Skip to main content

C'est la Z

Should every assignment be open ended?

This post by my friend Alfred caught my eye yesterday. It's a good post - some good examples of inserting student creativity even into small intro level assignments. While I like the post and agree with the sentiment of open ended, student driven projects, I had to take issue with the lead quote Alfred used:

“If you assign a project and get back 30 of the same thing, that’s not a project, that is a recipe.”

– Chris Lehmann

Open ended projects and student driven assignments are great and important but so are ones with similar or identical solutions and they're also not recipes.

First, there's a certain level of utility in a uniform assignment. If the assignment is "write a program to convert Fahrenheit to Celsius" you know that all the students know the conversion as you can provide it and it also allows them to easily check their work, either through an autograder or with each other. Yes, I know, some students students could be tempted to cut corners and copy solutions but usually attempts to control cheating do more to constrain students doing things the right way than to prevent the cheaters from cheating.

Also, some student, particularly early on need the structure of a set assignment and don't do as well with a "do anything assignment." Saying "just make a converter" can be much more stressful and difficult than convert from pounds to kilograms.

As to recipes? Not necessarily. Let's take this small assignment:

Given two variables a and b, write a code fragment that will swap the values represented by a and b. That is, if at the start, a represented 5 and b represented 10, after your code, a would represent 10 and b would represent 5.

You could even add a hint:

HINT: Consider using an extra third variable in your solution.

This isn't a recipe, it's an assignment. Even with the hint it doesn't tell the students exactly what to do, just the desired outcome. It's not a recipe but most or all of the solutions will look remarkably similar.

By contrast, a swap "recipe" might read:

  1. Create a new variable named extraBucket

  2. Assign the value currently in b to extraBucket

  3. Assign the value currently in a to b

  4. Assign the value currently in extraBucket to a

This is a recipe. A student would literally just rewrite the steps into the solution but even this might have it's place in a class.

A more advanced assignment that leads students through building and understanding something like a merge or quicksort could be set up similarly. Not in any way a recipe but rather a valuable exercise where all the students end up with very similar solutions. This is not necessarily a bad thing, in fact it was all the rage a few years ago as it could be set up as a discovery or experiential learning experience. I also poo pooed that as being the silver bullet but as with most teaching techniques, it has its place.

Then there are opportunities to introduce student creativity even when the assignments are, at their core, all the same. I recall a session at either SIGCSE or CSTA by Bootstrapworld. They were describing one of their classes major projects. It was a game. I forget the details but at the core, all the student projects were the same. They differed only in the graphics they used and the story around the game. That's where Boostrapworld interjected student creativity but the core CS and math of the student projects were all the same. Did the fact that all the assignments were essentially the same make this bad or a recipe? Certainly not. Students go to express themselves uniquely but at the same time the assignment was constrained and focused on what the Bootstrapworld designers wanted.

So, there you have it. There's nothing wrong with assignments where the student results are largely the same. There's also nothing wrong with student projects that are 100% open ended. It's all about the teacher having multiple tools and using them as appropriate for their students.

comments powered by Disqus