Heidi Ellis
Professor at Western New England University
Gregory Hislop
Professor at Drexel University
All Things Open
October 26-27, 2016
Raleigh, North Carolina
Student Pipeline to Open Source Communities using HFOSS
1. Building the Student Pipeline to
Open Source Communities
Using HFOSS
Heidi Ellis, Western New England University
Greg Hislop, Drexel University
2. We Are...
● Professor, Western New
England University
● Co-maintainer GNOME
MouseTrap
● Founding Member HFOSS
● Professor, Drexel
University
● Foss2serve team lead
● Software industry veteran
3. Where Are We Going?
● Student Participation in Open Source
● Challenges
● Helping instructors prepare to teach FOSS
● What Can Students Do?
● Research Results
● Next Steps
5. Current Status of FOSS in Education
Open SourceHigher Ed Industry
Students who learn FOSS Companies that use FOSS
Industry
6. What Students Know
● Open source software exists
– They can use the stuff without paying for it
● It's legal to do so
7. What Most Students Don't Know
● Development
– FOSS development tools
– Processes and techniques
● Organization and scale
– The extent of FOSS – number and scale of projects
– FOSS career opportunities
– Licensing
– FOSS culture
8. Survey Results
● Survey of Junior and Senior students at Drexel
– Multiple choice and fill in the blank questions
– Goal: get a baseline of student familiarity with FOSS
concepts
9. Survey Results
● Less than 10% of students
– Had ever heard of “The Cathedral and the Bazaar”
– Know that a lot of FOSS is written by people paid to do
so
● No concept of FOSS as a career area
● Most students think that the average number of
contributors to a FOSS project is > 50
11. Faculty Reward Structure
● Research emphasis
– Publish or Perish
– Education research not always
valued
– Peer-review is key
● Education innovation
– Not always welcome
12. Faculty Learning Curve
● Relatively few CS faculty have significant software
engineering experience
– Large code bases
– Active client communities
– Product management, evolution, packaging, distribution, and support
● Very few CS faculty have FOSS experience
– Tools and processes
– The Open Source Way
– Licensing, business models, etc.
14. Mismatch in FOSS and Academic Timelines
Fedora Project:
● Release Cycle: 6 mo
● Planning Cycle: ~1 yr
● Execution Time: 6-12 mo
Academia
● Release Cycle: 1-4 yrs
● Planning Cycle: 2-5 yrs
● Execution Time: 4 yrs
15. Pedagogical Expectations
● Instructor as expert
– Difficulty in being “productively lost”
● Meeting course learning outcomes
– Difficult with unpredictable events in FOSS
– Difficult to achieve the same outcomes for all students
● Short, fixed time frame of academic courses
17. POSSE v1.0
● Professors Open Source Summer Experience
● Red Hat outreach initiative for college faculty
– 3-5 days immersive introduction to FOSS
– Focus on open source culture, tools, processes etc.
● Challenges
– Funding for travel
– Covering instructional issues
18. POSSE v2.0
● Professors Open Source Software Experience
● Collaboration of college faculty and Red Hat
– Plus NSF funding
● Coverage of instructional issues along with FOSS
introduction
● Three stage model
19. POSSE v2.0
● Stage 1: Online
– Static, factual content
● Stage 2: In-person workshop
– Dynamic, situational,
human-centered content
● Stage 3: Small group
collaboration
– Community building
24. OpenMRS
● Development Environment
– Explore VM's and containers to package demo or
development environment
– Writing more installation instruction
●Translating system messages to multiple languages
– Updates and additions to French
– Additions in Polish
– Attempt in Arabic
26. Research Questions
● What is the impact of student participation in HFOSS on:
– Motivation to study computing
– Perceived learning related to computing
– Choice of major and career plans
Motivation SW Engineering Major and Career
Working on an HFOSS
project gives me a better
appreciation for the
usefulness of computing.
I can list the steps in
the software process
we used in HFOSS
project
The subject matter of this
HFOSS project is highly
relevant to my future career
plans.
27. Study Context
● 2006 - 2015
– Multiple institutions
– Different courses
– Additional data collection
● Builds on instructor
community
– Prior attendees at Professor’s Open Source Software
Experience (POSSE) workshops
28. Study Design
● Pre and post anonymous student survey
– Five-point Likert scale plus open ended questions
● Strongly disagree, Disagree, Neutral, Agree, Strongly
agree; Don’t know, Not applicable
– Mann-Whitney U used for testing significance
– Sample size
● Pre-course: 115
● Post-course: 94
–
29. Courses in the Study
Title Term(s) Students Length of Term
Software Engineering (WNE)Fall 2013,
Fall 2014
6,8 15 weeks
Software Engineering (URI) Fall 2014 40 15 weeks
Software Engineering (Mor) Spring 2014 10 15 weeks
Software Engineering (TCNJ)Fall 2013 19 15 weeks
Software Engineering (Mac) Winter 2014 15 13 weeks
Open Source Software
Development (WOU)
Winter 2014 20 10 weeks
30. Student Profile
● Major - 95% computing related
● Gender
– 17% female
– 8% not identified
– 75% male
31. Results – Motivation and Career
Q Description p-val < Diff Post Avg
H10 Participating in an HFOSS project made
me more comfortable with computing.
.001 2.1 3.72
G1 I am confident about working with
computing professionals
.038 0.24 3.92
G8 I have a high level of experience in the
HFOSS subject matter.
.000 0.88 2.89
32. Results – Computing Learning
Q Description p-val < Diff Post Avg
SE3 I can use a software process to develop an
HFOSS project.
.001 0.75 3.90
SE6 I can describe the impact of project
complexity on the approaches used to
develop software.
.001 0.50 3.75
SE10 I can use all tools and techniques
employed in my HFOSS project.
.001 0.69 3.65
SE11 I can participate in an HFOSS
development team’s interactions.
.038 0.28 3.83
33. Results – Negative Changes
Q Description p-val < Diff Post Avg
H8 Working on an HFOSS project has
increased my interest in computing.
.001 -.44 3.69
H9 Working on an HFOSS project has
increased my confidence in my
computing ability.
.009 -.36 3.79
34. Results – Gender and Ethnicity
● Comparing post-course scores
● Stronger agreement by the women
– H2. I have a greater awareness of the potential for computing to
benefit society due to working on an HFOSS project.
– p = 0.036, average for males = 4.04; average for females = 4.5
36. Key Activities
● POSSE - More workshops to be offered over the next
several years
● Teaching materials
– Re-construction of teachingopensource.org
– Teaching materials sprints
● Mapping teaching materials to existing curricula
Reason 1: Tenure-track instructors are under pressure to:
Publish, publish, publish
Peer-reviewed and there is a hierarchy of publication venues
There is NO venue for open source education
Attract grant funding,
Image CC-BY-SA by
Reason 1: Tenure-track instructors are under pressure to:
Publish, publish, publish
Peer-reviewed and there is a hierarchy of publication venues
There is NO venue for open source education
Attract grant funding,
Image CC-BY-SA by
Reason 1: Tenure-track instructors are under pressure to:
Publish, publish, publish
Peer-reviewed and there is a hierarchy of publication venues
There is NO venue for open source education
Attract grant funding,
Image CC-BY-SA by
Reason 1: Tenure-track instructors are under pressure to:
Publish, publish, publish
Peer-reviewed and there is a hierarchy of publication venues
There is NO venue for open source education
Attract grant funding,
Image CC-BY-SA by
Image CC-BY-SA by https://www.flickr.com/photos/blondinrikard/13927054414/
Warning: School specific:
The department (CS or SE) has to create a curriculum that fits together as a whole *and* that students can work through in 4 years. Given that, departments can&apos;t “afford” to add courses that don&apos;t fit clearly into a learning path.
Schools have a limited resource budget – both in terms of instructor hours and in terms of things like available classroom space. There&apos;s competition for that time/space.
When courses can be added – (and the process is documented) – it can take years.
Bump in enrollments is a plus, but interferes with new course development
Image CC-BY-SA by https://www.flickr.com/photos/blondinrikard/13927054414/
Reason 1: Tenure-track instructors are under pressure to:
Publish, publish, publish
Peer-reviewed and there is a hierarchy of publication venues
There is NO venue for open source education
Attract grant funding,
Image CC-BY-SA by
Reason 1: Tenure-track instructors are under pressure to:
Publish, publish, publish
Peer-reviewed and there is a hierarchy of publication venues
There is NO venue for open source education
Attract grant funding,
Image CC-BY-SA by