Lessons from teaching a hack class


As a college student, I spend most of my (sleep-deprived) days attending class. Rarely do I get the opportunity I had last Spring: to actually teach a class with one of my best friends, Charlie Marsh.

Before you get any ideas about my joining the faculty, I should provide some context. The class, titled ‘Introduction to Hacking’, was one of several student-run “hack classes” offered by the Princeton Entrepreneurship Club. We taught fellow students how to solve practical problems with programming, covering topics like data analysis with Python, web scraping, machine learning, and more. At a high level, we simply wanted to expose students to topics in programming that are typically absent from a course catalog. Around 20-30 students consistently attended the 7-week class, a good turnout given that the course was not offered for credit and wouldn’t be listed on any résumés.

In this post, I share some of the lessons we learned along the way. I hope this information helps future hack class instructors– and teachers anywhere– make the most of their classes.

1. Hack classes work

I’m enthusiastic about hack classes. They share interesting topics in a flexible format with genuinely passionate students.

Each hack class is motivated by an interesting topic– one that the university does not address, is difficult to learn alone, or both. When Stanford University began to offer an iOS programming class, the iOS hack class brought the same content to a physical classroom at Princeton. Unlike the iOS hack class, Introduction to Hacking was not motivated by a single topic, but rather a grab-bag of topics with technical buzz (“machine learning!”) or usefulness across many different career paths. In both classes, the hack class shared interesting knowledge that was not available in the university curriculum.

Flexibility is a key strength of hack classes. As an instructor, I can create a class or adjust the syllabus without the bureaucracy of a registrar. Students benefit from the freedom to attend whichever lectures they wish. Historically, student retention was a problem in hack classes where material was cumulative from week-to-week. Even if students missed just one lecture, they’d be unable to follow along when future material built on the previous weeks’. To reverse this trend, we designed the lectures in Introduction to Hacking to be modular– each session stood by itself and did not require students to attend previous lectures. This afforded students the flexibility to attend only the sessions they found interesting.

By their “optional” nature, hack classes attract only genuinely passionate students. This is an amazing benefit because a lecture hall of passionate students leads to rich discussions, which leads to better learning for everyone. Instructors also have an additional incentive to teach well. If Charlie and I didn’t make the material engaging, students would’ve simply stopped coming. When things are going poorly, this realization can be disheartening. But when things are going well, it’s enriching and motivating for students and instructors alike.

Hack classes aren’t perfect, but I believe they effectively fill a gap in the university curriculum. They’re flexible initiatives that invite students to teach interesting material to other passionate students. And they work.

2. Students take hack classes for different reasons

Charlie and I ask the ‘why’ question often. Before even deciding to teach Introduction to Hacking, we thought about why exactly we were going to do it. (More on this at the end.)

We realized that students too– often subconsciously– think about why they should participate in a hack class. Reasons vary from “I want to learn the material” to “the instructors are my friends”. As instructors, we tried to tease out these reasons early with a short survey.

To our surprise, we received over 50 responses! Either students were passionate about communicating their motivations or they loved taking surveys. Responses varied widely, but the two key takeaways were that: (1) students had diverse levels of experience, and (2) some students wanted to learn just one or two of the topics on the syllabus. We structured the course with these two considerations in mind.

Most students had taken two or three programming courses (COS 126/217/226), but some were beginners and others, experts. Because of the diverse levels of experience, we emphasized “survey over deep dive”. We used lecture to introduce “what’s out there”, and concluded with three slides listing next-steps for novice, intermediate, and advanced programmers. For example, in the data analysis lecture, we suggested that:

Conveniently, the nature of the topics in Introduction to Hacking made the class’s survey nature easy to implement. Many advanced programmers just hadn’t gotten around to checking out some of the topics like web scraping, and novices didn’t require special knowledge to understand how the tools and libraries were used. By explicitly communicating the survey nature of the course, we maintained focus while accommodating the students’ diverse levels of experience.

It’s perfectly fine for some students to attend one or two sessions. As I mentioned, we designed each lecture to stand on its own. In addition, we distributed lecture slides and example code in advance of the session day as often as possible. This enabled students to attend the lectures that mattered most to them. It also gave us feedback on which topics sounded interesting to students. “Git, Bash Tricks” was our least-attended, while “Computer Security: Passwords, Web, Wireshark” our most-attended.

Understanding students’ motivations helped us to structure the class most effectively.

3. Teaching is hard

Coming out of the class, Charlie and I have a greater appreciation for our teachers. Teaching is challenging for a number of reasons.

Teaching requires significant empathy for students. While we were comfortable with the content, we had to take a step back to identify how a beginner might understand the topic. Adequate preparation was key to answering this question effectively. One instance of this was in the security lecture, where I wanted to introduce packet sniffing tools to students who’d never heard of a network “packet”. I had to clear my mind of anything I knew about networks to realize all that mattered was an awareness for “hosts”, “routers”, and “packets” flowing between them. The result was a single slide demonstrating their relationship. Striving for simplicity over complexity, I realized: choosing what not to include was just as important as what to include.

Teaching requires significant preparation time. Each week, Charlie and I spent 3-5 hours writing the lecture slides. We then met for an hour to provide each other feedback and divvy up the task of writing the live demo, which usually required another 1-2 hours.

As much as teaching demands empathy and time, it is also incredibly rewarding. A well-prepared lecture feels like it just falls into place, and the joy of students’ understanding makes it worthwhile.

4. Communication is crucial

The quality of communication can make or break a hack class. Communication spans messaging, questions, and feedback.

Messaging promotes awareness for the class and maintains student interest from week-to-week. To our amazement, Introduction to Hacking attracted over 200 students to its signup list. I think some of this success can be attributed to “grassroots” methods– months in advance, we started adding individual friends to the list of participating students. During the class, we emailed the “what” and “why” of each lecture a few days in advance. Towards the end, we also promoted individual lecture topics to mass email lists around the school. That way, students who hadn’t signed up for the course attended a lecture just because they were interested in the topic. It’s amazing that a lot of people came to just the final lecture and got a lot out of it.

In the lectures themselves, we promoted communication by constantly encouraging questions. A trick I learned is to ask “What questions do you have?” instead of “Do you have any questions?”. Doing this communicates that questions are expected and are not a barrier to the class’s progress. We tried a “Hack Hour” following the lecture with an online form submission for questions, but this method quickly fell apart. We realized that the informal, one-on-one Q&As that naturally arise after lecture were more effective. (The majority of the help provided during that time was for development environment setup.)

Charlie and I were eager to hear feedback. Students preferred writing their feedback to saying it in-person. To facilitate that communication, we created an anonymous online form for anyone to send feedback just by clicking a link on the web site. While we received around five responses from that form over that the semester, the majority of feedback came by email from friends.

Closing Thoughts

A few months after Introduction to Hacking ran, I was chatting with a friend who had taken the course. A politics major, he was actually now studying for a software engineering interview. I was surprised. He explained that, over the past year, he realized software engineering was his true calling, and Introduction to Hacking was an influential reason for that change.

Across our projects, Charlie and I are most motivated by the potential for impact. It’s our answer to the ‘why’ question I mentioned earlier: we want our work to matter. My politics friend is just one instance of how hack classes are an incredible vehicle for impact. I believe that, in many cases, students learn more effectively when they’re taught by other students. And I’m grateful for the opportunity to have been a part of that.

December 2014