Tag Archives: learning analytics

Learning from the Tweetorial

I approached the Tweet Tutorial as an experiment in self-abandonment with some sense of a veneer of safety and protection within the realm of the #EDC circle. This was actually self-delusion, as there was no such protection, and all the tweets were out there on the net and are now forever part of my ‘digital traces.’ It was not my intention nor aspiration to achieve any level of notoriety, and if anything, I am embarrassed to be disclosed as a top poster. A first observation with this exercise is that it confirms that ‘quantity is not always quality.’

My recent personal rise of ‘influence’ is likely to be fleeting, analogous to Andy Warhol’s maxim about 15 minutes of fame. A maxim that I carry forward from my professional life is “don’t confuse enthusiasm with capability.” I believe that the Block 3 Summary provides an accurate representation of my ‘engagement’ during the Tweet Tutorial, but ‘engagement’ should not be confused with learning. As Siemens (p. 1387) alludes to, we should be circumspect about the ‘scope of data capture’ and bear in mind that the ‘quality’ of data needs to be considered.

I think that some of my ‘better’ Tweets were actually ‘snippets’ (Dr. Baynes’ Inaugural talk, 35:10-25) from our academic readings. I used Twitter as a learning device to capture key points from the readings which I felt summarized interesting concepts. I did, I hope, provide attribution to all the authors. But, now, was that educative, instructive and ethical, or just clever, unoriginal and self-serving? I did not tweet them to increase my numbers, more hoping to stimulate dialogue, of which there was some occasional Twitterverse interest (e.g., ‘engagements,’ ‘impressions’ and ‘mentions’).

PJ Twitter Summary

Twitter Tutorial

An extended time span of #EDC highlighting the ‘spike’ of activity during the Tweet Tutorial.

Data visualizations can be deceptive because they are ‘reifications’ of assumed objective realities, when they are really ‘chimeras’ of separate realities; for example, online students and teachers engaging asynchronously across time and space. It is more illuminating to look at trends over time. Also, the challenge with online learning with teachers and students that are physically separated by time and space, is how are assessments conducted to measure ‘authentic learning?’ Is learning measured by the teacher, the student, learning analytics, and some combination?

For example, regrettably, I failed to complete the Coursera ‘Scandinavian Film and TV Culture’ MOOC which was the focus on my Block 2 micro-ethnography because I did not not take time to complete the final 800-word written assignment on the popularity of contemporary Danish TV programmes and four peer assessments. I completed all the quizzes and learned a lot from the micro-ethnography, but that work was not visible to the course organizers. By objective measures of performance, I ‘failed’ the course. Ironically, I remained in the top 10 discussion forum posters despite the fact that I stopped posting comments halfway through the course and only posted 13 discussion forum comments.

forum discussion

In sum, by ‘exposing’ myself during the Tweetorial, I learned more about my own ‘comfort levels’ with this social networking service. There is an instant gratification appeal, but for educative purposes, I think we need to learn to be more reflective; again, not to gain ‘influence,’ but to understand how to contribute in meaningful ways to elevating the level of discourse for education and learning.

#EDC Top Words

Tweet Archive on #mscedc 2.24.15 – 3.16.15


Week Nine Summary: Algorithmic and Learning Analytics Sense-making

Lifestreaming this week has revolved around a deeper interrogation of ‘algorithmic culture’ and an introduction to learning analytics. The goal of ‘sense-making’ in these areas of exploration has proven elusive, but the week’s activities have been instructive and illuminating, and opened up a variety of paths for future research.

Ben Williamson’s video presentation offered many valuable insights as to what he envisions as ‘imagining’ of the digital university and all the attendant issues related to the influence of algorithms and learning analytics. The vision is of a higher educational institution that is ‘sentient,’ ‘data-driven,’ and mediated by digital technologies. He advocated for the need to develop a sense for the emergence of socio-technically-oriented “fabricated spaces.” His explanations of algorithms, though somewhat pedantic in the way he read directly from slides, provided new perspectives on ‘socio-algorithmic relationality’ (slides 4-6; citing David Beer, 2013) and the influence of “algorithmists.” I thought his investigations of the ‘epistomology’ of big data (slides 10-15) was most penetrating in describing how algorithms are increasingly becoming entwined with academic knowledge; what he refers to as the “algorithming of the academy.” Honestly, I am now more profoundly sympathetic towards the increasing demands on university professors, with their evolution from their traditional role as ‘performers’ in the lecture hall to the requirement to maintain a presence online, on social media, as well as in the lecture hall/classroom environments.



I believe that the ‘digital university’ is becoming a reality. For example, in the form of the Minerva School experiment in San Francisco, the Singularity University, and probably several other initiatives at higher educational institutions world-wide. The challenge will be adjudicating the potential of the learning analytics and algorithmically-driven technologies with the concerns for privacy, data access and transferrability, and monetarization of data in these exciting new learning environments.

Ben Williamson tweeted recommendation of a recent article by Frank Pasquale, “The Algorithmic Self,” which I strongly recommend. It convincingly articulates and place into context all the current themes surrounding algorithms. A poignant observation is that: “We need some common, clear awareness of whom the algorithms behind the screen truly serve before we accept their pervasive presence in our lives.”

PJ Klout My Klout influence rating, based 100% on my Twitter activity, has risen from a low of 18.91 to 34.90 in the past 90 days. So what? How do I ‘modulate’ my online selfhood? Is this my “data self”?

“…there is a delicate balance between appropriating new technologies and being appropriated by them.” (Harmut Rosa, cited by Frank Pasquale)

Dr. Jeremy Knox asked me to reflect on last week’s investigations of algorithms and my own personal Quill Connect report. My Tweeter activity swelled this week as we engaged in an intensive Tweet tutorial. I assumed a very, uncharacteristic, aggressive posture towards this activity. My intent was not to boost my ‘influence,’ but to experiment liberally with this media to determine the ‘reach’ of this modality. I repeated the Quill Connect Report this week, which revealed that I posted 50 tweets this past week, 31 tweets more than usual, but somehow my average tweets per week remain at the Twitter user average of 2 per week. So now I am more suspicious, cynical, skeptical of what is actually going on with the underlying algorithm of Quill Connect; as I am generally now of all algorithms.

Dr. Konx prodded me to consider: “What does that (‘average twitter user’) really mean for a service like Twitter? I think it means more ‘committed’ subscribers; that is, a consumption-driven imperative, not necessarily a thoughtful, or educative one). Dr. Knox noted that I focused last week on ‘statistical measures,’ and I confess that I probably feel into the Big Data ‘big fallacy’ trap elucidated by Ben Williamson (with credit to Rob Kitchin) that ‘de-contextualized statistical data analysis can be reductive, functionalist & unhelpful as it lacks embedding in wider debates, social theory & contextual knowledge.” Apparently, the “average Twitter” user is a middle-aged, American female with an IPod.

Frank Pasquale suggests: “the first step of protecting the self in age of algorithmic manipulation is to recognize such manipulation as a problem.” One needs “source of value,” “sources of the self,” and “anchors of integrity,” pertinent to each individual, to protect oneself from potential domination by powerful technologies. I think this is the role that education and educators can play, in transmitting such knowledge and values.

Surrendipitously, I happened to view a couple U.S. news broadcast that I felt were relevant to recent EDC themes; which I tweeted below. One broadcast featured the introduction of digital technologies into the classrooms of the very traditional, conservative, religious sub-culture of the Amish community in the Mid-west of the U.S.A. Surprisingly, despite maintaining a simplistic, technology-free lifestyle, the Amish have embraced digital, educational technologies. This was yet another example of how socio-technological relationalties are changing rapidly.

Another broadcast examined a rising interest incoding. There was some debate in our Twitter Tutorial this past week about whether everyone needed to learn to code, with some EDC peers citing Evgeny Morozov, who considered it a “most bizarre and repressive idea.” However, even Morozov conceded that he is “all for making us aware of the various technological infrastructures at work.” Morozov is also no fan of the notion of “program or be programmed,” disdaining the use of metaphors. But, the question arises what the average online user needs to know about algorithms, coding, programming? What do we (as educators) need to teach to the next generation, so that they can make sense of themselves and rapidly emerging technologies?


I look forward to “pulling it all together” over the remaining few weeks.

Pasquale, F. (2015). “The Algorithmic Self” in The Hedgehog Review: Vol 17, No. 1.

Siemens, G. (2013) Learning Analytics: the emergence of a discipline. American Behavioral Scientist, 57(10): 1380-1400.

Williamson, B. (2014) Calculating Academics: Theorising the algorithmic organization of the digital university.