Archive for October, 2011
The below article ran in the IEEE Spectrum Tech-Alert under the awful title, “With the Arduino, Now Even Your Mom Can Program.” The IEEE Spectrum editor immediately sent out an email retraction of the title as being offensive.
But even with the retraction, I don’t think that the piece adequately explores how different the populations are of Arduino users. The below picture is from Leah Buchele at this last May’s NCWIT Summit in NYC.
The graph on the left describes the gender makeup of the Arduino-using community. The graph on the right describes the gender makeup of the LilyPad-using community. The IEEE article simply describes the LilyPad as “waterproof.” Huh? Don’t they know about e-textiles? The red in the graphs are male, and the aqua are female. In statistics, this is called “inter-occular occlusion” — you don’t need a t-test, this just hits you between the eyes. Women like the LilyPad. The Arduino community has almost no women in it. The context matters.
If you’re going to make some crack about mothers programming, then you’d better speak to the significant gender issues. And if you’re going to write about Arduino, you should really know about the different communities. Arduino matters for women, because it led to LilyPad. Arduino itself plays no role in being a technology environment for mothers or just about any women at all. They’d better figure that out before they further explore “integrating it more deeply into the education system.”
To fuel greater adoption of Arduino, the team is exploring how to integrate it more deeply into the education system, from grade schools to colleges. Several universities, including Carnegie Mellon and Stanford, already use Arduino. Mellis has been studying how students and laypeople take to electronics in a series of workshops at the MIT Media Lab. Mellis invites 8 to 10 people to the lab, where they’re given a task to complete over the course of a day. The projects have included building iPod speakers, FM radios, and a computer mouse using some of the same components that Arduino uses.
But spreading the Arduino gospel is only part of the challenge. The team must also keep up with demand for the boards. In fact, the Arduino platform doesn’t consist of one type of board anymore—there’s now an entire family of boards. In addition to the original design, called the Arduino Uno, the new models include a more powerful board called the Arduino Mega, a compact board called the Arduino Nano, a waterproof board called the LilyPad Arduino, and a recently released, Net-enabled board called the Arduino Ethernet.
I like the way the McGraw Prize in Education is framed, and congratulate Mitchel Resnick on receiving the award for his work on Scratch.
“Technology in education is a great catalyst for change — for creating, managing, and communicating a new conception of learning,” said Mr. McGraw. “This year’s Harold McGraw Prize winners are the embodiment of the transformative impact of technology on improving education. Their innovations are enabling students to learn at their own pace and empowering teachers to inspire and coach.”
Mr. McGraw said, “Digital learning is the opportunity of the century. For many students around the world, technology makes education more accessible, adaptable and affordable. We applaud these exceptional leaders for guiding the way and enriching the lives of so many students.”
The College Board is increasingly using the Advanced Placement exams as a kind of intervention into high school STEM education, and even attempting to influence university education. The new Biology AP course and exam is explicitly aimed at reforming biology education and is based on the latest education research. Now, the College Board is thinking about moving into engineering, which supports multiple efforts to create more K-12 engineering education, e.g. Georgia Tech just won a grant to teach manufacturing principles in high school. I think it’s in this spirit that the College Board is working with NSF to create the new hope-to-be AP in CS:Principles. Can we use AP CS as a lever to change undergraduate CS education?
At U.S. News’s Making Science Cool event last Tuesday, he said the organization was “exploring the potential of more AP courses in engineering, energy, environment, and anatomy.” Don’t look for the new subjects anytime soon. Development of a new AP course can take up to six years as teachers and professors develop class curricula and the AP exam before smaller groups pilot the new course, according to Trevor Packer, senior vice president of AP and college readiness for the College Board.
I have come up today with an answer to several questions that have been vexing me for some time. Here are three of those questions, and the answer that I’ve come up with for all three.
- I tried to upgrade my desktop computer to the latest Mac OS X yesterday, but I couldn’t. My 2006 Mini-Mac has a Core Duo processor, and Mac OS X Lion won’t run on such an old computer. Georgia Tech, like most research-intensive universities, has no policy to upgrade computers for faculty. Faculty are expected to bring in research funding to pay for their own computers. All of my funding is from NSF which explicitly prohibits purchasing computers for faculty’s general use. (Overhead, which is supposed to pay for infrastructure like computers, is instead used to pay for new initiatives in healthcare and for wine & cheese “networking receptions.”) I complained to my Chair, who pointed out that other faculty don’t have this problem. They get corporate funds or defense contracts to pay for their equipment. Why don’t I? What’s wrong with computing education research that it doesn’t attract more corporate or DoD funds?
- I’m going to be interviewed for a podcast Friday, and one of the questions I’m going to be posed is, “Why doesn’t computing education research get the same respect as other CS research areas?”
- I’m not quite done yet with Abelard to Apple. I’m really enjoying it, but I’m realizing that it has nothing to do with me. Rich DeMillo talks about the three broad classes of universities: Elites, Middles, and the For-Profits. Elites succeed, and Middles can succeed, because they offer a compelling “value proposition.” That doesn’t necessarily have anything to do with learning. DeMillo also talks quite a bit about faculty and University reluctance to measure anything. Research is about measuring everything. My research is about improving the quality of education. Does quality matter to the economics of universities?
There can be no profit in education in America. There is no competitive advantage to better learning. What’s the first question that anyone asks of a new learning or teaching method? ”Will it work with the disadvantaged, the poor, the urban student with little preparation?” I agree with that sentiment, but that’s not how you make money. There’s no profit to be made by making sure that your best work goes to people who can’t pay for it. While I understand the arguments (and counter-arguments) about the fortune at the bottom of the pyramid, it’s still pretty clear that nobody knows how to do that well yet.
I expect someone to respond to this post with, “There can be profit in better education! You can find a competitive advantage if you just…” Yeah? Show me, please. Show me just one replicable example. Even the For-Profits aren’t saying that they get you better learning. They just say that they’re cheaper or more flexible (i.e., in on-line classes).
What would happen if we could teach computing better? First, how could you prove it? What could you offer that students and employees would accept as evidence? Second, who cares? The Elites draw students because they offer far more than simply learning — they offer a network, prestige, great ROI. There’s no advantage to teaching better. Will people pay more for better education? How would you know? There are schools that say that they lead to better learning (like the liberal arts schools) — they’re drowning in debt, and far more kids are going to the research institutions and the Elites, or the For-Profits. Better learning doesn’t buy you anything.
What if we could teach computing faster? All of American education is time-locked. Classes take 15 weeks, a semester (in most places — a quarter, 10 weeks in others). What if you could teach the same content in 12? Well, what would you do with those other three? Schooling is 12 years, and even if you could achieve the same learning in less, you run into huge obstacles of culture and economics to finish earlier. You can’t save time in schools, which means you can’t save money by doing better in less time.
Every other field in computer science offers a competitive advantage to somebody. If you can make operating systems better, you give your funder something to sell, you save people time and money. Security is all about protecting what you have. A better visualization system gives your analysts a competitive advantage. Better graphics get you more movie and video game sales. Health systems and technologies flourish because people can and will pay more for better healthcare. Why should Universities respect the field that brings no profit?
Education research can only succeed in non-profits. It’s a form of social work, “Computing at the Margins” as my Chair likes to say. But Universities aren’t non-profits — they’re totally in it to maximize profit, as DeMillo points out. I’m in the wrong job.
A significant problem of all object-oriented programming systems has been the learning curve of getting to know the object system. I’ve seen little work on helping the programmer to understand the object system. I had just these kinds of problems when I was trying to figure out the sound classes in Java, and suddenly had to deal with all these Consumer and Producer classes that I’d never heard of. The below is an interesting step in helping the programmer fill in that knowledge. An interesting follow-on question: What does the programmer remember after using Matchmaker about the object system? Does the programmer learn anything? Does knowledge accrue and transfer?
In the Eclipse framework, the window that displays code written in the new language is called an Editor; a function that searches the code for symbols or keywords is called a Scanner. That much a seasoned developer could probably glean by looking over the Eclipse source code. But say you want to add a new Scanner to Eclipse, one that allows you to highlight particular symbols. It turns out that, in addition to your Editor and Scanner objects, you would need to invoke a couple of objects with the unintuitive names of DamageRepairer and PresentationReconciler and then overwrite a function called getPresentationReconciler in yet a third object called a SourceViewerConfiguration.
With Matchmaker, the developer would simply type the words “editor” and “scanner” into the query fields, and the program would return the names of the objects that link them and a description of the modifications required to any existing functions.
I hope to get to PPIG someday, the Psychology of Programming Interest Group. PPIG dates back to the golden age of computing education research, back when the Empirical Studies of Programmers workshops were going strong. While I haven’t made it to the UK yet to attend one of their conferences, I remain on their mailing list and enjoy reading the reports.
The keynote from last August’s meeting had quite an interesting idea — that software can be gendered or gender-neutral. I’m surprised that they can find significant evidence of these differences, that the differences aren’t more subtle. I wonder what is the role of the genders of the developers in defining the gender of the software?
Margaret Burnett opened the conference with a keynote on Gender HCI and Programming. She reported on a series of investigations conducted by her and her students. They found that purportedly gender-neutral software tools do interact with gender differences, resulting in lower problem-solving effectiveness for female users. In particular, males were more prone to explore and attempt problem-solving by trial and error, while females did not explore as much and stayed with familiar functions. Female end-user effectiveness in programming environments like Excel could be improved by taking gender differences into account. This would not necessarily mean the tool would be less usable by males; in fact, many groups of people could benefit from the improvements.
The increased interest in CS education is terrific — computing is critical to our society and our infrastructure, and there are more jobs than graduates. Understanding computing is critical for innovation in every domain. It’s a shame that it has to come as an add-on. Computer science should be part of every discussion about science or mathematics education, or STEM education generally. Creating more and better educated computer scientists (and computing literate citizenry, across professions) is as important as having more and better educated scientists and engineers (and having a science and mathematics literate citizenry).
The Computer Science Education Act, according to Polis, would help train American students for the more than 1.5 million high-paying computing jobs expected to be created in the United States by 2018. The bill, he said, aims at helping states increase and strengthen their computer science offerings.
If passed, states will receive at least $250,000 in planning grants, according to Polis’ office.
At the University of Colorado, school leaders are proposing a second undergraduate degree program in computer science to increase the number of students in the field, and the Boulder Valley School District has several advanced computer classes available for students.
Polis’ legislation would require that states develop computer science standards and curriculum and form a commission to bring states together to address the shortage in computer science teachers.
Mike Goldweber sent me a link to this video at BBC News – Newsnight – Is lack of computer science teaching failing pupils?. I’m a bit worried about the overall theme of this story, that it’s important to teach UK kids programming in order to save the British gaming industry. (Are there that many jobs available in gaming? So much that we should teach every kid in school to program?) But the quote right at the beginning, around 44 seconds in, is a thought that I’ve been mulling a good bit recently.
We’ve got to teach kids to program, but not as a subject in isolation.
That sentiment is in keeping with Seymour Papert’s original vision for Logo in schools. In his book “The Children’s Machine,” Seymour critiqued the process of “schooling” that took Logo out of its natural home as a tool for exploration and expression in mathematics and science class, and instead “schoolified” it. Logo was shoved into the lab (far away from the interest of teachers), and made into a separate subject to be taught, after the pre-requisite “Keyboarding Skills” course. Seymour saw programming as going beyond “the 3 R’s,” in an amazingly prescient Wired piece from 1993, where one can see the seeds of the Open Learning effort and individualized learning support through technology.
This issue of teaching programming in the context of other courses helps to address some of the concerns voiced in the recent discussion about computer science as a liberal art. It’s a point I hear often from high school teacher advocates. ”The curriculum is too full. If we put CS in, what comes out?” Seymour was saying that that’s the wrong way to think about it, and I’m starting to think he was right on this count (among many others–some of his insights take me longer to grok). It’s not about doing something more. It’s about teaching what we have now, but in a new and more powerful way. It’s Andy diSessa’s argument for computing literacy — how much powerful are we when we are as literate with computing as we are with numbers or text?
Set aside the issue of programming being a weird (and arguably, badly designed for the general purpose of insight and expression) notation, and even set aside the technology issue. Imagine that we had a new form of writing, that wasn’t that much different than English (e.g., it’s not Esperanto), but was significantly more powerful (e.g., more accurate, more understandable, more usable for mapping to theory, more practical for solving problems) for expressing science and mathematics. And what’s more, once you used that language, you could easily visualize and simulate what you described. Wouldn’t that be worth teaching to every science and mathematics student, at least at the undergraduate level? Certainly, we would want that for every graduate student. Why is it not worth teaching, just because it’s a notation that runs on a computer?
I have advocated for some years now “Computing for Everyone,” that Alan Perlis had it right when he argued in 1961 that everyone on campus should learn to program. Perlis advocated a required course for everyone, and so have I. Maybe that’s the wrong end-goal. I understand the argument from CSTA, that it’s hard enough to teach enough CS teachers in high schools — how in the world can we teach every mathematics and science teacher how to use and teach programming? I totally get that we don’t know how to get there. But that doesn’t make it the wrong desired goal. (I’m reading in the middle of Abelard to Apple now, so I might be particularly receptive to learning-with0ut-courses these days.)
How do we get there, where students have true computing literacy and programming is part of science and mathematics? Are required courses in computing part of the trajectory to the desired state? Or is there a different path to integrating computing into the rest of the curriculum?
CS10K is the NSF effort to have 10,000 qualified high school teachers prepared to teach Advanced Placement Computer Science by 2015. 100Kin10 is a new one for me:
In January 2011, Carnegie Corporation of New York, Opportunity Equation, and NewSchools Venture Fund convened a diverse group of organizations to respond to this need. The focus: to prepare, deploy, and support 100,000 excellent STEM teachers over the coming 10 years in order to prepare all students with the high-quality STEM knowledge and skills needed to address the most pressing national and global challenges of tomorrow.
Interesting report in Education Week. They claim that the low performance of US students on international tests is really not that significant for competitiveness nor for getting more students into STEM. They cite data from the World Economic Forum and from the Bureau of Labor Statistics to support the argument that there is no correlation between test scores and competitiveness or STEM labor shortages.
For decades, our rhetoric and education policies have been based on the premise that the ranking of U.S. students on international tests will lead to a decline in our nation’s economic competitiveness and a shortage of American scientists and engineers.
It is ironic, then, that given the rhetoric and policies surrounding international test-score comparisons—much of it unsupported by evidence—little attention is paid to two of the most powerful findings of these comparisons: the strong negative effects on student performance of both family poverty and concentrations of poverty in schools.
Cool result! The numbers are small, but it’s still a positive direction. Unfortunately, the numbers aren’t showing up in the Taulbee report, so it’s likely that the surge is at the Masters level. Still — good news!
A report released last week by the Council of Graduate Schools (CGS) found that first-time enrollment in science graduate programs increased last fall even as graduate enrollment across all disciplines — including nonscientific ones — dropped slightly. But the study’s most intriguing number was hidden beneath the headlines: 33.6%. That’s the reported percent increase, between the fall semesters of 2009 and 2010, in the number of black and African-American students entering math and computer science graduate programs.
I just got my October 2011 copy of IEEE Computer and was pleased to see published my piece on “Learning how to prepare computer science high school teachers” (the last Word draft version is here, but the published version is severely proofread and much better). I tell the story of Barb Ericson’s “Operation: Reboot” (moral of the story: The best teacher preparation doesn’t help if there are no teaching jobs), Lijun Ni’s study of DCCE teachers (following the Disciplinary Commons of Josh Tenenberg and Sally Fincher, led by Ria Galanos and Briana Morrison) (moral of the story: It’s hard to be a CS teacher, but model teachers and a community really helps), and Klara Benda’s study of on-line CS courses for full-time teachers (moral of the story: We need a new model).
I’m also trying to make a meta-argument: Computer scientists have to get involved. If we just leave computing education to the existing education system, we’re going to end up with more applications classes — there’s a clear and evident need for such classes, and because we hide computing so well from the application user, it’s not clear and evident that there’s anything deeper (or that there’s any use for it). I used the analogy yesterday to agriculture and biology. Teaching agriculture is obviously useful, but it may not have been obvious 150 years ago that studying biology would lead to bigger, better, deeper ideas. We had a recent visitor, a Chair of another CS Department, who said that he wouldn’t let his faculty get involved in high school CS education efforts, because, “We don’t have that kind of expertise.” That’s not the point — the point is to inform the process about our expertise. To quote my own article:
In addition to more well-trained high school computer science teachers, we need more education research that is informed by understanding how CS is taught, what current practices are, and what’s important to keep as practices change.
Upcoming lots of travel: I leave tomorrow for the SIGCSE Board meeting in Charlotte over the weekend. I will be at Rutgers for a guest lecture next Tuesday 18th October. Friday the 28th, I’m visiting DePaul for another guest lecture. I am taking Barbara to Florence, Italy 2-9 November to celebrate a birthday with a zero in it. And Barb and I (with our daughters) will be giving talks in Melbourne, Adelaide (three in one day!), and Sydney, Australia, leaving 15 November and returning 26 November. (I will actually never see November 16 this year — it won’t be on Earth on that day. Weird!) I may not be able to keep up my daily postings during some of this travel (and I have that 3rd edition Python book to finish before 15 November, too). Thank you for your patience and ongoing interest!
First, I’ll give you the blog quote, then my response (below):
What if edu-passionate volunteers from around the world banned together to…
Build an open licensed, free, multilingual virtual school of education driven by the principles of peer learning.
Facilitate the free exchange of knowledge and ideas between educators in a context that positively transforms teacher and student practice, learning and engagement.
Codify existing educational best practices into a series of experiences that provide foundational knowledge of how to be an effective teacher.
Here’s my response (which isn’t showing up at Dot Learnt — maybe deleted as spam?):
Why would any teachers come? Yes, some might — the most active, the most excited, the most passionate. But they’re not the problem. How could you package teacher professional development so that it’s easily accessible, engaging, and no-cost (or as they’re finding in LAUSD, you have to pay the teachers to get them to attend)? You make a difference, at-scale, by engaging lots of teachers, not just the few who will voluntarily take the classes.
If you compare this to professions like law or medicine, the practice pays for the practitioner to seek out professional development (often at really nice locations!). But teachers are expected to work long hours at night, on their own time, for professional development? The sustainability problem with the volunteer school of ed idea is not to keep the volunteers producing, it’s to get teacher-participants.
Recently I blogged on an article by president of Bryn Mawr College, arguing for the value of single-sex college education. Now, here comes a meta-review published in Science saying that there’s no evident value. This is a significant issue for us, because we do need to find mechanisms to draw more women into CS. Single-sex programs are a potential avenue to consider — but maybe not?
The report, “The Pseudoscience of Single Sex Schooling,” to be published in Science magazine by eight social scientists who are founders of the nonprofit American Council for CoEducational Schooling, is likely to ignite a new round of debate and legal wrangling about the effects of single-sex education.
It asserts that “sex-segregated education is deeply misguided and often justified by weak, cherry-picked or misconstrued scientific claims rather than by valid scientific evidence.”
But the strongest argument against single-sex education, the article said, is that it reduces boys’ and girls’ opportunities to work together, and reinforces sex stereotypes. “Boys who spend more time with other boys become increasingly aggressive,” the article said. “Similarly, girls who spend more time with other girls become more sex-typed.”
I looked up the Department of Education’s report on Carnegie Learning, and the NYTimes article was actually kinder than it might have been. Three of the four papers that were reviewed actually showed that the cognitive tutor had a negative impact on the outcome measure! Their standards for what studies counted and which didn’t were a little odd to me. Like this comment from the paper:
Plano, G. S., Ramey, M., & Achilles, C. M. (2005). Implications for student learning using a technology-based algebra program in a ninth-grade algebra course. Unpublished manuscript. The study is ineligible for review because it does not use a sample aligned with the protocol—the sample is not within the specified age or grade range.
What does that mean? The ninth graders being studied might not have actually been in ninth grade? Or somebody should have checked?
I wonder whether either the “hype” of the salespeople or the whole “What Works Clearinghouse” make any sense at all. I raised this question in my piece for Greg’s Making Software. We have studies where Media Computation has worked well in terms of impacting student retention. So? That shows that it can work. That is no guarantee that it will work. The WWC says, “Let’s use randomized trials, of both students and teachers.” I’d make the same claim. Even the greatest teacher can be stymied by a class of poor and starving students, and even the greatest textbook can be completely ineffective with an unprepared teacher and unmotivated students. Is it possible to prove that any intervention will always work?
The federal review of Carnegie Learning’s flagship software, Cognitive Tutor, said the program had “no discernible effects” on the standardized test scores of high school students. A separate 2009 federal look at 10 major software products for teaching algebra as well as elementary and middle school math and reading found that nine of them, including Cognitive Tutor, “did not have statistically significant effects on test scores.”