Let’s read some books!

An article on how reading is important for leadership feels appropriate for the start of the semester, particularly with it’s mention in the second paragraph of the difference between literacy and the ability for deep reading.

A fun exercise is applying a bit of that “deep reading” to this article. You’ll probably notice that there’s lots of fine anecdotal accounts of great leaders also being great readers. When the evidence starts coming out, things get shakier. The supporting link for the claim that reading offers the best stress-reduction is to a newspaper article about the study that doesn’t make clear that they tested books versus reading web content and also reveals that the study was part of a chocolate marketing campaign. The supporting references for the links between reading and increased verbal skills are more solid but also do not seems to be specifically about the advantages of reading books. In fact, one of the cited papers breaks down the rate of rare words in various types of text and finds that books rank fourth below scientific articles, newspapers, and popular magazines. While this is all compared in that article to spoken language to explain why simply watching television or engaging in conversation doesn’t have the same effects, there is no breakdown of where web-based reading falls.

But in the end I also agree with the point of the article. I’m prepared to believe there are different effects from web versus book reading, probably based on issues of attention/distraction and the goals of the text, and recognizing that one would have to actually define what they mean by “web reading”. And, I do hope all my students coming back next week are prepared to do a lot of book reading!

Brick-and-mortar college

For some reason, this article about Best Buy as a showroom (from earlier this year but I only just read it) made me think about the conversation going on about MOOCs. I had to ask myself if my willingness to turn to online stores for a deal rather than spending more for a more robust experience (being able to try out products, get advice, have it immediately) revealed my feeling that small, in-person classes are still worth the money hypocritical or, at least, motivated by self-interest.

So I thought about what I do buy in-person, rather than online. Clothes, obviously – once I was out of grad school and had a bit more money that was the first thing I stopped shopping for online, being willing to pay more to get clothes that fit better. But in general, it’s items where I am picky about the features, particularly usability features. I could use a new computer bag, but no matter how many descriptions and pictures there are, I can’t pull the trigger on ordering one when I can’t try out the pockets myself (having a smartphone slightly larger than an iPhone means that you’ve got to test that “phone pocket” really does mean generic phone pocket and is not just shorthand for “iPhone pocket”). Any housewares where I really care what the color is – I’ll go to Target for my new dishclothes because I want to make sure the red matches instead of clashing with my toaster and trivets and such. And, items where I’m not sure I know enough about the product to buy it based just on text and images versus seeing it in person – during a toilet-rebuilding project this summer, browsing the Home Depot site on my laptop next to the toilet to figure out which part I would need to buy was way more frustrating than just showing up at the store with a crumbling gasket and looking for the one that actually matched.

So it seems like I want to go to a store when the physical form of what I am buying matters – either to ensure that it fits (my body, or my toilet), or to ensure that its appearance and usability meet my needs (having effectively placed pockets, or being visually consistent with my decor). Stores that make this easy encourage me to buy from them rather than online. And, at least for me, Best Buy doesn’t sell products where this is a factor. They would need to provide something else. When I think about what is challenging about buying electronics online, it is the risk that when you get it home, it won’t do what you want. The printer drivers won’t play nice with your wireless network, an adapter or cable is needed you didn’t know to order, or the minimum memory claim on the software assumed you didn’t mind getting a cup of coffee between commands. At least as much of my time exploring products is spent defensively – after I have decided the features, I want and selected a product at a good price point, trying to convince myself it will actually work as described in my particular setting. Finding the sweet spot in answering that need in a cost-effective manner (what I think Geek Squad was intended to do, and what some of my Mac-owning relatives think Apple stores do well) may be what is required.

And what does this tell me about my perceived value of in-person education? Well, it sounds like a similar situation. We can educate ourselves using MOOCs and other free resources (libraries ftw!), but how much additional time and effort is required to figure out how to educate yourself, what an education actually is, and determining if you are actually getting educated, versus spending time becoming educated, if your goal is the equivalent level of education? And, is it worth your time to have someone else have solved those meta-level problems for you, freeing your energies up just for the task of doing the learning?

Rethinking courses

I recommend both this article about plagiarism in Coursera’s courses but also the comment thread, which makes a few interesting connections between the specific issue of plagiarism happening in these courses, and the broader discussion about MOOCs and their role in the higher education universe.

The obvious question, posed but not answered in any of this, is why would students plagiarize in a free, non-credit course that they are taking entirely voluntarily? If you want to just watch the lectures, or just do the reading, there’s absolutely nothing in the structure of these courses to prevent it. And, completing the course gets you nothing more than a confirmation email with a lovely PDF “certificate” attached. My suspicion is that it isn’t that far off from why people cheat at games, even solitary, non-social games. As one comment puts it, we want to feel good about ourselves. Having signed up for a course, giving up on it because writing a paper is hard feels like failure, whereas plagiarizing is easier and makes us feel like we succeeded.

But then the broader questions come in. The article points out the vast cultural diversity within the courses and focuses on the “teachable moment”. Instructors in these courses are going to have to be much more explicit about underlying assumptions about how academic work should be done, and instruction on things like plagiarism and academic integrity probably needs to be integrated into every course, because every course may be the first one a student is taking. At heart, this seems to tie back to the lack of a prerequisite structure. At a traditional college, I can assume you took freshman comp and learned something about plagiarism. In Coursera, as it is set up so far, that’s not a valid assumption. For computer science or math courses, the prerequisite issue is more one of if the student is prepared, but if you sign up for cryptography ignoring the statement that you need to have some statistics background, the issue will quickly self-correct as you realize that you can’t understand the lectures let alone do the assignments. In a literature course, it is easier to muddle along and convince yourself you’re getting the idea, even if you don’t have the requisite background.

While it wasn’t the main topic of the article, this was also the first time I read anything about how essays are graded in these courses (which – they are peer-graded with, it sounds like, no instructor involvement). It immediately makes sense from a cost-savings perspective. And, I can see how it makes sense from a “wisdom of the crowd” perspective. Plus, peer review is a valuable exercise even in traditional classes – there’s a lot to be gained by reading and reflecting on another students’ work. But, the wisdom of the crowd only works if you either have the entire crowd react to each piece, or if you are only interested in the assessments being reasonable on a crowd-level (with high tolerance for a handful of inaccurate outliers). Both are problems here: you aren’t going to get each student to review the hundreds of other essays to get a crowd response to each work, and if you’re the student who gets an outlier review, that radically decreases the value of the course for you (and, perhaps, reduces your motivation to work hard, and avoid plagiarism…).

Finally, I found one comment particularly interesting (from “husky1”): “I am curious, can students cheat when there is no grading standard? And, why would you have grading standards in a non-credit course? Feedback on the material submitted by the student is one thing. If you really believe MOOCs are a poitive thing then I am not sure “plagerisim” should be a concern. I suspect even someone who cuts and pastes material from say Wikapedia is actually learning something about the topic they are addressing. And isn’t the point here for each person to get what they want out of participating in the course?”

What jumps out at me is that this comment assumes that plagiarism is a concept that is only relevant in the world of graded assignments – that it is a construct of academic assessment, not a real-world principle to be followed. More generally, it highlights the degree to which we’re still figuring out what these MOOCs are. Are they just about putting information out there to help people who want to learn? Ought they hold students to some level of academic standard? In what sense are they a “course”? I see a tension between people signing up to “get what they want” out of the course, and the idea of instructor-created learning outcomes for a course. If there is a fact you want to know, or a skill you need to develop, you can read a book, watch a video, or participate in an online “course” that will help walk you through the gaining of that particular collection of factual knowledge or well-defined skill. But, to my mind, a course is a more completely defined experience, and part of what you are getting is the determination by someone with more expertise than you as to what one should get out of the course. I am confident that the majority, if not all, of the content covered in these courses is available through other sources (books, articles, etc.). There is value in having someone with expertise filter the content, present it in a more appealing video format, and wrap it in explanation and context – that is certainly part of what makes a “course”. But if you lose the idea that the course has been designed to achieve some set of goals established by the instructor, and instead allow that any goal that a student might have in participating must be equally valid and supported, I think you may have lost the “course-ness” of of the course.

This entry readable in lynx 2.8.3 or higher.

This security critique of the Tesco website is a hoot. It walks through an increasingly deep, and increasing damning, look at what is wrong with their setup, and how you can tell. The critique is well peppered with links to additional content about the problems being described, so it’s not a bad starting place to learn something about web security. It is also an accessible illustration of the type of exploration and deduction that can be used to profile a system and its vulnerability. Finally, to me, it reads as a nice lesson in why you can’t just “throw some security on your site” without real expertise. I like the concept of “unconscious incompetence” being used to describe the situation where incompetence (here about security) is being compounded by a lack of awareness of the incompetence. If you at least know what you don’t know, you’re moving a step in the right direction!

Readability versus Realism

I’ve been reading a lot about games and game design over the past few months, and this recent blog post about when visual detail in games becomes overwhelming rung true for me. It’s responding to the difficulty that can emerge when trying to actually play the stunning, complex, 3D games that are coming out, when compared to less graphically “sophisticated” games. The idea of readability, and what makes a game readable, is nicely discussed. I liked the idea, hinted at, that if your game is only playable because you have added meta-labels that appear when you’ve successfully found or targeted an object, this may be a sign your games visuals are sacrificing readability for complexity.

Playing games about making games

While Gamestar Mechanic isn’t really a fit as a development tool for my course, it’s an excellent example of a teaching game, and I would highly recommend it for anybody with a middle-school aged kid (I think that is the right age range for it). The game is structured as a quest to learn to be a game developer, but what surprised me was how much of the focus was on good design, not just how to place blocks and enemies and make things go. You start out by just playing the various types of games that might get built (e.g. platformers versus top-down maze games) and becoming familiar with the differences, but soon you start getting walked through design concepts like how to use space or how to balance goals by playing the same game multiple times with a single aspect changed to see the effect. I loved the “quest” where you get to see how balancing a countdown timer and number of lives can lead to different types of game play. Overall, there’s a lot of showing not telling.

By the end, you can start building and sharing your own games in their “Game Alley, play others’ games, and it looks like there are occasionally challenges with prizes. The site does have a premium paid section as well as the free section, which may bother some, but from what I saw you can get a lot of value out of it for free – it’s not like some sites where you’ll find yourself almost immediately coming up against the limitations of what you get without paying.

Tsk Tsk Speedometer

I really enjoyed this assessment of how a speedometer both breaks a ton of good-visualization rules, and yet is a great visualization given its purpose and context of use. I particularly liked its discussion of why you would want to change scale halfway through a visualization in this setting. Obviously, don’t break the rules until you understand them disclaimers apply, but it’s a really elegant example of how blindly following rules alone also doesn’t make good design.

Sourcecode access wanted

GameSalad is a graphical game-programming tool available for Windows now as well as the Mac (though I had to install something called the Microsoft XNA Framework which doesn’t sound horrible at all), and supports HTML5 for deploying games on the web but also iOS and Android for tablet/phone games (in the pay version). It seems like the pay version focuses on integrating tools for monetization, ads, in-game sales, and social gaming. The core of the free version is very full-featured though. The built-in behaviors and attributes are broad. It’s nicely object-oriented, which I liked about GameMaker when I used it in a course because it sets up well for transitioning to Java.

Unfortunately, for all that I wanted to decide to use it in a course next spring, I just can’t. The interface could be more intuitive/explanatory (for example, it took distressingly long to figure out that you can delete an actor from a scene by clicking it and then pressing the delete key on the keyboard, since everything else gets deleted by clicking it and then clicking a button with a minus sign in the interface). Couple that with the fact that I’m finding most of the documentation is either for an older version of the system, or perhaps for the Mac version. Working through the tutorials, I finally had to stop because some of the features referenced aren’t just in different places, they don’t even exist in the version I’m running (and yes, I checked I had the most updated version).

I could have worked around those things, given how much I was liking the tool, but the killer is that it doesn’t look like you can actually get the HTML5 code out of the tool to view and use as you like. Rather, they publish it to their site and you can embed it from there. Perhaps there’s an undocumented way around it, but currently, the site is also failing to actually produce a published HTML5 version of the game for me (I keep being instructed to wait a few minutes). I can’t find documentation that the pay version would solve this problem either.

So, if you want to play around with producing games, and are willing to have them hosted by GameSalad or pay a bit to deploy to Android or iOS, this is a nice tool, worth checking out, and the Mac version is probably smoother to use. I’m disappointed it won’t work for my course.

Blocky coding

One of my projects this month is looking into tools I might use in a very-introductory course organized around the theme of games. I’m still circling in on the exact set of capabilities I’m looking for, but since one goal of the course is to warm people up for a more intense Java programming course, exposing them to simple programming in a visual manner is appealing.

One possibility is Blocky from Google. Web-based drag-and-drop programming where constructs are puzzle pieces. The maze demo gives a nice starting point for thinking about solving problems, using ifs and loops, debugging, etc. You could make a nice little one hour “so you want to know what it’s like to program” activity just out of that. I’m less clear on how easy it would be to go a step further and use it extensively though.

Next up – playing around with GameSalad.

Machine Learning in Usability Testing

It’s an elegant idea I haven’t run into before: gather data on site preferences by selecting what version to present on the epsilon-greedy solution to the multi-armed bandit problem and just letting it run. You’re looking at a setting where effectiveness can be easily measured, such as by clickthrough, but the contrast is with A/B testing where the effect of a single change is being measured for a time and then a switch is being made, if desirable. Comments suggest tweaks/details like ensuring that a single visitor sees a consistent view of the site, at least for small windows of time.

The technique builds in the idea that, if preferences change over time, the site could automatically detect that – which the blog author and the commenters note isn’t really things work – but it gets me wondering if there *are* choices that work that way. Perhaps not in key navigation, but how desirable a piece of content is might evolve over time – perhaps code like this could be installed under a rotating banner of featured items (we have a rotating slideshow of news items at the top of the College’s website) to figure out which ones get clickthrough and have those persist with less effective ones fading out more quickly. For a place like a College which may not get many repeat visitors nor have profiles on their visitors and their interests the way Amazon and other big eCommerce sites do, this might be a lightweight method for getting some preference learning built in.