visual programming: computing resources to unleash k-12 creativity joel adams, ph.d. department of...

37
Visual Programming: Computing Resources to Unleash K-12 Creativity Joel Adams, Ph.D. Department of Computer Science Calvin College 2012 Michigan Tapestry

Upload: hilary-norris

Post on 25-Dec-2015

213 views

Category:

Documents


0 download

TRANSCRIPT

Visual Programming:Computing Resources

to Unleash K-12 Creativity

Joel Adams, Ph.D.Department of Computer Science

Calvin College

2012 Michigan Tapestry

22012 Michigan Tapestry

A Problem

Many high school students believe: - computing jobs are boring- only nerds study computer science- computing = no social life- computing involves no creativity- all the jobs are going to Asia …

32012 Michigan Tapestry

CS Bachelors Degrees (U.S.)

42012 Michigan Tapestry

What Are The Facts?

According to the U.S. Bureau of Labor Statistics…

52012 Michigan Tapestry

62012 Michigan Tapestry

72012 Michigan Tapestry

82012 Michigan Tapestry

Solving The Problem

How can we attract students to computing and dispel the stereotypes? Research suggests we

need to engage students in middle school or earlier, before the negative stereotypes get set. If we wait until high school, it may be too late.

92012 Michigan Tapestry

CSTA

The Computer Science Teacher’s Association has defined K-12 Computing Curriculum SLOs:Level 1 (K-6):

CS and Me

Level 2 (6-9). CS and Community

Level 3 (9-12). CS in the Modern World CS Concepts and

Practices Topics in CS

How Do We Engage Students?

102012 Michigan Tapestry

Many of today’s students are visual learners

- We need visual tools to engage them

Demos

112012 Michigan Tapestry

Alice and Scratch at Calvin

• Imaginary Worlds Camps at Calvino Summer camps for middle school and upo Roughly 300 campers since 2003o 2003-07: Storytelling using Aliceo 2008-11: Games | music videos using

Scratcho Same concepts taught in both versions

(variables, selection, repetition, abstraction)oNoticeable differences in campers’ questions

• What are IWC campers learning?122012 Michigan Tapestry

Bloom’s 3 Lowest Learning Levels

1. Knows: Can recall or recognize ideas and information in the form they were learned

2. Comprehends: Can interpret or translate information based on prior learning

3. Applies: Can transfer or use principles or data to solve a problem or task

132012 Michigan Tapestry

IWC Projects

• Each IWC camper completes and demos an open-ended project at the camp’s Showcase Session

• We have a corpus of 322 projects…o 209 Alice 2.0 storytelling projectso 103 Scratch gaming projectso 10 Scratch music video projects

• All projects available at alice.calvin.edu

142012 Michigan Tapestry

Idea

• Study those projects to see what computing concepts campers are applying in themo If campers use a concept in their project, they

are reaching Bloom level 3 wrt that concept (variables, selection, repetition, abstraction)

o Count occurrences of variables, if statements, loop statements, subprograms, …

o Count animation constructs common to both Alice and Scratch (move, say/think, wait, …)

o Count specific objects (e.g., fire animations)152012 Michigan Tapestry

Research Question

Are there any significant differences between the different project genres (storytelling, music video, game) with respect to the concepts that campers use/apply?

We wrote scripts to count these constructs, and normalize the counts (per 100 lines)

162012 Michigan Tapestry

The Short Answer

We found statistically significant differences (p < .01) in the number of:oVariableso If statementso Loop statementsoDialog (say/think) messageso…

used in the different project genres.

172012 Michigan Tapestry

Variable Declarations Per 100 Lines

182012 Michigan Tapestry

Significance of Differences:+ Game vs Video: p = 5.4e-7+ Game vs Storytelling: p = 1.54e-7– Video vs Storytelling: p = 0.71

Percentage of Projects Using Variables

192012 Michigan Tapestry

Significance of Differences:– Game vs Video: p = 0.082+Game vs Storytelling: p = 2.095e-18– Video vs Storytelling: p = 0.629

If Statements Per 100 Lines

202012 Michigan Tapestry

Significance of Differences:+Game vs Video: p = 1.25e-7+Game vs Storytelling: p = 7.09e-37– Video vs Storytelling: p = 0.070

Loop Statements Per 100 Lines

212012 Michigan Tapestry

Significance of Differences:– Game vs Video: p = 0.951+Game vs Storytelling: p = 1.1e-5+Video vs Storytelling: p = 0.0064

Subprograms

• Alice 2.0 provides fully parameterized methods

• Scratch 1.4 provides parameterless message-handlers for broadcasts – A build-your-own-block mechanism is coming

in Scratch 2

We decided these abstraction mechanisms were too different to compare fairly.

222012 Michigan Tapestry

Project Length (Total Lines of Code)

232012 Michigan Tapestry

Significance of Differences:+Game vs Video: p = 0.0014– Game vs Storytelling: p = 0.083– Video vs Storytelling: p = 0.136

Dialog (Say/Think) Msgs Per 100 Lines

242012 Michigan Tapestry

Significance of Differences:– Game vs Video: p = 0.392+Game vs Storytelling: p = 1.1e-46+Video vs Storytelling: p = 8.24e-10

Age Differences

We compared projects of younger (11, 12) campers vs older (13, 14) campers:

– Variables– If statements– Loop statements– Objects– Lines of code – Use of particular constructs

We found just 2 significant differences…

252012 Michigan Tapestry

Alice Lines of Code By Age

262012 Michigan Tapestry

Significance of Difference:+p = 4.15e-5

Scratch PickRandom Per 100 Lines By Age

272012 Michigan Tapestry

Significance of Difference:+p = 0.0083

Gender Differences

We compared projects of male vs female campers for differences in the number of:

– Variables– If statements– Loop statements– Objects– Lines of code – Use of particular constructs

We found just 3 significant differences…

282012 Michigan Tapestry

Scratch Loop-Types Per 100 Lines by M/F

292012 Michigan Tapestry

Significance of Differences:– repeat n times (p = 0.28)+forever (p = 2.4e-7)

– forever if (p = 0.082)

– repeat until (p = 0.11)

Alice Dialog Msgs Per 100 Lines by M/F

302012 Michigan Tapestry

Significance of Difference:+ p = 3.67e-5

Alice Fire Animations Per Project by M/F

312012 Michigan Tapestry

Significance of Difference:+ p = 1.81e-9

Some Conclusions

• Visual tools like Alice and Scratch help students visualize and master programming abstractions– Age affects students’ ability to master

abstract concepts like randomness.

322012 Michigan Tapestry

Some Conclusions

• The game, music video, and storytelling project genres differ markedly in what they motivate students to use in open-ended projects (i.e., learn to apply at Bloom level 3).– Storytelling projects are good at teaching

sequential / algorithmic thinking– Games motivate students to learn to use

basic programming concepts like variables and control structures

332012 Michigan Tapestry

Constructs Per 100 Lines by Genre

342012 Michigan Tapestry

Some Conclusions

• Given an open-ended storytelling project, boys and girls tell very different kinds of stories, on average.– Stereotypical tastes begin early!

352012 Michigan Tapestry

Some Conclusions

• Alice and Scratch:– Both eliminate syntax error frustration, helping

students focus on logic, master concepts – Scratch has the easier learning curve– Scratch’s social networking site lets students

easily share their projects– Scratch’s 2D graphics let students create their

own scenes and characters– Alice’s 3D graphics + Sims models are cool– Alice’s objects and methods bridge to Java and

OOP

362012 Michigan Tapestry

Resources

372012 Michigan Tapestry

• Scratch: scratch.mit.edu– Educators resource site:

scratched.media.mit.edu– A full middle school Scratch curriculum is

available: colleenmlewis.com/scratch/

• Alice: alice.org• CSTA: csta.acm.org• CS Principles: www.csprinciples.org• Computing in the Core:

www.computinginthecore.org

Thank [email protected]