Podcast Lectures – hype, bane, or progress?

I never intended such a long delay in posting. My first semester using extensive elearning material is long completed, and it is time to reflect on the experience.

The subject was Discrete Mathematics for second year Computer Engineering students. I completely abandoned regular lectures, replacing them by videos published on a web page, supplemented by supervised exercise classes. These classes turned into interactive lectures, discussions, or individual tutoring, depending on the students’ needs of the day.

I have since learnt that this method is known as flipped classroom.

I have also learnt that the idea of flipped classroom is rather ambiguous. Newspapers are full of success stories about flipped classroom, elaborating on the technology and the use of video lectures. This promotion of flipped classroom is easily confused with the demand for flexible eduction, where students can consume their lecture at their leisure, whereever they can bring a smart phone.

Flipped classroom is, in my opinion, not about the technology, the videos, nor anything else happening outside the classroom. Flipped classroom is about what happens within the classroom. It is in fact, just a consequence of another educational buzz word, namely active learning, or even the much older slogan learning by doing. Learning rarely comes cheap, and new ideas and skills must be tried out by discussing with others, experimenting, or building something. Flipped classroom aims to spend as much as possible of the scarce contact hours on these activities.

Thus, every aspect of my module delivery has endeavoured to motivate, force, trick, and/or help the students to solve exercises, test their own understanding, and work on the subject. Below, I offer the techniques I employed as examples.

The method

The module design revolve around five components.

  1. Plenty of exercises for the students to work on individually. These are typical exercises of the format known from any mathematics textbook – no novelty about it.
  2. Class sessions, three 2h sessions per week.
  3. Fortnightly open-book class tests, graded as with pass or fail only.
  4. Video lectures, to replace traditional lectures, covering both presentation of new theory and sample exercise solutions.
  5. Closed-book, written exam (4h) at the end of term.

Thus, at the heart of it, we have plain old exercises, with the class sessions as the main forum to work on and debate the problems. The sessions were deliberately ad hoc, meeting the students where they are and answering their questions as they appeared. The sessions also provide an opportunity to monitor progress and receive feedback. The interactive form of the sessions led to a very open dialogue with the students, and their input was a great help to improve the learning materials.

The exact format of the class sessions varied depending on student needs and their level preparation. Sometimes the students would work individual, receiving help as required. More often, we would solve exercises on the blackboard. Only rarely did students volunteer to do the work on the blackboard, but very often it was possible to engage several of them in discussion, debating different approaches to problem solving. Sometimes there would be questions to be answered collectively.

Looking back, having added brief experience with a similar format on a different topic, the class sessions did not work as well as they should. The students were, generally, behind schedule, and it may appear that some end up staring at the problem for too long, without any progress. It is not certain that the ad hoc style is optimal. A more directed discussion on well-selected exercises might be able to secure a more steady progression.

The class tests were just a crude mechanism to force the students to keep up the pace with the regular exercises. They had to pass five out of six tests to be allowed to sit the exam, but the problems were all taken from the regular exercises. Being allowed to bring their notes as well as their books, it was meant only to obstruct those who did not attend regularly. Where someone failed, a second chance was given.

The video lectures were not motivated as an improvement over regular lectures. Rather, they would free up the contact hours to engage and interact with the students, allowing 6h/week where other modules may have only 2h. The philosophy is that since most of the lecture is one-way communications, there is no justification for wasting the scarce contact hours on them. Contact hours should be reserved for interaction.

Merely the statutory mechanism to justify credit awards, the written exam is of course the least important part of the module.

The benefits

Admittedly, I initially planned the supervised exercise sessions as a supplment to video lectures. It turns out in hind sight that it was the other way around, and the classroom sessions are the core. Subconsciously, I suppose, they were always the core, even in my initial conception of the semester. Post-rationalising the methodology, there are a multitude of benefits from the flipped classroom.

  1. More interaction during the contact hours give more feedback both from student to teacher and from teacher to student. Both student and teacher can be expected to improve their practice.
  2. Using contact hours for exercises raises their profile and priority.
  3. The focus is moved from presence to activity. Sleeping one’s way through the lectures is suddently a less natural approach to studying.
  4. Students can rewatch lectures which they found difficult, and have the lectures available to catch up after being absent.
  5. Students avoid a heavy cost of missing an important point early in a lecture preventing them from benefitting for the remaining 30-60 minutes. Video lectures can be rewound, and the interactive sessions actively encourage questions both in class and one-to-one.
  6. The video lectures are given in servings more suited to human attention span, usually 5-10 minutes.

I ended up with a class of students working a lot, throughout the semester; putting in much more effort than any other student cohort I have seen. They collaborated a lot, supporting eachother actively also outside the contact hours. I have never seen a similar student behaviour, neither as student nor as teacher.

Some have argued that the video lectures do not add anything beyond the textbook, and indeed, they are not a prerequisite for flipped classroom methods. The format of reading as homework and exercises at home has been known from primary schools decades ago. Maybe the only value added by video lectures is the excitement created by imaginary novelty.

The videos has a potentially useful feature compared to a standard textbook, in that each video serves as an illustration of a single key point of the syllabus. This makes it easy to navigate and the key points are high-lighted. The same effect could be created using text, in a non-book format, such as a set of hyper-referencing web pages.

Conclusion

Active learning remains an important educational method, and in the era of MOOC, it is going to be the a major selling point of campus based degree programs. The opportunity to meet the class in real life enables a greater range of potential activities, and it is much easier to inspire and engage people met face-to-face.

When the one-way lecturing can easily be moved into new media without loss, the scarce contact hours can much better be used for active learning and two-way discussion.

My onw class was small and easy to manage, starting with 16 students. Out of the 16 students one withdrew midway during the semester and one failed the exam. The 14 successful students made 1 E-s, 4 D-s, 7 C-s, and 2 B-s. Although these results are arguable mediocre, they were achieved on a degree programme used to terrible results, with very high failure rates, especially on maths modules. Maybe the teaching methods have not been optimal for the top-tier students who could have made A-s, but they seem to have raised a significant lower tier up to quite decent mid-level performance.

Posted in E-learning | Leave a comment

HOWTO: Wacom Cintiq on Linux

A stylus on screen is a wonderful design. It has all sorts of uses. My first experience with it was for teaching videos. With the stylus, I could clutter on the slides as I was speaking, in much the same way as I would with a pen on the old overhead slides of the last century. It is also handy to take notes on PDF papers while proof-reading or grading student work. It could even replace a common notebook.

 

cintiq-photo

First having learnt to apprecita the Wacom Cintiq 13 HD on a MacBook, it goes without saying that I wanted to have this wonderful device up and running on a decent, stable, operating system, such as linux. I am not a linux fan, but the serious competitors such as Sun and DEC seem to have priced themselves out of the market.

Generally, there is good support for Wacom devices, and a lot is written about debugging with different devices. Unfortunately little is written about the Wacom Cintiq 13 HD. It was not easy, but it can be done.

The good news is, in Debian jessie (testing) as of September 2013, the Cintiq works straight out of the box, mirroring the main screen. I tested the system on three different machines before reaching my final setup, so I dare say it was not pure luck.

The bad news is, not every distro is as helpful. I did quite a bit of research before I gave up on Ubuntu 13.04. Reportedly, Wacom support is not as good in 13.04 as it was in previous releases. I also tried it on the stable Debian release, wheezie, with no success, admittedly with less effort than I put into Ubuntu.

Let’s keep it simple and stick with Debian jessie. I, for one, found it easier to switch distro, than to make it work with the old one.

As far as linux is concerned, one Wacom Cintiq is two devices. It is a screen connected via HDMI and an input device connected via USB. Each one may require some tuning. Booting up the system and connecting the Cintiq, you should see the desktop on the screen and touching the screen with the stylus (pen), the mouse pointer should move, although not necessarily where you want it.

As a screen, the Cintiq is no different from any other. Presumably, you have installed appropriate drivers for your graphics card and are are connecting the Cintiq as a second screen. It can be set up either to mirror the main screen, or to extend it. If you run a desktop environment, it probably has a control panel to set up the screens. Even if you do not run Gnome (I don’t), it is possible to use its control panel, as follows:

gnome-settings-daemon &
gnome-control-centre &

Choose the display tool, and set it up as you want to. Placing the wacom in front of the main screen (as I have done in the picture above), one large desktop with the wacom under the main screen is logical.

Once you switch to a large desktop instead of mirror screens, you will probably see the mouse pointer moving far from the stylus, quite possibly on the other screen. By default, the working area of the stylus is the full desktop, even if it is larger than the wacom screen. Linux has no idea that the wacom input has any relation to the wacom screen. It could be traditional wacom device without a screen.

To fix the problem is a bit of manual fiddling, but not difficult. Firstly, we need to find the input devices, with the command:

xsetwacom --list

If the command is not found, install the package xserver-xorg-input-wacom. The output may look something like this:

Wacom Cintiq 13HD stylus         id: 9   type: STYLUS    
Wacom Cintiq 13HD eraser         id: 11  type: ERASER    
Wacom Cintiq 13HD pad            id: 12  type: PAD       

The interesting devices here are the stylus 9 (pen) and the eraser 11, which is the back of the stylus and recognised as an eraser by some software. To restrict the stylus to the screen, you do something like this:

xsetwacom --set "9" MapToOutput 1920x1080+0+1024
xsetwacom --set "11" MapToOutput 1920x1080+0+1024

Here 1920×1080 is the size of the Cintiq screen, and +0+1024 is its position relative to the full desktop. In this case it means that it is below the main screen, which has height 1024. You can find the screen sizes in your display configuration tool. If the Cintiq is above/left of the main screen +0+0 will do. If it is to the right, the first number is the width of the main screen, and the second is 0.

Once this step is complete, the pointer should be in the same ballpark as the stylus, but it may not track perfectly.

To get the pointer to follow the stylus exactly, it must be calibrated. Because of the glass between the screen image and the surface, there is parallaxis, and calibration must be done with respect to your working position and viewpoint. There are many ways to calibrate. In theory, it can be done on the command line, but I was not able to do it. On my best attempts, the pointer was 3mm off.

control-centre

My solutions was again the Gnome Control Centre. Start it as discussed above,
and choose the `Wacom Cintiq Tablet’ tool to get the window shown in the
figure. When you click the `Calibrate’ button, the rest is self-instructive.

There is one limitation which I was never able to get across. The Gnome Control Panel depends on the randr extensions in the X11 server. The randr extensions are very useful, but they do not seem to work with multiple graphics cards.
All my attempts to get the Wacom Cintiq up and running on my regular
desktop with four screens and two graphics cards failed.
I ended up with a secondary box with only one regular screen up alongside
the Wacom Cintiq. Maybe I should look for a quad head graphics card?

I am very happy with the above approach, and the Wacom Cintiq quickly became an indispensable working tool. It is highly recommended.

Posted in E-learning | Tagged | 6 Comments

Teaching videos for maths

The biggest obstacle for adoption of e-learning is to find and learn a set of tools, software and hardware, which does the job. This post is a simple account of a satisfactory approach I took on a mac.

Camtasia

screenshot

One of the most popular tools to record teaching videos is Camtasia from TechSmith. There are several tools in the Camtasia family. Personally, I tested out Camtasia for Mac. It provides two separate features in one tool. It is both a video recorder and a video editor,

Camtasia can record two synchronised video streams, one recording the screen view (screencast) and one recording from a webcam. A typical use would combine

  • A screencast of a PDF viewer with slides. Like most researchers in mathematics and computing I create my slides using LaTeX and beamer, but it really does not matter what tool you use to create them. Camtasia records whatever is on the screen.
  • A webcam recorder of the speaker’s face.

Once the recording is complete, one has to use the video editor. As a minimum, one has to figure out how to combine the screencast with the webcam recording. Maybe a small picture of the speaker on top of a full-screen screencast. This is straight-forward, with only one trick to remember. If one wants to change the layout, say moving the webcam image from one corner to another, during the video, one has to split the recording into shorter clips.

The video editor has many features which might take time to learn to use, and definately will take time to make useful. However, some useful operations are easy to adopt even for one with no particular interest in video editing. Sections can be removed, for instance if a mistake was made in the speech. Different clips can also be spliced. If you discover an error in the slides midway in the talk, you can abort, update the slide, and start over from them corrected slide. The new recording can be spliced with the first one in the editor.

In short, it takes little effort to get started with Camtasia software, and it does the job. If you are a happy mac user to start with, I think you will find it very good.

Some things annoy me, though. It does crash occasionally, and I had to give up using an external USB webcam, as I tended to lose contact withit midway in the recording. Rendering, or export, of the final video is a very slow process, and it would be much more comfortable if that could be done in batch and in the background, without blocking the GUI for the next recording.

Back to the Blackboard

However simple, the slide and webcam video is not necessarily the a very effective approach to mathematics. For starters, my students complained that the webcam image was more distracting than useful. I ended up with using a single screencast stream for the videos.

More importantly, maths teachers rarely use slides. Mathematics is a tricky thought process, and we need the students to follow as the argument develops. Therefore, we love the blackboard. Let the students follow as the ideas develop on the board. How can we effectively bring the blackboard approach on-line?

It is, of course, possible to use the web-cam to record the blackboard in a regular classrom, but it is not easy to get a good recording. It works best if the room is permanently rigged for the purpose. Quite doable if the instituation will make the investment, but difficult for individual initiatives. I found a different solution for my own part.

  • A Wacom Cintiq 13 HD is a screen (HDMI) with a stylus to allow you to draw directly on the screen. At about US$ 1000, or about 8000 NOK in this country, it is not cheap, but neither is it unreasonable.
  • Skim is PDF viewer and annotator. You can display your PDF slides and make handwritten notes on top. Skim is free to use.

Camtasia can still record the PDF view, while I am drawing thereon. This is a very flexible solution. I can have full slides, using the pen simply to highlight items as I speak. At the opposite extreme, I can start with blank slides, just like an empty blackboard. Or questions may be typed on the slide, with answers and notes added by hand. Or a typed slide at the end can summarise pages of handwritten solutions and proofs. I am still experimenting.

I am not entirely happy with Skim. To change drawing tool or colour, one has to search through menus. In another post, I will discuss xournal, and which gives me both tool and colour pallets on screen. Although I have only used xournal on linux, it appears to be available on Mac OS X as well.

Conclusions

At the end of the day, I don’t think Camtasia (for Mac) is sufficient to make useful teaching videos, at least not in mathematics. Combined with a cintiq and PDF annotation software like Skim, it is a good working solution. The cintiq is certainly well worth the investment.

I shall get back to why I rejected both Camtasia and the Mac for video teaching in another post. If you are otherwise a happy Mac user, you will probably not agree with my reasons.

Posted in E-learning | Tagged | 2 Comments

Intitial musings on e-learning

During my first decade as a higher education I have heard politicians and administrators hammering on about e-learning and digital teaching aids. It has been put as an obvious necessity, with very little discussion of the why, and even less about the how. Where would we have been if the powers that be had spent some of the resources on technical support, rather than all on bureaucrats and reports?

Developing a new module this semester on Discrete Mathematics for Level 2 Computer Engineering, I decided to take a dive into it, abandoning face-to-face lectures in favour of videos. In the next couple of weeks, I shall comment on some of my experiences and review a number of techniques and tools which I have adopted, or rejected.

Useful as it may be, e-learning technology is not the magic wand which miraculously transforms immature students to skilled graduates. We had one guest speaker from the university bureaucracy tell us of the necessity of e-learning. Lectures at set times in given locations do not suit modern students. Maybe they need to be somewhere else. They should be able to listen to the lecture whenever and whereever they find the time. The future is flexibility.

Truth is, the majority of students lack both the self-discipline and the self-motivation to complete a degree under unrestricted freedom. Their success depends on dedication and concentration, rather than flexibility and casual attention. There is a group of students who florish in a completely free and flexible environment, hardly requiring teachers at all. Most of us teach a different group of students. Where is the e-learning benefit for us?

The classic lecture suffers from two flaws

  • being transient, they do not help the students who miss it due to illness or who need repetition.

  • 45 minutes without break is at least three times the normal concentration span.

Video can resolve these to issues. Portioned into 5-10 minute clips, students can take reasonable breaks without the logistic overhead of a full class moving in and out of the room. They can take the time they need to do exercises before the next clip.

In my opininion, the opportunity to make 5-10 minute servings is the greatest benefit of e-learning. It helps me highlight key concepts, preferably one per clip, reducing a risk of an important material being passed by with the class half asleep, as often happens midway in a 2h lecture. Effective talks can be saved for reuse next term. Less effective ones can be polished and remade.

Little seems to be lost when video lectures replace the classic one-way communication of some lectures. I have a number of colleagues who have introduced video clips as an extra help for repetition over and above the convential teaching methods. For my own part, I wanted to go further, and reserve the scarcely scheduled contact hours for student interaction and responses to actual problems and questions. This personal interaction is critical. Not only do the students need continuous feedback to learn to solve problems themselves. Additionally, I need the continuous feedback from students to find the right pitch for the videos.

Without one minute believing that e-learning is the way to remote teaching, relieving rank-and-file engineerings students from the burden of on-campus presence, I continue with the project. There is no free lunch. The way is riddled with technical as well as didactical challenges. Doing it half-way is hardly worthwhile. With a massive effort one may hope to get a little improvement in learning outcomes, and recoup a some of the costs on the next course delivery another year. Maybe it’s worth it, maybe not.

Posted in E-learning | Leave a comment

Starting a blog

I have been contemplating starting this blog for some time. Mainly, it has been a debate over what tool to use. It seemed essential to use some decent content management system, to get features like keyword tagging and comment fields without wasting effort. In contrast, the prospect of having to use a web interface to type the contents was discouraging.

The solution displayed itself last week. I found the Blogit plugin for vim. At last I can get started, enjoying the benefit of a proper blog management system, while still typing my posts in a familiar editor.

There is also a python module to exploit the XML-RPC interface of WordPress. I trust it will solve my problem when I need to bulk process images or other material.

It looks promising. I will get started.

Posted in No category | Leave a comment