The Assessment in Higher Education conference 2015

I am writing this on a sunny evening, sitting in a pub overlooking Old Turn Junction, part of the Birmingham Canal Navigations, with a well-earned beer after two fascinating and exhausting days at the Assessment in Higher Education conference.

It was a lovely conference. The organising committee had set out to try to make it friendly and welcoming and they succeeded. There was a huge range of interesting talks and since I could not clone myself I was not able to go to them all. I am not going to describe individual talks in detail, but rather draw out what seemed to me to be the common themes.

A. It is all just assessment

The first keynote speaker (Maddalena Taras) said this directly, and there were a couple of other things along the same lines: the split between formative and summative assessment is a false dichotomy. If an assessment does not actually evaluate the students (give them a grade, hence summative) then it misses the main function of an assessment. This is not the same as saying that every assessment must be high stakes. Conversely, in the words of a quote Sally reminded me of:

“As I have noted, summative assessment is itself ‘formative’. It cannot help but be formative. That is not an issue. At issue is whether that formative potential of summative assessment is lethal or emancipatory. Does formative assessment exert its power to discipline and control, a power so possibly lethal that the student may be wounded for life? … Or, to the contrary, does summative assessment allow itself to be conquered by the student, who takes up a positive, even belligerent stance towards it, determined to extract every human possibility that it affords?” (Boud & Falchikov (2007) Rethinking Assessment in Higher Education: Learning for the Longer Term)

The first keynote was a critique of Assessment for Learning (AfL). Not that assessment should not help students learn. Of course it should. Rather, the speaker questioned some of the specific recommendations from the AfL literature in a thought-provoking way.

The 'couple of other things' were a talk from Jill Barber of School of Pharmacy at Birmingham, about giving students quite detailed feedback after their end of year exams; and Sally Jordan’s talk (which I did not go to since I have heard it internally at the OU) about the OU Science faculty's semantic wranglings about whether all their assessment gets called “summative” or “formative”, and hence how the marks for the separate assignments are added up, without changing what the assessed tasks actually are.

B. Do students actually attend to feedback?

The second main theme came out many times. On the one hand, students say they like feedback and demand more of it. On the other hand, there is quite a lot of evidence that many students don’t spend much time reading it, or that when they do, it does not necessarily help them to improve. So, there were various approaches suggested for getting students to engage more with feedback, for example by

• giving feedback via a screen-cast video, talking them through their essay highlighting with the mouse (David Wright & Damian Kell, Manchester Metropolitan University). Would students spend 19 minutes reading and digestion written feedback on an essay? Well, they got a 19 minute (on average) video - one of the few cases where some students thought it was too much feedback!
• making feedback a dialogue. That is, encouraging students to write questions on the cover sheet when they hand the work in, for their tutor to answer as part of the feedback. That was what Rebecca Westrup from the University of East Anglia was doing.
• Stefanie Sinclair from the OU religious studies department talked about work she had one with John Butcher & Anactoria Clarke assessing reflection in an access module (a module to designed to help students with limited prior education to develop the skills they need to study at Level 1). Again, this was to encourage students to engage in a dialogue with their tutor about their learning.
• Using peer and self assessment, so that students spend more time engaging with the assessment criteria by applying them to their own and other’s work. Also the suggestion from Maddalena Taras was that initially you give the student’s work back without the marks or feedback (but after a couple of weeks of marking) so that they read it with fresh eyes before they get the feedback (first) then the marks.
• There was another peer assessment talk, by Blazenka Divjak of the University of Zagreb, using the Moodle Workshop tool. The results were along the same lines as other similar talks I have seen (for example at the OU where we are also experimenting with the same tool). Peer assessment activities do help students understand the assessment criteria. It helps them appreciate what teachers do more. Students’ grading of their peers, particularly in aggregate, is reliable, and comparable to the teacher’s grade.
• A case of automated marking (in this case of programming exercises) where students clearly did engage with the feedback because they were allowed to submit repeatedly until they got it right. In computer programming this is authentic. It is what I do when doing Moodle development. (Stephen Nutbrown, Su Beesley, Colin Higgins, University of Nottingham and Nottingham Trent University.)
• It was also something Sally touched on in her part of our talk. With the OU's computer-marked questions with multiple tries, students say the feedback helps them learn and that they like it. However, if you look at the data or usability lab observations, you see that in some cases some students are clearly paying not attention to the feedback they get.

C. The extent to which transparency in assessment is desirable

This was the main theme of the closing keynote by Jo-Anne Baird from the Oxford University Centre for Educational Assessment. The proposition is that if assessment is not transparent enough, it is unfair because students don’t really understand what is expected of them. A lot of university assessment is probably towards this end of the spectrum.

Conversely, if assessment is too transparent it encourages pathological teaching to the test. This is probably where most school assessment is right now, and it is exacerbated by the excessive ways school exams are made hight stakes, for the student, the teacher and the school. Too much transparency (and risk averseness) in setting assessment can lead to exams that are too predicable, hence students can get a good mark by studying just those things that are likely to be on the exam. This damages validity, and more importantly damages education.

Between these extremes there is a desirable balance where students are given enough information about what is required of them to enable them to develop as knowledgable and independent learners, without causing pathological behaviour. That, at least, is the hope.

While this was the focus of the last keynote, it resonated with several of the talks I listed in the previous section.

D. The NSS & other acronyms

The National Student Survey (NSS) is clearly a driver for change initiatives at a lot of other universities (as it was two years ago). It is, or at least it is perceived to be a, big deal. Therefore it can be used as a catalyst or leaver to get people to review and change their assessment practices since feedback and assessment is something that students often give low ratings for. This struck me as odd, since I am not aware of this happening at the OU. I assume that is because the OU has so far scored highly in the NSS.

The other acronym floating around a lot was TESTA. This seems to be a framework for reviewing the assessment practice of a whole department or degree programme. In one case, however (a talk by Jessica Evans & Simon Bromley of the OU faculty of Social Science) their review was done before TESTA was invented, though along similar lines.

Finally

A big thank-you to Sue Bloxham and the rest of the organising team for putting together a great conference. Roll on 2017.

eSTEeM conference 2015

eSTEeM is an organising group within the Open University which brings together people doing research into teaching and learning in the STEM disciplines, Science, Technology, Engineering and Maths. Naturally enough for the OU, a lot of that work revolves around educational technology. Once a year they have an annual conference for people to share what they have been doing. I went along because I like to see what people have been doing with our VLE, and hence how we could make it work better for students and staff in the future.

It started promisingly enough in a way. As I walked in to get my cup of coffee after registration, I was immediately grabbed by Elaine Moore from Chemistry who had two Moodle Quiz issues. She wanted the Combined question type to use the HTML editor for multiple choice choices (good idea, we should put that on the backlog) and a problem with a Pattern-match questions which we could not get to the bottom of over coffee.

But, on to the conference itself. I cannot possibly cover all the keynotes and parallel sessions so I will pick the highlights for me.

Assessment matters to students

The first was a graph from Linda Price’s keynote. Like most universities, at the end of every module we give have a student satisfaction survey. The graph showed the student's ratings in response to three of the questions:

• Overall, I am satisfied with the quality of this module.
• I had a clear understanding of what was required to complete the assessed activities.
• The assessment activities supported my learning.

There was an extremely strong correlation between those. This is nothing very new. We know that assessment is important in determining the ‘hidden curriculum’, and hence we like to think that ‘authentic assessment’ is important. However, it is interesting to see how much this matters this is to students. Previously, I would not even have been sure that they could tell the difference.

The purpose of education

Into the parallel sessions. There was an interesting talk from the module team for TU100 my digital life, the first course in the computing and technology degrees. Some of the things they do in that module’s teaching is based around the importance of language, even in science. Learning a subject can be thought of as learning to construct the world of that subject through language, or as they put it, humanities style thinking in technology education. Unsurprisingly, many students don’t like that “I came to learn computing, not writing.” However, there is a strong correlation between students language use and their performance in assessments. By the end of the module some students do come to appreciate what the module is trying to do.

This talk triggered a link to back to another part of Linda Price’s keynote. An important (if now rather cliched question) for formal education is “What is education for everything is now available on the web?” (or one might put that more crudely as “Why should students pay thousands of pounds for one of our degrees?”). The answer that came to me during this talk was “To make them do things they don’t enjoy, because we know it will do them good.” OK, so that is a joke, but I would like to think there is a nugget of truth in there.

Peer assessment

On to more specifically Moodle-related things. A number of modules have been trying out Moodle’s Workshop activity. That is a tool for peer review or peer assessment. The talk was from the SD815 Contemporary issues in brain and behaviour module team. Their activity involved students recording a presentation (PowerPoint + audio) that critically evaluated a research article. Then they had to upload them to the Moodle Workshop, and review each others presentations as managed by the tool. Finally, they had to take their slide-cast, the feedback they had received, and a reflective note on the process and what they had learned from it, and hand it all in to be graded by their tutor.

Now for OU students (at least) collaborative activities, particularly those tied to assessments, are typically another thing we make them do that they don’t enjoy. This activity added the complexities of PowerPoint and/or Open Office and recording audio. However, it seems to have worked remarkably well. Students appreciated all the things that are normally said about peer review: getting to see other approaches to the same task; practising the skills of evaluating others’ work and giving constructive feedback. In this case the task was one that the students (healthcare workers studying at postgraduate level) could see was relevant to their vocation, which brings us back to visibly authentic assessment, and the student satisfaction graph from the opening keynote.

For me the strongest message from this talk, however, is what was not said. There was very little said about the Moodle workshop tool, beyond a few screen-grabs to show what it looked like. It seems that this is a tool that does what you need it to do without getting in the way, which is normally what you want from educational technology.

Skipping briefly over

There are many more interesting things I could write about in detail, but to keep this post to a reasonable length I will just skim over the posters with lunch. For example,

• The S207 The physical world module team giving more information about the worrying gender differences in achievement in level 2 physics. It is still early days as they try to work out what is going on there, and what they might be able to do about it, but it is being taken very seriously.
• An effort to analyse all the 2000+ figures and diagrams in S215 Chemistry: essential concepts to try to work out how best to make Chemistry modules accessible to the visually impaired.
• A study that captured and analysed all the emails we sent to a group of new students.

And, some of the other talks:

• a session on learning analytics, in this case with a neural net, to try to identify early on those students (on TU100 again) who get through all the continuous assessment tasks with a passing grade, only to fail the end of module assessment, so that they could be targeted for extra support.
• a whole morning on the second day, where we saw nine different approaches to remote experiments from around the world. For example, the Open University's remote control telescope PIRATE. I was left me with the impression that this sort of thing is much more feasible and worthwhile than I had previously thought.

Our session on online Quizzes

The only other session I will talk about in detail is the one I helped run. It was a ‘structured discussion’ about the OU’s use of iCMAs (which is what we call Moodle quizzes). I found this surprisingly nerve-wracking. I have given plenty of talks before, and you prepare them. You know what you are going to say, and you are fairly sure it is interesting. Therefore you are pretty sure what is going to happen. For this session, we just had three questions, and it was really up to the attendees how well it worked.

We did allow ourselves two five-minute presentations. We started with Frances Chetwynd showing some the different ways quizzes are used in modules’ teaching and assessment strategies. This set up a 10-minute discussion of our first question: “How are iCMAs best be used as part of an assessment strategy?”. For this, delegates were seated around four tables, with four of five participants and a facilitator to each table. The tables were covered with flip-chart paper for people to write on.

We were using a World Café format, so after 10 minutes I rang my bell, and all the delegates move to a new table while the facilitators stayed put. Then, in new groups, they discussed the second question: "How can we engage students using iCMAs?" The facilitators were meant to make a brief bridge between what had been said in the previous group at their table, before moving on to the new question with the new group.

After 10 minutes on the second question, we had the other five-minute talk from Sally Jordan, showing some examples of what we have previously learned through scholarship into how iCMAs work in practice. (If you are interested in that, come to my talk at either MoodleMoot IE UK 2015 or iMoot 2015). This lead nicely, after one more round of musical chairs, to the third question: "Where next for iCMAs? Where next for iCMA scholarship?". Finally we wrapped up with a brief plenary to capture they key answers to that last question from each table.

By the end, I really had no idea how well it had gone, although each time I rang my bell, I felt I was interrupting really good conversations. Subsequently, I have written up the notes from each table, and heard from some of the attendees that they had found it useful and interesting, so that is a relief. We had a great team of facilitators (Frances, Jon, Ingrid, Anna) which helped. I would certainly consider using the the same format again. With a traditional presentation, you are always left with the worry that perhaps you got more out of preparing and delivering the presentation than any of the audience did out of listening. In this case, I am sure the audience got much more out of it than me, which is no bad thing.

Is learning design like UML?

A couple of weeks ago, I attended the #design4learning conference, which was conveniently on my doorstep at the Open University. Jenny Gray has already written her summary of the conference (and she though she was a bit late writing it up!)

I would like to highlight the point the organisers made with the conference name. Calling learning design "learning design" is a misnomer. You cannot design learning. Learning is something that goes on inside the student's head, perhaps most effectively under the support and guidance of a teacher. Therefore, you can only "design for learning", whatever it is you are designing: a course, a activity, a learning community, …. I think this is more than just semantic pedantry. We should all remember this, particularly when thinking about educational technology. There is no magic bullet that guarantees learning will occur. Just things that are more or less likely to encourage students to learn. (Having said this, I am going to just write "learning design" in the rest of this post, since it is so much easier!)

The main thought I wanted to share here is, however, something else. After two interesting days at a conference all about learning design, I cannot recall a single diagram shown by any speaker where I thought, "that is a graphical representation of the design of a bit of learning." Was I right to expect to see that? I don't know, but I have seen other presentation about tools like CompendiumLD in the past so I know it can be done. Pondering this as I cycled home, I got to thinking about the type of design I do know about: design of software, and though of an interesting comparison.

Software developers have a well-established way to draw the design of their software, called UML (better description on Wikipedia). Let me say immediately that I am not trying to suggest UML as a way to represent learning designs. Rather, I think it is interesting to think about how developers do (or more often don't) use UML to help their work. Can that tell us anything about how and whether teachers might engage with learning design?

There are two different ways to use UML. There is the quick-and-dirty, back-of-the-envelope way, where you draw of a part of the system to help explain or communicate a particular aspect of your design. This is the way I use UML as can be seen, for example, in this documentation I wrote. You include the details that are relevant to making your point, and leave out anything that does not help.

The other way to use UML is much more elaborate. It is called "Model Driven Architecture" which I studied as part of OU module M885. In MDA, you try to draw complete diagrams of the design of your system using a very precise dialect of UML, dotting all the 'i's and crossing all the 't's. Then, using a software tool (that you probably had to buy at great expense) you press the magic button, and it creates all your classes and interfaces for you. Then you just need to fill in all the implementations. At least, that is the promise. As I say, I studied this as part of a postgraduate computing course. It was of some academic interest, but I have never seen anyone write software this way (though a some people do, if the references in the course are to be believed). I expect more people have bought expensive MDA tools than have actually used them. In a previous generation, the same was true of CASE tools that also failed to live up to their promises.

So what, if anything, can this tell us about learning design? Well, I can see exactly the same split happening. There will be hype about magic systems where you input your learning outcomes, and sketch your learning design, press a magic button, and hey, presto, there is your Moodle course. It won't work outside of research labs, but some vendors will try to commercialise it, and a some institutions will fall for it and end up with expensive white elephants.

On the other hand, it would be good to see a common notation emerge to represent learning designs. This would help teachers communicate with each other, and perhaps with students, about how their teaching is supposed to work. A good feature of UML is that it is really very natural. Most developers can understand most of a UML diagram without having to be taught a lot of rules. There are several types of diagram to represent different things, but they are the kinds of things people drew anyway before UML was invented. The creators of UML just picked one particular way of drawing each sort of diagram, and endorsed it, in an attempt to get everyone talking (drawing) a common language. If you want to draw highly detailed UML diagrams, you need to learn a lot of rules, but you can get a long way just by copying what you see other people do, which is a sign of an effective language. It would be nice to see such a language for communicating about learning.

Load-testing Moodle 2.6.2 at the OU

At the start of June we will upgrade the OU Moodle sites to Moodle 2.6. Before then we need to know that it will still perform well when subjected to our typical peak load of 100,000 page-views per hour. This time, I got 'volunteered' to do the testing.

The testing servers

To do the testing, we have a set of 10 servers that are roughly similar to our live servers. That is six web servers for handling normal requests, one web server that handles 'administrative' requests. That is, any URL starting /admin, /report or /backup. Those pages are often big, long-running processes, rather than quick page views, so it is better to put them on a different server that is tuned differently. There is one 'web' server is just for running the cron batch processes. Finally we have a database server and a file server.

In order to be able to make easy comparisons, we make two copies of our live site onto these servers. That is, we have two different www_root folders, which correspond to different URLs lrn2-perf-cur and lrn2-perf-upg. In due course we will upgrade one of the copies to the new release while leaving the other open running the current version of the code. This make it easy to switch back and forth when comparing the two.

In addition to the servers running Moodle, we have 6 virtual machines to generate the simulated load.

The testing procedure

We test using JMeter. In order to test Moodle, you need to send lots of requests for different pages, many of which include numeric ids in the URLs. Therefore, the JMeter script needs to be written specifically for the site being tested. Fortunately, our former colleague James Brisland made a script that automatically generates the necessary JMeter script. We shared that script with the community, and you can find a copy here. However, we shared it a long time ago, and since then our version has probably diverged from the community version a bit. Oops!

I say this tool automatically generates the necessary JMeter script, but sadly that is an oversimplification. It fails in certain cases like if a forum is set to separate groups mode. So, having generated the JMeter script, you need to run it and check that it actually works. If not, you have to go into the courses and activities being tested and modify the settings. We really ought to automate that, but no one has had the time. Anyway, eventually (and this took ages) you have a working test script.

Tuning the test script

Once the test script works, in that it simulates users performing various actions without error, one at a time, then you have to start running it at high load. That is, simulating lots of users doing lots of things simultaneously. After it has settled down, you let it run for 15 or 20 minutes, and then look at what sort of load you are generating. The goal is to get about the same number of requests per second for each type of page (course view, forum view, post to forum, view resource, ...) in the test run as in real use on the live system. If not, you tweak the time delays, or number of threads, and then run again. It took about four runs to get to a simulated load that was close (actually slightly higher) than the target request rates we had taken from the live server logs.

All that creation and tuning of the tests scripts is done on the lrn2-perf-cur copy of the site. Once that is OK, then you run the same script against lrn2-perf-upg. That should give exactly the same performance, and before proceeding we want to verify that is the case. It turned out at first that it was slightly different. I had to find the few admin settings that were different between the two servers. Once the configuration was the same, the performance was the same, and we were finally in a position to start comparing the old and new systems.

Upgrade to the new version of Moodle

The next step is to upgrade lrn2-perf-upg to the new code. This code is still work-in-progress. Final testing of the code before release happens next month, but we try to keep our code in a releasable state, so it should be OK for load-testing. However, this is the first time we have run the upgrade on a copy of all our data. Unsurprisingly, we found some bugs. Fortunately they were easily fixed, and better to find them now than later.

Also, a new version of Moodle comes with a lot of new configuration options. This is the moment to consider what we should set them to. Luckily, most of the default values were right, so there was not a lot to do. Moodle prompts you for most of the new settings you need to make as part of the upgrade. However, it does not prompt you to configure any new caches, so you have to remember to go and do that.

Compare performance

At long last (about one and a half weeks into the process) you are finally ready to run the test you want. How does 2.6 performance compare to 2.5? Here is a screen-grab of today's testing:

Good news: Moodle 2.6 is mostly a bit faster (5-10%) than Moodle 2.5. Bad news: every 15 minutes, it suddenly goes slow for about 15 seconds. What?!

Problem solving

Actually, there is a logical explanation. We have cron set to run every 15 minutes, so surely the problem is caused by cron, right? No. Wrong! We stopped cron running, and the spikes remained. We tried various things to see what it might be, and could not make any sense of it. One thing we discovered was that the spikes were about as large as the spikes you get by clicking the 'Purge all caches' button. OK, so something is purging caches, but what?

To cut a long story short, you need to remember that our two test sites lrn2-perf-cur and lrn2-perf-upg are sharing the same servers. Therefore they are sharing the same memcache storage. It appears that something in cron in Moodle 2.5 purges at least some of the caches. When we stopped cron on our Moodle 2.5 site the spikes went away on our 2.6 site. I am afraid we did not try to work out why Moodle 2.5 cron was purging caches, but there is probably a bug there. It turns out that purge caches does not cause a measureable slow-down in Moodle 2.5, at least not for us, which is worth knowing.

Why does Purge caches cause a slow-down in 2.6 but not in 2.5? I am pretty sure the reason is MDL-41436. When things slowed down, it was the course page that slowed down the most, and that is the one most dependent on the modinfo cache.

Summary

• Moodle 2.6 is about 5-10% faster than 2.5, at least on our servers, which are RHEL5 + Postgres + memcache cluster store. (MDL-42071 - why has that not been integrated yet?)
• In Moodle 2.5, doing Purge caches when your system is running at high load seems to cause remarkably little slow-down.
• In Moodle 2.6, doing Purge caches does slow things down a lot, but only very briefly. Performance recovered within about 15 seconds in our test, but then the test was only using a few courses.
• In Moodle 2.6, clicking Clear theme caches (at the top of Admin -> Appearance -> Themes -> Theme selector) causes no noticeable slow-down.

The bit about what happens when you clear the caches is important because sometimes, when you patch the system with a bug fix, you need to purge one or more caches to make the fix take effect. In the past, we did not know what effect that had. We were cautious and had people waiting up until after midnight to click the button at a time of low system load. It turns out now that is probably not necessary. We can clear caches during the working day, when staff are in the office to pick up the pieces if anything does go wrong.

Reflections on listening to conference presentations in German

I am at the MoodleMaharaMoot in Leipzig listening to people talk about Moodle.

First, the good news is that about half the words in English came from the same roots as German, so there are a fair number of words you can recognise, at least if you have time to read them from the screen. For words that seem really key, there is Google translate. Also, the Germans seems like using English phrases for eLearning-related things, like Learning Analytics, or Multiple Choice.

However, I don’t think I was even understanding 10% of the words. What really makes a difference to intelligibility is what is on the screen. If speaker just had powerpoint slides with textual bullet points, that does not help. If the speaker uses the screen to show you what they are talking about - screen grabs or live demos - that is much better. Of course, this is just: show, don’t tell.

It also makes a big difference whether you already know a little bit about what is being said. I talked to some people from University of Vienna two years ago when they started building their offline quiz activity, so I already knew what it was supposed to do. I followed that presentation (which contained many screen-grabs) better than most. What they have done looks really slick, by the way.

Regarding my presentation, I feel vindicated in my plan to spend almost all of the presentation doing a live demonstration of the question types I was talking about. Of course, I am sure that almost everyone in the audience has better English than I have German. Also, I apologies that I talked for the whole time, and did not leave an opportunity for questions.

Finally, I have been speculating (without reaching any conclusions) about whether the experience of sitting there, failing to understand almost everything that is being said, and just picking some scraps from the slides, is giving me any empathy for people with severe disabilities who need major accessibility support to use software? As I say, these thoughts are inconclusive. What does anyone else think?

By the way, Germans applaud by rapping on the table with their knuckles. Your trivia fact for the day.

Moving the OU Moodle code to Moodle 2.6.1

I spent today upgrading our Moodle codebase from Moodle 2.5.4 to Moodle 2.6.1. This is the start of work towards our June release of the VLE. We have a March release based on Moodle 2.5.4 to get on the live servers first, and testing that will overlap with the development of the 2.6.1-based version.

Doing the merge

The first stage of the process is to merge in the new code. This is non-trivial because even if you just do

git checkout -b temp v2.5.4
git merge v2.6.1

Then you will get a lot of merge conflicts. That is a product of how the Moodle project manages its stable branches. If your own code changes also lead to other merge conflicts, then sorting out the two is a real mess.

Fortunately, there is a better way, because we know how we want to resolve any conflicts between 2.5.4 and 2.6.1. We want to end up with 2.6.1. Using git merge strategies, you can do that:

git checkout -b merge_helper_branch v2.6.1
git merge --strategy=ours v2.5.4

That gives you a commit that is upstream of both v2.5.4 and v2.6.1, and which contains code that is identical to v2.6.1. You can verify that using git diff v2.6.1 merge_helper_branch. That should produce no output.

Having built that helper branch, you can then proceed to upgrade your version of the code. Our version of Moodle lives on a branch called ouvle which we originally branched off Moodle 2.1.2 in October 2011. Since then, we have made lots of changes, including adding many custom plugins, and merging in many Moodle releases. Continuting from the above we do

git checkout ouvle
git merge --strategy-option=patience merge_helper_branch

That gave a lot of merge conflicts, but they were all to do with our changes. Most of them were due to MDL-38189, which sam marshall developed for Moodle 2.6, and which we had back-ported into our 2.5 code. That back-port made a big mess, but fortunately most of the files affected did not have any other ou-specific changes, so I could just overwrite them with the latest versions from v2.6.1.

git checkout --theirs lang/en backup lib/filestorage admin/settings/development.php lib/form/form.js
git add lang/en backup lib/filestorage admin/settings/development.php lib/form/form.js

Simiarly, we had backported MDL-35053 which lead to more conflicts that were easy to resolve. Another case was the Single activity course format which we had used as an add-on to Moodle 2.5. That is now part of the standard Moodle release. The change caused merge conflits, but again there was a simple solution: take the latest from 2.6.1.

After all that, there were only about 5 files that needed more detailed attention. They were mostly where a change had been made to standard Moodle code right next to a place where we had made a change. (Silly rules about full stops at the ends of comments!) They were easily to fix manually. The one tricky file was in lib/moodlelib.php where about 400 lines of code had been moved lib/classes/useragent.php. There were two ou-specific changes in the middle of that, which I had to re-do in the new version of that code.

Verifying the merge

Having resolved all the conflicts, it was then time to try to convince myself that I had not screwed anything up. The main check was to comparing our ouvle code with the standard 2.6.1 code. Just doing git diff v2.6.1 ouvle does not work well because it shows all contents of all the new files we have added. You need to read the git documentation and work out the incantation

git diff --patience --diff-filter=CDMRTUXB v2.6.1 ouvle

That tells git to just show changes to existing files - the ones that are part of standard Moodle 2.6.1. That is a manageable amount of output to review. We have a strict policy that any change to core Moodle code is marked up like this:

// ou-specific begins #2381 MDL-28567
/*
        $select = new single_select(new moodle_url(CALENDAR_URL.'set.php',
                array('return' => base64_encode($returnurl->out(false)),
                        'var' => 'setcourse', 'sesskey'=>sesskey())),
                'id', $courseoptions, $selected, null);
*/
        $select = new single_select(new moodle_url(CALENDAR_URL.'view.php',
                array('return' => $returnurl, 'view' => 'month')),
                'course', $courseoptions, $selected, null);
// ou-specific ends #2381 MDL-28567

That is, the original Moodle code is still there, but commented out, alongside our modified version, and the whole thing is wrapped in paired begin and end markers that refer to a ticket id in our issues database and if applicable a Moodle tracker issue. In this case I can check that MDL-28567 has still not been resolved, so we still need this ou-specific change. What I am doing looking at the diff output is verifying that every change is marked up like that, and that any issues mentioned are things that are still relevant.

The other check is to search the whole codebase for ou-specific and again review all the issue numbers mentioned. These combined checks find a few ou-specific changes that are no longer needed, which is a good thing.

What happens next

Now that I think the code seems right, it is time to test it, so I upgrade my development install. It mostly works, except that our custom memcache session handler no longer works (the session code seems to have changed a lot, including an official memcached session hander in core). For now I just switch back to default Moodle sessions, and make a note to investigate this later.

Apart from that, the upgrade goes smootly, and, apart from thousands of debugging warnings about use of deprecated code, I have a working Moodle site, so I push the code to our git server, and warn the rest of the team that they can upgrade if they feel brave.

The next thing, which will take place over the next few weeks is to check every single one of our custom plugins to verify that it still works properly in Moodle 2.6. To manage that we use a Google Docs spreadsheet that we can all edit that lists all the add-ons, with who is going to be responsible for checking it, and whether they have done so yet. Here is a small section.

The state of OU Moodle customisation

Our regular code-merges are a good moment to take stock of the extend to which we have customised Moodle. Here are some headline numbers:

• 212 custom plug-ins: Of those 10 are ones we have taken from the community, including Questionnaire, Certificate, Code-checker and STACK (we helped create those last two). Of our own plugins, 58 (over a quarter) are shared with the community, though the counting is odd because ForumNG contains 20 sub-plugins.
• 17 ou-specific issues: That is, reasons we made a change to core code that could not be an add-on.
• Due to those 17 reasons, there are 42 pairs of // ou-specific begins/ends comments in the code.

So, we continue to be disciplined about not changing core code unless we really have to, but the number of plugins is getting a bit crazy. A lot of the plugins, are, however, very small. They just do one thing. Also, we run a range of very different sites, including OpenLearn, OpenLearn works, The Open Science Lab and our exams server. A significant number of our plugisn were just designed to be used on one of those sites.

Here are the numbers of custom plugins broken down by type (and ignoring sub-plugins of our custom plugins).

Plugin type	Number
Activity module	25
Admin tools	8
Authentication methods	2
Blocks	30
Course formats	3
Editors	1
Enrolment methods	1
Filters	6
Gradebook reports	1
Local plugins	44
Message outputs	2
Portfolio outputs	1
Question behaviours	4
Question types	14
Quiz reports	6
Quiz access rules	2
Reports	19
Repositories	3
Themes	9
TinyMCE plugins	1

Bug fixing as knowledge creation

There are lots of ways you can think about bug-fixing: it is just a job that developers do; it is problem solving; etc. Here I want to take one particular viewpoint, that it is generating new knowledge about a software system.

One was to think about software is that it is the embodiment of a set of requirements, of how something should work. For example, Moodle can be thought of as a lot of knowledge about what software is required to teach online, and how that software should be designed. Finding and fixing bugs increases that pool of knowledge by identifying errors or omissions and then correcting them.

The bug fixing process

We can break down the process of discovering and fixing a bug into the following steps. This is really trying to break the process down as finely as possible. As you read this list, please think about what new knowledge is generated during each step.

1. Something's wrong: We start from a state of blissful ignorance. We think our software works exactly as it should, and then some blighter comes along and tells us "Did you know that sometimes ... happens?" Not what you want to hear, but just knowing that there is a problem is vital. In fact the key moment is not when we are told about the problem, but when the user encountered it. Good users report the problems they encounter with an appropriate amount of detail
2. Steps to reproduce: Knowing the problem exists is vital, but not a great place to start investigating. What you need to know is something like "Using Internet Explorer 9, if you are logged in as a student, are on this page, and then click that link then on the next page press that button, then you get this error." and that all the details there are relevant. This is called steps to reproduce. For some bugs they are trivial. For bugs that initially appear to be random, identifying the critical factors can be a major undertaking.
3. Which code is broken: Once the developer can reliably trigger the bug, then it is possible to investigate. The first thing to work out is which bit of code is failing. That is, which lines in which file.
4. What is going wrong: As well as locating the problem code, you also have to understand why it is misbehaving. Is it making some assumption that is not true? Is it misusing another bit of code? Is it mishandling certain unusual input values? ...
5. How should it be fixed: Once the problem is understood, then you can plan the general approach to solving it. This may be obvious given the problem, but in some cases there is a choice of different ways you could fix it, and the best approach must be selected.
6. Fix the code: Once you know how you will fix the bug, you need to write the specific code that embodies that fix. This is probably the bit that most people think of when you say bug-fixing, but it is just a tiny part.
7. No unintended consequences: This could well be the hardest step. You have made a change which fixed the specific symptoms that were reported, but have you changed anything else? Sometimes a bug fix in one place will break other things, which must be avoided. This is a place where peer review, getting another developer to look at your proposed changes, is most likely to spot something you missed.
8. How to test this change: Given the changes you made, what should be done to verify that the issue is fixed, and that nothing else has broken? You can start with the steps to reproduce. If you work through those, there should no longer be an error. Given the previous point, however, other parts of the system may also need to be tested, and those need to be identified.
9. Verifying the fix works: Given the fixed software, and the information about what needs to be tested, then you actually need to perform those tests, and verify that everything works.

Some examples

In many cases you hardly notice some of the steps. For example, if the software always fails in a certain place with an informative error message, then that might jump you to step 4. To give a recent example: MDL-42863 was reported to me with this error message:

Error reading from database

Debug info: ERROR: relation "mdl_questions" does not exist

LINE 1: ...ECT count(1) FROM mdl_qtype_combined t1 LEFT JOIN mdl_questi...

SELECT count(1) FROM mdl_qtype_combined t1 LEFT JOIN mdl_questions t2 ON t1.questionid = t2.id WHERE t1.questionid <> $1 AND t2.id IS NULL

[array (0 => '0',]

Error code: dmlreadexception

Stack trace:

• line 423 of /lib/dml/moodle_database.php: dml_read_exception thrown
• line 248 of /lib/dml/pgsql_native_moodle_database.php: call to moodle_database->query_end()
• line 764 of /lib/dml/pgsql_native_moodle_database.php: call to pgsql_native_moodle_database->query_end()
• line 1397 of /lib/dml/moodle_database.php: call to pgsql_native_moodle_database->get_records_sql()
• line 1470 of /lib/dml/moodle_database.php: call to moodle_database->get_record_sql()
• line 1641 of /lib/dml/moodle_database.php: call to moodle_database->get_field_sql()
• line 105 of /admin/tool/xmldb/actions/check_foreign_keys/check_foreign_keys.class.php: call to moodle_database->count_records_sql()
• line 159 of /admin/tool/xmldb/actions/XMLDBCheckAction.class.php: call to check_foreign_keys->check_table()
• line 69 of /admin/tool/xmldb/index.php: call to XMLDBCheckAction->invoke()

I have emboldened the key bit that says where the error is. Well, there are really two errors here. One is that the Combined question type add-on refers to mdl_questions when it should be mdl_question. The other is that the XMLDB check should not die with a fatal error if presented with bad input like this. The point is, this was all immediately obvious to me from the error message.

Another recent example at the other extreme is MDL-42880. There was no error message in this case, but presumably someone noticed that some of their quiz settings had changed unexpectedly (Step 1). Then John Hoopes, who reported the bug, had to do some careful investigation to work out what was going on (Step 2). I am glad he did, because it was pretty subtle thing, so in this case Step 2 was probably a lot of work. From there, it was obvious which bit of the code was broken (Step 3).

Note that Step 3 is not always obvious even when you have an error message. Sometimes things only blow up later as a consequence of something that went wrong before. To use an extreme example, if someone fills your kettle with petrol, instead of water, and then you turn it on to make some tea and it blows up. The error is not with turning the kettle on to make tea, but with filling it with petrol. If all you have is shrapnel, finding out how the petrol ended up in the kettle might be quite hard. (I have no idea why I dreamt up that particular analogy!)

MDL-42880 also shows the difference between the conceptual Steps 4 and 5, and the code-related Steps 3 and 6. I though the problem was with a certain variable becoming un-set at a certain time, so I coded a fix to ensure the value was never lost. That led to complex code that required a paragraph-long comment to try to explain it. Then I had a chat with Sam Marshall who suggested that in fact the problem was that another bit of code was relying on the value that variable, when actually the value was irrelevant. That lead to a simpler (hence better) fix: stop depending on the irrelevant value.

What does this mean for software?

There are a few obvious consequences that I want to mention here, although they are well known good practice. I am sure there are other more subtle ones.

First, you want the error messages output by your software to be as clear and informative as possible. They should lead you to where the problem actually occurred, rather than having symptoms only manifesting later. We don't want exploding kettles. There are some good examples of this in Moodle.

• Moodle will tell you if your forgot to call setType when defining a form. If you forget, then you might create a security hole when users submit the form. (Because of a belt-and-braces approach in Moodle, MDL-42540 does not open any security holes, which is why I can safely link to it.)
• Moodle will tell you if the first column in a result set is not unique when it needs to be. In the forum post I just linked to, the developer had not set up their Moodle to display the helpful warnings, and so they encountered what appeared to be a mystifying bug. That is why I added the warning following a trying debugging session six years ago.

Second, because Step 7, ensuring that you have not broken anything else, is hard, it really pays to structure your software well. If you software is made up of separate modules that are each responsible for doing one thing, and which communicate in defined ways, then it is easier to know what the effect of changing a bit of one component is. If your software is a big tangle, who knows the effect of pulling one string.

Third, it really pays to engage with your users and get them to report bugs. Of course, you would like to find and fix all the bugs before you release the software, but that is impossible. For example, we are working towards a new release of the OU's Moodle platform at the start of December. We have had two professional testers testing it for a month, and a few select users doing various bits of ad-hoc testing. That adds up to less than 100 person days. On the day the software is released, probably 50,000 different users will log in. 50,000 user days, even by non-expert testers, are quite likely to find something that no-one else noticed.

What does this mean for users?

The more important consequences are for users, particularly of open-source software.

• Reporting bugs (Step 1) is a valuable contribution. You are adding to the collective knowledge of the project.

There are, however, some caveats that follow from the fact that in many projects, the number of developers available to fix bugs is smaller than the number of users reporting bugs.

• If you report a bug that was already reported, then someone will have to find the duplicate and link the two. Rather than being a useful contribution, this just wastes resources, so try hard to find any existing bug report, and add your information there, before creating a new one.
• You can contribute more by reporting good steps to reproduce (Step 2). It does not require a developer to work those out, and if you can do it, then there is more chance that someone else will do the remaining work to fix the bug. On the other hand, there is something of a knack to working out and testing which factors are, or are not, significant in triggering a bug. The chances are that an experienced developer or tester can work out the steps to reproduce quicker than you could. If, however, all the experienced developers are busy then waiting for them to have time to investigate is probably slower than investigating yourself. If you are interested, you can develop your won diagnosis skills.
• If you have an error message then copy and paste it exactly. It may be all the information you need to give to get straight to Step 3 or 4. In Moodle you can get a really detailed error message by setting 'debugging' to 'DEVELOPER' level, then triggering the bug again. (One of the craziest mis-features in Windows is that most error pop-ups do not let you copy-and-paste the message. Paraphrased error messages can be worse than useless.)

Finally, it is worth pointing out that Step 9 is another thing that can be done by the user, not a developer. For developers, it is really motivating when the person who reported the bug bothers to try it out and confirm that it works. This can be vital when the problem only occurs in an environment that the developer cannot easily replicate (for example an Oracle-specific bug in Moodle).

Conclusion

Thinking about bug finding and fixing as knowledge creation puts a more positive spin on the whole process than is normally the case. This shows that lots of people, not just developers and testers, have something useful to contribute. This is something that open source projects are particularly good at harnessing.

It also shows why it makes sense for an organisation like the Open University to participate in an open source community like Moodle: Bugs may be discovered before they harm our users. Other people may help diagnose the problem, and there is a large community of developers with whom we can discuss different possible solutions. Other people will help test our fixes, and can help us verify that they do not have unintended consequences.

Open University question types ready for Moodle 2.5

This is just a brief note to say that Colin Chambers has now updated all the OU question types to work with Moodle 2.5. Note that we are not yet running this code ourselves on our live servers, since we are on Moodle 2.4 until the autumn, but Phil Butcher has tested them all and he is very thorough.

You can download all these question types (and others) from the Moodle add-ons database.

Thanks to Dan Poltawski's Github repository plugin, that is easier than it used to be. Still, updating 10 plugins is pretty dull, so I feel like I have contributed a bit. I also reviewed most of the changes and fixed the unit tests.

I hope you enjoy our add-ons. I am wondering whether we should add the drag-and-drop questions types to the standard Moodle release. What do you think? If that seems like a good idea to you, I suggest posting something enthusiastic in the Moodle quiz forum. It will be easier to justify adding these question types to standard Moodle if lots of non-OU Moodlers ask for it.

Performance-testing Moodle

Background

The Open University is moving from Moodle 2.3.x to Moodle 2.4.3 in June. As is usual with a major upgrade, we (that is Rod and Derek) did some load testing to see if it still runs fast enough on our servers.

The first results were spectacularly bad! Moodle 2.4 was ten times slower. We were expecting Moodle 2.4 to be faster than 2.3. The first step was easy.

Performance advice: if you are running Moodle 2.4 with load-balanced web servers, don't use the default caching option that stores the data in moodledata on a shared network drive. Use memcache instead.

Take 2 was a lot better. Moodle 2.4 was now only about 1.5 times slower. Still not good enough, but in the right ball park. This blog post is about what we did next, which was to use the tools Moodle provides to work out what was slow and fix it.

Moodle's profiling tool

When your software is too slow, you need measurements to tell you which are the slow bits. Tools that do that are called profilers. One of the better profiling tools for PHP is called XHProf. The good news is that it has already been integrated into Moodle, and there is documenation about getting it working. Basically, you just need to install a PHP extension and turn on some options under Admin -> Development -> Profiling.

Since we already had the necessary PHP extension on our severs, that was really easy. The option I chose was to profile a page when &PROFILEME was added to the end of the URL, but there are several ways to control it.

Profiling output

Once you have profiled a page, the results appear under Admin -> Development -> Profiling runs.

This just lists the runs you have done. You need to click through to see the details of one run. That looks like a big table of all the function that were called as part of rendering the page, how many times each one was called, and how much time each function was responsible for.

Inclusive time is the amount of time taken by that function, and all the other functions it called. Exclusive time is the time taken by that function itself. Some people, like sam, seem to like that tabular view. I am a more visual person, so I tend to click on the [View full callgraph] link. That produces a crazily big image, showing graphically which functions call which other functions, and how much time is spent in each one. Here is the image for the run we are looking at:

You can click for the full-sized image. The yellow and red highlighting is applied automatically to try to highlight places where a lot of time is being spent. Sometimes it is helpful. Sometimes not. The red box in the bottom right is where we do database queries. No suprise there. We know calling the database is one of the slowest things you can do in Moodle code. The other red box is fetching data from memcache, which also involves connecting to another server.

What you have to look for is somewhere on the diagram that makes you go "What! We are spending how much time doing that?! That's surely not necessary." In this case, my eye was drawn to the far right of the yellow chain. When viewing this small course, we are fetching the course format object 134 times, and doing that is accounting for about 9% of the page-load time. There is no way we need to do that.

Fixing the problem

Once you have identified what appears to be a gross inefficiency, then you have to fix it. Mostly that follows the normal Moodle bug-fixing mechanics, but it is worth saying a bit about the different approaches you could take to changing the code:

1. You might work out that what is being done is unnecessary. Then you can just remove it. For example MDL-39452 or MDL-39449. This is the best case. We have both improved performance and simplified the code.
2. The next option is to take an overview of the code, and re-organise it to be more sensible. For example, in the course format case, we should probably just get the course format object once, and then use it. However, that would be a big risky change, which I did not want to do at this time (just before the Moodle 2.5 release). This approach does, however, also have the potential to simplify the code while improving performance.
3. The next option is some other sort of refactoring. For example get_plugin_list was getting called a lot, and it in turn was calling the generic function clean_param to validate something. clean_param is quite fast, but not when you call it a thousand times. Therefore, it was worth extracting a simpler is_valid_plugin_name function. Doing that (MDL-39445) reduced the page load time by about 2%, but did make the code slighly more complex. Still, that is a worth-while trade off.
4. The last option is to add caching. If you are doing the same thing repeatedly, and it is slow, and you can't avoid doing it repeatedly, then remember the answer the first time you compute it, and reuse it later. This should be the option of last resort because caches definitely increase the code complexity, and if you forget to clear them when necessary you introduce bugs. However, as in the course formats example we are looking at they can make a big difference. This fix reduced page-load times by 8%.

So far, we have found nine speed-ups we can make to Moodle 2.4 in the core Moodle code, and about the same in OU plugins. That is probably a 10-20% speed-up on most pages. Some of those are new problems introduced in Moodle 2.4. Others have been there since Moodle 2.0. We could really benefit from more people looking at Moodle profiling output often, and that is why I wrote this article.

Interesting workshop about self-assessment tools

About 10 days ago, I took part in a very interesting workshop about the use of assessment tools to promote learning:

Self-assessment: strategies and software to stimulate learning

The day was organised by Sally Jordan from the OU, and Tony Gardner-Medwin from UCL, and supported by the HEA, so thanks to all of them for making it happen.

People talked about different assessment tools (not all Moodle), how they were getting students to use them, and in some cases what evidence there was for whether that was effective.

Parts of the event were recorded, and you can now access the recordings at http://stadium.open.ac.uk/stadia/preview.php?whichevent=1955&s=1. There is a total of 3.5 hours of video there, so you may not want to watch it all. My presentation is in Part 3, which also includes the final discussion, all in 30 minutes, and provides a reasonable summary of the day.

Despite having spent the whole day at the event, and discussed various aspects of what self-assessment is, I don't think we reached a single definition for what is self-assessment. Actually, I think it is clear that it is not one thing, but it is a useful way of looking at many different things, from the point of view of what is the most useful thing to help students learn.

One of the tools discussed during the day was PeerWise. If you have not come across that yet, then you should take a look, becuase it looks like a very interesting tool. There is a good introduction on Youtube:

.

What I want to build next

Earlier this summer I finally finished the new Moodle question engine, which was released as part of Moodle 2.1. As you might expect with such a large change, a number of minor bugs were not spotted until after the release, but I (and others) have fixed quite a lot of them, and we will continue to fix more. I want to say "thank you" to everyone who has taken the time to report the problems they encountered. Pleasingly, some people, including Henning Bostelmann, Tony Levi, Pierre Pichet, Jamie Pratt, Joseph Rézeau and Jean-Michel Vedrine have not only been sending in bug reports, but also submitting bug fixes. I would like to thank them in particular. I don't know whether this means that the new Moodle development processes are working well and encouraging more contributors, or that I released the new question engine full of trivial bugs.

At the moment, apart from fixing bugs, we are about two months away from the end of the OU's one-year project to move from Moodle 1.9 to 2.x and implement a lot of new features at the same time. In the eAssessment area, we had about 30 work-packages to do, of which finishing the question engine was by far the biggest, and we have about 6 left to go. Most of the remaining tasks are at least started, but finishing them is what I, and the developers on my team, will be doing in the near future.

I have, however, been thinking ahead a bit, and I have an idea for what I would like to build, should I be given the opportunity. Honesty compels me to say these are not my ideas. I stole them from other people, and there are proper acknowledgements at the end of this post. I wanted to post about this because: 1. in my experience, if you post about your half-baked ideas, people will be able to suggest ways to make them better; and 2. I am hoping that at least one course-team at the OU will see this and say "we would love to use this in our teaching" because that might persuade the powers that be to let me build this.

Rationale

The Moodle quiz is a highly structured, teacher-controlled tool for building activities where students attempt questions. What I want to create is a more open activity where students can take charge of their learning using a bank of questions to practice some skill where the computer can mark their efforts and give feedback. For the sake of argument, I have been calling this the "Question practice" activity module.

The entry page

When a student goes into a Question practice activity, they see a front screen that lists all the categories in the question bank for this activity.

Next to each category, there are statistics for how the student has performed on that category so far. For example, it might say "recently you scored 19/21 (90%); all time you scored 66/77 (86%).” The categories are nested, and there is a subtotal for each category.

At the bottom of the page is an Attempt some questions… button. This takes the student to the …

Start a session form

… where they set up what practice they would like to do. Students can select which categories they want to attempt questions from. They may also be able to choose how many questions they want. For example "Give me 10 questions", "As many as possible in 20 minutes", or "Keep going until I say stop". The teacher will probably be able to constrain the range of options available here.

Once they are satisfied, the they clicks the "Start session" button. This takes them to the …

Attempt page

… which shows the student the first question, chosen according to the criteria they set. There will probably be a display of running statistics "In this session you have attempted 0 questions so far". The question will contain the controls necessary for attempting the question. There will also probably be a "Please stop, I'm bored" button, so the student can leave at any time.

When they get back to the front page, the statistics will have been updated.

If the student crashes out of a session, then when they go back in, the front page will have a "Continue current session" button.

Overall activity log

One batch of attempting questions will be called a 'practice session'. The system will keep track of all the sessions that the student has done, and what they achieved during each session.

The front page will have a link to a page that lists all of the student's sessions, showing what they achieved in each. This provides more detail than is visible on the front page.

Possible extensions

That is the key idea. Here are some further things that could be added to the basic concept.

Milestones

The system could recognise targets, goal, or achievement (I'm not sure of the best name). That would be something like "Attempt more than 10 questions from the Hard category, and score more than 90%". If the student achieves that target at any time, they system would notice, and the achievement would be recorded on the front page and in the session log in an ego-boosting way (e.g. a medal icon).

The whole point of this activity is to be as student-driven as possible, so should students be able to define their own targets or goals? Should students be able to set goals for each other?

Locks / Conditional access

The activity could also have locks, so that the student cannot access the questions in the Multiplication category until after they have scored more than 80% in the Hard addition category. Of course, unlocking a new category could be an achievement. We appear to be flirting with the gamification buzz-word here, so I will stop.

Performance comparison

Should there by any way for students to compare their performance, or achievements, with their peers? We are definitely getting to features that should be left until version 2.0. Let's get a basic system working first, but make sure it is extensible.

How hard would this be to build

I think this would not require too much work because a lot of the necessary building blocks already exist in Moodle. The question bank already handles questions organised into categories, and we would just use that. Similarly, the attempt page and practice sessions are very easy to manage with the new question engine.

The real work is in two places. First, building the start attempt form, and then writing the code that randomly selects questions based on the options chosen. Second, deciding what statistics to compute, and then writing the code to compute them.

Of course, before we can start writing any code, there are still a lot of details of the design to decide. Also one most not forget things like backup and restore, creating the database, and all the usual Moodle plumbing.

Overall, I think it would take a few months work to get a really useful activity built.

Credit where credit is due

I said earlier that I got most of these ideas from other people. To start with, things like this have been mooted in the Moodle quiz forum over the years. The discussions there usually start from Computerised Adaptive Testing, whereas this idea is about student-driven use of questions. I think the latter is more interesting. (As a mathematician, I think CAT is an interesting concept. I just don't think it would make a useful Moodle activity.)

The real inspiration for this came at a meeting in London at the start of 2011. That meeting was at UCL with Tony Gardiner-Medwin who has already built a system something like this, but stand-alone, not in Moodle; and David Emmett from University of Queensland, Brisbane (who was giving a seminar). David had been hoping to get a grant to build something like this proposal (in Moodle) but that did not pan out. We did, however, have a very interesting discussion, and that is where I got the key idea that this sort of question practice was most interesting if you could give the student control of their own learning as much as possible.

We have also discussed ideas like this on-and-off for a long time at the OU. There has, however, been a lot of other things we needed to deal with first. We had to do a lot of work getting the quiz system working to our satisfaction (a strand of work that eventually lead to the new question engine). We had to sort out the reporting of grades, including working with Moodle HQ on the new gradebook in Moodle 1.9, and integrating Moodle with our student information system. We had to make a new question types that our users wanted. Only now can we start to think seriously about the last piece of the jigsaw: more activities that use all the question infrastructure we have built. I hope this post is a useful starting point for discussing what one of those activities might be.

The Good, the Bad and the Ugly

It was the best of times, it was the worst of times, ... It has certainly been a mixed week.

The good

... was that I helped three OU developers submit their first bug fix through the new Moodle development process: MDL-27631, MDL-28517 and MDL-28620. Hopefully those fixes all get through integration review next week.

The Bad

... was that the time had finally come to deal with a hot potato that we have been tossing around for some months; and, to mix metaphors, when the buck stopped, I was in the the wrong place at the wrong time.

As part of some new question types we are developing, we want students to be able to type responses that include superscripts and subscripts. For example 3×10⁸ ms^-1 or SO₄^2-. We have an old implementation of this, done six years ago for OpenMark (for example this or this), but that never worked in Safari, and is a bit dodgy generally. We want a new, reliable implementation that works in IE, Firefox, Chrome and Safari.

Plan A: back near the start of spring, I quickly knocked up a partial solution using the YUI 2 Rich Text Editor library. It mostly worked, but there were issues. It did not work consistently across browsers, and it lets you nest superscripts inside subscripts inside superscripts which just gets confusing, so we want to prevent that.

I had a sneaking suspicion how hard it would be to get from my quick partial solution to a robust implementation. Therefore I moved on to other things, and tried to unload this job onto three other people in turn. There were plenty of other more urgent tasks on our todo list.

Time passed, and many of the other things got done, so at the start of the week I realised that creating this input widget could not be put off any longer. I also felt it was unfair to expect other developers to deal with a crappy job that I was not prepared to do myself, so I decided to have another go.

The other thing that had changed is that while attempting to implement this, Colin and Wale had both eliminated some blind alleys, and suggested some promising ideas. Therefore, I was continuing from a far better place than where I left off. Even so, it was a long week.

The Ugly

Plan B: Although we had a partial implementation in YUI 2, I did not want to continue with that. Moodle is trying to move away from YUI 2 and to YUI 3 as soon as possible. So, my first attempt was to use the YUI 3(.3) Rich Text Editor. As the docs make clear. That is not finished yet. It is also not terribly well documented. With hindsight, I now realise that it provides only a very thin rapper around the native editing facilities provided by web browsers. Therefore it does not really help with browser inconsistencies.

Plan C: Since the Rich Text Editor is only beta, I decided to have a look at what was new in YUI 3.4, which is due for release soon. The answer is that they have made quite a lot of progress - in the sense that if you are trying to walk from London to Edinburgh, you have made quite at lot of progress by the time you reach Milton Keynes. Compounded with the fact that it is hard to find any documentation for pre-release version of YUI, this approach also failed.

At this point, I decided to do a bit of reading. I found two excellent articles from Opera that explained exactly how contentEditable works in web browsers. I also found a good cross-browser compatibility table. That made me realise that YUI was hardly doing anything to help with cross-browser differences.

Plan D: Now that I knew roughly what the browsers were doing, I briefly toyed with the idea of implementing the widget entirely myself in plain JavaScript. Once again, getting something basic working was not too hard, but I had not even started to tackle the cross-browser differences.

Then I realised that TinyMCE, which Moodle uses, tends to work really well across browsers. It is a bit slow to load, because it is a huge mass of code, but perhaps all that code is there for a reason. A quick play with superscript and subscript in the Moodle HTML editor in various browsers confirmed that TinyMCE must be working around most of the problems. So I dived into the TinyMCE code with the original idea of stealing the bits I needed to make Plan D work. It did not take much looking for me to develop a new-found respect for how hard TinyMCE is working to keep different web browsers in line. I did not want to have to replicate all that.

Plan E: So I finally concluded I should just use TinyMCE directly. That is using a very large sledge-hammer to crack a nut, but at least it should work. Indeed, it was mostly just a matter of setting the right configuration options. What made it particularly good is that there is an option to limit which tags can be nested inside other tags. That robustly prevents people from nesting superscript inside subscript, etc.

I was very nearly there, but there were two more requirements. We did not want pressing enter to insert a line-break, and because we were only dealing with a single line of input, we wanted to use the up and down arrow keys as shortcuts for superscript and subscript. The only way I could find to do that was to write a simple TinyMCE plugin. Fortunately, that is well documented.

The end

I got there eventually. The code needs to be cleaned up, tested some more, and integrated into the question types we are building, but I don't foresee any problems doing that.

I would like to thank the Moxie Code, who make TinyMCE, even though they have completely ignored the patch I sent them some time ago in relation to MDL-27890. I would also like to thank Olav Junker Kjær, who wrote the Oracle blog posts, which were the most useful thing I read. Also, the team behind Firebug. I can't imagine doing JavaScript development without that debugging tool. Finally Colin, Wale and Jamie, who I tried to dump this on, and who in return gave me helpful ideas.

Keeping the discipline of not changing Moodle core

We have said in the past that at the OU we made too many changes to core code in our Moodle 1.9 system, and that as we moved to Moodle 2, we would make far fewer. The release of Moodle 2.1 provides a good opportunity to stop and reflect on how we are doing.

Exactly how many core changes we had made in 1.9 seems to depend on who you ask. It was something of the order of one or two thousand depending on how you count. As a result, every time there is a new Moodle 1.9.x release, someone (Derek) has to do a couple of days painstaking merging to upgrade to the new version.

Moodle 2.1 was released on Friday. On Monday afternoon we decided to try upgrading our development branch to it. The merge (literally git merge MOODLE_21_STABLE) only took a few hours, and that was most mostly a matter of thinking before typing git checkout --theirs to resolve most of the conflicts in favour of the Moodle 2.1 code. Then we had to test test install, upgrade, and basic functionality before pushing the merge to our central repository.

But, how many OU-specific changes do we have in core code right now? Well, the answer appears to be eight. Let me explain that number.

To control the core code changes, we use a simple approval procedure. Each change must be proposed by one of the leading developers. They do this by opening a special sort of ticket in our issue tracking system. That serves as a permanent record of the change, and is also a place to log any discussion. The other leading developers then review the proposal. For the change to be approved, at least one other leading developer must endorse it with a +1 vote, and there must not be any -1 votes. Votes are normally accompanied by an explanation of why that developer is voting that way.

After a suitable time for votes, the issue is declared either accepted or rejected. OU-specific changes can be rejected for two reasons.We may decide that it is not acceptable to change core to implement this feature, so we drop the feature; or we think of some devious way to achieve the feature without changing core code.

If a change is approved, then the code is written. Well, in some cases the code will already have been written, because you can have a much more informed debate about whether a certain change is a good idea if you can see exactly what the proposed change is. Once the code is written and approved, it is committed to out git repository and the issue moves into state 'Code committed'. Finally, we may find a way to get rid of the OU-specific change in future. The most common way that happens is if we contribute the change upstream to moodle.org. For example the new Moodle question engine was an ou-specific change as long as we were using it in Moodle 2.0, but now we have upgraded to Moodle 2.1, it is standard code. Therefore, that issue has now changed status to 'No longer required'.

Overall, our we, have 22 ou-specific change issues in our bug tracker. The break-down is:

Rejected: 2
New (under discussion): 4
Approved (but not yet implemented): 1
Code committed: 8
No longer required: 7

Most of the 'Code committed' changes are pretty boring. For example three of them are bug-fixes to the questionnaire module that we have submitted upstream, but which have not been reviewed and accepted by the questionnaire maintainers yet. Therefore, those three will almost certainly end up as 'No longer required' in due course. Another example is that we want to customise the "Database connection failed / It is possible that the database is overloaded or otherwise not running properly" page that you get when Moodle fails to connect to the database. If Moodle can't connect to the database, then it cannot load the configuration, and so cannot determine which theme to use to display the error. Therefore, the only way customise that page is to edit lib/setup.php.

The one 'serious' ou-specific change we have is some hacking around in course/lib.php to support one of our custom modules called 'subpage' (not released yet, but we hope to share it eventually). Given more time, we might be able to find a more elegant way to handle these changes, but we don't have that sort of time at the moment.

While we have controlled the core code changes, we do have written a lot of custom plugins. Those range from big things like forumng and ouwiki, to small things like a local plugin that just implements a single web-service function. I'm afraid I don't have a complete list, but we must have more than 50 plugins by now. As far as I am aware, the upgrade from 2.0 to 2.1 did not break any of them.

Etiquette for questions

I have been working hard at converting the new Moodle question engine to work in Moodle 2.0, aiming at a deadline this Friday (25th February). On Friday we should have a first OU version of Moodle 2.0 with all the key features so that we can start testing, even though students won't get onto the new system before July. I have have basically finished the question engine, give or take a few features that are not needed for testing, and this week I am just doing some final tidying up of the code.

Hopefully, next week I can start the process of getting it reviewed for inclusion in Moodle 2.1. As I say, there are some gaps in the functionality that will need to be filled in before it can actually be committed, but there is a lot of code to be reviewed (lucky Eloy!) and so I hope we can kick off the process.

So, my excuse for not blogging about the new question engine recently is that I have been too busy working on it to write about it. In the last few days, however, I encountered a couple of nice ideas that would be easy to implement using the flexibility the new question engine gives, and I want to describe them. First, I need to remind you of one key point about the new system:

Question behaviours

As I explained before a key idea in the new question engine is that of question behaviours. Whereas a question type lets you have either a multiple-choice, a drag and drop, or a short-answer question, a behaviour controls how the student interact with the questions, of whatever type. For example, the student may have to type in answers to each question in the quiz, then submit everything, and only then are the questions marked. This is known as the "Deferred feedback" behaviour. Alternatively, the student may answer one question, have their answer marked immediately. If they are wrong, they get a hint and can then immediately have another go. If they get it right on the second or third try, they get fewer marks. This is called the "Interactive with multiple tries" behaviour.

When I was first working on this, I did wonder whether it was perhaps over-kill to make behaviours fully-fledged Moodle plugins. It seemed to me that I had already implemented all the types of behaviour anyone was likely to want. It turns out I was wrong. Here are three ideas for new behaviours have I have come across since had that naive thought.

Explain your thinking behaviour

The concept here is that, in addition to presenting the question to the student for them to answer in the usual way, you also give them a text-area with the prompt "Explain your answer". When the submit the question is graded as usual. Moodle does not do anything with the explanation, other than to store it, and re-display it later when the student or their teacher reviews their attempt. The point is that the student should reflect upon and articulate their thought processes, and the teacher can then see what they wrote, which might be useful for diagnosing what problems the students are having.

I'm not sure that this would really work. Would the students really bother to write thoughtful comments if there were no marks to be had? However, this would be relatively easy to implement, so we should build it and see what happens in practice. The teacher could always manually adjust the marks based on the quality of the reflection, if that was necessary to incentivise students.

I'm afraid I cannot remember who suggested this idea. It was a post in the Moodle quiz forum some time ago, just after I had implemented the behaviour concept and was thinking that my initial list of behaviours was all anyone could possibly want.

gnikram desab-ytniatreC

This idea I only came across yesterday evening, in a blog post from people in the OU's technology faculty. It is a slightly strange twist on certainty-based marking.

With classic CBM, the students answers the questions, and also says how certain they are they got it right (for example, on a three-point scale). The student will only get full marks if they get the question right, and are very certain that they were right. If, however, they express high certainty and are wrong, they are penalised heavily with a big negative mark. To maximise their score, the student must accurately gauge their level of knowledge. This hopefully promotes reflection, and self awareness by the student of their level of knowledge.

The idea from the OU technology faculty is to do this backwards, for multiple choice questions. Rather than getting the student to answer the question and then select a certainty, you first show them just the question stem without the choices, and get them to express a certainty. Only then do you show them the choices and let them chose what they think is the right answer.

Again, I am not sure if this would work, but it is sufficiently easy to do by creating a new behaviour plug-ing (and a some change to the multiple-choice question type so that you can output just the question, without the choices) that it has to be worth a try.

Free text responses with a chance to act on feedback

This last idea I only heard about this morning. There was a session of the OU's "eLearning community" all about eAssessment, which naturally I attended. This is a monthly gathering with a number of different presentations on some eLearning topic. The first three talks were about specific courses that have recently adopted eAssessment, and how students had engaged with that, what effect the effect had been on retention and pass rates, and so on. That was interesting, but not what I want to talk about here. The final talk was by Denise Whitelock from the OU's Institute of Educational Technology who has just completed a review of recent research into technology-enhance assessment for HEA that should be published soon. Here, I just want to pick up on one specific idea from her talk.

I'm afraid that again, I don't recall who deserves credit for this idea. (Once Denise's review is published, I will have a proper reference, but I did not take notes this morning.) It was another UK university that had done this. It was in the context of language teaching. The student had to type a sentence in answer to the question, then the computer graded that attempt and gave some feedback. Then, the student was immediately allowed to revise their sentence in light of the feedback, and get it re-marked. The final grade for the question is then a weighed sum of the first mark and the second mark. You need to get the weights right. The weight for the first try has to be big enough that the student tries hard to get the question right on their own before seeing the hints, and the weight for the second try, though smaller, also has to be big enough so that the student bothers to revise their response.

Now, the OU is currently creating a Moodle question type that can automatically grade sentence length answers using an algorithm that my colleague Phil Butcher implemented the first version of in 1978! (When I say we are creating this, what I actually mean is that we have contracted Jamie Pratt a free-lance Moodle developer to implement it to our specification.) Anyway, once you have that, the idea of allowing two tries, with feedback after the first try, and a final grade that is a weighted sum of the marks for the two tries, is just another behaviour.

So, my initial thought that people would not have many ideas for interesting new behaviours seems to have been wrong. The flexibility I built into the system is worth having.