Posts tagged ‘Media Computation’
In the last couple of months, I have had the opportunity to speak to groups of Engineering Education Researchers. That doesn’t happen often to me, and I feel very fortunate to get that chance.
I was asked to speak about my vision for the future of Engineering Education, from my perspective as a Computing Education Researcher. What I said wasn’t wholly unique–there are Engineering Education Researchers who are already working on some of the items I described. The response suggested that it was at least an interesting vision, so I’m telling the story here in blog form.
For readers of this blog who may not be familiar with Engineering Education Research, the Wikipedia page on EER is pretty good. The most useful paper I read is Borrego and and Bernhard’s “The Emergence of Engineering Education Research as an Internationally Connected Field of Inquiry.” I also recommend looking around the Purdue Engineering Education department website, which is the oldest Eng Ed department in the US.
Engineering has had a long relationship with computing. Engineers made computing part of their practice earlier and more pervasively than scientists or mathematicians. I love how this is described in the motion picture Hidden Figures where Octavia Spencer’s character is part of the effort to use computing as soon as possible in the American space program. Engineering educators have made computing part of the learning goals for all of today’s engineering students, again more pervasively than what I can see in science or mathematics programs.
Much of my work and my students’ work is about embedding computing education (e.g., Media Computation which embeds computing in the digital media context that students value, or Brian Dorn’s work embedding computing in a graphic design context) and tailoring computing education (e.g., high school CS teachers need something different from software developers). Computing education can be embedded in Engineering classes and tailored for Engineering students, of course. My vision is about embedding and tailoring engineering education.
There are three parts to the story below:
- Engineering Education for everyone K-16, especially for STEM learners.
- Reaching a diverse audience for engineering education.
- Recognizing the differences between Engineering Education research and teaching, and the need for more research on learning outside of the engineering classroom.
In January 2016, President Barack Obama launched the “CS for All” initiative. When he said that he wanted students to be “job-ready,” he wasn’t saying that everyone should be a software engineer. Rather, he was reflecting a modern reality. For every professional software developer, there are four-to-nine end-user-programmers (depending on the study and how you count). Most professionals will likely use some form of programming in the future. That’s an argument for “CS for All.”
We also need Engineering for All. Engineering skills like designing, planning, collaboration on diverse teams, and trouble-shooting are needed across STEM. When I look at bench science, I see the need for engineering — to design, plan, collaborate, debug, and test.
Engineering education researchers know a lot about how to teach those skills. I’d love to learn how to inculcate some engineering perspectives in my CS students. When I see Chemical Engineering students designing a plant, or Civil Engineering students designing a bridge, they predict that they made mistakes, and they look for those mistakes. There’s a humility about their process. CS students often run their program once and turn them in. If you write a hundred lines of code, odds are almost 100% that you made errors. How do we get CS students to think that way?
Engineering for All is different than what professional engineers do, in the same way that what a high school teacher needs is different than what a professional software developer needs. Both need a mental model of the notional machine. A high school teacher also needs to know how students get that wrong, and probably doesn’t need to know about Scrum or GitHub.
I believe that there is a tailored part of engineering education which should be embedded throughout K-16 STEM. The American Society of Engineering Education’s mission is focused on professional engineers, and my proposal does not diminish the importance of that goal. We need more professional engineers, and we need to educate them well. But engineering skills and practices are too important to teach only to the professionals.
Engineering should play a significant role in STEM education policy. Engineering education researchers should own that “E” in STEM. There are many research questions that we have to answer in order to achieve Engineering for All.
- What is the tailored subset of engineering that should be taught to everyone? To STEM learners?
- All technically literate US citizens should know far more about engineering than they do today. Here’s a hypothesis: If all US citizens understood what engineering is and what engineers do, we might have less crumbling infrastructure, because we citizens would know that infrastructure is critical and professional engineers design, build, and maintain infrastructure. How do we get there?
- All K-12 students should have the opportunity to fall in love with engineering. How?
- Are there limits to what we can teach about engineering in K-16? What learning and cognitive disabilities interfere with learning engineering, and what parts of engineering? I also wonder about the kinds of bias that prevent someone from succeeding in engineering, besides race and gender. For example, here in the South, there are a lot of students who don’t believe in evolution. I’m pretty sure that belief in evolution isn’t necessary for designing a bridge or a distillation column. But someone who believes in intelligent design is going to face a lot of barriers to getting through basic science to become an engineer. Is that how it should be?
- Engineering should aim to influence K-12 STEM education nationally, in every state.
The American University (particularly the Land Grant University, developed in the late 1800’s) was supposed to blend the German University focus on research and the British focus on undergraduate education. My favorite history of that story is Larry Cuban’s How Scholars Trumped Teachers, but Michael Crow also tells the story well in his book Designing the New American University. We believed that there were synergies between research and teaching. It’s not clear that that’s true.
Research and teaching have different measures of success and don’t feed directly into one another.
Teaching should be measured in terms of student success and at what cost. Cost is always a factor in education. We know from Bloom’s two-sigma 1984 study (and all the follow-ups and replications) that the best education is an individual human tutor for each subject who works with a student to mastery. But we as a society can’t afford that. Everything else we do is a trade-off — we are trying to optimize learning for the cost that we are willing to bear.
Research should be measured in terms of impact — on outcomes, on the research community, on society.
It’s quite likely that the education research on a given campus doesn’t influence teaching practice on the same campus.
I see that in my own work.
- The best of Media Computation is no longer at Georgia Tech. Beth Simon and Leo Porter at UC San Diego have done better studies and are inventing cool interventions like MediaComp art galleries. Cynthia Lee at Stanford has created MediaComp for multiple languages. Celine Latulipe built on Beth and Leo’s work to implement lightweight teams in her MediaComp course.
- Subgoal labeling totally works (see Lauren’s dissertation or Briana’s dissertation). Coursera uses it in some of their videos. Rob Miler at MIT has picked it up. But there are very few CS classes using it.
We can see the transition for education research idea to impact in teaching practice as an adoption curve. Boyer’s “Scholarship Reconsidered” helps to explain what’s going on and how to support the adoption. There is traditional Scholarship of Discovery, the research that figures out something new. There is Scholarship of Teaching that studies the practice of teaching and learning.
Then there’s Scholarship of Application, which takes results from Discovery into something that teachers can use. We can’t expect research to influence teaching without scholars of application. Someone has to take the good ideas and carry them into practice. Someone has to figure out what practitioners want and need and match it to existing research insights. Done well, scholarship of application should also inform researchers about the open research questions, the challenges yet to be faced.
High-quality teaching for engineering education should use the most effective evidence-based teaching methods.
Good teachers balance teaching for relevance and motivation with teaching for understanding. This is hard to do well. Students want authenticity. They want project-based learning and design. I was at the University of Michigan as project-based learning for science education was first being developed, and we knew that it very often didn’t work. It’s often too complex and leads to failure, in both the project and the learning. Direct instruction is much more efficient for learning, but misses out on the components that inspire, motivate, and engage students. We have to balance these out.
We have to teach for a diverse population of students, which means teaching differently to attract women and members of under-represented groups. In our ICER 2012 paper, we found that encouragement and self-perception of ability are equally important for white and Asian males in terms of intention to persist in computing, but for women and under-represented group students, encouragement matters more than ability in terms of how satisfied they are with computing and intention to persist. This result has been replicated by others. Encouragement of individual students is critical to reach a diverse audience.
An important goal for a first year Engineering program is to explain the relevance of the classes that they’re taking. Larry Cuban tells us that a piece of the British system that got lost by the early 1920’s in the American University was having faculty advisors who would explain how all the classes fit together for a goal. The research on common first year Engineering courses (e.g., merging Physics, Calculus, Engineering in a big 12 credit hour course) shows that they worked because they explained the relevance of courses like Calculus to Engineering students. I know from my work that relevance is critical for retention and transfer.
Do students see relevance of first year Engineering programs? Most first year programs emphasize design and team problem-solving. First year Engineering students don’t know what engineers do. When they’re told “This is Engineering” in their first year, do they believe it? Do they cognitively index it as “real Engineering”? Do they remember those experiences and that learning in their 3rd and 4th years when they are in the relevant classes? I hope so, but I don’t know of evidence that shows us that they do.
Engineering education research, like most discipline-based education research (DBER), is focused on education. I see the study of “education” as being about implementation in a formal system. Education is a design discipline, one of Simon’s Sciences of the Artificial. Robert Glaser referred to education as psychology engineering.
We need more research on Engineering Learning. How do students learn engineering skills and practices, even outside of Engineering classes? How do those practices develop, even if it’s STEM learners and teachers using them and not professional engineers? How should we best teach engineering even if it’s not currently feasible?
That last part is much of what drives my work these days. We’re learning a lot about how great Parsons Problems are for learning CS. Very few CS classes use them. There are reasons why they don’t (e.g., they’re emphasizing the project side of the education spectrum). I’m figuring out how to teach CS well, even if it’s not feasible in current practice. CS teaching practice will eventually hit a paradigm shift, and I’ll have evidence-based practices to offer.
To focus on engineering learning requires work outside the classroom, like Multi-Institutional, Multi-National (MIMN) studies that we use in computing education research, or even laboratory studies. A focus on Engineering Learning creates new opportunities for funding, for audience, and for impact. For example, I could imagine engineering education researchers seeking science education funding to figure out how to teach high school science teachers the engineering that they ought to teach their students — not to introduce engineering, but to make their students better in science.
My vision for engineering education has three parts:
- K-16 STEM learners need Engineering for All. Engineering education has more to contribute than just for producing more professional Engineers. Engineering education ought to own the “E” in STEM education policy. Engineering skills and practices can be tailored to different audiences and embedded in STEM education.
- Reaching a diverse audience is critical for both research and teaching. For me, that diversity includes the people who need engineering education who aren’t going to become professional engineers, but also people who look different or even have different beliefs.
- Finally, research and teaching are different activities, with different measures of success. Teaching should be informed by evidence and be as efficient and effective as possible for a given cost. We need evidence for what we’re doing, and we should gather evidence if we don’t know if what we’re doing is working. Research should focus on what’s possible and on having impact, even if that impact isn’t in the on-campus classrooms. We shouldn’t expect research to impact teaching without explicit investment in adaptation to support adoption.
(Thanks to Barb Ericson, Beth Simon, Leo Porter, and Wendy Newstetter for advice on drafts of this piece.)
Every University Student should Learn to Program: Guzdial Arguing for CS for All in Higher Education
A colleague recently approached me and said, “It would be useful if Universities got involved in this CS for All effort. All Universities should offer courses aimed at everyone on campus. There should be a systematic effort to get everyone to take those classes.”
I agree, and have been making this argument for several years now. I spent a few minutes gathering the papers, blog posts, and book where I’ve made that argument over the last decade and a bit.
In 2002, Elliot Soloway and I argued in CACM that we needed a new way to engage students in intro programming: Teaching the Nintendo Generation to Program.
In 2003, I published the first paper on Media Computation: A media computation course for non-majors.
In 2004, Andrea Forte led the team studying the Media Computation class at GT:Computers for communication, not calculation: Media as a motivation and context for learning and A CS1 course designed to address interests of women.
In 2005, Andrea Forte and I presented empirical evidence about the courses we’d designed for specific audiences: Motivation and nonmajors in computer science: identifying discrete audiences for introductory courses. I published a paper in CACM about how the courses came to be at Georgia Tech: Teaching computing to everyone.
In 2008, I offered the historical argument for teaching everyone to program: Paving the Way for Computational Thinking.
We’ve published several papers about our design process: Imagineering inauthentic legitimate peripheral participation: an instructional design approach for motivating computing education and Design process for a non-majors computing course.
My 2013 ICER paper was a review of a decade’s worth of research on Media Computation: Exploring hypotheses about media computation
My keynote at VL/HCC 2015 was on how computing for all is a requirement for modern society: Requirements for a computing-literate society
My 2015 book is, to a great extent: an exploration of how to achieve CS for All: Learner-Centered Design of Computing Education: Research on Computing for Everyone.
In blog posts, it’s been a frequent topic of conversation:
- In 2011, I argued that it makes more sense to require CS at universities before pushing into K-12, because then all pre-service teachers have some CS which makes later PD much easier and cheaper: https://computinged.wordpress.com/2015/11/30/require-cs-at-universities-before-k-12-computational-community-for-everyone/ and https://computinged.wordpress.com/2011/05/17/if-you-want-cs-in-high-school-require-cs-in-college/
- In 2013, I pointed out that CS is becoming increasingly valuable outside of CS: https://computinged.wordpress.com/2013/12/10/why-are-english-and-lots-of-other-majors-studying-computer-science/
- One of my earlier Blog@CACM posts was on how students learn things in MediaComp that informs them about their world, not just about CS: http://cacm.acm.org/blogs/blog-cacm/26343-media-computation-for-creativity-and-surprises/fulltext
- On how CS is a value-added to a liberal education: http://cacm.acm.org/blogs/blog-cacm/101738-computer-science-as-value-added-to-a-liberal-education/fulltext
I don’t know how to convince University CS departments to do just about anything, but here are my contributions to the dialogs that I hope are happening at Colleges and Universities worldwide about how to prepare students to engage in computational literacy.
My Blog@CACM post for this month is on JES, the Jython Environment for Students, which at 14 years old and over 10,000 downloads, is probably one of the oldest, most used, and (by some definition) most successful pedagogical Python IDE’s.
The SIGCSE Members list recently had a discussion about moving from Python 2 to Python 3. Here’s a description of differences. Some writers asked about MediaComp. With respect to the Media Computation libraries, one wrote:
I’m sad about this one, because we use and like this textbook, but I think it’s time to move to Python 3. Is there a compatible library providing the API used in the text?
Short answer: No. There are no compatible Media Computation libraries for CPython 2 or 3.
We keep trying. The latest attempt to build Media Computation libraries in CPython is here: https://github.com/sportsracer48/mediapy. It doesn’t work on all platforms yet, e.g., I can’t get it to load on MacOS.
We have yet to find a set of libraries in Python that work cross-platform identically for sample-level manipulations of sounds. For example, PyGame’s mixer object doesn’t work exactly the same on all platforms (e.g., sampling rates aren’t handled the same on all platforms, so the same code plays different speed output on different platforms). I can do pixel-level manipulations using PIL. We have not yet tried to find libraries from frame manipulations of video (as individual images). I have just downloaded the relevant libraries for Python 3 and plan to explore in the future, but since we can’t make it work yet in Python 2 (which has more mature libraries), I doubt it will work in Python 3.
I complained about this problem in my blog in 2011 (see post here). The situation is better in other languages, but not yet in Python.
- I have been building Media Computation examples in GP, a blocks-based language (see post here).
- Jeff Gray’s group at U. Alabama has built Blockly-like languages Pixly and Tunely for pixel and sample level manipulations.
- Cynthia Lee at Stanford has been doing Media Computation in her classes in MATLAB and in C++
- The Calico project supports Media Computation in IronPython (based on Python 3) and many other languages, because it builds on .NET/MONO which has good multimedia support.
When we did the 4th edition of our Python Media Computation textbook, I looked into what we’d have to change in the book to move to Python 3. There really wasn’t much. We would have to introduce
Last week, I attended the Computing Research Association (CRA) Snowbird conference of deans and chairs of computing. (See agenda here with slides linked.) I presented on a panel on why CS departments should embrace computing education research, and another on what CS departments can do to support the CS for All initiative. I talked in that second session about the leadership role that universities can play in creating state partnerships and influencing state policy (see the handout for my discussion table).
Andy Ko was in both sessions with me, and he’s already written up a blog post about his experiences, which match mine closely (including the feeling of being an imposter). I recommend reading his post.
Here, I’m sharing a key insight I saw and learned at Snowbird. Before the conference even started, our Senior Associate Dean for the College of Computing, Charles Isbell, challenged me to name another field that is overwhelmed with majors AND offers service courses to so many other majors. (Maybe biology because of pre-meds?) Computer Science is increasingly the provider of courses to non-CS majors, and those majors want something different than CS majors.
The morning of the first day was dedicated to the enrollment surge. CRA has been gathering data at many institutions on the surge, and Tracy Camp did a great job presenting some of the results. (Her slides are now available here, so you don’t have to rely on my pictures of her slides.) Here’s the bottomline: Student growth has been enormous (across different types of institutions), without a matching growth in faculty. The workload is increasing.
But here’s the surprise: Much of the growth in course enrollment is not CS majors. A large percentage of the growth is in other majors taking CS classes. The below graph is for “mid-level” CS courses, and there are similar patterns in intro and upper-level courses.
Tracy also presented a survey of students (slides available here), which was really fascinating. Below is a survey of (a lot) of intro students at several institutions. All the differences described are significant at p<0.05 (not 0.5 as it says). The difference in what non-majors want and CS majors was is interesting. Majors want (significantly more than non-majors) to “make a lot of money.” Non-majors more significantly want to “Give back to my community” and “Take time off work to care for family.”
U. Illinois has the most innovative program I have heard of for meeting these new needs. They are creating a range of CS+X degree programs. First, these CS+X programs are significant parts of the “X” departments.
But these stats blew me away: CS+X is now 30% of all of CS at U. Illinois (which is a top-5 CS department), and 50% of all admitted first years this year! And it’s 28% female.
It’s pretty clear to me that the future of computing education is as much about providing service to other departments as it is about our own CS major. We have suspected that the growth is in the non-majors for awhile, but now we have empirical evidence. I’ve been promoting the idea of contextualized-computing education, and the notion that other majors need a different kind of CS than what CS majors need. We need to take serious the education of non-CS majors in Computer Science.
AccessComputing and DO-IT does terrific work. I get the question in the title a lot with MediaComp, since it’s a curriculum that lends itself towards producing just visual or just audio products, while exploring the same computing concepts.
With the increasing demand for computing professionals, it’s important that students with disabilities are included in computing courses. This video includes profiles of successful computing students and professionals who happen to have disabilities. Learn how accommodations, assistive technology, and universal design strategies can make computing courses accessible to students with disabilities.
Runtime: 10:24 minutes
JES 5.02 is now released at https://github.com/gatech-csl/jes/releases/tag/5.020. I have links to all the main downloads at http://mediacomputation.org .
Fixes in JES 5.02 (with many thanks to HenryStevens and sportsracer48 for fixes):
- Fixes a problem if you quit during raw_input that you can’t do another raw_input
- Makes the Sound explorer fully functional again
- Fixes colorizing
- Makes autosave work again
- Fixes a threading error if you hit return too rapidly in the Command Area
- Unicode characters in input file get flagged. (Jython can’t execute a line with unicode characters on it.)
You can find all the fourth edition Powerpoint slides (including a 68Mb zip of all of them) at http://coweb.cc.gatech.edu/mediaComp-teach/60. I’ve put some of the Peer Instruction question slides into the chapter PPT slide decks, but you can find more at the instructors-only website (see the Media Computation website for more on the teacher website).
Most of the changes are in the early chapters. Chapter 3 on text and language manipulation is all new. The latter chapter PPT slide decks have a few new slides in each deck, including:
- Creating state-preserving versions of picture manipulation functions in Chapter 16 on Functional Programming
- Subclassing Picture and Sound to move functions into methods in Chapter 17 on Object-oriented Programming
- Recursive turtle patterns, which are possible with an improved Turtle class in JES 5 in Chapter 17
Cool example of using JES to access external data!
I’ve been teaching CCT 374: Technologies for Knowledge Media course this term. It seemed a natural fit to use a Media Computation approach to teach Python programming. The students have a term project where they had to design an application that uses City of Toronto Open Data. Just about every team decided to make something that involved displaying something on a map. So, I had to figure out how to display arbitrary maps programmatically, as simply as possible. Using the Google Maps API would have been beyond most of the students. Then I found a blog post with a Python program to retrieve static images from Google Maps.
I have adapted the code from the blog post to work within JES (Java Environment for Students) using the Media Computation libraries. I’ve made the code available on a gist.