Here is a link to my Artefact. Since the MOOC I looked at was for a MOOC developed by the NHS, primarily for NHS staff and since I work as an e-Learning Developer for an NHS Trust, I decided to make my artefact in the style of one of the thousands of e-Learning courses that the organisation produces every year. Please don’t let this put you off, in this case the form of the artefact is in many ways part of the message I’m trying to convey 🙂
So, what did my intrepid expedition into the AI for Healthcare: Equipping the Workforce For Digital Transformation MOOC unearth in the first couple of weeks? As mentioned before, I’m using Week 1 of the MOOC for my microethnography. The MOOC runs over 5 weeks, so taking a snapshot of the first week, using data collected at the end of that week seems the fairest way to look at how the community started off.
In terms of community it is difficult to judge just from my own experiences of the MOOC. I can see that people Like and reply to some of my posts, and I’ve left replies for some others. However, the format doesn’t lend itself to prolonged conversations. To see if my experience is broadly similar to others’ I have collected a list of every known user (those who have posted or replied to a post in week 1 or on the initial Hello thread). I have then collected data on how many posts each known person has made on each of the topic threads, how many replies where made on each post and by whom, and how many Likes each person received on their posts and comments. I know that this methodology is a crude way of measuring community. It doesn’t take into account the nature or usefulness of the posts or replies (for instance, replies that just state “well said” or “I agree”), but hopefully we can get a rough idea of how much the students interact through these figure. There is also no way of looking at Lurkers (people on the MOOC who never interact) as there is no way of finding out the total number of students on the MOOC. I would suggest that there is a subcategory of Lurker – the Like Lurker, who only interacts with others through Liking posts or comments. I can’t prove the existence of these elusive people, as there is no way of knowing who has awarded a Like, but I think it is a reasonable assumption based on my experience of Social Media.
- Known Users (Week 1 and Hello thread) = 240
- Total number of Posts = 514
- Total number of replies = 74
- Total number of Likes = 503
- Users who posted at least once during Week 1 = 183 (76%)
- Users who received replies = 51 (21%)
- Users who left replies = 47 (19%)
- Users who received Likes = 123 (51%)
- Users who replied but never posted = 6 (3%)
- Users who’s posts never received Likes or replies = 59 (25%)
- Users who received replies but never replied to others (except on their own posts) = 38 (16%)
- Users who gave replies to others but never received any for their posts = 28 (12%)
These figures tell a story. While most known users (76%) posted in Week 1, relatively few replies where made, with most interaction taking place through Likes.
I decided to drill a bit further into the data to see if there where any patterns. For instance, are posts, replies and Likes spread evenly through the Known Users set, or are some students more engaged in forum activities than others? It could be that we have a small group of highly engaged students who largely spend their time on the MOOC Posting, replying to and Liking each other’s contributions, with a larger group showing some engagement and then the pool of Lurkers?
Analysis of Post numbers
As we can see, the majority of people made relatively few posts. 90% of students posted 5 times or less across the 12 topics looked at. So, 205 of the 514 total posts (40%) of the posts where made by 25 people (10%). This tends to support the idea that there is a core group of students who are more heavily engaged with the MOOC and that the emerging community may be driven by the comments made by this group of “influencers” who have a lot of control over the direction of conversation.
Analysis of replies and Likes
For this analysis I thought it would be useful to look at not just the number of replies each person received, but also how many they received per post. This should give us an idea of how successful our “influencers” are in controlling the conversation. A similar analysis of Likes was also carried out as this could be a better measure of how many people agree with the posts made by each person, while replies would tend to examine people disagreeing.
As we can see there are a couple of peaks in these graphs. Looking at the posts in question, one was a post that was “pinned” to the top of a thread by the course leaders. Since the posts are usually displayed in order of most recent, this blip could be caused simply by exposure, or it could be caused by the post being particularly insightful (which would reasonably be why it was pinned in the first place). The smaller anomaly was a post referencing a particular source of potential funding for projects, so it was naturally of great interest to the learners from within the NHS who would need to fund any AI applications that they intend to develop. I’ve reproduced the two graphs without these peaks so that we can look more closely at the trends.
As we can see, other than a small peak further along, the number of replies per post is relatively small and concentrated to those people who post often. However, the pattern for Likes is more uniform with those who post less frequently receiving a similar number of Likes per post to those who post often. In fact there seems to be a trend where those who post less frequently receive more Likes per post. There could be several reasons for this. Perhaps the people who post often are posing questions. Perhaps their views are contentious. Perhaps those within our proposed “influencer” group comment on each other’s posts, seeking them out to add to. Without considerable data gathering it is not possible to see which of these is correct, or if it is a combination. What we can see is that students are clearly reading the contributions left by their peers (remember, we don’t know if Likes definitely come from those who post, or if they also come from Lurkers). Not only can we see that students are reading a lot of the posts, but we can see that they are slightly more inclined to Like posts from those who are less prolific. There could be many reasons for this. Perhaps those who post less often only do so when they have a good point to make. Perhaps the usual way that posts are presented chronologically means that some posts get more exposure (on the front page) during times when more people are using the MOOC. Again, further data would need to be collected to arrive at any solid conclusions.
So what can I conclude from my examination of Week 1? It would seem that the MOOC has a large group of students who are engaged in the threads (although we can’t know this as a percentage of the number of users on the MOOC as we don’t know the total number of users). Within this group, there is a proportion (10%) who are heavily engaged in the threads (posting often and receiving replies and Likes in line with the number of posts). There is a larger group on the fringes of this community who involve themselves more rarely, but who’s contributions are either equally or slightly more highly regarded by the majority of users. The reason why we have this split is not clear, it could be down to personality type, experience of other online learning and MOOCs, familiarity with other online communities (for example social media), or it could be due to their reasons for attending the MOOC – it was designed for use by those within the NHS, and is very focussed on that organisation. Those attending who work for the NHS will naturally have more of an opinion on how the subject matter relates to their role, and may have insights that those attending the MOOC for personal interest do not. It would be interesting to see how these patterns play out in later weeks of the MOOC, or to see how they compare to a different type of digital course. It would also be interesting to see if there is any pattern in community engagement based on the background of the students (some data was collected on those users who are part of an NHS organisation or who have a medical background from the Hello thread before Week 1). I think that given the time constraints and the hours that it took to collect the data set so far, I will follow up on the last one of these ideas and look at patterns based on student history.
These are pretty key questions if we’re putting together a Microethnography to look at a “Community” of people on a MOOC. I suppose we could look at community in terms of interaction. If we imagine a scale between 0 = no interaction between Learners (for example a MOOC with no way for students to communicate) up to say 10 = some sort of Vulcan Mind Meld between all the students, then we could look at the features that make important landmarks along that scale (students can post, students can reply, peer review etc). That would be one way to judge “Community”, but it would still be a long way from being an objective measure (for example, would we consider an active forum on an LMS to be superior or inferior to student discussion taking place on Twitter or Discord? What are we basing this on?).
The MOOC that I’m taking part in for this assignment is quite limited in terms of possible interaction between users. The course is broken up into weeks, with a number of topics in each week. Each topic then has some “content” (text, images, video, links etc) and a single discussion thread where users can post their thoughts and reply to those posts (for a while I couldn’t see replies as being connected to the initial posts, but this turned out to be a setting that I had unwittingly changed). It is also possible for students to “Like” posts and replies. However, there are no guidelines about how to use the discussion part of each topic, so mostly they’re just a set of individual musings without any connection to each other.
To try and get a sense of the Community on my MOOC I’m going to measure 3 things. Firstly the number of individual posts made on each topic, by each known user. Then I’ll look at how many replies are made, by whom and on who’s posts. I will also look at the number of “Likes” each person receives and map this to each topic as well.
This is going to take some time and a great big Excel file with multiple sheets 🙂 To stop the whole thing from becoming unmanageable I’m going to restrict myself to collecting data on the Week 1 topics only. I’m also going to restrict myself to posts and replies made during the period 9th Feb to 15th Feb (the first week of the course). This is a MOOC, and people join after the first week. There would be a problem with constantly having to update my data if I don’t set a specific snap-shot of time.
If this method gives me interesting insights I can then return to the MOOC and either map a different week, or compare Week 1 as it was on 15th Feb with how it looks further down the line. Both of these could be potentially interesting as the first option would show how the course settles down (presuming that a lot of people who drop out do so after the first week) while the second option could show differences between people starting a MOOC on the start date and those catching up (for instance do people playing catch up seem to post more or less? How far back do they go to leave replies – is it just the first page or do they read the lot?).
Slightly delayed post as the first day for the MOOC I chose was today.
For our Micro-ethnography task I’ve chosen a MOOC that looks like it will tie in not only with the EDC course itself, but also my work role and personal interests, so triple the goodness 🙂
The title of the MOOC is: AI for Healthcare: Equipping the workforce for digital transformation it is on FutureLearn and is written and lead by a team from The University of Manchester and Health Education England (HEE). The MOOC is free for anyone, but is largely aimed at an NHS audience.
The initial look is quite appealing, but I’m not particularly convinced by the format for the first week. It looks a bit disjointed, and the discussion seems to be one long thread on each page, which makes it difficult to keep up. Admittedly this is only day 1 of the course, so maybe all I can expect is a set of randomly fired out comments from some of the students. A lot of comments so far quite definitely instrumentalist in their outlook, and also quite a lot of people happy to Black Box the whole subject area.
I have my butterfly net, sandwiches wrapped in foil and a flask of weak lemon squash for my expedition…
Here is my Artefact for Week 3. I’ll discuss the themes in a separate post (hopefully you’ll have some ideas about what the message is here).
Please leave comments here before reading the follow-up post. I’d like to see what people think of the movie coming to it cold 🙂
For a full explanation of what I was trying to get across in this short film, follow this link to the Artefact 1 page of this blog.
Film Festival – Part 2 and 1/2
This week we once again looked at some film clips (some of the same ones as last week, with a few additions, including Chappie which was nice) and considered not only what they tell us about people’s ideas about perceptions of technology, but also how these might apply to education using digital technology. The chat was interesting, but again seemed to focus on the negative (worries about control, fear of tech going out of control, cold and inhuman aspects of areas such as Learning Analytics and so on). While it is prudent to look for cautionary tales and draw conclusions from them, I think it might be equally as useful to look at our “White Hat” characters and see what they can tell us. Perhaps in their stories we can find the seeds of solutions to the sort of problems that we’ve foreseen? To avoid my tendency to ramble on forever, let’s just look at one of the types:
It seems that the general worries around cyborgs are to do with the idea that technology in some way disrupts or diminishes our essential humanity in some way. The influence of devices developed to run on lines of pure logic and reason having a detrimental effect on us, especially if they are physically part of our body. I think that this may be explored in “A Cyborg Manifesto” by Donna Haraway (2007), but unfortunately I’ve yet to force myself to read it all the way through as her style is incredibly annoying. I plan to get round to reading this and will update this blog entry afterwards.
So where do our cyborg characters stand on this and what can they tell us? As a reminder, we looked at Cyborg (DC Comics), Luke Skywalker, Picard, LaForge, Bashir, 7 of 9, Robocop and the reprogrammed Terminator. I consider that these fall into two categories (well, okay, many more than that, but for this argument let’s call it two):
- Cyborgs who are very human, just with some technological tweaks which may or may not be obvious (Cyborg, Luke Skywalker, Picard, LaForge, Bashir).
- Cyborgs who are mostly machine like, but are striving to become more human (7 of 9, Robocop, reprogrammed Terminator).
The first category gives us a great bit of news about becoming a cyborg (at least in terms of fiction) – it doesn’t make any difference to your humanity. Those characters embrace their humanity and live lives blissfully unencumbered by any machines taking over their mind to cause mayhem. The connection to education would surely be – it doesn’t make any difference, education is a human thing. Regardless of the technology used it is put together and received by humans. A traditional face-to-face course can be cold and inhuman (especially if we think of really traditional classes in say a Victorian Public School). The medium isn’t the message when its online any more than it is when its death by PowerPoint.
The second category shows us how even something that starts off seemingly inhuman can grow and develop in ways that make them more acceptable to our human sensibilities. 7 of 9 and Robocop where originally human, and manage to re-engage with that side of their nature. The Terminator on the other hand started off as a machine that has been reprogrammed to bodyguard a teenager, who then sets about making him display human like behaviour. I think the first two tell us that when we talk about Digital Education we are really just talking about Education. The Digital aspect is something new that seems a bit scary and inhuman to some people, but eventually these tools and techniques will just be considered to be the norm – education will have the human touch regardless of how it is delivered (it may take a period of adjustment, but we’ll get there). Sian Bayne covers this interestingly in “What’s the matter with ‘technology-enhanced learning?” (2015).
Terminator has a slightly different story to tell though. Even the most inhuman, logical, put together for sinister purposes piece of technology can be given a veneer of humanity that can fool us. The character is a “White Hat” not because it chooses to be so, but because its programming compels it to be so. While education can shift to encompass and incorporate the possibilities of digital, leading back to being a human experience, the worries we have about the abuse of data, leading learners down specific learning paths and so on will still be relevant. The public “face” of these systems can be more approachable and human, but in the background they were built for a purpose, to specific design specs, and they will carry out those tasks with maybe less oversight and intervention if we don’t scratch below the surface and see them for what they are.
I would love to bring in some of the Robots and AIs, especially Chappie and his journey from being built to frighten and injure to being a person in his own right who enriches the lives of those he meets, but this post is already about five times too long. Perhaps I’ll do something in a separate post 🙂
In the first week of the course we took part in a Film Festival, watching several short films and clips featuring cyborg, robot and AI characters (films linked at the bottom of this post). Our reading revolved around the idea of what it means for someone to be a cyborg, and how this links with Posthumanism.
What struck me about the film festival was the almost universally negative image that the clips had of our technological friends and helpers. Synthetic (and part synthetic) entities were depicted as being either emotionless (and thereby likely to make hideously immoral decisions based on purely logical steps to achieve their set goals) or actively malevolent (displaying the worst traits of the humans who created them). It was all a bit “Black Mirror” (an excellent series. If you’ve not seen it then you should probably watch it immediately). The one that didn’t fit into this theme was Retrofit, which in the best traditions of Sci-Fi told us more about what it means to be a human in our time than it did about some high-concept, high tech future.
I went in search of examples of good, decent, helpful synthetics and part-synthetics from popular culture, to see if we could balance this out a bit. It also gives us the chance to see if these characters (and the “baddie” ones) can be fitted into any sort of framework and categorised. Part of my search was just wracking my brains, and part of it was pulled together by posting a question to Facebook and letting my splendid friends remind me of (or introduce me to) characters. Unfortunately my experiments with trying to get Facebook posts to show on this blog with the comments in place have hit a brick wall, so I’ve collected the results and categorised some “white hat” characters here:
- Cyborg from DC comics
- Reprogrammed Terminator
- Luke Skywalker (Star Wars)
- Picard, LaForge, Bashir and 7 of 9 (various Star Trek series)
- Optimus Prime, Bumblebee and the rest of the Autobots (Transformers) alien rather than human, but still a species bonded with machines
- Data (Star Trek)
- Bishop (Aliens)
- Baymax (Big Hero 6)
- Johnny 5 (Short Circuit)
- Chappie (Chappie)
- Kryten (Red Dwarf)
- C3P0 and R2D2 (Star Wars)
- Holly/Hilly (Red Dwarf)
- Emergency Medical Hologram Mark I (Star Trek Voyager)
- Orak (Blake’s 7)
- Jarvis (Ironman/Avengers)
Looking at our “White Hat” characters, several themes seem to stand out:
- Synthetics and Cyborgs who strive to be more like humans (7 of 9, Data, Baymax, Johnny 5, Chappie, Kryten, Emergency Medical Hologram)
- Synthetics who have incompetence built in for some reason (C3P0, Holly/Hilly)
- Synthetics and Cyborgs as tools or servants despite being better at what they do than their masters (Terminator, Robocop, Bishop, Jarvis, Orak, R2D2)
- Synthetics and Cyborgs as disposable for high risk situations such as combat (Terminator, Robocop, Bishop, Chappie)
- Cyborg parts not affecting how humans are treated (Picard, LaForge, Bashir, Cyborg from DC Comics, Luke Skywalker), and conversely human like characteristics of synthetics being ignored by most characters in the story (C3P0, R2D2, Orak, Jarvis).
I’m going to have a bit of a ponder on these and try to refine them a bit. I expect there are some other categories I could group them into, and that these categories could tell us something about our relationship with machines. There already seem to be some trends based roughly on the decade when the films were made (the 80’s in particular seem to have had a grim idea of the future).
The great thing about blogs is that I can go away and reconsider parts of it, make alterations based on reader comments, round up some references to back up ideas/add in ideas I come across in the reading. For instance I’ll almost certainly write something about the different types of cyborg discussed by Miller, (2011) “9. The Body and Information Technology” from Understanding digital culture. Perhaps it would be interesting to see if the types of cyborg are reflected differently in fiction (are Enhancement types usually villains or heroes?).
Film Festival Links