Teaching Algorithm Design: A Literature Review | Hacker News Reader

Teaching Algorithm Design: A Literature Review

arxiv.org

87 pointsbelter8 comments

The vast majority of this paper discusses the methodology.

The actual part that you would use in teaching is miniscule.

It's education research, of course there are no randomized trials... you could never believe them anyway.

Honestly, this preprint is not worth reading. If you want the takeaways, jump to section 5 on page 11. You will find things like "intentional problem selection" and "problems with a variety of solutions".

Possibly of note "introducing visualizations provides no statistically significant learning outcomes".

Papers are quoted but there is no digest version of the results (as far as I could see).

You would be better off just reading an evidence-based book on teaching, like "How People Learn" [1] or the related "How Students Learn" [2]. At least those books go somewhere and present results. Updated references would be welcome.

[1] https://nap.nationalacademies.org/catalog/9853/how-people-le... [2] https://nap.nationalacademies.org/catalog/11102/how-students...

I had a similar reaction. I plan on using this paper mostly for it's bibliography.

Active learning and problem selection is what I came to as well going through the literature. I was thinking of teaching algorithm design by having students build their own algorithm laboratory where they create visualizations and experiments that motivate interest in problems and the design of solutions.

A build your own algorithm laboratory sounds fun.

I think “motivating the problem” (a phrase I hear a lot in American lectures) is often the weakest part of an algorithm and data structure course. The teaching approaches seem quite abstract, or make passing mention to practical problems.

Personally I find algorithms a bit boring in the abstract. I’ve always wondered why DSA projects are so rarely things like “Here’s a simple database that doesn’t support indexes. We’re going to query it and it will be painful. Then you’ll extend the database with a b-tree index.”

> “Here’s a simple database that doesn’t support indexes. We’re going to query it and it will be painful. Then you’ll extend the database with a b-tree index.”

Because this is in my opinion a too complicated project for many computer science students (of a non-elite university). In other words: this looks like a great project to "weed out" students who should better not study computer science. Lectures/projects to weed out unusuitable student seem not to be accepted in the academic environment in the USA.

Not really. I had projects like that at Georgia Tech, I know Carnegie Mellon, MIT, Stanford, etc have similar.

That is why I wrote "(of a non-elite university)".

That sounds like a really fun way to learn!

Reply on news.ycombinator.com