Learning Analytics

Video in TIB AV-Portal: Learning Analytics

Formal Metadata

Learning Analytics
Title of Series
CC Attribution 3.0 Germany:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor.
Release Date

Content Metadata

Subject Area
Uniform resource locator Integrated development environment Digitizing Digitizing Analytic set Planning Integrated development environment E-learning Digital signal Disk read-and-write head E-learning Form (programming)
Context awareness Gradient View (database) Source code Dimensional analysis Measurement Medical imaging Hypermedia Videoconferencing Series (mathematics) Multiplication Descriptive statistics Social class Physical system Algorithm Theory of relativity View (database) Bit Price index Measurement Type theory Process (computing) Prediction Quicksort Physical system Spacetime Social software Observational study Digitizing Lemma (mathematics) Mathematical analysis Streaming media Content (media) Power (physics) Wave packet Network topology Integrated development environment Traffic reporting Form (programming) Modal logic Context awareness Information Mathematical analysis Content (media) Dimensional analysis Analytic set Multilateration Cartesian coordinate system Limit (category theory) Word Integrated development environment Extension (kinesiology)
Analytic set Student's t-test Open set Group action Trigonometric functions Open set Field (computer science) Connected space Data model Integrated development environment Prediction Basis <Mathematik> Universe (mathematics) Physical system
Context awareness Group action Greatest element Theory of relativity Integrated development environment Endliche Modelltheorie Student's t-test Open set Social class Number Physical system
Point (geometry) Logical constant Statistics Observational study Feedback Multiplication sign Information systems Median Student's t-test Plastikkarte Mereology Theory Number Data model Architecture Frequency Sign (mathematics) Circle MiniDisc Design of experiments Physical system Area Dependent and independent variables Trail Software developer Projective plane Gradient Feedback Analytic set Basis <Mathematik> Maxima and minima Line (geometry) Cartesian coordinate system System call Open set Inclusion map Latent heat Integrated development environment Prediction Basis <Mathematik> Universe (mathematics) Interface (computing) Self-organization Website Quicksort Fingerprint Resultant
Web page Presentation of a group Observational study Multiplication sign Computer-generated imagery Visual system Price index Student's t-test Information retrieval Inverse problem Internetworking Forschungszentrum Rossendorf Videoconferencing Bus (computing) Information Search engine (computing) Source code Multiplication sign Slide rule Projective plane Operator (mathematics) Coma Berenices Line (geometry) Term (mathematics) Subject indexing Data mining Word Personal digital assistant Multimedia Key (cryptography) Forschungszentrum Rossendorf Diagram YouTube
Information retrieval Subject indexing Beta function Word Information Twin prime Projective plane Videoconferencing Right angle Quicksort Mereology Inverter (logic gate)
Server (computing) Group action Service (economics) Overhead (computing) Observational study Open source Information systems 1 (number) Student's t-test Open set Mereology Theory Number Wave Data model Roundness (object) Atomic number Different (Kate Ryan album) Process (computing) Endliche Modelltheorie Office suite Traffic reporting Position operator Amenable group Physical system Area Software developer Projective plane Moment (mathematics) Mathematical analysis Theory Analytic set Basis <Mathematik> Lattice (order) Netiquette Information privacy Mathematics Data management Voting Integrated development environment Universe (mathematics) Buffer solution Self-organization Normal (geometry) Logic gate
Word Multiplication sign Data storage device Self-organization Analytic set Student's t-test
Information Finitary relation Analytic set Control flow
Area Service (economics) Analytic set Bit System call
Information Archaeological field survey Analytic set Self-organization Student's t-test Mereology Abstraction
before I start many of you know something about landing analytics sense location so I hope you
will not get bored my my plan is that you like an overview of learning analytics give
you some examples and some challenges and since I am the last speaker I try to vary of
as much as I can so you can your cold drinks but you're probably not surprise that's so you're living in digitized aged and is the digitalization also showed up and in the classroom in form of digital learning new environment online class from all meant to be a true reality and in my
mind this is 1 important impact and the impact these that now using all the data what we can collect in class from we consume into the learner instead of dealing with the average learn of which has happened in the Pearl past 2 centuries so we will we use this average learner to to build an education system around now we can really concentrate on individual and this is this is a great thing we should really embraces and I think this is where the impact of learning analytics space definition was there's a very good definition from solar which is a society of learning analytics it's called it is the measurement collection analysis and reporting of data about lemmas and their context for purposes of understanding and optimizing the learning and the environment in which it appears I think it's a great definition I once small problem with it and that's the way it was pointed in optimizing so I don't like the word optimizing I prefinal on very positive person so I would improve learning but in I can I can also accept that in some environments you you need to optimize instead of that learning analytics basically has 4 layers and the 1st layer and the basic description leaders to crucial to understand and describe what's going on and then the next later it's about diagnoses so what is this and why is it there what can we do about it and most of the applications right now are at these 2 layers the 3rd 1 is predict so we can we can use this technology to predict what's going to happen in the future try to predict and then the last but not the least we can try to recommend learners what to do and in the future in relation to their studies and so what are the critical dimensions of learning analytics 1 of the important thing is that you should I always put this technology into some sort of didactical pedagogical context and when you sort of defined this context then you can start thinking about the objectives like white why do you want to use analytics what do you want to analyze and you have to think about the stakeholders like who should be involved in this analytical process then 1 has to consider internal and external limitations in a constant of of the process also the instruments what are available in the series The algorithms other technology what you want to use and last but not at least think of the data so learning analytics is
very contextual so it's very hard to to say what data is available because it could be nothing or everything really depends on the learning contexts context but what type of data we can use and we can use of course performance data these upgrades assignments Wikinews behavioral data like clicks clicks streams and content views both video and more power or document use social media we can think of using multimodal data which can be forms of videos images text physiological data there's a lot of research on that these things geographical data and can last but not the least we should think about the context and start collecting data about the learning context and in my previous work I road use quite a bit of vacancy data always of us trying to put that analytics in learning antics in the concept of workplace that training will sedate of was very rich a source of information but you can think of search data or economic in indicators so so this was a definition for so now we can move
on to landing and a Dixon connection that with a few examples so the higher
education the holy grail of learning analytics is student retention so hard to keep students in the schools and really worth looking at the UK the United Kingdom which is 1 of the country's leading this field besides the US and the Australia and in the united United Kingdom the Open University has a system called whole you analyze this system is actually very nice because it can predict which students are at risk based on their richer learning environment and graphical data and also explains why idea at risks and gave activities fathers that will like it looks like this this is the student therefore here you see the student activity in the world to a learning on the
environment than here they put their student in context in relation to other students in his or her class here they gave grading sulfur his or her assignments
and here in the bottom there are a number of actions a number of recommendations for the students the take and action It's a very nice system and it also open so that they're
all the data is open so if you're interested in learning and its research I can't really recommend you to visit the website and download the data and play with the other 1 that what they we have recently learned from minimum you in Germany but I already know a very good example of a simple but the efficient learning analytics is coming from Stuttgart has formed the Horschel again median and they have a very simple methods to predict truck out with 96 comma decimal 8 per cent accuracy which is very high so the highest I have ever seen and they do it on the basis of very simple statistics so they look at ECTS point they look at failed exams I think in the 1st few months of the of the studies was a look at historical data what the students have to submit when they apply to this higher education institutions this is the result of it is really impressive so if you are into Dropbox research again I seriously recommend to look at this so if the Germany I go back to England and in England there is an organization called just can just is love like responsible of to develop ICT solutions for universities and they started a learning analytics analytics project they develop a learning learning antics infrastructure for the UK universities right now at least 50 universities are part of this initiative and it's growing there they invest a lot and into this infrastructure which is physically available all users all stakeholders of the of the UK higher education it has 2 major complement 1 the green 1 is basically responsible for collecting and tracing students in their original learning environments then there is an analytic school this this is where actually the analytics happens and this analytic score is connected to a number of dollars for their sports like the student constants of his student L. at an information system student again this is a very successful initiative I here is the link up please have a look I think there's a lot to learn from this and
now I jump to the ideas activities of a show to projects what we are working on 1 is actually my project and it's about I was sitting in education so corseting theory and methodology of the very powerful basically they say that it someone said schools that will almost automatically improve performance without actually measuring by the course if you if you start measuring your achievement to what's goal it's even better but it interestingly this gossiping theory is hardly ever used in education and read the article it's in on Saddam and and the University of Western Australia we were wondering about what ways that is strange and such a powerful theory and is not used in educational readily used in education we have to let's let's 1st about some experiment soviet design the goal setting application of both for higher education and for MOOCs and we made students the circles and get feedback on Goals share their balls publicly with others and we made teachers to grade those course so there was some sort of quality assurance system behind it and once it was done we connected to a discourse thing that through the Leningrad cost or of the university to see and to compare how students the so basically lines through the period of this experiment so that the brain uses all the sentences that in the higher education settings that amounts of them this experiment failed completely and so it was it was really a disaster and what happens over the researcher if there is a disaster in the she does so should was a qualitative inquiry like other what happened why why why did why did this happen so we did interviews with the students and we found out that 1 they did not accept calls again surprising for a 20 21 you old but the 2nd surprising fact was that they didn't even want to settle they said we don't need this why why do you want to do this with me is just extra work I come to the university the university's here to tell me what to do and I was like whoa this is very scary because once you leave the university you will need to be able to set calls but you don't know it and you don't want to know if this is this is restrained luckily into Mochrie area at that data we have and analyzing it looks different but I think this is of sort of warning sign for higher education that maybe we shouldn't tell the students all the time what to do
the other project is is not mine it's a it's a colleague of mine who couldn't be here today are of and it's about surges learning and he does this project together with their loudness Institute the provision made into being in and also the asterisk Research Center and they want to scaffold the lending activities what line as to why this happened do searching what does that mean imagine that your and I think it happened with all of us the you're searching the Internet for a time what you did know in this case is inverted index and you get this the page with the relevant documents and you click on the presentation the you read about the inverted index and then suddenly a new concept come which called here the stop word but what is this stuff words no 1 is telling you about so you
move on with your search now you find video about inverted index you watch this video there is no words a bus stop words so what's going on here how can I how can the student organize his or her studies this is this is difficult luckily we already have some solutions so that the TID
above it and had the TID AB portal if you say search for the same word inverted index and you find the video and all our videos are
annotated and based on the concepts they find in the video so the learner can simply john to the right part of the video to the right section and get the information would be what this is great and this is what the project was to sort of of right and was with so now after the
examples over to the great challenges II listed here sweet challenges however I think there are are many many more so this that this is where subjective I think the the or the importance some people think meeting differently I and the first one
is no surprises pricing and attics physically wherever you go in the area of learning analytics on alone in its buffer and of someone was the what is the biggest problem of learning analytics the people was say this this highly political of course right now we have to GDPR everyone is discussing GDPR at the moment as it seems we have to us constants that from the server from the student so people started developing constant services think about the purpose of the study which is always very critical I'm still what's people discuss a lot is the openness of all atoms because oftentimes which learning environments which are not open source we don't know what do they do in the background and this is definitely a problem another problem and what we also faced at the University of onto them I worked before the ID was a very complex infrastructure so we did a study on Saddam and there were 6 this 3 different information systems containing learning related data some of them are isolated so some of them are connected but it's a huge mess if you if you want to do research for you want to design services based on the 63 systems because then comes into an apology part who owns the ones that system there's a lot of gatekeeper resistance and so and so forth for many people lost OK but since you want to do using the individual data is it actually etiquette to use that individual data but many other people say way at if we are not using individual data that we are reducing the quality of education is this an article issue there's a lot of debate in the community about this and if you think about this is also if people start asking giving constant on an individual basis at the end universities may find themselves in the position where they provide different services to every single individual this is a kind of an interesting it would be quite an interesting development and then comes to organizations there is an issue around organizational readiness so my colleagues in Australia developed this learning an attic sophistication model and basically most of the universities around the world are only at this stage a learning analytics venison experimentations there are very rare large-scale deployment projects so far but why is that I haven't had a good example so again announces on when we started working on on on the university of 1st from 1st learning analytics projects of some of the university we were wondering like all of you have about how to position a project like this who are our stakeholders the Gambia we researchers what we do we try to publish and would make research and publish somebody there stakeholder analysis with my master's student and that after a few rounds of report the real interviewing like 20 30 people at the university in different departments they came up with this and when he 1st showed up in my office with this I was like Who what is this at but had sexually very simple so this is the learning analytics project this is the overhead the university overhead all the different like layers of management this is the legal body is completely no disconnected from learning analytics and these other users again completely disconnected with the amenities right OK so altogether we established 23 stakeholder groups and know what to do so something university project with 23 different partners internally This is a nightmare I think everyone will have a voice at the university a vote I agree with this so we decided to again do some research work and use the normative stakeholder theory to classify stakeholders on 23 stakeholders and we wanted to start with the definitely stakeholders who follow with legitimacy and who had burgeoned claims on learning analytics so what do you think how many stakeholders did we classify as definitely stakeholders who says below 5 between 5 and 10 book attempt last emplacers the rest and you were here the rat altogether 14 14 stakeholder groups were classified as a definite which is a very high number and I think it shows quite clearly the importance of learning analytics in a university Of course we had a lot of problems with the project involving all these 14 and different organizations but I think no using what we learn
is maybe we should start educating people because most of these was of this internal organizations didn't even know that learning and its success right so it's very important for the organization to learn about analytics to to feel to touch the data and how to get around with this and the what amount
of time to escape it's over nearly half of you very clearly the search challenge is but basically we have to educate store words that are students the what skills which are radically different than what we are educating that is a more transversal skills and
lifelong learning related skills more of the learning would be more about information informal learning and I believe blending and analytics can contribute to lost to this agenda we will discuss this
maybe in in them coffee break the last
little bit of the last but not the least of the summary I think there's a lot of to expect from an analytics we we have access to Big Data for me that means that a research and maybe better individual learning and there are many opportunities and ideas in this area what we can hold but we are still lacking expertise and experience and that's why I you should
also go and try out for learning analytics yourself I give you 2 examples can you do that once we are organizing a conference call learning insulin
analytics conference and Hagadone announces on the 22nd 20 5 October please come and join the discussion is in the disciplinary so there will be practitioners there would be also policymakers and OS researchers it's a very nice conference I can I'm organizing if it could be on and the other 1 is at least a part of our is learning analytics risk mitigation survey so basically what we're doing here here we are trying to collect what risks all organizations see when they deploy learning analytics and how do they want to solve those 1st and made but make this risk that I mean cluster this risk analyze this was an offer it to the community so we can learn we can all learn from them and with this I would like to close if you're have a if you need more information on this subject I would be around after this book and probably tomorrow as
well and if you have any questions let me know thank you and and


  306 ms - page object


AV-Portal 3.20.1 (bea96f1033d39fbe77f82542458e108105398441)