 Yn ymhwy yw wedi bod wedyn yn cilio'r diwrapeth felly ac wedi cael ei fod yn ei lidd yn holl o'r gwrth gwrdd! Rwy'n meddwl i'r domch, Cepham, yn ymdweud os ymdegwyr yng Nghymru o'r falch bwysig ond os ymlaen o'r colleg Brynhyr, yn mynd i'w ddiweddol am y ddylog i ddigon i yn eu gwbl cwmpun o'r tradd, d adolescenteth i gyd ynhawer am y cyfle yn gilydd! Cymru yn rhaid i chi rhywbeth i'r colleg Brynhyr. Roc천hurst Coleg is one of the most successful sixth form colleges in the UK. It also has links right out into the community.. ..providing access to learning for adult learners.. ..and a wide range of other students like apprentices.. ... investigated very, really looking forward to what you're going to say to us today. Thank you. Good morning folks. My name is Dominic Ch légman. I'm direction learner at Roc천hurst Coleg. I'm going to take you through the work we've been doing over the last year or so, implementing the exceptional student experience projects with IBM and Portal. I'm really, really excited to be here because it's been a very big part of what I've done over the last year or so, and we've actually gone live with a lot of this stuff in the last few days, so I'm really, really good to be able to talk in the present tense. Just to give you a bit of a flavour about the college to start with. So, we have about 8,000 learners all in all, about 1,000 apprentices, about 600 staff. However, our core business is sixth form. We have about 3,016 to 19 learners, primarily doing A-levels and vocational programmes, and that is really the heart of the college, and that's where we've focused everything that I'm going to talk about today, initially, although we've got some quite significant expansion plans over the next year or so. So, we are a successful college. However, we find ourselves like a lot of FE colleges with a number of challenges, and we set out on the road around what we've done over the last year, really because we knew and we understood that we were not in a position to be able to stand still. And we started to look at the ESE projects, really because we were looking to do something around exceptional learning, do something about changing the student experience. And it was attractive to us because of both the technologies that were in there, which we tried and trusted with IBM and Portal, but also very innovative stuff. And the real key thing for us is that within our catchment area, nobody is doing anything like what we are now doing. In terms of us, we do have a very strong issue. We've had a great one-off since 2004, but we do have quite a forward-thinking management team, leadership team, and that's why we've embarked on this journey. It's really a case of getting ahead of the competition rather than standing still and waiting for something to happen. So, what we have been doing is looking at three significant areas, and it's probably just worth contextualising a little bit about where we are. Brock College is right in the middle of the New Forest, and when I say it's in the middle of the New Forest, I mean it's right in the middle of the New Forest. So, we have a very wide catchment area to Southampton and the East, Bournemouth and Port Collabation in the West, north to Salisbury, and even we have students come across from the Isle of Wight. But the key thing is that almost all of those students have to pass somewhere else that they could study in order to come to us. So, we are in an incredibly competitive environment for sixth-form students. Furthermore, we are very close to the Hamshould Orsic border, and schools of sixth-form are very aggressive in terms of their growth. And the other thing that we know is that in the next five years, we will see a 12% drop in 16-year-olds coming out of schools within our catchment area. So, we knew we had to do something about it to maintain our position. And the other thing is that our financial landscape has changed significantly over the last five years. And we knew we had to do something to ensure that we are continually to be financially viable. So, we've embarked on the projects, and we are looking to make sure that we are able to not only address that 12% shortfall, but also look to grow our learner numbers. Retention is really, really key in the sector. And I'll talk a little bit more about what we're looking to achieve in that in a minute. It's really, really important for our financial health that we are retaining our students. And within the projects that we are doing, we're also looking for a significant upturn in students' satisfaction, the happiness factor, which we hope will, in turn, lead to a lot of additional students going and leaving us, progressing to wherever they need to progress with really, really high grades and really good outcomes. So, we're looking to something that's student-centric. Everything that we do is very much at the heart of the student, and that's really where we've ended up, in terms of the way we've chosen to implement the different technologies. So, we have run two projects over the last 12 months. We've gone from concept to implementation in about 14 or 15 months. One is a predictive analytics project, which is now live, and the second of the projects is our social media platform. So, if we talk about the predictive analytics first, we are looking to ensure that we can maintain our student body. Now, within the FE sector, we have something called a 42-day rule that those of you that are in the FE sector will know very, very well. But essentially, within the first 42 days of students joining us, they are non-students in terms of being countered towards retention and indeed towards funding. So, we were targeting very much on these 42 days where students may arrive with us. We can do the hard work around recruiting. They arrive with us, but they may well leave within those first 42 days. And for us, that means that we have no funding for them, and for them, it means that they've obviously arrived with us and had some kind of bad experience and chosen to leave again. So, we're looking to really address that through the analytics, and we've been able to make some significant changes on the back of it. The first thing we did is we sought to understand what we knew about our learners. And what we have done is we've actually built two analytics models. We've built what we call a start-of-term model, and then we have an intern model. And what the start-of-term model is seeking to do is to try and identify those learners who are most at risk of non-retention before they even start to learn with us. So, obviously, as they apply and they choose to become a student at the college, we have an amount of information about them. We know where they live, we know what their demographic information is, we know what their ethnicity is, some health background, we know where they've been at school, and we know what their academic background is. And so, we've used that to build up a picture of our learners that we can use before they've arrived so that we can identify which students, as they walk through the door on day one, do we need to make sure that we are looking after, make sure that we are doing additional things for and come up with strategies. Subsequent to that, we have an intern model which is looking at the behaviours that they start to exhibit once they are with us. So, as they start and they've arrived with us, then they will start doing things. They will start attending classes, they will start handing in work, and we're using that data to build up a picture. And what that will do is it will compare how they're doing now against the kind of picture that we thought of when they first arrived. So, we can ascertain, are their risk levels dropping, are their risk levels increasing? Now, we're live on this now. We went live on this literally in the last few days. We are very much at the beginning of our understanding of this. So, we have plans to make this very much more sophisticated over the next couple of years. So, we're starting to feed in relatively basic data at this point. So, attendance data, performance data, where we're looking to get to are things like, you know, how often are they interacting with our online learning platforms? Is that more or less than their peers? Does that appear to be affecting their grades? And we can, we hope, become much more sophisticated in terms of the way that we're understanding the risk of non-attention for our learners. We built models over a number of months earlier on in the year. And it's using IBM SPSS software as its base. And the way in which we've constructed the models is based on a regression model over five years of history. So, we've got quite detailed history of where our students have come from and how they've achieved as they've moved through the college with us. So, we've used that five years' worth of data to build and refine the models. The refinement, it's not one of those things that you can put data into, and it magically spits out numbers at the other end. It really is looking and understanding at the data, trying to work out what is significant, modelling, remodelling, and refining those models. When we finally got to a point where we were happy, what we did was we ran it on last year's data. And we have actually got an 87% success rate, which we're really, really happy about. So, effectively, what the system is doing is it's predicting the likelihood of non-retention with an 87% success rate. So, this is the model that we're using to go forward with now into next year. And what we're doing is we're using that data in a number of different ways. And the key and most significant thing that we're doing that's enabling us to do is it's enabling us to completely change the way that we handle interventions, completely change the way in which we deal with students that have begun to have problems. So, traditionally, our model was that if the first final problems were to occur with a student, it would typically be the lecturer that would deal with that. So, Jimmy hasn't handled his homework in, they'd sit down, we have an agreement system, they'd get set targets, you've got to get your work up to speed by the end of next week. End of next week comes, nothing happens, it then gets escalated to a pastoral manager who then takes on the case and deals with it. The difficulty with that is that often it's two, three, maybe four weeks down the track from the initial problem before any kind of intervention strategy at a higher level has started to take place. And in a 30-week programme, that's a significant amount of time. So, what we are using the data for is we're putting this data out to our curriculum managers, to our lecturers, and to our pastoral managers so that they understand who they need to know and who they need to look at very, very early on. So, what we've been able to do is completely shift the way in which we're handling those interactions. So, we're moving from a reactive to a proactive model. We've also redesigned the organisation on the back of that which has actually allowed us to take some cost out as well. So, we are starting to get those interactions much, much earlier than we have done before. So, we're using... We use the MIA system to do that. We're also putting the information into our student ILP system so the lecturers are able to see that. It's interesting just picking up on something that Rebecca said earlier. One of the things that we're doing is we're also putting the constituent parts of what's making their risk into that information. So, if I'm a lecturer and I see a student at a high risk, I can understand what factors are making the student at a high risk. And some of those will be environmental factors and some of those will be in performance factors. And it's that kind of information which will help them make decisions. And crucially, what we've also decided to do is we've also decided to put this information in front of the students. And we debated very long and hard on how we might do that because again, just picking up on something that was said earlier, what we wanted to make sure is well, how do we do this in such a way that we don't demotivate the students if they suddenly see that they're at very high risk and actually, you know, we only think you've got about a 20% chance of getting to the end here, then why should I bother? So, we've done that by using a dashboarding system which has got speed dials in it where regardless of their initial risk, every student starts off at the same level. Students may be at high risk because of factors that they can't control. They come from a particular area, a particular deprived area if they've got a particular background. That might just place them naturally at higher risk and that isn't their fault so we didn't want to pass that information on. So, what we've done is we've shrouded it in such a way that they can see if their risk levels are moving up or down without knowing what their initial risk level is. It's very, very early days on this but what it's starting to do which is what we hoped it would do is actually spark conversations. Students are curious. They're going, what's going on here? Why am I higher risk in biology than I am in computing? What it allows us to do is have conversations because if we can have conversations with students then it will allow us to find out if there are issues and it will allow us to take those all-important interventions. We are looking to move that into a more sophisticated regime later on next year where it will actually try and identify strategies automatically for the student. So it might be, hey, you know, you're higher risk in computing because you didn't turn up to class for three Thursdays in a row so we can actually push that information to them. We're not there yet but that's where we're looking to go to next year. So a second major project has been our social portal. So we've built a social learning platform that's based using IBM Connections technology and this has been a really, really interesting journey. These are very different but they're really interlinked in so many ways. This is an IBM Connections as a social learning platform which if you are a Facebook user any kind of social media platform linked in Twitter, you will know immediately how to use the platform. What I like about this and initially when I looked at it I would be the first to put my hand up and say I was initially a little bit sceptical about how we would use it but with six to nine months of being embroiled in it I've become a complete convert. Beauty of the system is it takes a lot of the familiar functions that we have within a other social media platform but it puts them into one system and that system is a controlled environment. So what effectively we have is a closed environment within the collage which allows us to interact with learners and with each other using the tools that we've now become very familiar with and it's proving to be a very powerful experience and we've seen tremendous uptake in the first few days as we've gone into implementation. So what we've done with the system is we've turned off a lot of our communications so it's becoming our primary communications tools so things like our student bulletin student information is going straight out through the tool it's going to status updates with the students and what it will allow us to do is really gather huge amounts of data about how students are interacting and will allow them to start to interact with other students and develop new periods in the same way that they do in the wider social space. It will give them a very personalised learning experience so a lot of connections is built around the concept of communities bit like Facebook groups and the communities will allow individual classes, curriculums enrichments to build those virtual communities and that allows students to pick and choose what they want to get involved in we've got some things that all students are involved in so we're using it for our pastoral system we're using it for our student bulletin we're using it to do things like travel but beyond that they can choose who they can connect with much the same as LinkedIn it will do things like make recommendations to you you can see how people are connected things like that you can find people that have similar interests to you the other thing about the system is there are mobile apps and there are push notifications with that so that's quite a powerful thing for us so we've started to see already in the first few days of implementation things like students posting information up on forum asking questions about initial homework other students are starting to reply to them great thing about it is you've got push notification it comes through in your pocket you've certainly found that you can't ignore and we're starting to see these interactions coming through what we're hoping to achieve with this is that by aligning the learning and the experience that they're getting at college with the experience that they're choosing at home they're going to become far more engaged and because they're far more engaged that will then push them on to greater success and we've gone live with this system we've been testing it over the last few months we introduced it to the staff in the latter part of the last academic year they've been doing their levels of preparation we went live with 3,000 students on Monday it's Thursday now one of the things that I have been massively surprised about is the sheer volume of activity that we've seen in the first few days both in terms of students interacting with each others and the staff getting to grips with the new tools that they've got which will allow them to interact with their students in different ways one of the things that I would say about that is that from a staff perspective it's been really really interesting naturally with social media you have those people who in these days they've probably actively chosen not to take part in social media in a wider context it's been a little bit more of a challenge with them helping them understand that this is a closed environment because there are some people that are very nervous however one of the benefits that we haven't quite anticipated was that actually a lot of the staff were really keen on it because it allowed them to actually retract back from using things like Facebook groups where there is an element of personal risk with running social media groups in the sort of wild west of Facebook and those so actually that's been an unplanned benefit I think from the students a little bit about the portal and IBM partnership we've worked very closely with both companies over the last year or so one of the things that it has helped us do one of the ways in which we've put together the contract was that it enabled us to build a bid to put in a new stem centre through the local enterprise partnership we were successful in that thanks to the work that we were doing here that's enabled us to get 3.7 million pounds worth of funding and there is a brand new stem centre being built on the back of the campus which is opening next September so in some ways everything that we've done around looking to benefit we've already got the benefit to that there are some other benefits as well some of which we are yet to explore fully one of which is that we have access to a very wide global network within the technologies that we've got but the other thing I would say about that is that we have two sister schools in China that we run and operate and they are already on board with this so they have gone live with that with this straight away so they are working on connections straight away and we can see the beginnings of our UK and Chinese starting to interact together we've just done our initial work on that it's going to be very, very exciting and the other thing is that we've been able to harness the expertise of IBM and Portal and they've actually come in and worked with our own students to help build their own skills so just to finish off a lot of what we're doing is really around putting at the centre of what we do we are trying to achieve an environment which will be exciting for the students we are trying to work on making sure that when they arrive that they want to come to us and that they are happy while they're here and we can spot those problems before that they are actually starting so that we can make sure that they have a positive journey student that I've just highlighted here it's really about their success one of the things we're really, really proud of in the college, regardless of our exam results and whatever is the level of progression that we get we have really, really positive progression through to universities through to apprenticeships increasingly through to the world of work so one of the things that I haven't talked about here is in the long term is harnessing our current group of students as alumni and connecting them back through connections so that over the course of the next few years we can actually start to build up a greater connection we're quite good at holding on to our students while they're at university and understanding what they're doing four or five years down the track when they leave uni they go into jobs we're not so good at harnessing that obviously we've got this huge student body who are doing very successful things so interesting worldwide opportunities as I've just talked about and we're hoping that the students will take on what we're doing to reach new goals in terms of higher goals that we don't have thank you very much I didn't realise that you had only gone live this week and so it's been a really hectic week for the whole college it's nice going to talk in the present tense right questions, comments colleagues what was it that you took away in terms of load on the teachers that made it possible for you to be able to see this as a benefit to them in doing their job are you talking about the the social media? yes, I'll be about it that's been an interesting that's a really good question one of the things that I didn't touch on is we already have quite a well developed model e-learning platform so when we were putting this in what we've actually done is we've co-branded the two systems so if you look at the two systems you can flip between them within one click and we were very clear on what we were looking for Moodle to do which was all the things that Moodle is very good at and all the things that Connections to do which is all the things that Connections there is some overlap in terms of functionality think like forums for example so we were very explicit about that in terms of the staff we solved it very much in terms of this is additional tools that you can use in able to change the way that you're doing things within your classroom we're quite fortunate in terms of our teaching body it's pretty good at taking on new things so that was the way that we did it so we've probably branded it and solved it as an extension to our e-learning platform hi Roger Emery ironically from Stampton Solent University first question can I invite you to come up and talk to the rest of our team at Solent at some point because this is really interesting it's a long way to come to meet each other but the second thing is just looking at the HE side of it because we're doing similar things and we're looking back at where our students are coming from in FE is do you have any plans or aspirations for this to prepare students for HE so predict further beyond their journey through Brockenhurst so that you can support them through Brockenhurst to be better at better is the right word but achieve better at HE in terms of our analytics it's one of those things that is really going to be a growing part of what we do our initial focus was we don't have a retention problem at the colleges you're probably aware of but we know we can do better and in particular within those first 42 days we know there are students that come and go so there is our initial focus and the reason that we went for that first is probably a financial one because that's where our greatest immediate financial payback will be if we keep another 25 students this year it's 100,000 pounds of funding that we wouldn't have had but we do have plans for additional models and one of those models will be upon effectively looking at predicted outcomes in terms of performance so possibly we will need to do more in terms of our base data before we can do that so obviously we know student comes from this particular school this particular area this particular set of GCSEs they went on to achieve X but I think before we do that we'll probably become more sophisticated so we'll be looking at other factors environmental factors how are they doing on their course in terms of performance are they interacting with our systems more or less than their peers how does that affect their performance because that's where you get that level of sophistication that says well actually you might want to go and make sure that you're reviewing your homework but actually they went back and they did this so that I think is where we're going to head it's probably two years out if I'm being honest so we need to understand what we've got and make sure we get really exciting Any more questions, comments? Yes we have a question up there if you could go up to serve something so anybody on this? I'll take one more question up Thank you I really enjoyed your presentation so thanks for that just a wee question about your you mentioned an accuracy rate in terms of your predictive model of 87% so if I've understood you correctly what that means is that 13 out of 100 students who perhaps are at risk are not being identified so what's your approach to dealing with those students and the second part or the corollary to that question is how do you deal with students who are identified as at risk by the system but who actually in reality are not at risk at all? That's two good questions Let me take the second one first One of the things that we have these two models so what we do know is that student aid is high risk as they walk into the college now a lot of that might be environmental so we pull in students from some quite deprived areas for example and that will have an effect on their level of risk but as you well know there are always exceptions that prove the rule and there are some really really good kids that you think are never going to achieve and they're the ones that just defyre and do it so we will be looking at them we've got light touch on this so we've done things like we've put those students with particular pastoral managers who we know are very good at looking after a particular type of student so in doing that they're getting the level of attention that they need what we should find is that as we run and start through the terminal models and we start to see performance data and attendance data and those kind of things their risk level should start to drop off because it's giving us a high level of confidence that they are able to achieve on the course so hopefully that's answered that question with regard to the 13% that's a really good question if I'm being honest the answer is I'm not quite sure yet however 87 this is the second time we've looked at analytics and we ran a project with a US company a couple of years ago which we never implemented because the accuracy level was 6% and it was literally I could have predicted better by throwing darts into a class list so we're a long long long way forward from the first time that we did this and what I would also hope is a similar answer to the same question which is as we get into term and we start gathering the performance data we should see because we're running this every couple of days at the moment so we should see risk levels for those students start to rise and obviously that will then start to give us lists of students who we may need to then look at and pay some attention to right and again I think the conversation will carry on into the break Dom thank you very much can I give you that on behalf of all thank you so we're moving into a break now until 12 o'clock but don't forget the robot wars it's about to happen just as you go