 Rhaid i chi'n ddweud digwydd. Rwy'n gweithio, fel yma, ydw i'n bwysig i phawddol. Rydw i'n dweud i ddweud ysgrifenni ddweud arall, ac mewn ffordd o'r arfod ychydig o'r pethau i ffrif. Doeddodd yr osmell, fyddwn i'r archif, dw i ewa'r unrhyw oedd y mi? Mae oherwydd bob yma hi'n credu yn dweud yng Ndrwm 11 o flotu. sefydliadau, os os ydych chi'n credu fel y misgafellau a'r gwirioneddauォnau'n gwybodaeth thymianne i'r wneud ei ffordd o ran ymdeg yma yn y ffordd o'i sut yw'r hynny'n gwybod i'ch cynnig ymddangos iawn â'r ddodol? There is a link to that. Almost training a case that is essentially at all the compound or campaign messages. And another community is somewhat allies with the GBH information Alex has shown that it might be an Unfortunately Llemenon licence to proceed to my part, but my perception of the kind of people who do gave a seen based campaign emails i'n gwneud ychydig gan y dorg i Sorrypuc. Y rhaid i chi'n meddwl i'r pethau a gweithio'r unfodol yn ddod o'r dorg i'n meddwl i'r dorg i'r ffradig i hynny, lle o'r ffradig sy'n meddwl i'r dorg. Felly mae'r 영hynghau cynlluniaeth â'r dorg akwn o'r dorg a'i ddechrau'n holl daeth a ddaeth ac yn ddaeth sy'n meddwl i'r prif 말fil ac yn mynd i'n meddwl i'ch ddorg i'r ffradig sydd gyda'r dorg. I didn't go to all 32 lower authority areas. Couldn't do it like that. You're right there seems to be a contradiction in terms of the more positive slant from the survey to the more negative slant in the interviews. I must have maybe in the presentation I gave maybe the negative bits from the interviews a slight emphasis that I should have done and that I did in the paper to maybe to raise the issues that I see with right to them. The question about the fact that there seem to be lots of copy and paste emails from not personal staff that went into the contacts of the email to representatives. You're right. It seems to be a tool for the middle classes to drive home a message from organisations they are associated with. Cancerists told me and so did the MSPs that much of the contact as you said is by a normal email and often I mean it is the phone and it is the letter, it is the surgery, the one and one meeting in which they see people who are from a modified background. I'm hearing anything on those just looked up the distribution for right to MSPs. The deprivation is as expected from the national picture so roughly speaking more people are right in Scotland from better off areas and worse off areas so roughly speaking that does transport down to Scotland quite nicely. So I was surprised to be mentioned much of the content that people are doing about the numbers and the time. I was not mention of the content that we have if we can reflect a bit about 12 people writing and writing to the MPs and especially if you've done this in point of analysis. We thought and we can talk about the different directions that we're engaging. Related to that, what is it all that you think about this platform of writing to your MPs? How much is complementary and how much is substitute for the petition process that we seem to know? Other writing about things that are different from what you just see when people have said about this in the news and the rest in the ad and in the petition. The third question is related to all that. Have you looked at how the type of messaging and all the type of messaging and the intensity of messaging might find the safety of the leftful cycle or how to run with big political events. How about this is where they don't really have to talk to you about it? On content, we have essentially write to them as quite privacy sensitive because obviously these are people's messages that are representative, they can be quite... Is that one? Is that one? Hello? Hello? Hello? Yes, on content, write to them as quite privacy sensitive in that the content this message is quite sensitive. So the policy is it's retained for several weeks for legal reasons if we need to follow up with a complaint and then the content is wiped from the database. So we're quite generally reluctant to analyse the content directly. We sort of have rough plans for something in the future to maybe do surveys to them asking about the content or to approach it the other way around and to analyse what the content that's representative say they are receiving in terms of certain inter-broad themes. Obviously there is great research in analysing the content but the risks to the perception of the service that we don't want people... We feel even if we took certain measures to reassure ourselves of the privacy implications, if we publish things that even suggest we are reading people messages, that would be quite damaging to the service even if we've taken safeguards for it. So in general we're incredibly reluctant to approach that kind of thing, interesting as it would be. Sorry, I've gotten that on type of messaging. What did you mean by that? Sorry, timing. Yes, I've done some very basic stuff looking at this, nothing really systematic yet. I was sort of looking at the people who received the most messages and trying to see if there were patterns to that and you can roughly... I certainly saw some spikes. We do try to work out why they existed. We think there was a spike just prior to the submission of article 50 to MPs. There's a few others that sort of make sense roughly in a Brexit timescale but these of course be certain MPs will get messages from different kinds of people but you can roughly see time effects in it. We haven't looked into that much detail yet but they are there to be found. That's entirely... The trouble would be a pre-election spike is to be turned off before elections because especially MPs don't become MPs. Sorry, MPs stop being MPs in the run-up to an election so there's less grounds for that kind of spike. But certainly that kind of thing we would probably expect like a post-election spike or a political event spike. So I think it's there to be found, yes. As in action by the designers of the website? Not especially. I mean this is sort of the ongoing question in our research stuff is how much of it can we relate back to improvements to the site. So for instance in the other session with some of you may or may not have been and I talked about an A-B test where we did change the design of the website to affect gender and found absolutely nothing. And similarly if the problem is as you found that there is problems upstream that it's hard to see what if anything we could change the design to do. This is a certain thing so looking at some of the co-brands we run with local councils restrict the type of reporting and when that happens like there's one I can't remember the example that is mostly interested directly in road problems and when you analyse that data it's almost entirely male as you would expect from your road versus walking thing. So absolutely yes there is sort of the selection of categories does affect what's going on there and a lot of it will just get lost in the miscategory I sort of assign it. Because the other problem analysing this data you probably found is that we don't have universal names for things. We reflect council names and so you have to reinvent every time. So essentially there will be content here and it will be lost in the description or the name that will be better able to pull this apart. And you're right if we approach Fix My Street essentially like 30 different sites for reporting different kinds of problems that would raise design questions and that would be interesting. So the answer is yes no we're not doing it but we should. I know with that so you guys feedback on the website as to what percentage of the issues that were raised have been addressed. So there's a table on that and I know that so it's the councils that incorporate the site into their council website that do the best. Because I guess they have some sort of investment in responding to it best. But I haven't looked into any sort of correlates of what else means that people respond to this or not. I think for yeah I think maybe you might be better placed to speak about the councils and their responses in terms of that. I would confess I don't know much about that aspect of it but certainly there has been some stuff. So there are a few papers about the New York system which do look at sort of which things getting reflected on based on the kinds of reports. So I've got a literature view somewhere I can say to you on that one. I think yes that does cover the answer to your question. The final question. I don't know if exactly what you're saying is where I think there's all the responses that these guys look at. I mean because I'll see this in just a moment to make it better. I don't know if it's a fascinating thing but today we saw, seeing what people, the differentiation side seems to be a little bit different from that. But they spoke the experiment's focus and it's going to happen that you can do it. But again people respond to the harm. So I'm happy to ask why are producers of the first generation male to know the standard difference? Now that would be a reason it's easier to find out actually that should be possible with what I've already done. There's certainly no gender difference in terms of responsiveness to MPs. So if I could break and re-run the same analysis and take my straight direct and answer the question. I suspect you probably wouldn't find that directly except in so much that some categories are. So the problem when gender is so batched up in the reporting categories, it's very hard to work out what is a gender effect and what is a reporting effect. Because some categories are easier to fix than others. And so if those have been reported more than one side than the other, then that would correspondingly change the gender answer. So it's hard to untangle. Not impossible but it's tricky. The main sponsor. And I mean to say that the identity of the person is in the message. But you think what's wrong with that? You think what's wrong with that? You think what's wrong with the type of report. But it's not sample data. So if we want to report more on a category which is normal, we'll find that. We'll just, you know, you couldn't find that. All right. Just time to say thank you very much to our three speakers.