 Cynllun. Hy, rai ydych chi'n cael y cwnteg. Rwy'n mynd i'n ddweud y cyfraith o'r syniadau a'r cyfraith. Rydw i'n mynd i'r ysgolio ar y last 18 mlynedd fel gyda'i gofyn o'r gweithio'r gweithio'r gweithio i fi'n ddangos cyflwyno a'r cyflwyno'r cyflwyno'r gweithio, a'r cyflwyno'r cyflwyno'r cyflwyno ar y byddai'i gwellain o'r cyflwyno. Roedd yn hoff y wneud unig o'r data i'r ffordd yr oedwch yn gyfer o'r hyn yn drwg. Roeddwn ni'r uchydig i'r gwaith cyfraisi gweithio ar oedwch yn gyfer y bwysig ac mae'r gymryd hyn yn cyffredigog ni'n dda i'r ddatat o'r cael eu cyfrifio ar unig a wes i'r dddur i'r cyfrifio y dylau'r ddechrau. A ydych chi chi'n 10 oed yn gweithio ar hyn o'r ddweitrreid ac rhaid i'r dylau. Roeddwn i'n gymryd neu ynghylch gan gweithio ar y cyfrifio? is we asked them two questions. For Fix My Street we asked them if this is the first time reporting a problem, making a problem, and if it's the first time, and if the problem was fixed after about a month. So this sort of gives us information in two directions. One is what kind of person is making this request, is this someone who is already reporting lots or is this someone who is being activated by the service? And the other is sort of a measure of the responsiveness of the authority. So the first question sort of gets at the sort of old debate in sort of general the civic tech in the internet sphere is, does the technology mobilise new people to take advantage of it by lowering the costs of accident services, or does it reinforce people who already had greater access but now lowers the cost in general? And the answer we generally find by subdividing this data is of course it's both in different times. So to look at the, so in this question I'm not so concerned about the authority question, I'm going to look at just if a person has previously responded. So just to give some context on Fix My Street, this is the number of reports going on by year. So generally it's clearly a bit towards the present day but the early years are very, very insubstantial compared to the present day. So the next graph shows us how many for each year people report, sorry respondents to the survey said they had previously reported. So in the first few years we have massive amounts of people, 60% of people are saying they had previously made a report and then as the number of people increases we are getting more and more people who have never made a report get into the system and that slowly creeps back up. So in general I tend to ignore those first few years because they obviously stand out and say between 15, 60% of people have never made a report before. That's not just for Fix My Street but that's to their councillor in general, that's how the question is phrased. We can separately examine if they've used Fix My Street reports. So essentially roughly speaking half of the people in any year have not reported a problem to the council before. But this is sort of unevenly among categories. So certain categories are more likely to report it as a first time problem and others are more likely to come up by repeat users. So first looking at people who are... Yes. So similarly by sort of analysing which ones were more likely to be reported by first time people, parking problems, abandoned vehicles, dog graze, street lighting problems. So these are sort of things that people are likely to encounter and report once with parking standing out absolutely enormously. So that's like issues with parking being blocked, parking being inaccessible. That is the number one thing people who are reporting for the first time. And then often because of the power drop off never reporting again. The average across the site is about two reports and that is mostly a normal number of people are only reporting one thing ever and it will likely be one of these. The next thing is that certain things are reported quite serially by graffiti, litter, flytip and strictly. And these are people who are often reporting these quite a lot. So they sort of have it in there. One of the interesting things, the jobs at the data set is like there is a ability to report a miss bin collection. But actually look at the data. It's a small amount of users like reporting. So you're bin collection every week. If you report it when it's missing like five times a year over 10 years, it stands out in the data. But that's because you as an individual are reporting your bin being missing. A lot to that can stand out in the data quite a bit. So basically people who are new to the service, new to reporting problems act in quite a different way. So using the same sort of thing looking at right to them, we can sort of, in this case we can look at both questions. So the first was, so just to give you an example of the scope of right to them. It's hard to get a rough idea of how many people write to their MP in the UK. The Hansard Society sort of estimate in the last year 12% of the out of population wrote to their MP, which seems very high to me, but that's the number we have. And so working from that, write to them accounts for just under 1%-ish using incredibly rough numbers. So almost all people write to their MP are not doing it if we write to them, but enough are that it matches to an extent and we get a reasonably wide sample on it. So we asked them if their representative wrote back, and that gives us a measure of responsiveness on the representative, and they asked them if it's their first time writing, which gives us again a look at how they're approaching, if this is their first time engaging with the system, or if we're just capturing people who were already writing all the time and now migrating to digital ways. So one of the first things we can do is sort of break down this by different institutions that certain things are more likely to be first contacts. People writing to the Lords are very unlikely to be the first time they're corresponding with a representative. They tend to be people like 70% of those people have written to their MP before or written to a councillor before. These are engaged people by the time you get to writing to the Lords. On the other end of the spectrum, most people who write to their local councillor are contacting for the first time, and there's similarly sort of, no, that's the only one where that's the majority. Everybody else, most people have already contacted in some way with MPs and then the Lords being at the bottom of that thing, including roughly in there, the MEP, which sort of works in a strange way in the system. So the other way of looking at it is the other question, asking if your representative wrote back. No, I'm still looking at that. Sorry, I've missed the slide. So the other way is approaching it. Within those systems, there are different kinds of representatives. So the Scottish Parliament National Assembly for Wales and London Assembly get two kind of representatives. One elected directly through constituencies and the other elected via IPR lists. And these, it's a bit different behaviours in the, in terms of, in response in this communication. We've got a paper exploring this out shortly. Hopefully it's just been approved. But essentially people who are representing constituencies rather than areas and so are elected directly by voters rather than indirectly through parties are more responsive to communication. And in the area in the Scottish Parliament where we have enough data to isolate that from other factors. So we can written say it's not a factor of being part of the government. It's not a factor of various things about tenure of the representative that still stands out as a factor. So there is different kinds of representative are communicating with the constituents in different ways. And this isn't necessarily a judgment on that they're in a good or a bad job. It's simply that they see their role differently. Certainly in the Scottish Parliament, the regional representatives tend to sort of see their role as being more on policy, more on committees, and so de-emphasise the constituency nature of their work. So this sort of validates the aspects of the political system can be extracted from responsiveness, which is an angle you usually get. You have to usually start the other way around. Similarly, we can apply the idea of deriving gender from name to the right to them database. And we can extract roughly speaking demographics. Again, it's vague, but it's good enough that over a very large set of data we can make judgments on it. So when we look at the when people write to their MP, we can see this is roughly speaking the story we usually tell about civic tech. There is a balance, but it is a mostly male thing. And similarly, when we apply, because for writing them data, we have postcodes, we can map postcodes to the low super output areas and get measures of deprivation. And similarly there, generally speaking, there is a gradual increase in more deprived areas. There are less people writing to their MP in well-off areas. There are more. There is a spread. It's not completely one way or the other, but you can see the increase in the data. So here is a sort of weird thing. I haven't quite got a good theory on this one yet, but look on the hour of the day people are reporting, are writing to their MP and then splitting it by gender. You get this weird effect that women are simply more like less likely to write to their MP during the night, which means men are simply more likely to write to their MP during the night. Obviously most people are writing during the day. It's just an odd thing that popped out of the data. So not entirely sure what to make of that yet, but there is clearly a sort of this will speak to a certain pattern of people writing to their MP. And this will sort of be explained by that sort of difference in the two genders. These will in turn be breakable down into different kinds of behaviour when people want to contact their MP. When we look at the local councils though, there is far more equally balanced. So roughly speaking, this is still very slightly more male, but it's within the fuzziness of the gender analysis that roughly speaking we can say when people write to local councillors roughly speaking we get an equal number of men and women writing. And when we look at the demographic breakdown there are more people writing from more deprived areas when they write to their local council. And when we move away from the index multiple deprivation which is combining multiple different kinds of issues and look just at crime, there is an incredibly clear pattern going down once that this is the kind of thing. I mean obviously this maps against other issues as well, but roughly speaking this seems a quite clear pattern of not just that this is perhaps one of the issues people are writing to their representatives about. That is a leap beyond which the data can really support, but it's quite suggestive that people are concerned with their local areas and that's why they're writing. So this is generally speaking getting at some of the issues we're sort of finding with civic tech websites is that when we sort of survey our users and sort of find what kind of people are taking part in them, if we don't subdivide by use we might end up with the wrong answer. For instance if you write to them was just about writing to your local councillor we'd be very very happy in general with the demographics of the websites. The fact that you can also write to your MP skews it the other way and so we had to sort of think about the different uses, the different kind of patterns of use we can get out of this stuff and similarly to sort of backtrack that it's difficult to work out how to resolve some of these demographic issues. So for instance the fact that I don't fix my streets different genders report different kinds of problems. We sort of consider it as a bad thing that the site is not demographically balanced, but that sort of reflects that our different genders encountering different problems at different stages in their day and are they willing to encounter some problems more than others and if so the genderedness happens far before they ever reach the website and so what can we do to correct that problem at that stage? It's not entirely clear if there's an easy answer to that, it's a more general social problem that we then pick up as a result on several bits down the way. So this is just to talk a bit about how I'm approaching investigating this at the moment. So one of the things essentially I've been building a series of data exploration websites that allow me to sort of pour all this sort of demographic data in and then various other factors in to quickly tell me what areas are and are statistically significant to sort of get a bit of idea of where we should go do more quality of investigation of what's happening. So in this case this is looking at yes the different kinds of interactions in the variety of them data, I've got a similar thing for Fix My Street that covers a lot of the same ground as your talk and it's sort of a very nice way of being able to click around a lot and my hope for these is we published one last year looking at the RFOI website, what do they know and essentially the hope is to build these exploration sites to a quick report on what we're finding inside the data, what we're finding that's interesting and then release these sites alongside their reports so mostly so we can explain the weirdest stuff where the interpretation, the office interpretation might be slightly misleading and then have more of this data available publicly for people to draw conclusions from and further research to happen on and that's that, thank you.