 Just to chime in with what Jillian said about don't wait till your date is perfect. I Just wonder what your plan is for sharing that data. Are you going to be? Sharing the data that you're gathering and that you're generating as you go along rather than waiting till the end of each grant because You know, I'm sure there will be other researchers who'd love to To combine their data with some of the data that's that's being generated here I'm very happy to tell you that we have been sharing the data the data are available as a part of our data service and We Have been putting out calls for expressions of interest in which we list our sort of our data catalog I think the best way might be to send us an email so we can put you in our mailing list Perhaps there are other ways or as well by means of which we can ensure that this community Gets the news of the kinds of data we have and can apply on time so that they can access our data But it's very much available some of the data. I was talking about Phil Kauser from NHS National Services Scotland We didn't really talk much in any of these the speakers talks around the public and the public trust And there's a there's a view that particularly when it comes to public services that there's a trust deficit around the use of data And and obviously being able to routinely Share personal data is going to be critical to improving public services And we heard Roger talking about the example of the local authorities Which is something that that my organization's personal is involved with around trying to get linked health and social care data And that reticence to share is a reflection probably of organizations interpretation of that trust deficit I just wonder if any of the the panel members has a view On how we address the trust deficit around the sharing of personal data for public benefit This issue of trust is really important because I don't think it's as straightforward as simply trust There is research that would suggest that Most people trust their NHS To look after their data their medical records because they understand that Public bodies public servants do appreciate and understand that sensitive data particularly is personal That's blown out the water when the front page of the Herald talks about somebody just being fined 250,000 pounds for leaving medical records in a skip in a car park But if we I think look wider than only personal data and Look at how the public Generally value information We do an annual poll about access to information Rights to information awareness is really high in Scotland 85% of people know they have a right to ask But when we say Does it give you more confidence in decisions? actually Well over 70% of the public in Scotland say Access to information gives you more confidence. Now if you stop in the decisions of public bodies So if we think about building trust not just in relation to personal data But into responsible and good use of data Generally and then overlay the personal we understand the different risks on top of that I think we'd go a lot further and I'd also say that Key to all of this is communication Because I I think how something is communicated in terms of what's going to be done with it We saw that example with the care data Had it been communicated differently? I don't think we'd have been where we were So what I would say to data owners is Don't forget all those good things you've learned about good customer relations and how you build general trust And then apply it just a little bit better to data Just a couple things to add to that if I may I think there's there's three things And I've completely agreed with all that Rosemary said there For me there's something about building the expertise in knowledge of of of handling data And that's both from sort of the of the legal issues But also in ensuring that there's there are trusted systems that people can stick to for Handling that data securely considering the ethics Of hand of managing that data So we've got the principles of having safe people save research and safe data and To do that, but I think there's also something else which is as mentioned earlier. It's about ensuring that people know what data is about them have have about them under you know, this this is the Underpinning this is transparency for people and Transparency is a way to to engender that trust. So I think there's there's something about Ensuring people know what they do the God but also Engaging with the public at scale that we haven't done so far in what the benefits of Sharing data and the risks of not doing so And that you know well what we know that the vast majority of people Expect us to be sharing data already and it's like it spies when when they realize that we don't There are a significant You know not not insignificant my minority that that don't feel that way and I feel that So take taking Taking that head on and sort of explaining to people so that is not the it's not the media headlines that Rosemary Mention there that they get prominence. It's about Some of the great things that we talked about here and some of the Avoiding some of the bad things as well in data sharing so public engagement certainly there's a real issue about health data as to who feels most threatened by the issue of access to health data I think there's one debate about if you like patients And there's a different debate about doctors and medical staff And I think a lot of the the issues about access to medical data are actually more about the rights of the professionals And that they're actually about the rights of the individuals So I think we need to be careful about, you know, who's protecting what and on what basis because because these things get confused And that's that's something that really needs to be needs to be looked at I think the other thing is actually we need a proper honest debate about Privacy issues, you know, where government is actually honest about what it wants and what the benefits of sharing data are and The the privacy lobby is a bit more honest about what it what it's trying to achieve We've got a very false debate going on at the minute, you know, where everybody's kind of shouting at different ends We're not actually making the kind of progress we need to because I don't think there's a sufficiently realistic debate There are great benefits to be got from For example, if you take the management of management of care from having Social work data being shared with health data being shared with with other kinds of things with the proper safeguards for the individuals The reason that's not happening is not because Individuals are objecting. It's because the professionals in each of these silos are not sharing or feel threatened by the process of sharing So we need to work out where the real barriers are as opposed to the theoretical barriers and then I think There's also the sense in which government needs to say why it wants to do these kinds of things more explicitly and in a more Reason way than perhaps it's it tends to shy away from it because it's difficult But I think if we can get out there and make a realistic explanation and and and make sure that the appropriate Safeguards are put in place rather than, you know, sort of theoretical safeguards will actually get much much further And I think in in Scotland in particular We've got such data potential because of the way in which records have actually been constructed in the past It's a real source of potential competitive advantage for Scotland if we get this right I think it's something that if we can take forward a proper debate on it, it would be to the advantage of everybody Freedom of information environmental information, it's domestic Scottish law We have complete control over it Data protection, it's UK law Scotland does not have complete control over it We don't even have a seat at the table when it comes to European debates in relation to GDPR and things like that Now I'm going to know further than say you might want to give that food for thought in terms of If if your aim is policy if your aim is public sector improvement if your aim is the citizen of Scotland Are we at a disadvantage in Scotland because we're one removed from the lawmakers In terms of things that impact on how we use data Couple of comments as well. I think trust is a difficult one. I think it's difficult because First of all is very difficult to define and secondly It's not something I don't think we can address overnight or immediately trust Requires having system of laws Technological solutions to make sure that your bank account doesn't get hacked and so on and so forth It's not only about privacy. It's really about information security as well But I think I'm most passionate about Things relating to consumer awareness digital literacy the idea that definitions of digital literacy has to change because Everything is digital. You're producing that you're always constantly in a in a data environment now Where can you where are you not producing data if you're in a city like this or wherever? So I'm Creating the pipeline of rights and responsibilities and understanding Sort of the real world implications of people's behavior. I think all that is needed as well in order to sort of Start a culture of thinking about trust in a more open way So it's more than just laws and regulations all the stuff I was saying earlier about the big challenge of value chains, that's the easy bit compared with the So cultural legal Technological societal upheaval we need to actually come to sensible conclusions about that question and I think what Des was saying about what we need an open debate really start that but The steps we need to take to the point where an individual can understand what their Data is being used for and have some input to and control over that and balancing Individual preferences and needs versus communal needs of maybe sharing your medical data versus having a big collection of data to do research on that kind of question and Letting people educating people on what is being done and what They should be done and obviously completely up turning our legal system to support it Yeah, difficult to word start, but at least having an open debate about it. I think is what we need This is an area about citizenship as well. Most of the data around most of the administrative data is given freely by citizens in order to access a government service and I think most people who provide Information freely to internet services Realized that they're doing that in order to get a better service out of that out of that company They don't want Amazon to sell Your date or they don't want Amazon to sell data to big pharmaceutical companies They certainly don't want their health data sold to big pharma But giving data to government ought to be part of a Or giving data to government and allowing government to reuse it under very clear Codes of practice and also under very clear legal constraints ought to be part of the pact of of being a citizen and operating within society and This is much too broad an area for I think it's much too broad an area to to think about Now, but it is an important area where Educators we've heard the word education. I think spoken by everybody Educate educators are not picking up this challenge as part of helping people Understand especially younger people Understand where their rights are within a digital economy and a digital economy that is A slightly more confused by the role of government or Or the role of their data within government. I'm looking for for some more questions, please Matthew I think that was a great introduction to my question my next question actually talking about education I'm dr. Elena Karajanidu from Strathclyde University Department of Education specifically and thank you so much for that Introduction you're bringing in mind a lot of work that is done by educators in Scotland in the moment For example, you Scotland involving young children and older children young people and Discussing educating themselves and educating as adults about their rights and their obligations when it comes to the Internet So I'm hearing a lot of questions here about privacy and trust and nephics and the legal framework around it and The only fair to biases and we're talking about reliability and validity of data And I think it's quite telling that these questions are amongst the first questions to come up from the floor So my question is really linked to to what you were saying about reusing the data Only I will reward that if you don't mind the way. I like to think about it is not reusing data, but Repurposing data using data for a different purpose And I think that's quite important for me It means asking different questions about the same data and for different purposes and What I'd like to hear from our panel today is your take on this perhaps. What do you think is important? To consider what frameworks ethical legal etc are important when it comes to a repurposing data One example of this and it's a huge. It's a huge question. It has many aspects, but as an example say for example that I'm I represent a tobacco company and I'm using I want access to date data that has that have been gathered By researchers at a university to explore adolescence experiences with smoking and their habits as A tobacco company in getting access to public data like that. I might want to repurpose it So what I want what I would like To hear today is a little bit about you know the frameworks that you think are important when it comes to not just reusing Repurposing what I call repurposing data. Thank you Okay, thank you very much. I'm You're going to try and kick this one off. Well, I'm gonna say say one thing that which is Piece of work a couple of years ago was looking at trust in data and sort of asking the What are the factors that affected? Whether they were comfortable or not for they dated to be using different things and quite an underlying theme that came from that and from from some groups across Scotland was this the concept of data can be used for improving the public good and that That is sort of pretty much a driving force behind the way the the arrangements that we've got for sharing linking and reusing data in That we've got now got a public benefit and a privacy panel that is set up to To take judgments about specifically what is the public benefit of the particular proposition to to share in link data We've and so I you know think for me that is at the heart of it So in the particular situation that that you described with a tobacco company You know the I would say that the case for public benefit in that situation will be harder for them to Harder for them to make then it would be in in many other cases But so I think for me that's that that's the heart of it is is having that as a principle Someone else want to have a knock at this ethical Question go ahead just a quick one I mean I understand that the kind of examples you're thinking over are ones probably with significant personal data And and those are clearly tricky ones from a privacy point of view, but just to say in more general terms, I think Repurposing of data kind of assumes that there was only one purpose to start with and a lot of our most useful data collections were collected with a more general idea in mind, so say things like the the census you is it's designed to meet a huge number of different purposes and a lot of the time if we're looking at maximizing impact then the more repurposing and reuse we can support the better because There is a huge number of questions out there that might be helped by your bit of data, but you hadn't realized it when you collected it obviously we need to deal with kind of Disclosure controls and personal data and all that kind of stuff, but the in cases where Those risks are acceptable Then I think there's a kind of process and technology stuff around You documenting your data and how you collected it well enough that someone else can reuse it in a different context Reliably or nowhere whether it makes sense or not to do that. So yeah, the more repurposing the better within privacy limits, I would say People in the room are familiar with reuse regulations to 2015. Sorry. I sound like a lawyer. I'm not Did you know they exist most public bodies in Scotland are subject to reuse regulations 2015 and In a nutshell what that says is anything that's not personal data that's anonymous and that is accessible. I It's published. It's already made available. You can get it through FOI There is an automatic right to reuse it for a different purpose for which it was created So I would say the first challenge for public bodies in Scotland if we want to leverage Value out of our data is to try and make as much of it anonymized as we possibly can And if you're going to do that it ties in with something Roger said earlier about Design that in from your collection and the way you hold it and manage it to start with There's a different debate about Reusing personal data, and I think that's a more ethical sense and within a legal framework As well and that is about this is about ownership and consent and control Because I'd say the thing that underpins lack of trust is lack of Perception of control of our own data So we I think would have an opportunity with a new general data protection regulation When it comes to things like Consent owning your own data the number of times that you get are something Are we tick here if I can contact you about Christmas cards about this about this about this? Even a starting point for consent is Would you be willing for your first someone to contact you about personal data to be used for? Health research this research this research, and I think some of it is Being smart about giving confidence that we are taking Personal data and an individual's right to control it Seriously, and we put in place things to deal with it But in terms of some research the more we can make available without personal data in it even better Now I think that's I think that's a really important distinction, I mean I I've been doing some work on the ethics of the re you ethics in the reuse of reuse or repurposing of New and novel forms of data or big data and one of the things which I think we tend to forget is this distinction between the protection of people's rights, which is a legal Matter and the respect that we should also have for People, which is which is the ethical side of it and the respect tends to get forgotten when we're dealing with Things that are supposed to be absolute black and white in terms of the legal side I think if a big To but or if a tobacco company I'm full disclosure. I smoke But a tobacco company if they can make a case that using these data is in the public benefit Then yes, they should be allowed to use it. There should not be an ethical Reason for them not to use it But if they use it for for that purpose of public benefit only Then that's acceptable, but if they use it for a purpose, which is clearly not in the public benefit Then perhaps they shouldn't or not even perhaps they shouldn't but that's why the I'm sorry. I've forgotten what it's called the public benefit and privacy panel the national statistician in England and Wales has an ethics review group The the idea of the approved researcher under the Statistics and Registration Act allows for public benefit for a research proposal to be assessed before people can have access to Consented for research personal data But it's a it is a tricky question and there are no black and white There are no black and white answers your definition of public good or public benefit may end up being different from mine So it ends up being decisions by committee Who are trying to do the right thing in respecting what they believe? the the individual Would have consented to or needs to consent to I've got one more Got time for one more Question, I'd love to I'd love to hear somebody ask About data quality Von Nuss mentioned bias and in in data and We I'll ask I'll ask the panel if if you can make some remarks about the the integration of What I'll say is now pretty standard statistical evaluation of that of the quality of the data and How do we deal in terms of big date? It's not a direct question, but it's how do we deal with data that's missing? There is so much missing. There is so much that is imperfect Within some of the larger newer novel forms of data that we're dealing with Is there ever going to be a sort of a standard statistical type approach to dealing with this or is this going to be left in the hands of The young the young data scientists who who are all going to be miraculously trained somehow Yeah, I'll start with that one, but I'll start with just some Humor or lighting like light enlightenment If any of you don't or I haven't done I have a listen to Tim Harford's radio four program more or less where he Do we say blows holes and lots of statistics and reports and back to the tobacco company? Frankly if the tobacco company tell anyone in the room that smoking is good for you Then I'm not sure we would believe that particular conclusion But watch that and there's there's a fascinating organization in the US just now Fact-checking on the politicians statements in the presidential election And you can actually send in quotes from Trump or Clinton and they will rate them in terms of from absolute truth to liar liar pants on fire And it is really quite entertaining and they've published information In fact neither of them get above the halfway line from the liar liar pants on fire levels, so I Think data quality is an issue. There is absolute holes in data and it's back to that point I mean you the year but data will never be perfect start where you are what is is very encouraging and Some of the work that we are doing with academics is that there is more and more research and understanding and new models and new techniques that handle missing data and correct data to I guess Those new techniques those new capabilities those those new types of deep deep learning networks, etc Cognitive vi all of those things are actually Helping in in the area where the data isn't perfect But but those new capabilities that that new research is being applied in ways that that you can still defer Insight and and hopefully drive impact from from data that is imperfect Thank you With big data There is all you know, there's this new mantra of doing analysis, which is data driven modeling and Of course, the so-called old world was that it was research driven modeling I don't know what the differences really are. I think they're all very fuzzy And and and the idea that and you know people get really upset about this, right? I mean data driven modeling and research driven modeling But I think this I think if you look at if you start from a set of If you start from a point of view that you want to address some complex problem and You need different types of data to answer that problem No, like you said, no data are perfect. I've spent year Years in some cases fixing problems of case deletion in gold standard social science survey where With the people who didn't respond because they're not representative and blah blah blah, you know You can't do anything very much with it and so on so forth. We know this this stuff is existed forever So I think they're really in a sense We started from that world in the social sciences and then we went through a very data driven world where we pull together all this very imperfect sources of data supposedly and Then now we are I think going back full circle in a sense that we are going back to Answering the questions starting from a phenomena and then thinking about how do you bring all these forms of data? perfect or imperfect and tell Stitch it all together to tell you a story I think that's going to be important not Individual forms of data because individual forms of data are always going to be problematic When I hear people talk about this and I do a lot When I hear about the distinction between Research driven and data driven the research driven model always results in in research Whereas the data driven model tends to End up with producing insight and it made that may simply be a semantic difference But it does tend towards two cultures one which is more academic and one which is more Business and the insight world has spread from business Into the third sector into government much more than it's spread into academia Academia seems to still prefer research But you need both yes And they are in fact the same thing. I mean insight is a product of research And research delivers an insight