Archive for January, 2017
Higher education should be about more than lectures: What students do is more important than what they hear
I was reminded of this work by Ken Koedinger in a recent faculty meeting focusing on Georgia Tech’s future strategic directions in education. We got to considering the quality of different courses, and the analysis centered around “materials” (e.g., slides, textbook), “lectures” (the dynamic presentation of the materials), and “assessment” (e.g., exams and homework). That feels like the wrong set of categories to me. The most important category is “What students do to learn.” “Lectures” simply aren’t an important part of student learning.
So it was difficult for me to open my mind to fresh data analysis, from Professor Ken Koedinger of Carnegie Mellon University, which adds more weight to the argument that lectures aren’t an effective way to learn, despite our nostalgic memories of enjoying them. Koedinger didn’t study live lectures, but recorded ones that were part of a free online psychology class produced by the Georgia Institute of Technology. He and a team of four Carnegie Mellon researchers mined the data from almost 28,000 students who took the course over the Coursera platform for Massive Open Online Courses (MOOCs).
They found that video lecturers were the least effective way to learn. Students who primarily learned through watching video lectures did the worst both on the 11 quizzes during the 12-week course and on the final exam. Students who primarily learned through reading, or a combination of reading and video lectures, did a bit better, but not much.
The students who did the best were those who clicked on interactive exercises. For example, one exercise asked students to click and drag personality factors to their corresponding psychological traits. A student would need to drag “neuroticism” to the same line with “calm” and “worrying,” in this case. Hints popped up when a student guessed wrong.
Thanks to Greg Wilson for sending this to me. It takes a while to get to the point about computing education, but it’s worthwhile. The notion is related to my post earlier in the month about engagement and motivation.
I’d been socialised out of using computers at high school, because there weren’t any girls in the computer classes, and it wasn’t cool, and I just wanted to fit in. I wound up becoming a lawyer, and spending the better part of twenty years masquerading as someone who wasn’t part of the “tech” industry, even though basically all of my time was spent online.
And I can’t begin to tell you how common it is. So what if your first experience of “code” is cutting and pasting something to bring back replies because Tumblr took them away and broke your experience of the site.
Is that any more or less valid than any dev cutting and pasting from Stack Exchange all day long?What if your first online experiences were places like Myspace and Geocities. Or if you started working with WordPress and then eventually moved into more complex themes and then eventually into plugin development? Is that more or less valid than the standard “hacker archetype”? Aurynn gave a great talk recently about the language we use to describe roles in tech. How “wizards” became “rockstars” and “ninjas”. But also, and crucially, how we make people who haven’t followed a traditional path feel excluded. Because they haven’t learnt the “right” programming language, or they haven’t been programming since they were four, or because, god forbid, they use the wrong text editor.
Scientists Looking at Programmers’ Brains see more Language than Mathematics: The Neuroscience of Programming
I’m not convinced that our ability to image brains is actually telling us much about cognition yet. I did find this result surprising, that our understanding of programming languages seems more linguistic than mathematical
Scientists are finding that there may be a deeper connection between programming languages and other languages then previously thought. Brain-imaging techniques, such as fMRI allow scientists to compare and contrast different cognitive tasks by analyzing differences in brain locations that are activated by the tasks. For people that are fluent in a second language, studies have shown distinct developmental differences in language processing regions of the brain. A new study provides new evidence that programmers are using language regions of the brain when understanding code and found little activation in other regions of the brain devoted to mathematical thinking.
I drew on Cognitive Apprenticeship a lot in my dissertation — so much so that Carl Berger asked me at my proposal, “Are you testing Cognitive Apprenticeship as a model?” I had no idea how to respond, and 25 years later, I still don’t. How do you test a conceptual framework?
Cognitive apprenticeship, like situated learning, starts from the assumption that apprenticeship is a particularly effective form of education. Then it asks, “How do you offer an apprenticeship around invisible tasks?”
What I like about the essay linked below is that it places cognitive apprenticeship in a broader context. Apprenticeship isn’t always the best option (as discussed in the post about the Herb Simon paper).
Active listeners or readers, who test their understanding and pursue the issues that are raised in their minds, learn things that apprenticeship can never teach. To the degree that readers or listeners are passive, however, they will not learn as much as they would by apprenticeship, because apprenticeship forces them to use their knowledge. Moreover, few people learn to be active readers and listeners on their own, and that is where cognitive apprenticeship is critical–observing the processes by which an expert listener or reader thinks and practicing these skills under the guidance of the expert can teach students to learn on their own more skillfully.
I wonder if this result explains why the second semester students in Briana’s studies (see previous blog post) didn’t have the “W” effect. If you do enough code, you move down the power law of practice, and now you can attend to things like context and generating subgoal labels.
Different subjects start the experiment with different amounts of ability and past experience. Before starting, subjects took a multiple choice test of their knowledge. If we take the results of this test as a proxy for the ability/knowledge at the start of the experiment, then the power law equation becomes (a similar modification can be made to the exponential equation):
That is, the test score is treated as equivalent to performing some number of rounds of implementation). A power law is a better fit than exponential to this data (code+data); the fit captures the general shape, but misses lots of what look like important details.
Insightful new report from ACCESS-CA on who is taking AP CS in California and on the challenges (quoted below):
Despite the strong outlook for the technology economy in California, there are major challenges in meeting the growing demand for skilled technology workers and preparing Californians to participate in the workforce of the future:
The lack of computer science standards, courses, and teachers and the lack of alignment between computing pathways and workforce needs. Roughly 65% of high schools in California offer no computing classes and the state has yet to develop a statewide plan for computing education.
The lack of diversity in the computing education pipeline and within the technology sector, particularly given the rapidly-increasing diversity of California’s population. 60% of California’s student population is Latinx or African American, yet these students comprise just 16% of students taking AP CS A and 15% of the technology workforce
California is now starting a process of developing computer science standards for K-12, explicitly using the new K-12 CS Framework. California is huge and has a huge influence on the rest of the country’s education policy and practice. This will likely be one of the most important outcomes of the K-12 CS Framework process.
Computer Science Content Standards Development
The CDE, Instructional Quality Commission, and State Board of Education (SBE) are commencing the process for developing new California computer science content standards. Per California Education Code. Section 60605.4, “on or before July 31, 2019, the Instructional Quality Commission shall consider developing and recommending to the SBE computer science content standards for kindergarten and grades 1 to 12, inclusive, pursuant to recommendations developed by a group of computer science experts.” Information and updates concerning the development of computer science content standards for California public schools will be posted here.