This post is mainly aimed at people who run a lot of Moodle sites (e.g. Moodle partners). Would it be possible for you to run the query
SELECT qtype, COUNT(1)
FROM mdl_question
WHERE hidden = 0 AND parent = 0
GROUP BY qtype
ORDER BY qtype
on as many of the sites you host as possible, and report the totals?
(This is a bit of a pain to do, because you have to connect to the DB for each separate Moodle, and run the query. Obviously, that is scriptable, but someone would have to write the script.)
I suppose I ought to start by giving the numbers from the OU:
| qtype | count |
|---|---|
| calculated | 1 |
| ddwtos | 2077 |
| description | 1227 |
| essay | 1329 |
| match | 4090 |
| multichoice | 35779 |
| numerical | 9909 |
| opaque | 1783 |
| oumultiresponse | 5012 |
| shortanswer | 11680 |
| truefalse | 4161 |
3 instances, first one is m2, others are 1.9
| qtype | count |
| calculated | 5377 |
| calculatedmulti | 13 |
| calculatedsimple | 43 |
| description | 42 |
| essay | 11714 |
| match | 3217 |
| multianswer | 7886 |
| multichoice | 183721 |
| numerical | 46 |
| randomsamatch | 3 |
| shortanswer | 27240 |
|
truefalse |
17920
|
| calculated | 12545 |
| description | 322 |
| essay | 1741 |
| match | 1322 |
| multianswer | 1490 |
| multichoice | 29907 |
| numerical | 524 |
| shortanswer | 4297 |
| truefalse | 2973 |
| calculated | 5984 |
| description | 280 |
| essay | 14941 |
| match | 10885 |
| multianswer | 7868 |
| multichoice | 262234 |
| numerical | 1428 |
| randomsamatch | 2 |
| shortanswer | 28457 |
| truefalse | 10977 |
Do you want totals or averages for all the sites?
Attached is a CSV file containing the number instances of each question type for 1494 Moodle sites. Each row is a Moodle site. Each column is a question type.
I notice you have three sites with a significant number of Random short-answer matching questions.
I had a bit of an ulterior motive for calling this survey. I was hoping to demonstrate that no-one used that qutstion type, and so it would be safe to move it out of the standard release in Moodle 2.1, and put it in contrib instead.
Would it be possible for you to work out which three of your sites are using that question type; have a bit of a look at what they are doing with it; and perhaps ask them how annoyed they would be if it was not supported in Moodle 2.1? Thanks.
If anyone else has strong opinions about removing the randomsamatch qtype from Moodle 2.1, please speak now.
Making this question type work at all in 2.1 will be really hard (actually, I strongly suspect it does not work at the moment in 2.0. If the randomly selected short-answer questions have images in the question text, then I expect that students will see broken image links. However no-one has reported this bug, which is further evidence that not many people use this qtype).
I think the right way to implement functionality like this is through a separate plugin that can randomly generate a category of matching questions from other questions in the question bank. Like Joseph Rézeau's Export Glossary to Quiz plugin.
Those three sites are actually all the same client (two of them are just test sites). We actually have a staff member who used to work for that client. I'll see if she can chime in on this conversation.
Thanks for the update.
My co-worker has contacted the client to which I was referring, but unfortunately, they have been unresponsive.
Well, thanks for trying.
Can I say a big thank-you to everyone who has replied to me, either here, or through other channels. I really was not expecting, when I put out this call, to get information about more than 50 million questions from 2500 Moodle sites within two days.
I have aggregated the data I have so far on https://spreadsheets.google.com/spreadsheet/pub?hl=en_GB&hl=en_GB&key=0AvW3H1TnOxK-dEROU2dVQ2lkWDU3SlEzaXJqU2lSRUE&single=true&gid=0&output=html, and will continue to add any further data I get.
From University of Strathclyde:
+---------------+----------+ | qtype | COUNT(1) | +---------------+----------+ | calculated | 440 | | description | 16 | | essay | 3569 | | match | 3018 | | multianswer | 6125 | | multichoice | 50076 | | numerical | 8 | | randomsamatch | 1 | | shortanswer | 13942 | | truefalse | 1773 | +---------------+----------+
(I just spoke to the creator of our only "randomsamatch" question - it's only a test question.)
Hi Tim,
The results are really fascinating: I hadn't realised so many question types were possible. I really must get over my reluctance to using Quiz which I tend to see as a tool for testing rather than one for learning. I wonder though if the fact that most teachers (87%) use either multichoice or truefalse means that it is mainly being used for testing.
Cheers,
Glenys
I don't think that is the right conclusion to draw.
I think the key driver is what is going on is ease of use - and I am thinking more at the pedagogical end of things, rather than which button you have to press. When you create a multiple choice question, you can be sure exactly now Moodle will grade every possible student response. With short-answer, you have to start worrying about students misspelling words, and how you will grade that.
Of course, if that was the only consideration, then everyone would use true/false, and we don't see that. So, really it is about pedagogic bang for buck. MC clearly hits a sweet spot of letting teachers ask all sorts of questions, while still being conceptually very simple.
Another thing driving MC use (I guess, and particularly in the US) is the number of standardised tests that use MC - and teachers have to prepare students for those tests.
I am surprised how many essay questions there are. I suppose that once you have decided to bit the bullet and grade manually, then again essay is conceptually very simple and flexible.
In terms of whether all this is formative or summative. What I see at the OU is much more diagnostic use of quizzes - or very low stakes assessed use, awarding just enough marks to encourage students to do the activity, rather than have a major effect on their total course grade.
The point here is that if you are setting up a test for high-stakes assessment, you have to be very careful and check everything several times. On the other hand, if you are just making a self-assessment test for your class, you just need to make something good enough to be useful. On the other hand, if you are doing serious assessment for learning, you need to start writing lots of good feedback, which is more work again. Perhaps I should admin I really don't know what people are actually using the quiz for. I would be interesting to find out.
Appreciate you've got a lot of answers already but here's ours (South Devon College):
description 8
essay 133
match 187
multianswer 38
multichoice 12956
numerical 129
randomsamatch 2
shortanswer 364
truefalse 407
We've just finished bringing in our courses from WebCT, and are getting to ready to roll it out to production in September. I don't know if this data will skew or help what you're looking for, but right now we're sitting with these numbers across 38 courses. I'm pretty sure the numbers will change as our teachers begin to use Moodle over the next few months.
QTYPE COUNT
calculated 474
match 779
multianswer 12335
multichoice 18619
shortanswer 20072
Cheers!
| qtype | count |
|---|---|
| essay | 31 |
| match | 890 |
| multianswer | 27 |
| multichoice | 42066 |
| shortanswer | 2 |
| truefalse | 12240 |
Using Moodle 1.9 in the corporate environment.
description--------------43 essay---------------------182 match---------------------113 multichoice-------------1026 shortanswer------------96 truefalse-----------------884Using Moodle 1.9.6
from the University of Canberra (v1.9)
qtype count %
calculated 5984 1.7%
description 289 0.1%
essay 14986 4.4%
match 10888 3.2%
multianswer 7868 2.3%
multichoice 262482 76.4%
numerical 1437 0.4%
randomsamatch 2 0.0%
shortanswer 28466 8.3%
truefalse 10988 3.2%