Poorly Structured
16 min read
I received a few boxes from my Mom recently containing high school detritus. One in particular was packed solid with papers generated by classes I took in high school, neatly organized into manilla envelopes labeled AP Comp 5, AP Stats, AP US History.
Thinking back on high school made me wonder: Why is the instructional model of college so different from high school? And is it possible that we could help students succeed in college by simply making college courses more like high school courses?
I was a good high-school student. I was a terrible college student. Some of my troubles in college were surely due to trying to punch above my weight. I was an above-average student in a decent upper middle-class public high school who, probably at least partly due to a lucky break on a standardized test, ended up at Harvard, surrounded by people who were genuinely smart. And nice too.(1)
But part of my problems were also structural. I led a highly-structured existence in high school. This was partly due to my parents who erected and enforced very strict rules about what I was allowed to do. It’s a lot easier to focus on your homework every night when hanging out with friends, watching TV, and surfing the internet are not options.(2) It set me back socially. But it did keep me on track academically.
To the degree that my parents helped keep me on track, the fact that there was a track at all was because high school itself was much more structured than college. Most classes met daily, and I remember most nights having something that I needed to complete for the next day: a problem set, or a response paper, or assigned reading. My parents really just had to create time for me to do the work. I always had something to do. And I always knew what needed to be done.
Classes where I learned something—particularly math and science courses, but also the better English and literature classes—expected you to complete your homework and employed a lot of strategies to make sure you did. If the assignment itself wasn’t directly graded, there might be a short quiz in class, or a group activity at least partly designed to embarrass unprepared students. Classes where I didn’t learn much—such as four years of fruitless Spanish—tended to accommodate students that hadn’t completed the homework.
A good portion of class time was usually spent engaged in some kind of supervised activity. Not everything was that elaborate. Teachers would regularly just give us 10 or 15 minutes at the end of class to start our homework while they circulated, returning graded work and providing feedback and support.
When I arrived at college I found all of these structural supports suddenly kicked out from under me. Not only were my parents not around to help make sure that I did my assigned work each night and didn’t just while away the hours downloading MP3s or playing hearts with my roommates—it also wasn’t always clear what I should be doing on any given night.
Some nights it was clear. Those nights tended to be rough, since they meant that there was something due the next day and usually something that really needed a lot more than single night I had allocated. My room freshman year sat astride the Red Line tracks leading into Harvard Square Station. Not everybody knows this, but at around 4AM the MBTA sends a first round of empty trains down the tracks to test the lines before service begins. The reason I know this is because I took the calculus course intended for math majors.
Clearly I struggled in college at least partly because I was terrible at self-discipline and time management. I take full responsibility for this. It really wasn’t until the end of graduate school, a full decade later, that I was mature enough to return to some of the productive habits that had been imposed on me as a high-school student. I have very few regrets about my life. But I do find myself wishing I could go back to college and start again, armed with the ability to take it seriously and get the most out of experience.
My struggles to succeed in college were entirely my fault. But aren’t they also fairly predictable? I was surely not the first college freshman to be living on their own without parent supervision for the first time. Nor was I the first to be distracted by the many temptations and new experiences presented by college life.
On campus we usually describe this process as adjusting to college. Some places now try to ascribe a more generalized nature to this transition by referring to it as learning how to adult or adulting.(3) There is definitely something healthy and inevitable about this change. For many students, college is the first real chance to establish adult independence. College represents both a chance to break free from family patterns of acting or thinking, but also opportunities to meet people from different backgrounds with different ideas, values, and belief systems. This opportunity itself is one of the most valuable things that college provides, and it happens completely outside of the classroom and independently from student coursework. It may actually be more valuable than anything students learn in class.
But adjusting to college is also problematic, witnessed by the fact that many students fail to complete this transition. Some fail out or drop out, often not before incurring significant costs. Others, like me, simply don’t get nearly as much out of college as they could.
While failure to adjust to college may involve difficulty establishing effective routines and habits that allow students to succeed in the classroom, it is expressed entirely in terms of academic performance. Students fail college because they fail their classes, even if they might be in the midst of a very productive personal transition into independent adult life. So even while students are learning how to adult, they better keep those grades up.
Which returns us to our original question: Why is college structured so differently than high school? And is this structural change actually beneficial, particularly given its interaction with the other significant changes that students are going through as they enter college?
Let’s pause for a minute to try and define the high-school instructional model and the college instructional model.
It’s worth noting that the high school instructional model is also fairly similar to what is used throughout the rest of K–12 education, even if it reflects a gradual transition towards additional independence and expectations appropriate to older students. But given that my point is to highlight the substantial break between high school and college, it’s worth focusing on the model used in high school, rather than in earlier K–12 years.
In addition, while colleges offer many different types of classes, I’m choosing to describe large lecture-style courses. At most universities, many if not all of the courses taken by first-year students will fall into this model. This is because many courses taken by freshman are introductory, many introductory courses are large, many large courses use this model. However, I will argue below that large courses do not have to be structured this way, and probably should not if they want their students to succeed.
I see two substantial differences between the high school and college instructional models. The first involves how class time is used. In college, class time involves largely passive participation by students: lectures that they are expected to attend, although attendance is frequently not enforced. In high school class time is more likely to involve a mixture of passive and active participation: not only listening to the teacher talk, but also participating in the discussion, engaging in group activities, completing homework or an assessment, or other forms of active engagement.
The second involves frequency of assessment. Many college courses utilize a small number of assessments: maybe one midterm, one final, and a paper. Assessments are few, far-between, and necessarily high-stakes, since each must be worth a substantial portion of the overall grade. In contrast, most high-school courses utilize a larger number of assessments: nightly homework, assigned readings followed by quizzes, short response papers, and so on. With fewer assessments, each one is usually lower-stakes.
Giving frequent assessments has another important impact on a class and leads to another key difference between the high-school and college instructional models. Each assessment requires telling students what they need to know to prepare, and so frequent assessment naturally requires breaking the course material down into much smaller pieces. The timing of the assessments combined with their instructions and expectations together produce a fine-grained schedule that students can use to plan their work.
In college, it wasn’t uncommon for a professor to distribute an enormous reading list and simply expect students to pace themselves to be ready for the high-stakes assessments at the middle and end of the semester. In contrast, in high school you’d be assigned one or two chapters from a longer book and discuss or be quizzed on them the next day.
I should pause and point out that the latest educational hotness referred to as active learning or peer instruction is not much more than a rebranding of the high-school instructional model. Expecting students to prepare for and actively engage in class only seems novel after you’ve spent decades lecturing at them. It’s also pretty much what they were doing for a decade before they arrived at college. Which is why it’s not surprising that active learning works, particularly compared with the alternative.
Clearly what is presented above are stereotypes. Not every college course is a large lecture-style course. Not every high-school course involves classroom participation and frequent assessment. There are high-schools that intentionally try to emulate the college model of instruction in order to help prepare students for success at college. Frequently these high-schools are private. And there are colleges that, while they tend not to describe it this way, emulate the high-school model. Usually this is advertised as small classes and more contact with faculty, and accompanied by higher tuition. Frequently these colleges are also private.
But I think that these stereotypes do a reasonable job of describing the most common transition from high-school to college. Specifically, a move from senior year at a public high school to freshman year at a large university. What results is a sudden and dramatic shift from a highly-structured highly-engaging instructional environment to a poorly-structured less-engaging instructional environment. Managing this transition becomes one more challenge that first-year students face, alongside all of the other ones previously described.
Why do we put our students through this? Is there actually something better about the college instructional model?
College faculty and administrators will usually claim that they are preparing students for the real world and life after college—teaching them how to adult. I find this laughable. If anything, good working environments strive to provide employees with far more structure than a typical large university course. There isn’t anything completely analogous to class or assessments after college. But as a software developer, I see open-plan offices and stand-up meetings as ways of creating daily engagement, and unit and integration testing as ways of providing small frequent assessments. I’m sure there are analogs in other fields as well. Leaving someone to complete a large make-or-break task with no support or intermediate check-ins is usually considered bad management. So there’s no real reason to do so in college.
It’s also telling that many professional schools seem to actually run their courses in a much more structured manner, combining daily engagement with frequent assessment. Or at least other ways of ensuring that students are performing assigned work outside of class—like the fear of being cold-called as a 1L.
Even if you accept the claim that learning to succeed in the college instructional model helps students in the long run—which I don’t—there are two other problems. First, we actually need to teach them how to do it, not just suddenly expect it from freshmen on Day 1. Suddenly throwing a bunch of students into the deep end isn’t a swim class—it’s a swim test.
The other problem is that teaching students how to learn without structure or engagement needs to be balanced against other objectives. Like, for example, actually teaching the course content. But too frequently courses taught in the college model fail to do either. They expect, rather than teach, that students can work with a high degree of independence, and in so doing they limit the ability of students to actually learn the material. But assuming the former, you end up with neither: a complete lose-lose.
I suspect the real reasons for the ubiquity of the college instructional model are harder to defend. One is simple inertia. Do you want to lecture on MWF or TR? Actually I’d like to teach daily to help keep my students on track and engaged. Do you want to lecture on MWF or TR? That’s what’s in the drop-down menu. Nobody knows why. And if you try to do something different, you will cause massive scheduling headaches for the students in your course. You wouldn’t want to do that, now, would you? Just be a good professor. Do you want to lecture on MWF or TR?
Another reality is that many college faculty don’t want to teach at all.(4) High-school teachers are teachers first. Most college faculty are researchers first. Highly-structured and highly-engaging courses require more time and energy to teach. If your goal is to avoid teaching, then the college instructional model might seem like a great fit. Because it’s less work.
But what about size? Clearly one difference between large lecture courses and high-school classes is that large lecture courses are, well, large. Sometimes much, much larger than any course a student would take in high school. Do large college courses inevitably require utilizing the unstructured, low-engagement college instructional model?
It’s certainly easier. Assembling even 100 students together for one class makes it very hard to engage the entire class in a dialog as a good high-school teacher might with a class of 30. Lecture halls are also designed for—you guessed it—lecturing! That frequently makes them awkward or even downright unsuitable for other kinds of activities like discussion or group work. Grading is another challenge. Requiring nightly assignments from a huge number of students creates an enormous amount of grading, which might easily overwhelm even an appropriately-sized course staff. Given these challenges, it might seem impossible to replicate the core engagement and structure-through-assessment features of the high-school instructional model.
These difficulties are real. And, as an instructor of a large course, I can confirm that there are few resources and little flexibility provided to instructors that want to challenge the status quo. This may start with “Do you want to teach MWF or TR?”, but the rigidity extends to staffing, space, computational resources, and limits you in all kinds of ways that make it hard to offer a large course in a highly-structured and engaging fashion.
But I’ve found a few workarounds that have worked for my class, and might work for yours as well.
First, let me acknowledge that I have a huge advantage here. I don’t grade. Neither do my teaching assistants, or my undergraduate course staff. All of my assignments and assessments are autograded. 100%. I’m sure you hate me now, at least a bit.
This alone allows me to recreate the high-school instructional model regarding frequent, low-stakes assessments. My students complete daily homework, weekly quizzes, and a longer-term project complete to weekly checkpoints. Throughout the course, they invariably have something to complete every single day.
We introduced daily homework in Fall 2018. Students have tolerated it extremely well. While many other CS1 courses utilize autograded assignments, few require daily homework. I’ve sensed reluctance to do this even from people that are in a position to do so easily. My suspicion is that this is due to fears that students will rebel against something so highly-structured, and so different from the instructional model used in their other courses. All I can say about this is that I have not found it to be true. We receive few to no complaints about the assessment structure.
It took me longer to restructure the content delivery portion of the course. Moving away from the lecture model is really, really hard, given the degree to which it’s embedded in the policies and physical plant of the university. Honestly, I wasn’t brave enough to try it. I also feared that I would get a lot of pushback if I did, and I wasn’t completely sure what else to do. There aren’t a lot of choices on the menu. Would you like to teach MWF, or TR?
But then the pandemic hit, and new options were suddenly available. Did I want to teach synchronously, or asynchronously? Nobody had ever asked that before! Even if a crisis doesn’t represent both danger and opportunity, it does augur both danger and change. This was an opportunity to try something new.
For Fall 2020 I restructured our course content into a series of daily lessons, each accompanied by a programming problem, in a sequence borrowed from the series of homework we had been using since Fall 2018. Our daily lessons are designed to be highly-engaging, and combine text, runnable code playgrounds, video, and a new animated editor component that we refer to as interactive walkthroughs that combine the best features of video and playgrounds for live code examples.
I’m extremely proud of the transition that we performed. But let me be clear about what I did: I simply redesigned my course to move away from the college instructional model and toward the high-school instructional model: Daily learning, daily practice, highly-engaging, highly-structured.
Did it work? I think so. Drop rates went down, while the success rate for students that complete the course remained the same, even during the strange and challenging pandemic year. In addition, I was able to single-handedly teach almost 2000 students over two semesters.(5) There are also a bunch of other really wonderful things about this new instructional model that I’ll comment on elsewhere.
About the only thing that I’m uncomfortable with is the sense that I may not be preparing students for the rest of college. But I think I’ve made my peace with that. Because, if anything, the rest of college should probably be a lot more highly-structured.