It was very interesting to see the results of the Tweetorial on Tweet Archivist and Keyhole, and my initial reaction was a sense of satisfaction at being mentioned and ‘visible’.
PJ and Nick were the top tweeters by far but the proportion of tweets didn’t unbalanced on Twitter. In fact it felt like everyone was present and participating, and I felt my own contributions were fairly represented.
The top words were ‘algorithm(s)’, ‘learning’, ‘digital’ and ‘data’, which makes perfect sense and is representative of the discussion and discussion questions.
It’s interesting that ‘trouble’ was so high on the list coming after ‘data’ and above ‘education’ but on further investigation I realise that this was because of the title Sian’s talk and the fact that her tweet was retweeted so many times.
There were only 34 tweets on the 12th of March but 118 on the 13th as people started replying to tweets and this accurately conveys how the Tweetorial unfolded.
Sian had the highest number of mentions and came out in the middle of the image, which highlighted how she appeared to be at the centre of the discussion and was the point through which many comments were made.
Interestingly only 36.2% of posts were original posts, 39.3% were replies and 24.5% retweets. I do wonder how typical this is of Twitter.
From a geographic perspective Asia didn’t even figure, despite several of us being based here. America (14%) and Canada (1%) look over represented because of their size and the U.K. (48%) looks under-represented. The location description states that the map shows where in the world the posts originate from but is this in relation to tweeters or number of tweets? – analytics mean nothing if they are not clear and specific.
90% of the posts originated from males. Only 10% from females. The Tweetorial certainly didn’t feel male dominated and I wonder if this is in fact accurate – although we are only 3 females for 8 males on the course and the males were the top tweeters…
86.8% of tweets came from desktop, which is representative for me. Even if I saw the tweet on my phone or iPad, I waited and responded from my computer (although I’m not exactly sure why).
Overall a reach of 21,789 was impressive for a 48 hour Tweetorial based on 26 users.
Although I initially oohed and ahhed at I the data and the way it was presented after more careful analysis I’m not sure that I was any the wiser for it – I was being informed of everything and nothing at the same time. My overriding sensation was that the LA seem to level and flatten, and remove the ‘colour’ of the Tweetorial. There is definitely a loss of perspective as small details are brought to the foreground and overshadow more salient data.
LA is purely quantitative and from an educational point of view teachers are only aware of who is participating rather who is really engaged and producing meaningful and relevant tweets/posts, responding to what is being said and developing the discussion. For LA to be truly useful it would need to also evaluate quality not just quantity.
Knox, J. (2014). Abstracting Learning Analytics. Code Acts in Education ESRC seminar series blog.