Class competitions: Motivating or damaging?

As I’m writing this, I’m enjoying the quiet solitude that comes from working on campus over a holiday break…  I’m preparing for what is my “signature” course: Networks: Structure & Economics, which I’ve called “The ideas behind our Networked World” until this year. It’s a very unique course, and I’ll probably post a bit about it as the term starts.  But, for now, I want to write about something I always struggle with as a teacher: class competitions.

I’ll start by saying that ever since I started teaching this class, I’ve always had at least two class-wide competitions, and this year, for the first time, I’ll have three.

The two that I’ve had going for a few years now are:

  • Rankmaniac: The task is to, given a very large unknown graph, compute the nodes with the top ten pageranks as fast (and accurately) as possible using Amazon Elastic MapReduce.
  • Clickmaniac: The task is to generate as many clicks/likes for the course facebook page given an advertising budget of a few dollars a day.

And this year I’m adding:

  • Pandemaniac: The task is to, given a very large unknown graph, pick a small set of seed nodes that will start a cascade/epidemic/pandemic that takes over more of the graph than any of the other teams (which are also picking seed nodes).

The weeks during the competitions are always some of my favorite ones of the term because of how engaged the class is (and they also seem to be loved by many of the students, who rave about them in the course evaluations); but, I still always feel a bit guilty about them… So, as I prepared for class this week, I read a little bit of the academic research on teaching.

Are class competitions damaging?

My guilt stems from two things… First, because many of the students get engrossed in the competition, they end up spending a considerable (maybe ridiculous) amount of time on them!  This is despite the fact that grades are not affected by the competition…

I typically have the grade for the student teams come from beating various TA teams, which are designed to differentiate the thought that went into the various algorithms.  But, the students rarely stop at just beating the best TA team.  For example, last year in rankmaniac, the best TA team took ~ 2hrs (a naive solution took 4-5 hours), and yet every student team took less than 40min!  …and to get there many students spent more than 30hrs in one week for this one class.  That’s clearly too much!

But, at the same time, many of those teams were the ones that enjoyed it the most — and it certainly was a memorable learning experience for them.  I know that the courses from my undergrad that I remember the best (and most fondly) had some sort of competition element — 15-213 at CMU was one of my favorites!

So, this first worry doesn’t bother me that much.  As long as I ensure that it’s a useful learning experience, then the time investment is in some sense “worth it” … especially if the students are doing it of their own volition (and not because of grades).

The bigger worry for me is that, a lot of academic research on teaching that is pretty negative about class competitions…and not just because of a negative view on competition itself.  The negative view seems to stem mainly from two reasons:

  • There may be gender differences in how students respond.  For example, studies have found that young boys often prefer games where one succeeds at the expense of another, while young girls often engage in games where success does not come at the expense of others.
  • Cooperative learning (where the whole class works together to solve problems) has tended to have improved learning outcomes to competition-based learning in academic studies.

Sharon Nichols and Jeremy Sullivan have a nice survey that highlights these issues and suggests reasons why cooperative learning may be preferred — it really pushes the idea that class competitions can be damaging to classroom learning…

Can class competitions be healthy?

Despite the academic critiques, I find it hard to give up on class competitions…especially since they played a very positive role in my education and I haven’t seen the negative effects in my classes that studies worry about.

So, after the lit review, my feeling is that, as with many teaching tools, one just needs to be careful in how class competitions are used.  For example, I’m very careful to make sure that the students view it as a learning exercise.  Most of the grade comes from a report about what they did in the algorithm design and what they learned from the experience.  I always make sure that the groups share the ideas with the whole class after the competition, so that they view it as a global learning experience. And, I make sure the students are working in groups during the competition,  so that they still can have a cooperative experience, exchanging ideas within the group and learning from each others’ strengths during the competitions.

I was happy to find that exactly these points were highlighted in a number of the papers as key features for ensuring “healthy competition” in the classroom…

In any case, I’ll be using competitions again this year, so I hope it goes well again…  In particular, I’m really excited to see what the students come up in terms of algorithms for competing cascades (pandemaniac) — since this is a very hot and important area of research currently!


4 thoughts on “Class competitions: Motivating or damaging?

  1. Adam, Thanks again for a very nice post. I wanted to share my experiences on class competitions. Whenever I teach a fourth-year networking class, I split it into three or four groups, each with eight to ten teams. Groups compete and teams collaborate on a grand challenge problem. In 1998, for example, groups built an entire soup-to-nuts Internet telephony system that allowed both telephones and desktops to participate in a teleconference. In 2007, groups implemented a peer-to-peer video distribution system using smart phones, developing algorithms for disconnection-tolerant data transfer and routing. Students love these challenges and many of them have gone on to continue to work in computer networking as a career. I find that by combining competition and collaboration, this approach seems to avoid the hyper-competititveness that plagues pure competitions.

    thanks keshav

  2. Pingback: Rigor + Relevance | Releasing Rankmaniac

  3. Pingback: Rigor + Relevance | Competing epidemics: Crowning a pandemaniac

Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s