of the Dunbar number and how many meaningful connections can be maintained at once certainly came up in MOOCMOOC in discussions about the formation of sub-groups within the MOOC.
The ADE and NCTE have long recommended a maximum writing-intensive load of 40-60 students in 2-3 sections, with no more than two preps. Of course, that’s based as much on paper-commenting/grading time as anything else, but I think it’s about right, and it comes intriguingly close to the numbers you’re reporting. Of course, I’m an introvert — which means that I expend energy on interacting with people,and gain it from being alone — but it sounds like it’s not only introverts who feel there’s an upper limit on how many students they can effectively interact with each semester.
The distinction between an ongoing and a one-semester relationship is important, too, I think. I teach almost exclusively a single core course, which means that, with the occasional exception of someone from a lower-level course I occasionally teach, or someone who repeats my class after failing or dropping it (yes, it happens; go figure), I’m faced with c. 90 new people each semester. That’s definitely too many to truly get to know (although, interestingly, I get to know their research projects, which are completed in stages, pretty well. That may be a matter of temperament; I really do get more excited about interacting with ideas than with people, which probably isn’t such a terrible thing for a college professor, and may not even be a noticeable distinction for the students who put something of themselves into their work). It would definitely be different to work in a program where I might see the same student in several classes, at different levels, over the course of several years.
(I apologize if this is a duplicate; the comment form doesn’t seem to be working entirely smoothly)
The ADE and NCTE have long recommended a maximum writing-intensive load of 40-60 students in 2-3 sections, with no more than two preps. Of course, that’s based as much on paper-commenting/grading time as anything else, but I think it’s about right, and it comes intriguingly close to the numbers you’re reporting. Of course, I’m an introvert — which means that I expend energy on interacting with people,and gain it from being alone — but it sounds like it’s not only introverts who feel there’s an upper limit on how many students they can effectively interact with each semester.
The distinction between an ongoing and a one-semester relationship is important, too, I think. I teach almost exclusively a single core course, which means that, with the occasional exception of someone from a lower-level course I occasionally teach, or someone who repeats my class after failing or dropping it (yes, it happens; go figure), I’m faced with c. 90 new people each semester. That’s definitely too many to truly get to know (although, interestingly, I get to know their research projects, which are completed in stages, pretty well. That may be a matter of temperament; I really do get more excited about interacting with ideas than with people, which probably isn’t such a terrible thing for a college professor, and may not even be a noticeable distinction for the students who put something of themselves into their work). It would definitely be different to work in a program where I might see the same student in several classes, at different levels, over the course of several years.