makes use of cc licensed ( BY NC ) flickr photo by bitchcakesny: http://flickr.com/photos/bitchcakes/5187928447/
Now that the 2012 Summer of ds106, Camp Version is over, I spent some time trying to pull some numbers from the machine, given how crazy folks are about analytics and massive.
First of all, I want to thank all 172,000 people who registered for the course (just kidding, we want to make the BIG MOOCS nervous).
This is the number of blog posts by the 11 registered UMW students over the 10 week course. I first ran a database query to find the number of posts in ds106, where all the aggregation happens, that were tagged for our summer section (umwsum12)
SELECT count(p.ID) FROM wp_posts p INNER JOIN wp_term_relationships tr ON (p.ID = tr.object_id) INNER JOIN wp_term_taxonomy tt ON (tr.term_taxonomy_id = tt.term_taxonomy_id) INNER JOIN wp_terms t ON (tt.term_id = t.term_id) WHERE tt.taxonomy = 'post_tag' AND t.slug ='umwsum12'
And I got 435. We have been having some issues with Feedwordpress missing a few blog posts (the developer is working on a fix for us), so in the interest of pure analytics, I went to each students blogs and counted their posts. This os not quite as tedious, if you page through their archives Until you get a not found response and go back one), and find that their last paged url is 5 and there are 3 blog posts on that last page, that means a total of 43 (10 per page plus the last 3).
The activity per student ranged from 22 to 55 blog posts for the class with an average of 41.2 (that would be 4 per week). The distribution actually might be closely proportional to the grade distribution:
Some consideration should go into that some students did a lot individual posts per daily create where others did weekly summaries, and if I really was going to get into it, I should analyze post lengths and amount fo media embedded.
Total number of blog posts from open participants, here by looking at posts aggregated to the ds106 site tagged ‘openonline’ and after the start date of the class:
SELECT count(p.ID) FROM wp_posts p INNER JOIN wp_term_relationships tr ON (p.ID = tr.object_id) INNER JOIN wp_term_taxonomy tt ON (tr.term_taxonomy_id = tt.term_taxonomy_id) INNER JOIN wp_terms t ON (tt.term_id = t.term_id) WHERE tt.taxonomy = 'post_tag' AND t.slug ='openonline' AND p.post_date > '2012-05-14 00:00:00'
This again is an approximation given some trouble we have had from some blogs not being picked up by Feedwordpress, although the discrepancy fro the UMW student blogs was only 4%.
But this would indicate the activity of the non-registered students exceeded that of the registered ones, but likely spread over more individuals.
Combined, we had well over 1000 blog posts come into the site over the 10 week course.
Number of unique tags/categories used across all blog posts (we convert all tags on incoming feeds to categories. I have no idea what this means. But among the tags for assignments, we find from random sampling:
- Under pressure
- just plain blogging
- Doctor Who
- lifelong learning
- dew player
- Zack Parsons
- ray harryhausen
- the magic of zazzy
Draw your own conclusions.
Number of blogs/users syndicate and published to the Magicmacguffin site. This is based on the number of users created on the site, which is done by Feedwordpress as new feeds are added. Our signup form and ds106 registations were the source for us assigning 84 users to our groups (or bunkhouses). So this means that 16 people supplied information but never posted.
Does this mean our “dropout” rate (a term I think has no meaning in open courses) was 19% (or 81% of the people who signed up participated by blogging at least once).
But you know how I feel about the use of the word “dropouts” (“hence” “scarequotes”)
The use of the term”dropout” in an open course is a signal of a major lack of understanding of open learning.
— Alan Levine (@cogdog) July 25, 2012
This is the total number of assignment blog posts aggregated to the site (based on the number of occurrences of the post tags, found by searching for the tags in the admin dashboard). I know this is an under estimate because end of the semester we found one student had mistagged all of her posts (she forgot the comma between the general and specific tag).
By looking at the distribution, we might infer that the visual and design assignments are more popular, but it should be noted that the audio and video assignments carry more stars (difficulty rating) so students generally do fewer of the number of assignments because they are more challenging)
or by the numbers
- VisualAssignments 116
- DesignAssignments 89
- AudioAssignments 52
- VideoAssignments 40
- RemixAssignments 11
- MashupAssignments 18
This is six plus hours of the 18 videos created by Martha and I just for the weekly announcements. Tn the first few weeks, we filmed them in DTLT and used the green screen, editing and uploading to YouTube. Because I was traveling for the rest of the course, we went the easy route and did our videos via Google hangouts live, and using the awesome feature that archives them directly to YouTube… This means they may not be edited cleanly but there was no post processing.
And we did all of these without any scripts, after discussing what we would do before hand for maybe 10 minutes, they were totally improvised (the last video had some serious editing, but the audio was done improv).
Without a doubt, working with Martha on these was one of the highlights of the summer experience.
One of my to do lists is to add some capability of the Daily Create site to track the number of submissions each challenge (since these are drawn from third party sources, we would need to poll via their APIs).
Also, not figured out yet is some way to keep track of the comment activity across the blogs (#pipedream)
Also what I’d like to get at are some ways to visualize the activity in this busy network on the front of the site, as the chronological view moves so fast and is not very useful.
So there you go, the numbers. Coming up sometime soon, my own reflections on teaching.
Camp Magic Macguffin By the Numbers by CogDogBlog, unless otherwise expressly stated, is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.