Friday, April 22, 2011

Paper Reading #25, "A Code Reuse Interface for Non-Programmer Middle School Students"

http://angel-at-chi.blogspot.com/2011/04/paper-reading-19-tell-me-more-not-just.html
http://csce436-hoffmann.blogspot.com/2011/04/paper-reading-24-using-language.html

A Code Reuse Interface for Non-Programmer Middle School Students

Paul A. Gross, Micah S. Herstand, and Caitlin L. Kelleher, Washington University in St. Louis
Jordana W. Hodges, University of North Carolina

IUI’10, February 7–10, 2010, Hong Kong, China

Summary
This paper describes a tool to assist in code reuse for novice programmers, especially middle schoolers. It is specifically associated with a format called "Looking Glass IDE" that enables the users to create animated stories. This is considered a desirable area of study because it is believed that the middle school stage is a critical point in the process in attracting boys & girls to the computer field.

The project, which was never given a name beyond code reuse interface, functions by allowing users to save a script for an action, say one object running into another and knocking to over, generalize it, and then re-insert it elsewhere or in another program specifying new characters to assume the roles. It includes protections against compilation errors, say ordering a table to run.

The results of the study, which are provided in some detail in the paper, appear to have been reasonably satisfying. Out of 47 subjects, 77% were able and motivated to create a program with more than five lines of code. Also successful was the "social propagation" of code and ideas among the students.

Discussion
This paper did a very good job of presenting the work in an understandable fashion, although whether that was because the authors were good writers or because the material was inherently simpler since it was ultimately intended for middle schoolers is an open question.

The paper is significant because producing more good programmers is always significant.

Given infinite time and resources, the best next work would be to give this to a bunch of students and see if it increased either their interest or ability in coding relative to their peers, measured five or more years after the initiation of the test.

The potential flaw with the whole concept is that the program exists. I am not convinced that the need for novice stages in things like reading and coding is necessary. Rather than learning a cut-and-paste pre-assembled language, perhaps we should start folks out in real programming languages, just like we start 'em out in math with real equations, rather than reassembling other's. Just a thought, I'm no expert.

From the paper.

2 comments:

  1. This is a really good idea. I did not have a single programming class until I came to college, and the idea of training young people's brains to think like a programmer at a young age would greatly benefit our profession.

    ReplyDelete
  2. This kind of project is really important in order to get a new generation to like this type of industry. I hope they include a little more of what a programmer would really be doing in this however, but it is still interesting.

    ReplyDelete