Archive for the 'Analytics' Category

Learning and Knowledge Analytics (LAK11) Week 1

So we are now into week 2 of the open course in Learning and Knowledge Analytics LAK11. Whilst I’m already doing better at this course than PLESK10 I would still only class my involvement as periphery participation so I’ll be no doubt be revisiting the LAK11 syllabus again at a later date. A couple of things I’ve picked up from week 1 you might be interested in:

The only paper I had a chance to properly read was Elias, T. (2011) Learning Analytics: Definitions, Processes, Potential. It was more luck than anything else that I started here but I was very glad of the fortune [it was only later that I read Dave Cormier’s MOOC newbie voice - a slackers entrance into lak11 post which reassured me that although I wasn’t doing much at least it was the right thing].

Things I took away from the paper were:

  • Some examples of learning analytics systems already being used.
    • Purdue’s Signal’s block for Blackboard“To identify students at risk academically, Signals combines predictive modeling with data-mining from Blackboard Vista. Each student is assigned a "risk group" determined by a predictive student success algorithm. One of three stoplight ratings, which correspond to the risk group, can be released on students’ Blackboard homepage.”  [this reminded me of University of Strathclyde’s homegrown STAMS VLE which appears to have disappeared when the University moved to Moodle – bit of a shame as it was developed by staff in Statistics and Modelling Science so imagine behind the scenes it had a dusting of analytics – that’s progress for you]

    • University of California Santa Barbara’s Moodog Moodle module - “In addition to collecting and presenting student activity data, we can proactively provide feedback to students or the instructor. Moodog tracks the Moodle logs, and when certain conditions are met, Moodog automatically sends an email to students to remind them to download or view a resource.”  Zhang (2007) (p. 4417) [I was a little disappointed to only find references to this is academic papers]
  • Something on collective intelligence - Woolley et al. (2010) identified the existence of collective intelligence which “is not strongly correlated with the average or maximum individual intelligence of group members but is correlated with the average social sensitivity of group members, the equality in distribution of conversational turn-taking, and the proportion of females in the group" (p 686)
  • Some terminology/theories for recommendation systems – “recommendation methods based on different theories such as collaborative filtering algorithm, bayesian network, association rule mining, clustering, hurting graph, knowledge-based recommendation, etc. and the use of collaborative filtering algorithms (Cho, 2009)” [at this point in the paper I thought about Tony Hirst’s Identifying Periodic Google Trends posts, mainly in underscoring the shear scale of the field of learning analytics]

Overall the paper was very useful in highlighting how much I didn’t know, but was an indication of the things I might need to know [whilst it might not sound like it this is a positive outcome to let me self-regulate my learning].

Some things on participating on the course in general

Google Reader 'magic'There were other things I did during week one including playing the the recommendation search engine hunch. This experience was juxtaposed to the course moodle site, which was blindly sending me hundreds of emails from the course discussion forums. In the end I decided to unsubscribe to the email notifications and pull the forum into Google Reader via RSS. My hope was Google Reader would ‘sort by magic’ to pull interesting things to the top, but the algorithm is struggling to do anything other than chronologically order the feed [my guess is Google don’t have another data for my personal or group preferences – ho hum ;)]

Where are you coming from: Search referrer and contextual related post/information

Just before Christmas Brian Kelly wrote a post on Trends For University Web Site Search Engines, which gives an overview of which search engines [Edit: Russell Group*] universities are using on their websites (75% using Google products) . This combined with my interest in Learning Analytics (plug: it’s week one of the open course LAK11 Learning and Knowledge Analytics), got me wondering how many institutions uses information about their visitors to customise content. There are a number of ways you could potentially do this from using media campaigns to using one of the Facebook Social plugins.

*Brian Kelly has kindly pointed out this was a survey of Russell Group institutions and all all universities as originally implied

The question of what is already being used is probably best answered by someone else, like Brian, instead I’m going to highlight one other simple way that you might customise the visitors experience (and at the same time improve how information on this blog is presented), by using Search Referrer information.

For the majority of users when they navigate around the web they leave a ‘referrer’ trail. When they land on a page that site can usually see where the person came from (if you read the HTTP Referrer entry on wikipedia you’ll see why this information isn’t always available). Monitoring tools like Google Analytics can track referrer information so that you can track where your traffic is coming from. I use Google Analytics to monitor traffic to this site and whilst I don’t use this data extensively (although I did modify the Google Analyticator WordPress plugin to display top posts based on Analytics data), it is useful information to check the general health of my blog.

When monitoring referrer information it is possible to record the whole web address of the page with the click through link including the query string (junk at the end). For example, if you were to open this page http://www.google.co.uk/search?&q=jisc+rsc+mashe and click on the link for this blog, if your browser is passing referrer information I can see you got here from a Google Search for ‘jisc rsc mashe’.

In fact I know from my Google Analytics data that in 2010 over a third (37%) of my visitors arrived via a search engine, almost all (97%) using Google and as the table shows I even know the main search keywords used.

Table of top search keywords used in 2010

So if I know over a third of people end here having searched for something, wouldn’t it be good if I could highlight more of my content based on their search? Hopefully our answer is yes otherwise I’ve wasted a hell of a lot of my own time chasing my tail on this.

Rather than completely reinventing I had a quick look at some existing WordPress plugin’s to see if anything would do this. The one that came closest was WP Greet Box which as well as providing a custom greeting message based on where your visitor is coming from if the user comes from a search engine it uses that search query to optionally display some related posts.

The problem I had with this solution is, as well as thinking the greeting was a little tacky, it only suggests related posts. As this blog has evolved I have more bespoke pages and tools which is why I use a Google Custom Search Engine (CSE) (combined with some of my own ‘instant’ magic to let people search all of the material in the MASHe directory.

Finding nothing else and as I already use the Contextual Related Posts plugin I thought it would be fun to modify this so that if someone lands on one of my posts from a search engine, their search query is used to pull related material using my sites Google CSE.

Now using my modified contextual-related-posts.php if you go to this TwEVS post the related post information at the end remains the same containing links for:

You also might like :

But if you end up at the same page by for example clicking a link to it from this Google Search you get the following instead:

You also might like (based on your search for ‘twevs’):

So what do you think?

Personally I’m not entirely convinced that this will have any impact on driving traffic internally within this site because my volume of visits is relatively low and the placement at the end of the post isn’t optimum for leveraging extra content. The other factor is normally in posts if I’ve written or have a related tool that might be of interest to the reader I include a link in the body of the post. But if nothing else maybe you’ve learned a bit about referrer information and you’ll come up with a better idea than me.

[A couple of other ‘techy’ things I learned along the way worth sharing:

About

This blog is authored by Martin Hawksey e-Learning Advisor (Higher Education) at the JISC RSC Scotland N&E.

JISC RSC Scotland North & East logo

If you would like to subscribe to my monthly digest please enter your email address in the box below (other ways are available to subscribe from the button below):

Subscribe to MASHe to monthly email updates

Loading...Loading...


The MASHezine (tabloid)

It's back! A tabloid edition of the latest posts in PDF format (complete with QR Codes). Click here to view the MASHezine

Preview powered by:
Bluga.net Webthumb

The MASHezine (eBook)

MASHe is also available in ebook and can be downloaded in the following formats:

Visit feedbooks.com to manage your subscription

Archives

Opinions expressed in this blog are not necessarily those of the JISC RSC Scotland North & East.

JISC Advance Logo

JISC Advance is a new organisation that brings together the collective expertise of established JISC services:

For further information visit www.jiscadvance.ac.uk

Creative Commons Licence
Unless otherwise stated this work is licensed under a Creative Commons Attribution-ShareAlike 2.5 UK: Scotland License