Why My MOOC Is Not Built on Video
class-central.com
class-central.com
I imagine it has to hurt their Google rankings as well, especially when they don't include a transcript.
If this opinion is popular, maybe there's a good reason for it?
A great example is news sites creating ~1 minute videos instead of just printing out the story. News is almost always better to read than to wait for a person to say it to me in a video.
I frequently will download all the videos for an online course. Subsequently, when I watch them in VLC [1], I'll bump the playback to 1.25x - 1.75x.
Fwiw: for me, playback speed strongly correlates to how technical the topic is, less technical videos tends toward the 1.75x; while highly technical tends toward the 1.25x.
I'm also a big proponent of subtitles and transcripts.
In contrast, lectures accompanied by well-produced written material allow me to skip areas I have a grasp on, or provide a resource to reference much more quickly when I need to review a topic. Having to guess which lecture, and where in that lecture, a comment occurred is a significant spend of time.
If I'm browsing the web, I'm already reading textual content so reading is about as low effort as can be. Watching a video means "context switching", reading to watching, along with making sure I've got headphones on, pausing any music I'm listening to etc.
That's not to say the above is true for everyone, but for some people, watching a video is substantially less convenient and entails more effort than reading text.
I wish there was a more fluent way to control the playback speed (Fn Key and Mouse Scroll) but otherwise it works well.
Really handy for podcasts that while interesting tend to be a bit slow.
EDIT: Just checked new version, you can display the speed bar by showing status bar, clicking it and then hovering cursor and using scroll wheel, missed that before.
So in the time it takes to watch a video, I could have scanned throough several documents online. Couple that with copy-able text and it seems apparent why one might avoid videos.
[citation needed]
I agree that on average we're much more willing to learn things ourselves and interested in the subjects you listed, but I reject the idea that the community prefers text content because we're smarter than average. One problem with HN being connected to YC (though it comes with a lot of upsides) is that people will always go out of their way to appear smart. Everyone knows that can help their chances of getting accepted.
Sure, literally everyone that has replied to my original comment has said they're in favor of learning by reading plain text, but the fact is less HN users would have followed along with Sam Altman's "How to Start a Startup" course had it been in that format.
It's especially surprising since having your MOOC captioned is already mandatory if you want to comply with accessibility laws.
Of course, with the popularisation of using videos for "all the things" lately, I don't feel at home at many a (popular) website any more. But that's another matter entirely, I suppose.
Similarly, when I'm reading text and a message (e.g. email, IM) comes in, I can switch over to the other program, reply to the message, switch back to my browser, and the text is right where I left it. With video, there's no such guarantee. Also, I've never come back to a text document and needed to sit and wait while it buffered again.
I can easily alternate between four or five different text documents, taking each a paragraph of two at a time, and still follow what it happening in each one. This may be a personal deficiency, but I've never been able to do that with videos.
But so many videos are unedited. Some guy rambling about stuff, advertising himself or his social media presence, while slowly setting up whatever the actual point of the video is. Possibly this is because editing takes skill (which most people don't have) and tooling (video editing software is hard and/or expensive).
The last time I had to get information out of a video because there was no other source was for a minor car repair issue. It was useful to see which screws to undo and how the parts come apart and go back together. But that was one minute of a ten minute video.
The "rule of 30%" (immediately skip the video to 30% in) helps quite a lot.
Video is one of the worst ways to learn. Anecdotally, think of times at high school when a teacher has put on a video: it's typically to distract the students so that the teacher can have some time out from interacting. It has a similar function to leaving a baby in front of The Happy Little Elves. It's better than having no stimulus, but no serious learning is expected from that.
Although the article says "Watching videos is not better (or worse) than sitting through lectures", I would even disagree with that. A good lecturer has the opportunity to gauge the audience and adjust their lecture depending on how many students are throwing paper aeroplanes (or looking lost). This just doesn't happen with video.
I'm not a big fan of video as a tool for conveying information in general. Not only can I read faster than I can listen, but I can skim and skip around as well. There are situations where video is the best way to impart certain information, but those seem to be the rare cases. I hate it when websites make me sit through a 3 minute video in order to get an answer to one simple question, or to understand "What does this thing do?"
I've never taken a MOOC course because I can't imagine having the time to watch X hours of someone talking at a camera. If this video-less trend catches on I might actually be able to benefit from online education.
I've always feel like I tend to gloss over the details whenever I'm watching the videos, along the opposite that I'm being bored and distracted because it's too slow. Since both could happen in the same speed, moving speed up and down is not an option.
That's not necessarily a bad thing. If you think about how many times your reading is slowed down because you feel you have to re-read a passage to 'get it', audio/video sources totally help do away with that. It may sound like that weakens your learning, but if you've ever researched anything on improving reading speed, you'll know that it's better to not focus on 'getting' things on the first pass anyway, and instead continue plowing through the entire content just trying to get the jist of it, relying on multiple passes to really understand things better instead (this probably works for the same reasons spaced-repetition works). Obviously you can do that with reading as that's where the tip originated, but clearly auditory/visual input helps facilitate that if only because rewinding is less tempting than re-reading due to it being a chore.
> along the opposite that I'm being bored and distracted because it's too slow
Seems like a personal preference thing. To me, I get this with reading as well. It's just not stimulating enough for me, even when I read fast, cause it's easy for competing thoughts to take over and distract me.
Audio/video on the other hand, work for me precisely because I can't think with that much stimulation going on. I'm too busy listening and watching to really have the resources to process any distracting thoughts, so it's easier to focus.
Just my $0.02
It seems like everything I search for these days, I find as a YouTube video. Why? I was trying to take the center console out of my truck and needed to find out what size the bolts were on the back of the driver's seat. Only results were YouTube videos. So I'm sitting in my garage streaming a five minute YouTube over my limited LTE just so I can find out it's a 17mm wrench. Most of that five minutes is the guy telling me who he is, how to subscribe, what website he's with, and why I need to watch the video that I'm already watching. Google can tell me the up-to-the-minute March Madness scores but the only way to find out what size wrench is needed for a Chevy Silverado, one of the most common trucks in the US, is a YouTube video from a guy with an accent so thick I can't understand him.
Useless. Some days I feel the world would be better off without YouTube. Until the day when we can index and search video and voice the way we can text, YouTube is nothing but entertainment to me. Stop making "educational" videos, I don't want to watch them.
I agree with the frustration, but we've got a system now that rewards video makers for this stuff, and doesn't reward people posting this information in text format.
So they expect me to watch a commercial and 3 minutes of video, when all they had to convey was a very concise point like "democrats offered to make 90% of the requested cuts, and the other side is shutting down the government over the last ten." But I "absolutely" must watch the video, with no summary of the point because "[my congresscritter hero] totally proves you wrong".
... and have the full support of the other reddit posters. F--- that.
Of course, if the answer to something is on Wikipedia, it's a win. But Wikipedia isn't designed to be an encyclopedia of all human knowledge (sadly).
First-order:
* At a computer ==> Text * Commuting ==> Audio/video
Second-order depends on things like type of content, cognitive load, etc.
In theory, VLC allows queuing of online content for playback, though in practice it's flaky enough that it crashes, loses the playlist, and/or gets confused during playback. I much prefer mplayer as a general rule.
Once you've got a given lecture prepared, it's easier to edit and modify it. But getting there is the hard part.
I still prefer the mostly-webpages approach, but Khan definitely did a lot more to make use of the format than "dump a camcorder in a lecture hall".
It is also the reason why I prefer MIT's OCW video lectures over their edX ones. I like how most of the lecturers often take the time to write down the terms and ideas on the blackboard in the old-school way.
I find learning that way helps drive home the concept in a more concrete way than flipping slides with voice-over. I believe this is especially important in case of subjects like math where 'doing' is perhaps the only effective way I am aware of learning the material.
But then again, it may just be me.
I am finding the best approach is to watch the video at 2x speed, take notes at the same time and then catch up with a book a little later on.
With a regular book for example, I'd take about nearly a year to finish just 1 probably, if I can even stand to read it consistently, and even then I wouldn't remember most of the information from it, making the entire time spent on it seem like a large waste.
With audiobooks on the other hand, I can read a large book in about 2-3hrs, and have read about 25 books per year because of them. I even use text-to-speech on ebooks for titles that don't have audio versions -- it's just that much more efficient for me. I can go out for walks while listening, do laundry, cook, whatever, and still get my learning fix, all while retaining about the same [albeit poor] amount of information as I would with reading normal books.
Now whenever I can, the first thing I do when trying to learn about a topic, is go to youtube and watch long lectures on it at 2x speed, or download them and watch them at 3-4x speed. That seems pretty efficient to me if what I want is a good overview of something. If I want to reference back to specific factoids, then text is better of course, but for a completely new subject I'm not familiar with at all, audio/video is a godsend for me. Dense texts like wikipedia might as well not even exist as learning materials for me, cause they just cause me endless headache trying to parse them if I don't know what I'm looking for ahead of time.
[1]: Roughgarden's Algorithms part I, Programming Languages by Dan Grossman and Machine Learning by Andrew Ng.
No disrespect meant for Prof. Grossman. His course was very interesting and, as you said, the course notes were so good the videos were completely redundant.
As others have stated, the reason that video is worse than text for learning is because
A. The rate at which you can read is faster than you can listen
B. Video technology is often a hassle when you need to skip around.
C. It's cheaper to write and revise text than video.
D. Videos consume large amounts of data.
But I'm going to add another point.
E. Lecturers most often don't know what I as a student am struggling with. They may focus on things I understand easily and spend a majority of time discussing things I already know. However, if I decide to skip a portion of a video I could potentially miss a really important detail necessary for understanding the subject matter.
With a book I can skim. Important information is often marked in a certain way to make it stand out from other data. It's also organized in sections and chapters so I can easily skip around. With videos this isn't possible.
Also, what happens when all of this information is outdated? Now you have a bunch of videos with outdated and potentially false information out there and nobody will know.
My ideal alternative would be to remove lectures entirely and to have interactive labs where teachers and TAs help students complete assignments based on assigned readings. The instructor for the course would be responsible for overseeing the teaching assistants and preparing the course outline. The TAs would help guide the students in completing the assignments or mini projects. Then a couple of time a semester there would be larger projects that tie together what the students have learned.
_yes_.
This seems so obvious to me. I am completely mystified as to why all universities seem to be ignoring it. I worry that it's just because videotaping lectures and putting them online is so much easier than actually re-evaluating the pedagogy.
Often nobody knows what's going on. You can argue that confusion is good, but it's not fun and it feels wrong. You can argue that class should be challenging and active, but this method led to a minor student revolt.
Discrimination and prejudice happen. Women are underrepresented (with respect to the ratio in the class) in the conversations that happen.
We haven't figured out the right incentive structure to make people actually do the readings yet. Sometimes people don't do them.
Personally, I love this new method. It has problems on the scale of a 60ish person class in a technical subject (nonlinear dynamics). When we're working in small groups, that helps, but whenever we try to do something as a class it devolves into an impromptu lecture (most of the time by a student).
If we want to learn this way we need to know how to scale it. It seems drastically more effective, and I think that college would be so much better if this method was normal.
What our teacher did was: 1) Write a small piece of pseudocode on board and called it a "unit" 2) Present a problem to the class. 3) Ask students to come up with solutions that used the unit (or modified it a bit. 4) Allow Students to present their solutions to the rest of the class. 5) Allow the class as a whole to debate on competing algorithms. 6) (If applicable) Reveal the name of the algorithm the class has just derived. 7) Repeat
The teacher rarely ever lectured. He just presented problems and let the class figure a way out. Occasionally he would nudge us in a direction we weren't considering, but mostly he was there just to ensure things run smoothly and that's all.
On the whole, it worked out great. Before the class I knew most people had trouble designing algorithms. At the end, even the weaker students had a pretty good grasp on the subject because the format itself encouraged everyone to contribute and thus learn.
As a teaching tool, video doesn't lend itself to iterative refinement or discussion. It's an excellent audiovisual experience, but also a bet that the content is best delivered in a read-only format. (Practically, you can't update a video, only redo it -- presumably at full cost.)
I see video as a supplement to lesson formats that can be improved over time (text, or whatever medium comes next).
I looked at a course like you describe for this semester. It looked fantastic, but I live in Europe and it was taught on the East coast of the US; I would have had to wake up at 2am my time to participate.
Last year I took a course that was almost entirely written -- the professor felt more comfortable if his lectures were written out, so he did so. The videos added flavor, but >90% of the content was there directly in the lecture for me to refer back to.
Having many ways to learn available -- lecture, reading, question/response, group work -- maximizes the chances a student's particular strengths will be matched in the course.
What struck me this time was how much the emphasis is on practical software engineering rather than Lisp. That's what Ableson is so intensely passionate about in the video, not the theory. And so picking up the book and looking at the parallel text, I reread it in a whole new light. It was like rereading Eco's Name of the Rose after 25 years and not as the piece of pop fiction I thought it was the first time.
Which brings me to my point that video is just another communication channel. It conveys particular information efficiently and other information less efficiently. Text is no different. Without the video I would never have picked up that SICP and Code Complete have many similarities in subject matter and one just starts with a clean slate and the other in the messy middle.
Video does not need to be fancy to be effective. See Jeffery Ullman's Finite Automata on Coursera. It's PowerPoint and a small talking head and a red pen: And it's better than reading the book he wrote with Who because though dense, it's an order of magnitude less dense than the text.
On the other hand, University of Phoenix operated its online courses with NNTP and was reasonably interactive though asynchronous. It's a matter of curriculum development I suppose. Dan Grossman at University of Washington compares developing a MOOC to writing a textbook and that probably captures the domain of possibilities.
My experience is based on 3 programming courses from Coursera: Functional Programming Principles in Scala, Reactive Programming, and Programming Languages. In all three cases, the videos did little to improve my understanding of the material, and I preferred the slides. Unfortunately, because the videos existed, the slides were (in general) less complete; I'd rather the videos did not exist and the written material was more complete instead.
Videos frustrate me because the lecturer goes too slowly or chooses to focus on the parts I find less interesting. And because it's pre-recorded, I cannot ask him/her questions. Yes, I can skip or fast-forward the video (in one course, I was so used to playing every lecture at max speed, I was surprised to find the lecturer's voice at normal speed sounded completely different :P ), but I find it so frustrating I'd rather be reading detailed text instead.
The embedded interactive quizzes are fun and motivating, but I wonder if the videos are needed at all.
Also, the deadlines from Coursera and most MOOCs conspire against watching videos at normal speed. Most of us have day jobs and the only time we can read the lectures and do the exercises is a few hours a day, late at night. In that case, speed-watching and focusing only on the most interesting parts are the only options, especially if you want to get full credits for the course. (Of course, you can take the more relaxed approach of simply not caring about the credits, and getting things done at whatever pace you find comfortable, and that's completely valid!)
One of Grossman's observations is that the length of the class correlates with completion rate. The longer the class, the more likely something will come up and throw everything off schedule. Knowing that fact means the longer the course, the more likely cutting quality will be necessary to earn a certificate.
The underlying premise for skipping around to just the interesting parts is that the entirety of the subject does not interest the student. There's nothing wrong with that. But that's not really a good target audience for the design of a class. That student is already committed to partial disengagement.
My take on programming has become that watching a programming video is like watching Jimi Hendrix play the guitar. A beginner learns almost nothing applicable because they don't have the mechanical technique as a basis for informing their seeing. A virtuoso will see new techniques and ways they could do them better.
Watching Grossman type text into a box or Ullman do a proof is less informative until I had the mechanical experience of following along. We all tune out video because it is so ubiquitous. The courses on Coursera are real courses. The results of not taking notes during lectures are pretty much the same as not taking notes during a brick and mortar lecture: some geniuses will ace the test anyway and everyone else will do average, mediocre or poorly according to their nature.
Um, I think no-one finds every part of any given subject interesting to the same degree. That's human nature. But also, maybe you already understand some parts and want to skip ahead to the more difficult parts. Sadly, in the excessively constrained format of video, sometimes the teacher will choose to focus on the topics you already understand, and neglects others. Such are the limitations of video format -- and indeed, of brick and mortar classes! Only the full, unabridged text gives you a (relatively) unconstrained explanation of all relevant topics. Of course, it also comes with some disadvantages: it can overwhelm the student, or present topics in the wrong order, or -- even -- focus on the "wrong" things, because after all written notes are also space-constrained. But I find it a way less constrained format anyway.
I find taking notes to be an uninteresting artifact of "live" classes. What truly matters is that you do the exercises and understand them.
If I gave the impression that I favored the "lazy" approach, I apologize. To be honest, I think watching the videos is both the lazy approach (only reading the notes seems like the serious approach to me) AND a time-waster. You are both more efficient AND gain a better understanding of the source material if you read the written notes/book and also do the exercises.
BTW, I found the format of the exercises for that class tedious and overly dependent on SBT. I walked away knowing some Scala and more functional programming but configuring the IDE to start a new very simple Scala project wasn't covered, unfortunately.
I will also add that Odersky's videos were a bit confusing because they were disconnected from the bricks and mortar class where they are also used, apparently. Curtailing the subject matter for the MOOC may make sense, but the those videos show how tempting it is to avoid the effort that editing video entails.
I think Grossman's class was much better designed. The slide based format without human body parts allows more flexibility with regard to editing. Taking body parts off the screen allows more flexible combination.
In both cases, I found the exercises very engaging and for me they were the main "thrill" of taking the course, but I understand that other people found them very frustrating -- they were very vocal on the forums -- often for the reasons you mentioned: the tools were hard to configure or confusing.
I just found the exercises for Functional Programming in Scala to be drudgery. That wasn't the case in Grossman's class. Odersky's class was full of klunky complexity - SBT, IDE's, and Scala's compiler performance and the warts of working on top of the JVM. Formatting the exercises partially completed code just added to the mess [though the partially completed code idea is a useful pedagogical approach].
On the other hand, Grossman's use of Emacs was one of the reasons I took the course. I still use Emacs. I don't still use Eclipse or ScalaIDE.
PS: SICP is not written either by or for Coursera, or even particularly adapted to it. If you use it as the main supporting material, what is the point of Coursera at all? In contrast, Grossman's notes were written by him and directly support his course.
Hints and solutions can be time consuming to obtain at university (office hours) and might be impossible if you can't find instructor manuals for self-study.
Khan Academy makes the most sense to me. You work the problems until you can't, at which point you use the hints or watch the videos. Then continue.
The only lectures that were useful to me at university was the first one, where they discussed how the class would run, and the last one where they discussed the exam.
Labs and tutorials were useful if I wanted to get hints on how to work problems.
I agree with TFA. Videos should by no means be the focus. They can be very helpful demonstrations. If the goal is to be able to solve problems, that's what you should be spending the most time doing.
Oddly enough, I found this to be the case, even in my humanities courses. I learned more about a subject by having to write about it, than from merely reading or memorizing for an exam.
I'm this way too, but I don't think everyone is. I think a successful program should include problem solving, reading, and audio/visual.
Because people learn differently.
Another advantage is the subtle leakage of the instructor's opinions and thought processes. At least I perceive video to be slightly better in this regard.
That's great when you're refreshing or referencing, but when you're learning something from scratch, text presents an easier opportunity to "skim" and miss things than video which sorta decides the speed of the treadmill for you.
I'm the author of a book and the amount of queries I get that were answered on the previous page.. I think many people just don't read every word in books they read (but I sympathize, because neither do I).
(PS - I'm actually looking for programming authors or bloggers who would be interested in publishing online with questions and programming challenges.)
To a big part it comes down to the fact that ideas and chapters presented in books are most often structured and sequenced in a way that suits the author in writing it. Less about what would actually be the best way of presenting the material from a learner's perspective.
I'm actively seeking feedback and testers in the education space!
At the same time as this I took another course on a different website (coursera) Cloud Computing Concepts Part 1 which was entirely lecture based. The topic was extremely interesting to me and I have moved on to the next course in the series but I do find the all lecture format with quiz based homework to be less than ideal. I try to make up for the lack of a textbook by reading the documentation of concepts covered each week but often times it is hard to pick the most essential material.
The edx platform is really stunning in terms of engagement and feedback from a student perspective. It is easy to imagine a course presented on it with little to no lecture if a strong companion text was provided.
Apart from 1971 having only nascent experience with educational videos, it's really hard to accept that the science videos of the 50s and 60s were no more effective at imparting information than a book. Seeing things move around as the presenter talks about them can really help to understand the topic.
It really depends on the topic that's being discusses - sometimes a video is better, sometimes a book is better, sometimes you absolutely have to get your hands dirty. You can read a book all you want and watch videos 'til the cows come home, but neither will make you proficient in tying a ligature.
I guess my problem with the statement is that it's absolutist. Some things are better in text, some things are better shown, some things are better done in person, not to mention other methods as well. Choose the appropriate tool for the job. And know your audience as well - the 'general public' is different to 'self-motivated autodidacts', for example.
Some part of this is that text books tend to be written as reference material, and don't come with a "first learner's cut", a path through the material which is suited for someone just starting to learn the subject.
Videos are a medium which seems to be easier to master for university instructors than really good teaching texts. Don't ask me why...
--BEGIN POST-- Every profession and every professional engages with teaching and learning in some capacity. Yet many individuals go through university or higher education degrees and fail to learn to their full potential.
Why?
Because the learning environment is focused on content delivery. In both online and offline educational settings, the structure of teaching and learning is primarily instructional: the lecturer delivers content and the student passively receives it. Lectures and tutorials are teacher-centred and teacher-directed; the teacher possesses the knowledge and power within the learning environment. There is rarely peer-to-peer learning, nor student handover or empowerment within the learning process.
This of course incites some level of learning – if it didn’t, we would all still be stuck in first-year university – but it doesn’t promote that deep, conceptual understanding of a learning area that is the hallmark of rich learning.
Despite the pervasiveness of the instructional model within higher education, how people learn best isn’t through passively receiving content. Rich learning occurs when students actively construct their own knowledge and understanding; when they connect new information with their own world and when they are intrinsically motivated in the process. This deep learning also happens through interaction, discussion and collaboration with others.
Think about a topic you’re most excited and passionate about. Do you want to be actively involved in your own learning process? Do you want to discuss and share your experiences of the topic with your peers? Do you want to connect new knowledge with your current awareness through real world examples, to develop that deeper understanding? Do you want to be intrinsically motivated to learn? Do you want to be empowered?
Or do you want to passively listen to an expert deliver a lecture?
OpenLearning wants to change the current model of higher education, to align with how students really learn. We’re about empowering students in their learning process; where learners can actively construct their own knowledge that is relevant and meaningful to them.
We provide an online learning environment that fosters deep learning, intrinsic motivation and collaboration with other students. We promote dynamic communication through fun and interactive learning communities; where students can discuss concepts, share relevant and meaningful media, and provide peer review, support and feedback in a safe and positive learning space. We are both the Learning Environment and the Student Hangout. Traditional educational models see these as mutually exclusive entities; we see them as the same thing.
Students love doing courses through our platform, because we speak in their language. Individuals and organisations love running courses through our platform, because students are engaged, intrinsically motivated and productive.
The OpenLearning philosophy promotes a lifelong love of learning that facilitates students reaching their full learning potential. --END POST--
Full Disclosure: I work at OpenLearning.