[BLANK_AUDIO] So what learning analytics means to me is the use of data that you collect, about the events that occur in a learning environment. And how you can you use that data to guide your design decisions such that, the overall experience is improved. Learning environments can both be online or in class face to face traditional learning yes. So the data that I am collecting is basically or it could be summarized as: The interaction or all the interactions that occur in a learning environment. And by interaction what I mean could be students with the students, if they answer, each others' posts for a discussion forum. It could be the students with the material, so you detect when a student is using certain resource. Students and tutors it could be when a, a tutor asks a question in a forum and the students answer that or the other way around. So any sort of interaction that occurs in that environment, if it is technology mediated. Then, you typically gain access to recordings of the events that occur. So from my point of view, the way I see it, is the first step that a teacher should consider for using learning analytics is actually thinking, what kind of problem or aspect do you want to detect and act on, your learning environment. You need to have some sort of objectives. So for example, suppose is I want to make sure that my students don't drop out of my course. Or I want to make sure that they sustain their engagement all throughout the course. or I want to make sure that they, I don't know, they participate actively in team activities. So when you tackle one specific aspect, then you start again working backwards and trying to deduce first, what kind of actions would be helping me to achieve that. And what kind of data would give me an insight if those actions are working or not. So then, when you put everything together. If you have this scheme in place. You would have, data that is being collected. You're looking at the data and see how does that relate to your objective and then deciding what kind of actions or adjustments you need to deploy in the environment. Such that your objective or your outcome is achieved. [BLANK_AUDIO] So for example, sources of information that you can get directly from the students. You can ask them about the dedication they had, with some activities. Was it too intense? Was it not? What kind of activities did they end up solving, or not solving? Participating or not. You can ask them also about when do they get these activities done. How are they, how are they going about doing this? suppose, for example, that you want to foster teamwork. You can ask them and say, so do you do these alone? Do you do these with your team? You can even ask them directly. Is your team working perfectly? Would you change something on your team? So that type of information already offers you a view or an insight about what happens in terms of interactions and you can react on that. So, the variety of data sources you can have is, is huge. So you shouldn't be obsessed with getting electronic data only from online platforms. You again should be very creative and, and open-minded and say, what I need is information, what I need is insight about what happens in the environment. And anyway I can see to get that information is correct. Once you get that information then the delicate, part comes. Which is the analysis or the sense making. You have to make sense of that data you collected. So, if you have the number of times that a student's logged into the platform. Plus [SOUND] the number of times they post in the forum. You have to make sense of that. And you have to try to see if it gives you an insight on the level of engagement or not. And once you make sense out of that data, then, you, decide, what kind of adjustments you would like to deploy on your, on your environment. So for example suppose that, you get the data that, half of the students barely connect to the, platform or barely participate in the forum. To give you a simple example, what kind of action would you, [SOUND] decide. It could be something as simple as sending them an email saying by the way, it's been two three weeks in the course. I see that you haven't participated in the discussion forum. It is important for our course, and therefore I would like you to participate or tell me what kind of issues, or what kind of difficulties you're finding for participating. And that type of action could produce an effect in which you see either the student. That begins to participate or do they come back to you with some reasons by which the forum is not actually working the way it should which points you to another aspect you should go for improvement. So I think the crucial part is not so much capturing the data, but it is more like sense making and deciding what kind of action would you deploy. So a concrete example that I'm using in terms of learning analytics is detecting. So in my course the students are supposed to use certain amount of tools to perform certain tasks. Fairly procedural, but we give them those tools and some of the tools are optional. But we want them to get used to that type of environment. So one of the things that we are observing is the level of use of those tools, so how often do you come in contact with those tools. And what we have detected is a correlation between the set of tools that are used more often out of three or four, and academic achievement. So we see that certain students that do not use a specific tool within the portfolio. They correlate with low academic achievement. So, the actions that we derive from there is, try to lower the barrier for them to use that type of resource, providing additional support, tutorials, hands-on type of guides. Kind of like scaffold a little bit more. That type of activity using those tools such that they get exposed to that environment. And hopefully, they will translate into better academic achievements. Yes so another example is to provide the students with brief questions about certain topics that you plan to cover in the class. And you ask them to read before coming to class certain basic documentation. And, what did is we embedded the questions like next to the document. It's basically, you can, take, make no difference between the document and the questions. And the questions are grouped by topics. And then what we know before going to class is what kind of questions were answered more often. And, which one of those were answered incorrectly more often. And that informs us on how to approach the lecture. We can go there and try to tackle certain issues that we have identified previously that are difficult. Or that are, or students are struggling with coming to terms with that type of procedure or concept or topic. And therefore, we emphasize a little bit more on the lecture based on that. So my, my final comment about learning analytics is that it is a very promising area. But at the same time it comes with challenges. The promise is that by knowing exactly when it happens in detail in a learning environment, very likely we'll be in a much better position to improve it. But when you try to deploy it in reality, it requires a lot of multidisciplinary work a lot of people involved. It's not only something you can do on your own, only in your class. You will need support, technical support. You probably need some strategic view of the level of the institution. So it is a high potential, but also tricky implementation type of tradeoff for learning analytics. That's the way I see it. [BLANK_AUDIO]
