Apr 272015
 

This afternoon I am attending a talk on the Privacy of Online Social Networks which has been arranged by the Social Network Analysis in Scotland Group (SNAS) and is taking place at the University of Edinburgh. The speakers are Jordi Herrera-Joancomarti, Cristina Perez-sola, and Jordi Casas-Roma, all from Universitat Autonoma de Barcelona (UAB). I’ll be taking notes throughout, although I think that the talk is also being recorded so may be available later. As ever this is a liveblog so corrections, comments, etc. welcome. (I will also be adding some images from the event later today as some of the processes discussed were quite complex and require illustration!)

We are opening with an introduction to the SNAS group, which meets at the University of Edinburgh on the last Tuesday of every month… They have a mailing list and I’ll add a link here later. Dr Jordi Herrera-Joancomarti is leading the talk, and is an expert on privacy and security.

Dr Jordi H-J: This is collaborative work with my colleagues Cristina and Jordi. My background is not social sciences but mathematics, so it is a good challenge for me to speak to a non technical audience here… Hopefully there are no scary mathematical equations here! I’ll open with an introduction, talk about Online Social Networks and graph theory, talk about the data you can mine, and I will talk about Online Social network Data anonimisation, and how you can release data from networks without compromising privacy, before coming to my conclusions.

So, to start with the definition of Online Social Network I am using is an “online service, platform or site that allos to create a user profle which can be connected with other user profiles of the network… ”  – a very computer science definition.

So this can be about specialisms like Flickr, LastFM, WikiLoc… specialised format (e.g. Twitter); Scope limited (e.g. LinkedIn); General purpose (e.g. Faebook, Google+) etc. The denomination of connectivity can be network dependent (e.g. Facebook: friends; Twitter: followers). An dinteractions between user profiles are also network ependent (e.g. Facebook: “like” action, post a message; Twitter: tweet, Retweet etc).

So, why are OSN interesting or important? Well they have become an important part of people’s everyday communications, with huge volumes of users. But there is also a book, Big Data (Viktor Mayer-Schonberger and Kenneth Cukier) which includes chapter 5 “Datafication” talking about the quantification of the world along the time from differnt aspects. So, when words became data (Google books in 2004); when localization becomes data (GPS); and when relationships become data (OSN). For instance Facebook datafied relationships, and most notably with the introduction of “Social graph”.

To graph theory then. A graph is a mathematical tool used to represent objects (nodes) that can be connected by links (edges). OSN can be modeled using graphs and analysed with graph theory. So… You can represented connections between individuals etc.

There are different OSN propoerties that dertmine the type of the corresponding social graph:

– Undirected graphs are those with no meaning on the incidence of an edge in the node. Facebook social graph is an undirected graph. So, no arrows between individuals, no value to that edge.

– Directed graphs (digraph) are those in which the edges have a direction associated with them. Twitter social graph is a directed graph. For instance you can follow someone, they don’t have to follow you… So we have arrows here to indicate connection and direction.

– Weighted graphs assign a weight to every edge in a graph.

So, when you add direction to a graph you can borrow many analysis tools from graph theory. So if we try with a degree of a node in an undirected graph… The degree of a node is the number of edges incident to that node, denoted as deg(vi).

In a directed graph the same concept applies but it is more complex… We have In-degree of a node and that is the number of head endpoints adjacent to that node denoted as deg-(vi). Similarly we can have out-degree for number of tail endpoints, denoted as deg+(vi).

So, in a facebook social graph the degree of a node is the number of friends of that user. In Twitter social graph, the in-degree can be seen as the number of followers of that user. High in-degree may indicate a popular user. And the out degree can be seen as the number of users that person follows.

We can also talk about the clustering coefficient. We see local clustering coefficient of a node – the proportion of edges between the nodes within its neighbourhood divided by the number of edges that could possible exist between them… So it measures how far are the neighbourood of a node to become a clique. So this is how well the friends of a node are connected. These kinds of technical techniques can be used to understand user connections and relationships.

We study OSN privacy from an information-fundamental point of view, analysing OSN privacy from a graph mining perspective. We do not study specific OSN services, configurations or vulnerbailities. In some cases we do make some assumptions about the type of OSN: open vs closed profiles. For instance Facebook is more difficult to extract data from than Twitter, an open social network.

So there are two kinds of users information that can be extracted:

1) Node information – data about a specific user, details contaied in the users profile on a specific OSN

2) Edge information – data about the relationship between members of the network – and that is what we are most interested in.

Edge information can, however, directly disclose node attributes – e.g. an edge representing a sentimental relationships between two individuals of the same sex would be revealing about their sexual orientation. It is more difficult to protect edge information than node information – as it depends on behaviour of connected people whereas node information is controlled by just one user. Relations between users can also detect communities, and more node attributes.

So, I wanted to explain about data retrieval. How do you ontain social network information? Well you can ask OSN providers – but many are not that cooperative or put a great deal of restrictions/agreements to do that. They provide local and/or anonimised data. OR you can take the data from the OSN providers – that is not always possible adn depends on the open degree of the OSN service. And it is very important to take care on the mechanism used to obtain information as that may determine the bias of the data you collect.

You can gather data several ways. You can use a web crawler to gather daya from an open OSN (like Twitter). Web crawlers are computer programs that retrieve web pages starting from a single (or multiple) page and exploring all its linked pages and also the pages linked to those ones and so on. Since most of OSN interact through the web, you can use web crawlers for OSN data retrieval… The process is iterative…

A download is the interface between the OSN and the crawler – it downloads the users profiles and passes it to the parser, which then parses that data. You draw out the friends of that user and add them to the queue, which contains all the users that are awaiting to be explored, found when crawling every user. And the scheduler selects which user, from the ones in the queue will be explore and sends the decision to the downloader. The scheduler impacts on both performance and data quality.

If you are exploring the whole network then it is not so important to consider the crawler details… if I am crawling every member I will find all of the connections at the end… the order you gather data in doesn’t matter in that case. BUT you cannot crawl all of the network available now… So you will have to, at some point, decide to take a partial view of the network. So to do that we have to think about notification and assumptions…

Users can be crawled (one that all his profiles information and all friends are known to the crawler (v E Vcrawl). A discovered user (connected to the user crawled), and an explored user  (discovered by relationship to discoverd user)?

So… for instance a Breath-First Search (BFS) Algorithm would start with one user (h)… you find they have two friends (d and j)… I crawl j and then discover they connect to users l and k and g (and I’ve already crawled d and h)… Then I crawl user d, finding connections to f, e, b, c… others are already found… Then I crawl l, find connections etc…

So, that is your schedule, the order you crawl. And the idea is that you can end up with all the elements of the network… This is quite a linear process. So, this is one approach, and this BFS algorithm produces graphs quite dissimilar to other algorithms you could use.

An alternative approach is the Depth-First Search (DFS) which works as a traditional stack, the first nodes to be crawled are the last ones that have been discovered (LIFO management). So, in this approach… If you start with user h… you discover j and d… But the next node you explore is d… then when you find connections to f, g, e, b, c… and you next explore node c. At the end you will end up with all the nodes as well… But in a different order than you had before… So, again, if you do this with a group of users (example here being 162 flickr nodes) it looks quite different…

Then you can do more intelligent things… You can use “greedy” algorithms:

– Real-degree greedy (hypothetical greedy or higherst-degree-crawler) takes its decisions based on the real degree (which may be unknown to the crawler before the node is crawled) of the nodes in the OSN. So a user has degree 5, degree 7 etc. based on the edges between different nodes… You can gather the whole network, or you may have restrictions and only capture part of the network…

– Explored-degree greedy (greedy) uses the actual known degree of the nodes in the OSN… So if you graph that you see many many connections, you look more conciously to the mode connected nodes.

You can also choose to select more variance in the network, to randomise your sample to an extent. This can be done with a lottery algorithm…

So, if you take information from a social network or a social network graph you have to be really well aware of what you are getting. When you do your sampling from different profiles, etc. that you understand what your sample is of. As far as you can see you can just adjust the scheduler to get what you want… you can do that to focus on particular users, types of users.

Schedulers have implications on privacy… depending on the level you select that has different implications… So your scheduler can have different objectives for the crawler – taking the privacy attackers point of view. So you can then understand which scheduler algorithm fits those objectives most appropriately…

You can also do more tricky things… For instance the classification of users from a graph point of views. So, I want to classify users, identifying the set of categories a new observation belongs to. The decision is made on the basis of a training set of data containing observations whose category membership is already known. When you try to classify users within the network, you can see link information which may help you to classify a user – connections to a community for instance.

The idea is that you can see classification as a privacy attack – user classification allows an attacker to infer private attributes from the user. Attributes may be sensitive by themselbes, attribute disclosuer may have undesirable consequences for the user. So the design of a user (node) classifer that uses the graph structure alone (no semantic infomation needed)… So, for instance… We may classify the user, with a neighborhood analysis to better classify the user… So the classifer analyses the graph structure and maps each node to a 2-dimensional sample using degree and clustering coefficient. The output is an initial assignation of nodes to categories…

And you can make that neighborhood information to classify the node… You can also have a relational classifier, which maps users to n-dimensional samples, using both degree and clustering coefficient and the neighborhood information to classify users…

So coming to the issue of data and data release… When you obtain a collection of data… you may have a more anonymised data view… You may see connections etc. but without user names, for instance. The intention is to preserve the privacy of users. But is this enough? Well no… this nieve anonimisation potentially reveals huge amounts about the user… if you know other data (other than names), you may be able to deduce who is in the network, you might find one user in the network and thus expose others. Removing the identifiers is not enough… So, you have to do something more elaborate…

One approach is to modify the edges – adding or deleting edges to hinder re-identification… But the problem is that you have two opposite objectives: On th eone hand you want to maximise the data utility and you want to minimise noise in that data. But you also want to preserve users privacy…

So, there are different ways to quantify the objective…. There are generic information loss measures (GIL) – measures like average distance, diameter, harmonic mean of shortest distance, etc… You want to preserve that in your data. So… you have the original network, you do one metric… and end up with a different network that is anonimised, and you can apply a similar metric afterwards to use it… In statistical databases you can preserve the mean of all the registers that sold boots (say)… If you know the questions to ask of that data, you know the process to keep that anonimised data close to the original data set…

You can also use specific information loss measures (clustering process)… Similar problem here… You have the original clusters, you use a clustering method to get to an anonimised (perturbed) version.

So, some measures behave in a similar way independently of the data in which they are gathered.

And then you have the idea of k-anonimity. A model that indicates that an attacker can not distinguish between different k records although he managed to find a group of quasi-identifiers. Therefore the attackers can not re-identify an individual. So, node degree can be the quasi-identifier… We can presume the attacker may know some of the nodes in the network… We can preserve the degree sequence, and the ordered degree sequence. And you can measure the k degree by understanding how many nodes have the same degree. So if two nodes in the network have degree 4, then the k-degree anonymity is 2. You can then make use of this to preserve the graph…

To modify the graph you can use edge modification (adding and/or deleting); node modification (adding and/or deleting). You can use uncertain graphs – adding or removing edges “particially” by assigning a probabiity to each edge. The set of all possible edges is considered and a probability is assigned to each edge.

Edge modification can include edge rotation, random perturbation, relevant edge identification, k-anonymity orientated anonimisation. These can allow you to keep data you want to keep, whilst preserving user privacy.

So, in conclusion, OSN can be modeled with social graph and analysed using graph mining techniques. Web crawlers may retrieve sensitive information from OSNs but the quality of the collected information will depend on the scheduler algorithm specitifities. Relational classifiers may provide relevant user information by just analyzing the graph structure information… Data anonimisation is needed for releasing OSN data without compromising the user’s privacy. This is a research field that is quite new and quite difficult… unlike statistical databases, where you can change one user without impacting on others, any change here does effect the network. And anonymisation algorithms need a trade-off between information loss and user anonymity loss.

Q&A

Q1) You talked about how much stuff is being datafied… Soon with smart watches we’ll have health data available. Because crawlers take some time… things could change whilst you are crawling.

A1) One of the problems in social networks and graph theory, is that algorithms for this sort of data are complex and time consuming… And that is a problem… Especially at scale. And sometimes you have the information, you make a lot of computation but the information is not static… so not only a lot of work not only on algorithms but also on understanding different and changes in the network – what happens when a node is removed for instance. There are people working on algorithms for dynamic data… But much m

Q2) What kind of research questions have you been using this with?

A2) There are two different issues for me in terms of social sciences… We don’t start with research questions… we start with problem and try to start it… So when AOL released data about lots of servers… you could identify individuals from the data… but you shouldn’t be able to… That happens because they don’t understand or care about anonymising data. So we are trying to provide tools to enable that anonymisation. We also have ideas about the crawling approach… So as a social network provider you might want to avoid this type of crawler… you might use this approach to trap or mislead the crawler… So the crawler end up in a dead end… and cannot crawl the network.

Q3) Some of the techniques you showed there were about anonymisation… do you use removal of nodes for that purpose

A3) There are several approaches for adding or removing nodes… Sometimes those approaches collapse those nodes… So you anonymise all the nodes too… But the general techniques that are more used are those that perturb and move the nodes.

Q4) One of the last things you said was about that trade off of utility of analysis and user privacy. My question is who makes that decision about the trade off? Would the people being studied agree with those decisions for instance, in the real world?

A4) The real world is much more complex of course. The problem is about deciding level of usefulness of the data… At the present time these methods are not used as far as they could be done… For statistical data this is often fixed by government… for instance in Census data you can see the method by which data has been anonimised. But for OSN there is nothing of that type, and nobody is telling… and basically no-one is releasing data… Data is money… So if we can try to give good algorithms to enable that, then maybe the OSN companies can release some of this kind of data. But at this moment, nobody is putting that idea of privacy there… Generally privacy level tends to be low, information level is high…

Q5) I didn’t totally understand how you set the boundaries of the network… Is it the crawling process?

A5) The idea is that there are no boundaries… Crawler goes… Maybe it completes within 1000 nodes, or 3 hours… or similar. You won’t crawl everything and you want some data. So 10 million users might be the boundary for instance… Then you have data to look at… So I have 10 million users out of a pool of 500 million… But which ones do I have? How representative? That needs consideration…

Q6) The crawler gathers a model of relationships and behaviours, and I’m sure that marketers are very interested. Is there potential to predict connections, behaviours, intentions etc.

A6) Yes, there are lots of techniques of graph theory that allow that sort of interpretation and prediction. OSN use these sorts of approaches for recommendations and so on…

Q6) How reliable is that data?

A6) Understanding similarities there can help make it more reliable… similarity rather than distance between nodes can be helpful for understanding behaviour… But I will say that they are quite accurate… And the more information they gather, the more accurate they are…

Q7) I was wondering when you were talking about measuring the effectiveness of different anonymisation methods… Is there a way to take account of additional data that could effect anonimisation

A7) In computer security in general, when you model someone you have to define the adversary model… What the adversary is able to do… So, what is the attacker able to have… The available information… So the more information is available, the harder it is to protect the individual. It is a complex scenario.

Q8) Is there a user friendly web crawler that can be used by non technicians…

A8) No. Sorry about that… No, because there are some frameworks… But you don’t have one solution to fit all… But the idea is that there are some frameworks that are more suited to computer science people… Tomorrow in the workshop we will explain extracting information from Twitter… And those techniques will let us explore how we could develop a crawler on Twitter… So exploring connections and followers, etc.

Q9) What are the ethics of web crawling in social sciences? And what are the positions of the OSN on that?

A9) You can crawl OSN because the information is public. So you can crawl Twitter, as information is public. If you want to crawl Facebook, you have to be authorised by the user to look at the profile… And you need to develop an algorithm to run as an app in Facebook… and authorise that… But that doesn’t mean the user understands that… But for instance in last US Election, Obama campaign did an application on Facebook that did that… graphing their supporters and friends… And use that in the campaign…

Q9) I was wondering about the crawling of discussion forums… where you cannot get authorisation. But you also mentioned that providers not keen… is it legitimate to do that…

A9) I think that it is… If you are crawling public information… There is another thing of the OSN not liking it – then they can make some restrictions. If I do things that avoid OSN restrictions that is fine… You can do that

Q10) I wanted to follow up on that… There are legal and ethical issues associated with crawling websites. You have to consider it extremely carefully. If I use a website that says it does not allow crawlers, I don’t expect it to be crawled and that would not be legal under data protection law. And there was some research about 10 years ago a research project found that bloggers, although posting in public, didn’t expect to be analysed and interpreted… And you do have to think about the ethics here… And you need to think about the user’s expectation when they put the data up.

A – Christina) Everyone uses Google, you can’t expect that when you put something on the internet you have to expect it to be crawled

A – Jordi) From my perspective, as a researcher doing cryptography what you say is quite strange… My work is about protecting information… It assumes people will be trustworthy with your information…

Q10) No, I’m saying an ethical researcher should not be breaking the law.

Comment) There can be an expectation of privacy in a “public” space…

Comment – from me) I would recommend the Association of Internet Researchers Ethics Guide for more on how you can mediate expectations of users in your research. For your cryptography work that may not be as relevant, but for others in this audience that guide is very helpful for understanding ethical research processes, and for thinking about appropriate research methods and approaches for ethical approval.

And with a gracious close from Jordi, we are done! There is a workshop running tomorrow on this type of analysis – I won’t be there but others may be tweeting or blogging from it.

Apr 232015
 

On this very sunny Thursday I am at the IAD in Bristo Square for the elearning@ed forum’s 2015 conference which is focusing on Designing for 21st Century Learning. I’ll be taking notes throughout the day (though there may be a gap due to other meeting commitments). As usual these are live notes so any corrections, updates, etc. are welcomed.

The speakers for today are:

Welcome – Melissa Highton, Director, Learning Teaching and Web Services, IS.

Thank you all for coming. It’s a full agenda and it’s going to be a great day. Last year Jeff left us with the phrase that it is “exciting times” and that’s reflected by how fast this event filled up, sold out… you are lucky to get a seat! Being part of this community, to this forum, is about a community commitment we will see throughout the day, and we are very lucky and very appreciative of that.

Designing for 21st Century Learning is our theme for today. As someone who did all their formal learning in the 20th century, I started with a bit of Googling for what 21st Century might be – colourful diagrams seems to be the thing! But I also looked for some accounts from the university of what that might mean… some things that came through where that it is about teaching understanding of difficult things in all subjects, do a little to remove the inequalities of life, practical work and making things with one’s hands “the separation of hand and brain is an evil for both”. But these words are from 1905, they are from the University Settlement. But actually many of those remain common values. But there are are also issues of technology, of change…

“It’s not ok not to understand the internet anymore” – Martha Lane-Fox delivering the Dimbleby Lecture at London’s Science Museum, March 2015. That is certainly part of what we are talking about. Most in this room will feel they understand the internet, but we also have to be thinking about the challenges raised, the trends. And I’m going to finish with a graphic from the New Media Consortium (which the university is part of) tracking some of these changes and trends here/coming soon.

Chairs session – Individual short presentations, followed by open panel discussion (chaired by Jessie Paterson)

Designing for 21st Century earning: the view where I sit Prof. Judy Hardy, Physics Education, (Physics and Astronomy) Profile

I was asked to give the view from where I am, in 10 minutes, which is fairly tough! So I will be sharing some of my thoughts, some of what is preoccupying me at the moment.

Like Melissa I saw the concept of 21st Century Learning and thought “gosh, what’s that”. So I tried to think about a student coming here in 2020. That student will probably be just about coming to the end of their first year at secondary school right now. So what will it look like… probably quite a lot like now… lectures, tutorials, workshops etc. But what they will have is even more technology at their fingertips… Whether that is tablets or whatever.

We have been working on a project tracking students use of technology. We didn’t tell them what to use or how. They used cloud based word processor saying it saved times, seeing each others writing styles benefitted the flow of the report they worked on together. They used Facebook and self organised groups to compliment and coordinate activity. They just did it. I think many didn’t mention it as they just took it for granting…

Interactive engagement in learning performs something like double the learning gain (see R.R. Hake 2007). But wht is that? We did research (Hardy et al 2014) on academic staff teaching in UK university physics departments. Many want to teach, many focus as much on teaching as research. So what are the challenges? Well time and time as a proxy for other things… We can’t ignore that if we really want to move from a dedicated few doing great teaching work, to mainsteaming that. Deslauriers, Schelew and Wieman (2011) in Science found that it took 20 hours preparation to teach with a flipped classroom – that reduces after the first run but it is a substantial investment of time. Pedagogically there is also confusion over the best tools or approaches to take..

What is preoccupying me quite a bit at the moment… It is not about the “what” and “how” but about the “why’. There is awareness of what we should or might do. How to do that is very important – you need to know what to do and how to do it. But you also need to understand the principles behind that, why you are doing that, what the purpose is. You need to know what you can modify, and why, and what the consequences of that might be. When we are doing teaching, when we are thinking about teaching, we need to have this in mind. Otherwise we end up using the same formats (e.g. lectures) just surrounded by new technology.

Prof. Sian Bayne, Digital Education, (Education) Profile

It was a bit of a wide brief for this session, so I thought I would talk about something happening this week. Some of you will be aware that the #rhizo15 MOOC is running again this week, the Rhizomatic Learning “cMOOC” idea. And I saw lots of tweets about a paper I’d written… Which got me thinking about what has been happening… and where things are going…

That paper looked at the Deleuze and Guattari (1988) concept of striated space (closed, hierachichal, structured, etc.) vs smooth space (open ended, non hierachichal, wandering-orientate, amorphous). And that these spaces, these metaphors, intersect… And this paper was using these metaphors in the design of learning itself. So, back in 2004 the VLEs and LMSs was pretty much what there was in terms of online learning – very striated spaces. Emerging at that time in a more smooth space – were ideas like scholarly hypertext, multimodal assessments, anonymous discussion boards (which went, but are kind of back with YikYak), wikis and blogs.

So, what has changed around 10 years later? Well in the striated space we have VLEs and LMSs, Turnitin, e-portfolios, and we have things that may be striating forces including personalisation (flexible but to rules), adaptive learning, learning analytics, gamification (very goal orientated), wearables.  In terms of the smooth spaces… we have Twitter (though some increasing striation), YikYak, real openness. And we also see augmented realities and flipped classrooms, maker spaces, and crowd-based learning as smoother spaces.

So, what’s next? The bigger point I want to make is that we have a tendency in this field to be very futures orientated. I was also googling this week for elearning and digital education trends 2015.. huge numbers of reports and trends which are useful but there is also a change acceleration, trends and practices to respond to and keep up with. We need to remember that we are doing those things in the context, to look back a bit, to consider what kind of teaching do we actually want to do, what kind of university do we want to be. And ultimately what is higher education actually for? And those kinds of considerations have to sit alongside that awareness of changes, trends, technologies…

Using Technology to support learners’ goal setting – Prof. Judy Robertson, Digital Learning, (Education) “Using technology to support learners’ goal setting”.  Profile

I am also talking about what I am working on this week, which has mainly been data analysis! My work looks at technology use by children (and sometimes university students). I design and evaluate technology for education and behaviour change, often designing learners in the design process. There are aspects of behaviour change and concepts from games that can be particularly useful here, but games tend to have set goals built in (even if you can choose your goals from a set), and I look at learners setting their own goals.

So my research vision is about working with users to develop technology which enables them to set and monitor appropriate goals for themselves in the context or education and healthcare – that could be working with children and teachers to develop software which enables goal setting around problem solving and physical activity, or to work with new undergraduates to help them to plan and monitor their studying, or even working with older adults to assist them to change their patterns of sedentary behaviour. But there is a risk of becoming like the Microsoft paperclip… How do we actually make technology useful here?

So I have been working on an exergame (a game where physical exertion is the input medium) called Critter Jam (aka FitQuest) which is looking at whether it is possible to motivate children to increase their activity. So the game might have you collecting virtual coins, or being chased by a virtual wolf… It is all about encouraging mainly running activities, with mainly playground game type activities. Within the game children can pick from different goals… For those with intrinsic motivation tendencies you can aim for your personal best… For some children you might set a custom points target – and how children (or indeed university students) pick that target is interesting. Some children may want to top the leader board  – that motivates some, but competition can be negative too…

So, we are also looking at fine grained log file data from around 70 kids over 5 weeks as part of a wider RCT data set. I’ve been looking on the sort of goals kids set and how they achieve them. And also looking at how self-efficacy relates to goal setting. And as you look at the data you can look at the high performing kids and see where there are patterns in their goal settings.

It turns out that kids achieved their goals around 50% of the time, which is a bit of a disappointment. And those who expect to do well, tend to set more ambitious goals – which raises some questions for us. And in terms of how goal setting relates to high performance gains we have some interesting qualitative data. We interviewed some students – all of our kids here were 10 years old – and they reported that if they had set too hard a goal, they would reset to a lower goal, but then aim to keep improving it. This seems reasonable and thoughtful for a 10 year old. At 10 that’s not what all students will do though (even for undergraduates that doesn’t even work). Speaking to another child they aimed fairly low, to avoid the risk of failure… again something we need to bear in mind with university students and how ambitiously they set their own goals.

Prof. Dave Reay, Carbon Management and Education, (Geosciences) Profile

I completely misunderstood the brief… or perhaps took it differently… I wanted to tell you a bit about what we do, and the work I do in digital education. I’m based in geosciences and I work on climate change. But seven years ago – in this very room – we started a new masters programme on carbon management, aimed at helping our students understand how we tackle the holistic challenges of climate change. And part of the challenge for us as lecturers was how we can make this issue apply, feel practical, that included applied experience. So we started to think about how we could develop online learning to do this. So we started by developing tools on “hot house schools” using Labyrinth to let students take the role of teacher, headmaster, etc. to understand decisions taken to keep students safe, to make changes, etc. And I got a real passion for online learning.

The interactive stuff worked well, the interactions with students online worked well… And we launched that online masters four years ago. As you will all know that interaction online can be at least as rich as face to face programmes. And we now have a new programme with both face to face aspects and a core course running online. We are also creating a course on sustainability, the idea being for our on campus face to face students to really understand sustainability in their field (whatever that is) and an online course was what we felt could deliver this. The vision is for every student on campus to have the opportunity to look at this, to think about sustainability in their fields. They will leave this institution understanding not only sustainability but also a positive experience of online education, that they think of Edinburgh when they think about lifelong learning, of retraining – a very 21st century learning issue. So, I think in a few years time I will have exciting slides to share on that.

Finally I wanted to talk about my research which is on climate change and carbon footprints. In the last few years I have been looking at digital education, ICT, etc. from the perspective of their environmental impact. So we have quantified all of the emissions associated with the programme – we are calling it the greenest masters ever! The face to face programme is great but travel of students is significant, estates and buildings have a big carbon footprint, so we can actually put a number on every aspect of the online masters and its carbon footprint – and we can offset it too! So, if you are interested in the kinds of innovations taking place, and how they relate to emissions and carbon footprints. We want data, we want to quantify online as a greener way for our students to learn, so please get in touch.

Learning Analytics – Prof. Dragan Gasevic, Learning Analytics, (Informatics and Education.) Profile

I am based in both the Schools of Education and Informatics. And I will talk a bit about what we are talking about when we say “learning analytics”. Usually we mean that we are looking at data from learning technologies. But before we get to that we need to talk about why we might do this. We have already heard about our learners as non traditional, heterogeneous… but we cannot personalise the entire learning experience for every students manually. Feedback loops are, however, so important to the learning process.

So, most educational institutions today have student information systems – from before enrolment, courses taken, financial information etc. And then we also have learning environments – LMSs and VLEs like Blackboard, Moodle, etc. But we also have so much more out there… From social networks, to searches, to blogs and other collaborative and reflective tools, and then we also have slides and resources. And wherever we go here we are always creating a digital footprint. And that is irreversible. Today we have the computing technology to analyse that data too. What we want to do with learning analytics is to use those digital traces, for use by instructors, by organisations. And that enables the provision of personalised feedback back to the learners.

We are touching, most of our research, on most of these nodes… But the guiding force here is that learning analytics are about learning. We must not forget that. It is not just data capture without questions. It is a reminder that we have to think about the critical factors that learning analytics need to account for. We have to remember that learners are not black boxes, they are individuals and they have traits but those traits change – background knowledge, understanding, technology and cognitive tools. To really deliver on the expectations of learning analytics we need to understand that.

So, one example here is a piece of technology, for video annotation, to enable reflective practice. Students can view a video and can then leave comments at a particular moment at the video, tag that comment, etc. But if learners are unaware that technologies or tools might be beneficial, they won’t be motivated to use it. So we have a responsibility to scaffold our learners use of these tools, and convey that to our learners so that they are motivated, and so that they understand those benefits rather than just be presented with the tools.

We ran a study in British Columbia we tried too approaches to creative reflective activities and tools. In one group they were not graded, in another they were graded and received feedback. But we also ran a third course which was similarly graded, but these students had previously used this tool and they started to internalise those benefits – they doubled their use of their tool. When those same students (who had initially been graded on their use) undertook a non graded task, they continued to use it… which tells us a lot about these students motivations. We did see some quality reduction in their annotations… So that tells us that we need to provide additional scaffolds for their work… So for instance simply encouraging students to share annotations with each other can do that.

Learning analytics are only useful if we know what we need, what conditions we work in – counts don’t count much if decontextualised. We need to think of this and approach it as a scaling up of qualitative analysis in some ways, and for that to be part of learning analytics as well.

I also wanted to say that pretty visualisations can be harmful. We have to be very careful when sharing visualisations with students. University of Melbourne showing visualisations of performance to a group of students that was quite demotivating – both for those doing less well, and for those performing well who saw they were doing better than others.

One size does not fit all in learning analytics and institutional policies and practices have to reflect that. And with that I will end for now.

Virtual Edinburgh – Turning the whole city into a pervasive learning environment – Prof. Jonathan Silvertown, Technology Enhanced Science Education, (Biological Sciences) “Virtual Edinburgh: Turning the City into a pervasive learning environment”.

The thing to know about the future is that the seeds of the future are already here… Perhaps in your pocket through your smart phone. Many of the devices you carry around with you already have huge potential, and may be starting to be used in education but there is more that can be done.

I’m talking about  a project we are calling “Virtual Edinburgh” which is looking to harness that existing technology and use the whole city as a learning environment. This picture in my slides is taken from a bus enabled with wifi – that’s part of what I mean by the future already being here… And there are already apps seeking to do this… Walking Through Time – lets you explore historical maps of the city, LitLong (formerly Palimpsest) – shares literature in the context of the city, MESH – looks at social history in the city, BGS’s iGeology 3D lets you explore the geology around you, FieldTrip GB lets you create your own research data collection form, iSpot lets you identify aspects of the natural world, and Wikipedia has a nearby function that can be used with students… There are already a lot of stuff we can use in this environment…

So I just want to show you an idea of how we could put this whole idea together… So a trip on a bus from Calton Hill to Kings Buildings… You might identify some wildlife on Calton Hill with iSpot – discovering what a plant species is, looking it up on Wikipedia… The missing link here is back to the university and what we do at University of Edinburgh – if you searched for that plant you’d get back to the scientists researching these plants at Kings Buildings… So, Virtual Edinburgh is looking to connect these aspects together and to expose these elements more widely.

Looking at the University’s ‘Emerging Vision of Learning and Teaching” I wanted to draw out the elements that call for students having greater agency in co-creation of learning, and of being part of the wider community and learning with them. So, I see Virtual Edinburgh as engaging in various modes of student participation – within pre-baked VE apps there will be elements of data retrieval and engagement; as well as more interactive aspects including students creating new data, new apps, new ideas as well. And the Infrastucture will be about a teaching and learning infrastructure, a data infrastructure and a technical infrastructure…

The ultimate objective is to make Edinburgh the city of learning.

Q&A (all speakers)

Q1) One of the running themes here was about digital literacy. Judy’s comment that students barely commenting on the use of Facebook, as not worthy of mention by them… So what baseline of technologies do we expect from students these days, and what do we expect staff to keep up with?

A1 – Judy R) That’s a really interesting question. Although children and secondary school learners are exposed to technologies we cannot assume they understand how to use them appropriately. We cannot assume that.

A1 – Judy H) One thing to add to that is that we have to understand how institutional and personal technologies are intermixed. In that study there were centrally provided technologies but most moved swiftly to their own personal choices of technologies, and we have to understand that and what we do with that.

A1- Dragon) We know that there are no such things as “digital natives”, that we cannot assume understanding. Students may be more exposed to technologies but young kids are not neccassarily exposed to creating things in these spaces… They may even be at a lower level of skills than in the past simply because of the affordances of the types of tools they are using.

A1 – Dave) I have an embaressing confession to make. When we first ran this course we looked to use Google Hangout… I was all set up… I was waiting… The time ticked over… and noone joined me but my email went wild with students unable to get in… And we learnt that we have to understand and pre-set up those spaces ahead of time…

A1 – Sian) What Dragon said is really important here in terms of our expectations of students and the realities of their knowledge and understanding of these tools.

[Apologies, at this point my sore throat kicked off so I was unable to type… We had some interesting questions about the gap between students in first and second year, the innovations there, and what happens later on in a programme… ; and on learning skills and how they relate to learning outcomes]

Q2 [in my numbering, about the fourth or fifth in the room]) Internationally we have MOOCs, we have students from across the world

A2 – Dave) Part of what is so exciting about teaching online is that so many students internationally could not attend in person – due to location, family commitments, immigration restrictions. And online learning not only has environmental benefits but also opportunities to really help make the university the brilliant place it can be.

A2 – Sian) I think that it is useful to distinguish between learning and education – where education is the formalised accredited aspect of what we do. It’s not that we shouldn’t be part of that wider space of learning but that that distinction matters.

A2 – Dragon) Sian’s distinction is very important here. But we also have to remember that students don’t just attend for course content. It is about the knowledge and skills of those they will be engaging with. To learn online students also need exceptional organisational skills and discipline to fit their learning around their lives. But we also see different types of learning – capabilities and competency based learning which can have negative connotations but are also quite useful concepts.

Q3) I’m always quite interested in the gap between primary and secondary school education in terms of technologies… And how we keep up with that…

A3 – Judy R) There are quite different expectations around technologies. We have primary schools using Microsoft Office – which seems kind of weird given that it’s a professional productivity tool – and some use of blogs appearing although there is something of a horror at the use of anything social, and of any tools beyond the walled garden.

A3 – Judy H) We also have to remember that not all our learners come from Scottish schools… There is a great range of backgrounds that our learners have come through…

A3 – Dave) I do see what my own kids encounter, how they are learning… But I would also refer to the oracles at Moray House as well to get an idea beyond what I see in our undergraduates…

A3 – Jonathan) Perhaps next time this event runs that is a talk we should see here in fact.

And with that Jessie thanks our wonderful speakers for a stimulating session, and we are off for tea, coffee, or in my case a lot of Fisherman’s Friends and a quiet glass of water.

“Co-Creation: Student Ownership of Curriculum” (Workshop) – Dash Sekhar, VPAA, EUSA and Tanya Lubicz-Nawrocka, EUSA

Tanya: The panel session today was a great way to kick off this event. And it certainly made me think about Ron Barnett, and his book Realizing the University in an Age of Supercomplexity. I’m going to be taking you through some of the theory I am looking at – as I am both a member of EUSA staff and a PhD student at the Moray House School of Education. 

Kuh’s definition of student engagement is “the time and effort students devote to activities that are empirically linked

Cathy Bovill (Cook-Sather, Bovil, Fenton 2014) also talks about Co-creation of the curriculum being about “partnerships based on respect, reciprocity and shared responsibility between students and faculty”. That has great opportunities but can also be difficult – students don’t always know they can share in a lecture, and that co-creation idea can seem scary to both staff and students.

Thinking about co-creation and representation, we just had our teaching awards last night. Students are the experts in their own learning so student representatives are not only invaluable as sources of feedback, but also as proposers of solutions as well. Co-creation of the curriculum is about recognising student expertise, their goals, where they want to go, and how the learning outcomes of the course relate to that. It opens up the boundaries of what we can expect of education.

Dash: We’ve talked about the concepts and radical ideologies and of moving governance of the university so that students are active at all levels. But I’m going to talk about examples, in a range of universities.

For instance student led community projects are already part of a number of courses, for instance in the Geosciences project presented at senate. The students create the project, they design that, they carry out that project. This puts students in charge of creating their own goals, their own content. Obviously there are technologies that make co-creation more possible. But the area that I want to focus on are about assessment.

This exampe is about student partnership in assessment (in Social Policy?). Students met early in the course with academic staff to discuss assessment options, weighting different forms of assessment. Projects, exams, etc. with students able to vote on options/weighting – so not all students got what they wanted. Students welcomed the opportunity of choice, reflection, to discuss those options.

Another example, in the US, enabled students to be involved in the grading criteria. They were able to create or influence the grading criteria, and to reflect back on that process as well.

I also want to talk about social bookmarking. This example is from a Statistics course. Here the lecturer asked students to tag 10 sites related to the course, handed back to professor, then they were presented in the VLE, trends were shown, professor referred back to those examples found within the course. It is surface level to an extent but it is students creating content, influencing the course.. It is a radical shift.

So, what we want to do now is to have some discussion about what these changes mean. We want you in groups to discuss:

– How can you integrate these examples within your work?

– How can new technology enhance this partnership further?

– What support may staff/students need to implement these?

[cue break whilst we discuss]

Comments back from groups:

Group 1) Advanced students, honours levels etc. quite well set up for those broader learning objectives

Group 2) I am teaching on an MSc where students have a choice over the units that they take, the students really thrive in that environment and the students really push themselves and achieve

Group 3) One of the things my colleague Peter Evans is seeing through accreditation for the MSc in Digital Education is a 20 credit course within which students can create their own 5 credit activities, giving students a lot of autonomy within a structure there.

Group 4) We were talking about assessment and how students can engage in that, and anonymity in that process. Getting students to write questions and challenges against which they evaluate their colleagues – particularly talking about Peer Wise

Dash: There is another example with peer assessment, students had to justify not just if they met that criteria, but also to justify why that was the case.

Tanya) One group I sat with was the issue of not all students wanting to assess or be assessed by others. They see the lecturer as having greater authority, that they may not like peer assessment at first.

Group 5) We were also talking about anonymity and tools like Textwall which allows students to share anonymous comments on a wall (like a Twitter wall), also clickers, etc.

Comment) We tried a Twitter wall with one of our large undergraduate classes. It was sort of 50% brilliant and engaged. And 50% really inappropriate. There wasn’t much self-policing.

Group 6) We talked about beaurocratic barriers, getting something through the board… That there is reluctance to change, that perhaps only 5-10% of what you can do can be novel. So it’s how to get the beurocrats who sit on the board to approve something new and innovative. And how do you then pass on the work to the external examiner.

Dash: Luckily we have an assistant principal pretty much responsible for that.

Ian Pirie, assistant principal) I would say that my background is art and design, where we already provide videos, images, etc. to external examiners, so I would say that that can be done. That’s a disciplinary culture issue, and do please talk to me if you meet those sorts of barriers.

Dash: There you go. We are at time but please do come and find Tanya and I about co-creation etc.

“Using e-Portfolios to recognise our student and graduate attributes” – Simon Riley (CMVM) and Prof. Ian Pirie, Asst Principal Learning Developments

I’ll be talking about a number of uses of portfolios in art and in medicine. In both fields portfolios enable students to capture and evidence competencies. Everything is documented in that portfolio. And the students will update and prune, and reflect on that – sometimes we have to stop students from pruning too much! I couldn’t take you into a lecture and talk to you about playing the piano, and an hour later you can play it. You have to assimilate that, to practice and engage, to construct the essential knowledge. That’s the reason portfolios come in to these disciplines.

Portfolios are already well established in Art, Design and Architecture, in Medicine, and in other fields such as engineering, healthcare, etc. And often that is associated with professional competencies and evidencing those.

In Art, Design and Architecture portfolios are central in visual arts education (for ECA that is since 1760). That is from admission to higher education, for further study, for professional purposes. Once someone has committed to study in these subjects, they maintain that portfolio. And already school leavers engage with portfolio concepts of enquiry, reflection, etc.

In 2008 there was a change in submissions, so applications for ECA now run to 7000 applicants for 150 places. The logistics for physical portfolios were impossible. We have moved to digital portfolios. But we have looked at this, checked the robustness, and the digital submissions are assessable in the same way as physical portfolios were, the same decisions are made.

Simon: I’m talking about medicine here. When Ian first showed me that set of slides of those portfolios I thought those were exit rather than entry portfolios. That standard is amazing.

I am talking about medicine here and we are governed by the General Medical Council. They convey their requirements in this document called the “Tomorrow’s Doctors”. I came to this through my running of the “student choice” element of the programme. Students have genuine choice over about 20% as long as it covers skills in the right way. Post graduate students already have a long history of a log book, a portfolio of their work and practice that runs alongside this.

So, the GMC gives us a set of learning objectives. And we have tightly mapped our curriculum into what the GMC requires. We have themes running through the curriculum… And we need to tie themes together in competancy, thematic ways rather than switching all the time. So, how do you do this? Well we did this with eportfolios. This is currently on bespoke VLE system (EEMC). So, what goes in? Well students do case reports on specialist tasks and activities. They do a range of projects and one of the characteristics of Edinburgh is that we use our research rich environment as part of teaching medicine – the students work on research projects, seeking new information, generating their own data sets, etc.

We are also getting students to reflect on their learning, and that is critical. How good are we at doing this? Well we are getting there but there is probably more we could do. And there is that maxim of “see one, do one, teach one” and whilst we’d like to think there are more gaps than that, we do have senior students and members of staff teaching junior colleagues.

There are some other elements to the portfolio – and this is where we are changing things as we move from EEMC to something open source, probably PebblePad. But the parallel strand here is the professional development portfolio – CV, reflection, etc.  If we look at our portfolio here, it looks a lot like Learn (though it is a precursor) but it lists competencies, evidence, etc.

So to give an example here is the SSC2 Group Projects are projects which generate portfolio items they use WordPress, and they are open to potential applicants etc. And the material produced here are absolutely brilliant. They look at novel areas of medicine, they take real ownership, and working with a not very senior colleague they create really excellent materials.

These portfolios capture competencies, they prepare students for professional life after studying, they allow us to assess reflective skills.

Now, as Ian and I put this presentation together, from our two disciplines which seem poles apart… We see that we actually share so much…

Ian: Based on Koh’s model, visualising stimulus, input, action… as a cycle of Action, Creation, Selection, Reflection and all aspects feeding into the eportfolio. That is a shared pedagogy between our subjects. The format of the lecture leaves us unable to understand what the student is learning, what they understand, what is going in… Fundamentally it is the understanding and reflection area where students can find themselves frustrated, wanting better feedback, etc.

ePortfolios have huge potential here but, for a while, our colleagues in England were required to do this. Student didn’t take to them but that is perhaps because they did not understand the benefits of them. When our students move onwards their degree might get them an interview but employers are really looking for everything else, all that stuff that would be in that portfolio. That is what will count for them. And what is really important in the eportfolio is that we really have to properly value each students portfolio and recognise it formally, as well as thinking about how they take that forward, how they make onward use of these portfolios they have spent so much time creating.

Designing for Open- Open Educational Resources and new media for learning – Melissa Highton Director, Learning Teaching and Web Services, IS.

One of the things we have to ensure we do at this institution is to close the feedback loop. And I’m very pleased that I’m able to do some of that. Last year we had a passionate plea from Alex at EUSA about opening up the institution so I’m going to report back on that…

When Alex told us we should be more open as an institution, he said there was an opportunity to open up all learning materials as an ethical issue, as a sustainability issue. The University set up a task group, the OER Short-Life Task Group to explore ways to take forward an OER strategy for the University and to report findings and recommendations to Learning and Teaching Committee. Open Educational Resources are about opening up resources, making them discoverable, reusable, etc. So, we had a very good think about an OER vision for the University of Edinburgh and we proposed three strands that extend the strengths of the university.

Since 2007 a number of institutions have signed up to the Capetown Open Education Declaration (2007) around philanthropy and practice in education. About sharing large collections of rich resources, shared to parts of the world where there are perhaps less. But there is also the issue of how one adopts, adapts, tweaks that material is also important. Often that can be a barrier, unless we understand how we can tweak that material. Or you can find a black market in reuse, where we reuse but try to hide our reuse of others materials…

There are also some pretty strong opinions about publicly funded institutions not sharing materials they have been funded to create, seeing this as a moral issue. But there is also a reciprocity issue – if you take from the internet, you should also give back. But one of the problems of the word “open” is that it has many different meanings… Some thing online is open, some think open is not open until there are no restrictions. But there is a website for this, opendefinition.org, provides a helpful definition:

“Open data and content can be freely used, modified, and shared by anyone for any purpose”

And that is particularly helpful as it moves away from thinking about open educational resources, towards thinking of our resources in the context of open content more broadly, and to the wider understanding of openness.

For us to share openly we also have to understand what we mean by open. We also need our colleagues, our students, etc. to understand what we mean by open as well… To understand the implications of openness, licensing, sharing and use of online materials – whether those you have found or those that you publish. And this is very much aligned with the University’s mission as a global institution engaging globally.

Creative Commons licensed work are increasing, and these licenses are very relevant to how we use and create and share materials. These licenses were invented within the academy – law faculties from the US and UK looking for new ways to license content for the web. These have been available since 2001, and more varieties since 2007. And these licenses come in different formats – lawyer readable, user readable, but also machine readable. And you can share content with that license attached, which is hugely useful.

Some countries have made legislative commitments to open education, including Scotland and the UK (separate countries in this list, probably because of the varying legal systems). And looking at where these CC-licensed works are published the majority are from North America, any from Europe… So for example we wanted to create some new learning materials on the LGBT experience and looked at how that might be developed, but as we calculated the potential time and cost of that.. and then we found OER resources from a North American university that could be easily adapted at a fraction of the cost and the time. That’s hugely useful for us, and for diversifying our teaching for that course where we felt we had this gap to address.

Open.Ed is a website, a vision, and a strategy with three strands… “for the common good” – teaching and learning materials; “Edinburgh at its best” – showing what we do best; and “Edinburgh’s treasures” – making a significant number of our unique learning materials available.

In terms of managing assets the licensing on materials make it possible to do this stuff. The license to adapt and change allows us internally to adapt and change materials, to store and keep and move and share and reuse. Without those types of licenses we risk great unsustainability. And Edinburgh has a great tradition of sharing – think of the common stair. So the license lets us keep material clear, available, clean, sharable, etc.

Lunch (where there’ll be some posters to explore) then Labs/practicals chaired by Marshall Dozier (this is where I may be at meetings and you may wish to switch to watching #elearninged) including:

 “Designing teaching spaces for the 21st century learner: The story of the nostalgic Dad and the horrified Son” – Victoria Dishon (School of Engineering), Stephen Dishon (IS Learning Spaces Technology)

DYNAMED: Student Led Development of a Dynamic Media Library for the R(D)SVS – Brian Mather and Rob Ward – (CMVM)

Experience with Cogbooks pilot on personalised learning. – Eduardo Serafin (Geosciences) and Mark Wetton (IS)

Offshoots and Outputs session chaired by Marshall Dozier:

CMC Vellore India partnership – online MSc in Family Medicine – Liz Grant (CMVM) and Jo Spiller (IS)

Digital tools for lighting education” – Ola Uduku and Gillian Treacy, (ECA)

Research, Teaching and Learning” – Michael Begg (IS)

 And I’m back… just in time for most of Sue Rigby’s talk… 

“Developing the Vision for 21st century learning” – Prof. Sue Rigby, VP Learning and Teaching

We have come up with a six point vision for where we want to go with learning and teaching. This has gone to every academic department, and to every support unit, within the university which we are bringing together our bottom up vision for learning and teaching. And I am going to talk about some of the ways that technology that will enable us to do… But this is about technology as enabler in learning and teaching, not just about use of technology.

1. A portfolio approach for an unpredictable future – making the most of the Scottish degree

That longevity of degrees can be a real benefit of our degrees – longer exposure for our students that benefits potential employers, novel approaches… But we want that portfolio of content to also reflect much more dynamic approaches to learning, a portfolio if learning styles.

2. Giving students agency to create their own learning – students at the centre, not degree programmes

This is about giving students the space physically and digitally to follow their own journeys, to craft their own narrative… They may do the same degree but have very different experiences… Every students experience are different but there are commonalities that matter here of skills, or experience. Things like the Wikipedia Editathon in ILW is about learning what makes a good Wikipedia entry, what warrants inclusions…

You also see things like one of our undergraduates working with the Girl Guides to explain physics and meterology to teenagers with common materials – and that reached many girl guides.

3. Extend learning beyond the traditional knowledge-centred course – e.g. international experiene, service learning, self-defined projects, entrepreneurship

As a scientist you can have a clear idea of the core of your skills and experience. By extending knowledge as undermining that centre, but as adding to that corona… So a colloquial example – chemistry students go on placement as students, but come back as chemists, actually doing their subject. And often that sort of experience isn’t in our course descriptions, and it matters that that is captured.

We also see students from civil engineering working on the rails – so they understand the work before supervising others. We have students giving TEDx talks – those presentation skills are hugely valuable.

And we can open up opportunities online, and our community online. And encourage and recognise that our students can be creative – students are sometimes more daring online than in our physical university spaces.

4. Every student a researcher or practitioner – joined at the hip to a research group from year 1, offered a higher degree place on attainment of a good degree

If we don’t do that, why should our students come here rather than to a teaching led institution? We need our research to be central to the learning and teaching practice…

So here we have a box of shells… Our student found a collection of old shells to exemplify evolution and the work of Charles Darwin… This was first class work but

5. Course design for 21st century learners – appropriate use of technology and student centred learning

Cue a plug for Fiona Hale’s Learning Design Project, which will clarify the requirements, both for IS and University partners, for learning spaces and technologies.

An example to share here – the Vet students are contributing to a virtual anatomy museum… you can help to break the boundaries of the university, and of what we share, and

6. Focus on multiple learning styles and learning for life – at least one online course taken by all students, explicit reflection on learning style and capacity

And that’s starting with Dave’s sustainability module, and an online big data module. And there will be more. But we also have our MOOCs… and we can start about aggregating MOOCs into our existing courses, by using them as learning objects, or to be used in credit bearing units.

So, I wanted to give you a context… What I would suggest is that we have to experiment for a while. When we find things that work, we have to bring them into the mainstream. We’ve been good at experimenting. I think we can be even quicker and even bolder, but also bring this into the mainstream!

Q&A

Q1) Do you really think that large scale face to face teaching is entirely dead in the future?

A1) No, but we should aim for it. And we can keep them when this is the best possible pedagogical model… At the moment it works the other way around…

Q1) How would you host an event like this without these big spaces?

A1) But all of us have started to give presentations at conferences that I am not attending – virtual presentations. If there is a sliding scale we are stuck at the lecture end… I’m saying push the other way… and then find the right place – probably in the middle… Flipped classrooms worth well

Q2) Student views on this?

A2) We had schools ask students. And also workshops through EUSA… If you give students questions, they want what they have… Often predicated on response of their schools… So more conservative schools create more conservative students… But if you preface questions with ideas and alternatives, students do present new ideas, they are interested in new approaches.

Q3) Our students come from very different backgrounds. Some will be really used to having some agency…

A3) We have a somewhat damned if you do, damned if you don’t situation… Some come in from high tech environments and our teaching looks comparatively old fashioned. Others come from very strict, hierachichal, traditional places and we have to move students along from that. So we have to scaffold students in induction, in programme design… Really careful induction I think. BUt at the moment we are already moving towards a place where our early years education at the University is probably more conservative than what our incoming students are used to from school…

Q4) We’ve talked about community a lot today. We have to understand the importance of a large lecture, networking, serendipitous meetings of people… And we have to understand how best we utilise and capture that.

A4) I agree with that… But we have to understand that as part of the purpose of the lecture. Student halls used to be about housing, with accidental communities. Over the last few years Pollock Halls have actively supported and encouraged the building of community… So if we want a lecture for that purpose, lets say it as that and that we use the time in that way… And make sure that that is what happens in those spaces.

Conference closing – Wilma Alexander, Convenor, eLearning@ed Forum

I just want to say some huge thank yous to all my colleagues on the elearning@ed committee… And I’d like to thank you all for coming and to all our speakers for there fantastic contributions to the day. And we now have time for you to meet each other, to explore the posters further, ask questions, etc.

And with that, I’m done blogging for the day. Remember that you can catch tweets from the sessions I couldn’t make on the hashtag from today, #elearninged.