 Hello and welcome to this webinar organized by SESDA. Our topic today is data on political behaviour. I am Jen Buckley and I work for the UK Data Service and I'll be joined by Hanesh Vards, a data processing specialist for the Comparative Study of Electoral Systems. She's based in Cologne at the GASIS Leeds Institute for the Social Sciences. We also have Alexia Katzenidou, who is Director of the Data Archive at GASIS and also an active researcher specializing in political behaviour and with an interest in European politics. Okay so here is the overview of the webinar. I'll start with a short background to SESDA and social science data services and then I'll give a broad overview of the types of data available and how to find and access the different data resources. I will then pass you over to Hannah who will guide us through the content and structure of data from the Comparative Study of Electoral Systems. Then we have Alexia who will give us the research of perspective looking at how we can and also might continue to address questions about citizen representation. There will be an opportunity to ask questions. You can ask questions by typing them into the question box on your webinar panel. This box is normally on the right of the screen. If you can't see it you may need to maximise the box by clicking on a red box with an arrow. You can type questions at any point but we will pick them all up at the end. So we've got an exciting programme. There'll be a lot of information but we'll provide a copy of the slides after the webinar. We've also put together two information sheets containing details about the data resources we mentioned. Therefore don't worry too much about catching all the details. We will also add a recording of the webinar to the CESDA website. So CESDA. CESDA is a consortium of European Social Science Data Archives and it aims to provide a research infrastructure to enable the research community to conduct high quality research. Key tasks underlying this vision include developing standards and best practices around the management and archiving of social science data and facilitating research access to important resources. The core operational bodies in CESDA are national data services. National data services provide access to extensive collections of data useful for social and economic research. You may be familiar with the service in your country or some of the larger data services such as GASIS or the UK data service. There are many national data services in Europe. So this image is from the CESDA website where you can find information and links to all the national data services in Europe. So fortunately there is an array of social and economic data useful for research and political behaviour. However the upshot of this availability is that we have a somewhat complicated data landscape with data located in many different places and available through different ways. However most of the individual elements as we hope sure are not too sort of challenging to navigate. Election studies are one of the richer sources of data available. So many countries have well established academically directed election studies with for example those in the Scandinavian countries, Britain, Germany and the Netherlands all going back to at least the early 1970s. And they're largely designed for testing competing theories about electoral behaviour such as who won and why. So typically the core part of these studies is a cross-sectional survey carried out a few months after the election. And this survey is usually based on probability samples and includes varied questions about the election vote choice as well as political orientations and attitudes. Many national election studies also include other elements. So for example some studies include a panel element collecting information from the same individuals in waves either across the election period and or across different elections. There are also a number of completed and ongoing projects producing comparative election data. Most notably is a comparative study of electoral systems which Hannah will discuss. And here's an example. So this is the British election study and I have to admit partly because we have an election tomorrow and also because it's one of the longest running in Europe. So the webpage shows up for the 2015 electionist cross-sectional data information collected during the election campaign and internet panel and expert survey data collected from Twitter. And many of these also go down to cover the recent EU referendum. So data from election studies is sometimes distributed by a dedicated website such as this one and or through national data services. Though access points can also vary across elections often because the organizing team or institution changes. And the guide that we've put together will include some names and links for individual countries. So social surveys are one of the major sources of data and political behavior. And European countries have featured in many of the seminal surveys of political behavior such as almond and verbose civic culture study or the political action study by Barnes and Cass and more recently the citizenship involvement in democracy study. And you can access data from all of these studies and still through the GACES data archive. A source of more recent data can be found in the major cross-national surveys such as the European social survey and international social survey program. So these both have questions about electoral participation and party affiliation and their core questionnaire with further questions found in the thematic modules. There's also the long-standing Europe parameter series and the European values study. So access to some of the cross-national studies can be through infrastructures and websites dedicated to distributing the data such as this one for the European social survey. These portals are a very useful resource of information about the survey content and variables alongside information about some of the substantive findings and some methodological discussions. But access to some cross-national studies is via one of the national data services. So for example the international social survey program data is distributed by GACES and who also offers some useful web pages for exploring the survey content. So in total data services provide access to many thousands of data collections including data from major academic projects, government and outputs from smaller research projects. So some particularly relevant examples include the many general social surveys that occur in Europe such as the British social attitude survey and the German albus. These surveys tend to include a few questions on political behavior in their core questionnaire and then more in their less regular survey modules. There are a number of longitudinal studies in Europe which includes some questions about political participation and attitudes. So this includes some household panel studies such as the German social economic panel and the Swiss household panel and the UK data services also provide access to a number of cohort studies. Also although less common than quantitative data several European national data services give access to qualitative data such as in-depth interview transcripts, field notes and answers to open-ended questions. So all the national data services have websites with online catalogs for searching or browsing. The example on the screen is the catalog from Dan's in the Netherlands. These catalogs allow you to search using terms such as political behavior or to search for particular studies. As you can see here you can find information and links to the major national data services in Europe on the system website. Also an important announcement to make is that Cesar are in the process of developing a new catalog which is due to go into service in 2018 and when operationally you'll be able to search across all Cesar service providers. The final source of data to highlight is that there are a range of data resources relating to context and outcomes of political behavior. So for instance the supply side of politics there are data sets from expert surveys on the position of the political parties and data sets with coding from political party manifestos. For political outcomes and policy there are data sets containing information about parliament and government composition, policy agendas and social expenditure. Again these resources are all available from different sources and the guide you've put together includes some information for these key examples. When it comes to accessing data you'll find that some of the arrangements differ between the different sources but usually you should be able to download it from the respective website though often you will need to register and when doing so you will usually be agreeing to some terms and conditions. Common terms and conditions include things like trying not to identify individuals and not distributing data to others and restrictions about the use of data will also come in so most typically that data might be for non-commercial use only or the use in research or teaching only. And the final point in when using existing data it's good practice to cite the data so a data citation gives credit to the data creators and it also allows for the researchers to find the data. In general a citation should include enough information to locate the exact version of the data and you may find as well a recommended citation is given by the data provider. So at this point I'm going to pass you over to Hala to discuss the comparative study of electoral systems. Thank you Jen and welcome to the part of the presentation that will be about the comparative study of electoral systems and the exciting possibilities that we see in it for cross-national electoral research. First of all let me give you an overview of the project. So the CSCS is a decentralized and collaborative project. It's a network of independent post-election studies of different countries across the world. Our collaborators in these countries all agreed to implement the CSCS question battery into their own election studies which is a 10 to 15 minute questionnaire. And the data are then deposited with us the secretariat and we merged them into a single data set along with demographic district level and macro level variables. The researchers in our planning committee devise a new theme and a new questionnaire every five years and I will introduce you into these different modules shortly. The rationale for the CSCS when it was founded in 1994 was to connect national election studies to make electoral research more global or more specifically to enable cross-national comparative research. As there is a wide variation across countries in context factors such as electoral rules, systems of governance or lines of political conflict, we can study how these variations impact individual attitudes and behaviors and in this case especially voting behavior and turnout and this is the main research objective of the CSCS. And we enable this kind of research by not only making available the themed module variables so the individual level survey variables but also context variables containing information about the context in which individuals are nested. Namely data is available on two further levels on the electoral district level and on the country level. And I will go into more detail as to what these levels of data contain in a minute. For now I want to stress that this multi-level data structure allows researchers to conduct cross-national analysis as well as cross-level analysis. Let's look at the individual level data in more detail. These are our modules. There are certain questions which are common to all CSCS modules and they can also be used for longitudinal analysis. These variables are respondents turnout in the current as well as in the past elections, respondents vote choice, their satisfaction with democracy, their sentiment of political efficacy, whether and how attached respondents view to a party, respondents evaluation of parties on like-dislike scale, as well as their ratings of parties on a left-right scale. And furthermore a number of demographics have been consistently asked across modules among them age, gender and education. Please note that a comprehensive overview of all of our variables across all CSCS modules is provided on our website and you will find a link that takes you directly to a table containing this at the end of this presentation. So our first module was in the field between 1996 and 2001. Its central theme was system performance and let me give you some examples for the kind of research questions and topics that the module was designed for. So it's set up for research on constitutional and institutional effects on democratic performance. It also allows the investigation of social underpinnings of party systems and is suited also for research into attitudes towards parties' political institutions and the democratic process more broadly. For module one we finally had 39 election studies across 31 countries implemented. In the next step, let me give you some more concrete examples of variables that are contained in this module. For example, respondents were asked whether they think the last elections were conducted fairly. They were asked for their opinions about the legitimacy of politicians and political parties. In questions such as do you think politicians know what ordinary people think? Do political parties care what people think? And in your view are political parties necessary? Moving on to the second module, this was run between 2001 and 2006 and focuses specifically on accountability and representation. Central research themes for which the module was intended are dealing with whether elections are successful and holding governments accountable or whether they are successful as a means to represent citizens views. Furthermore, the module is set up for research into citizen engagement and participation and data for this module is available for a total of 41 elections across 38 countries. And let me again give you a flavor of what are the particular items in this module. For example, respondents were asked whether they were contacted by a political candidate or party during the campaign. Their political participation was closely assessed by asking whether they had persuaded other for a certain vote choice, whether they had participated in campaign activities themselves or participated in protest demonstrations. Also, respondents were asked how well they generally saw voters views to be represented in the elections. As accountability was central in this module, respondents were also asked to assess government performance on what they thought was the most important issue in these elections and also about the performance of the party that they had voted for previously. The next module was run between 2006 and 2011 and had its focus on electoral choices. And this module gives insights into how respondents perceived the variety and quality of political choices in an election and how their satisfaction also varied with these choices. It is suited for analyzing what could be consequences of perceived limited choice. For example, this could be looked at in terms of voter turnout, the emergence of new parties, or threats to democracy. And in this module, we had 50 election studies and finally across 41 participating countries. So the perception of having electoral choices was captured by several items. For example, asking respondents whether they perceived there to be differences between parties and candidates during the campaign. Also, respondents were asked if they considered voting for other parties and candidates, or whether there was maybe some parties and candidates that they would never vote for. Other items concerned respondents' perceptions of the most important issues facing the country over the last legislative period. And respondents were also asked to judge the competence of different parties in dealing with this exact issue that they perceived as most important. The module we are currently finishing up is a module 4, which was around between 2011 and 2016. And there's one final release of data outstanding for this module, which we expect to happen in 2018. Its theme is distribution of politics and social protection. Against the background of the global financial crisis, respondents were asked for their views on government expenditure and on redistribution, as well as their perceived job security. And the module also captures aspirations of upward mobility, and there's a battery that particularly measures respondents' wealth. There's a secondary theme to this module, which is mobilization in terms of campaign contact, also particularly via social media, and campaign participation. And so far, we have 38 election studies, but we expect to add another 10 or so for the final release. Some examples of particular variables for this module include attitudes to public expenditure in different areas, such as education, health, welfare or defense. Also, respondents were asked how they expect their standards of living to develop over the next 10 years. In terms of mobilization, in this module, respondents were asked whether they had been contacted by a party, but this time they also asked for the means of contact. And a similar question battery was implemented for mobilization by personal contact in the sense of whether an individual in the respondents' close surroundings had made an attempt to mobilize them. Let me talk a bit about the geographical coverage of the CSES. So this world map illustrates the coverage of module two. You can see that most of the so-called western world is covered, and additionally Russia, parts of Latin America and some Asian countries. But we are still underrepresented, especially in Africa and the Middle East. However, from this year on, we've had our first African country included, which is Kenya. But as this webinar focuses more specifically on Europe, I want to show you this table that will also be continued in the next slide that shows the coverage across different modules in European countries. I'm showing this to you, so maybe you can already spot a country that you are interested in here. And this is the continuation of the table. I want to mention that, however, you can find the disinformation about which election studies are included in the CSES across the different modules on our website, and the direct link to this will be given at the end of the presentation. What you can also note in this table is, and you should be aware of, is that countries are differently represented. Some participate consistently, and others only partially. Coming up is our module five. It has already gone into the field in some countries, and we expect the first advanced release of data for 2019. The theme is democracy divided, people, politicians, and the politics of populism. At the core of this module are a number of attitudinal questions, particularly regarding political elites and so-called out-groups, but also towards representative democracy and majority rule, as well as towards direct involvement of citizens in decision-making. Another theme of the module is corruption. As for some specific variables of this module, we have a political interest question, and a question on how widespread respondents perceive corruption to be in their country. And then there's numerous attitudinal statements targeted at political elites, for example, such as most politicians do not care about the people, or the people should make the most important decisions. And then there are those looking at attitudes towards minorities, such as minorities should adapt to customs and traditions of this country, or whether the will of the majority should always prevail, even over the rights of minorities. So all of this so far was about the micro-level survey variables, but I also want to briefly introduce the district and macro-level data to you. Firstly, district data is available in all five modules and for most countries. What we collect here are the number of seats contest in the district, the number of candidates or party lists that are running in the district, as well as the percent vote for the top six parties and the turnout per district. In module five, an additional district level variable is introduced, namely the size of the electorate. As for the macro-level variables, they are quite numerous, but I will mention some of them to give you an idea of the kind of data that is contained. First of all, there are system-level macro data, such as official national election results, voting rules, party characteristics, such as party family, or expert judgments on the most salient factors in the election. And another type of macro-level data that we offer are aggregate data from public sources, for example, the population total, or other indicators such as GDP growth or the human development index, which are also offered for three time points, so the election year and then two years prior to that. Generally, I should say that should there be any macro variables that are not in the CSCS dataset that you would like to use, it's rather easy to merge them onto our data. We have also recently increased our set of country identifier variables to facilitate the merging with macro data from other external sources. So after I have hopefully given you some insights into the contents of the CSCS, let me provide briefly two examples that illustrate the kind of research questions and analysis that our data enable. The first paper I want to quickly present is by Karpen Banducci and it looks at whether the election of women candidates in national legislatures influences the political engagement and efficacy of women. The authors use CSCS module two and a total of 35 countries. The research design is such that the outcome variable is on the individual level, it is the political engagement and the campaign activity of women. Their main predictor variable is a macro variable, namely the percentage of women in parliament. So they conduct both cross-national comparison and a multi-level analysis which go together in these kinds of contexts because the variation at the country level is captured by the upper level of the multi-level model. Karpen Banducci's results show that women are not mobilized by the fact that there are women representatives but that the percentage of women in parliament is correlated positively with positive evaluations of the quality of the democratic process. Another paper that's using a similar design is the following one by Valls from 2008 and he looks at whether differences between the degree of globalization shape individuals' perceptions about whether politicians can make a difference. He uses data from modules one and two and includes a total of 72 elections across 40 countries in his analysis. In terms of design, this is again kind of a typical CSCS research so to say in which the effect of a macro variable in this case trade dependence and financial integration on a micro-level variable which is here respondents' perception of responsible party government is assessed. And his results show that there appears to be no such link. Next let me give you a brief note on methodology of the survey and the CSCS philosophy of data. So it's important to note that our survey is a decentralized project and we don't just lack the kind of control over the design and implementation of the studies that centralized surveys have. However certain preconditions have to be fulfilled in order for the CSCS to accept studies into its dataset. In terms of sampling these are that they should be national samples of all age eligible citizens and that random sampling procedures have to be used at all stages. And we ask our collaborators for detailed documentation of their sampling procedures and we also make these documents publicly available for users. The requirement in terms of sample size is that no fewer than 1000 interviews per country should be conducted. The CSCS philosophy of data, I think it's important to get an impression about that because it can help you handle the dataset once you decided you might want to work with it. So we are of the opinion that imperfections of a study should not be hidden but rather highlighted and this not only enhances the credibility of the project but it enables our users to make well-informed decisions about how to handle the data and that can improve the quality of their resulting analysis. And our major aim is for our code book to note anything that the individual election studies that has a possible impact on analytical outcomes. So this results in large code books but then enables our users to decide how they want to use certain variables if they conduct a comparative research project and let me give you an illustration of the code book. This is what it looks like below the variable name at the top and the answer options. You can see this section with election study notes for the separate countries and years. So yeah this section I want to draw your particular attention to because here we note all possible deviations from CSCS norms that we know have occurred in the single election study and that might have an impact on how you analyze these data. So please look at these election study notes closely before doing your analysis as they might really change the way you treat or you record certain variables. Generally the CSCS data philosophy is that more documentation is better so that practitioners can make decisions in the context of their particular analysis. And next to the code book we also provide on our website further documentation such as the original and English language questionnaire as well as a macro report that contains information about the election results and electoral institutions and the report about the sampling design and data collection so kind of methods report. Our data are available free of charge and they can be downloaded from either the CSCS website or via the Gises data catalog and the data are archived at Gises and ICPSR and a full release is published every five years but however we publish advanced releases always containing new election studies on a yearly basis. And these last points are the important links that I pointed your attention to before leading you to a table of all election studies in the data and all variables across modules and you can check them on the slides that will be uploaded after the webinar. Last but not least I want to point out a handy tool to do some preliminary analysis that's provided by Gises. CSCS is available in the online analysis tool SACAT. If you register here for free you can look at variable descriptions, tabulations, but you can also already run correlations or regression analysis and directly export the results from the website. So that's it from my side and I hope I could give you some useful insights. Thank you Hannah that was great. I'm going to pass you over to Alexia now for the research perspective but just a reminder that if you have any questions for Hannah please type them into the questions box and then we'll pick them up at the end. Okay so Alexia. So hi from me as well I'm just trying to get my screen working there. Can you see the new slides? Yes that's perfect thank you. Thank you very much. So I'm here to give you the research perspective so I suppose I am pretty much in the same situation as our audience. I'm a researcher of political science and at the same time I'm wearing two hats because I'm also director of a data archive so I'm like the best of both worlds in any way. So what I'll do I will just tell you what I do in my research and then I will discuss a little bit about what needs I have as a researcher from data archives and from data in general. So I'm a political scientist focusing on political behavior just like you and I do so on a comparative perspective so what I'm trying to do is compare the same research question answered the same research question across Europe so European Union of 28 29 countries depending on a possible before Brexit could be my sample. So issues that I in my research questions can focus on public opinion for example especially now on environmental issues so like questions like what brings people become more environmentally friendly does for example the new Trump policies or withdrawal from from the Paris Agreement will have an effect on environmental issues environmental attitudes in the European Union but other other topics maybe the most close to my heart let's say is the overarching question of political conflict in Western Europe what do we understand from political conflict here we have the concept of representation behind it so individuals vote political parties and political parties govern in their turn if they are elected in Western Europe so party competition is one idea and individual voting behavior is the second idea the specific topics within this overarching topic that I focus on is your skepticism which is very trendy these days lack of trust so the drop the success of the slow drop of trust towards political institutions in Europe especially because of the crisis the radicalization of individuals that are going goes on also because of the crisis and of course the quality of representation now all of these questions have one thing in common in order to answer them we need data so what data do I use overall I need three types of data we have already heard from the previous presentation in this webinar that there are different types of data out there and the first source is individual level data individual level data are types of data that help us understand the behavior of individuals ideally they are representative samples so their surveys for example but these days because of the problematic nature of surveys we have to look elsewhere so why is the nature of service being becoming problematic you have noticed maybe that the population is becoming saturated and they don't answer they don't answer they choose not to pick up the phone they are not to be found on the phone or they just flat out refuse when they are asked to be part of a survey the alternatives that have been used so far is internet surveys for example that we know that a little bit of different quality in some countries they are better than in others depending on internet penetration in the population but we also have these other type of data that are more they are less representative they have they have a particular type of bias built in them these are the voting advice applications you have heard maybe of them in almost all European countries they have different names but there are also some that are have the same name across Europe especially for the European elections you have heard maybe of EU profiler and EU and I and EU Vox for the last European election and these are generally tools where individuals can go online and click through a set of questions and find themselves in a multi-dimensional policy space and find the proximity to specific other political parties in their area but also in other European countries so these data they have the positive thing that first of all they are not limited in 1000 participants per country but you can have thousands in the last European election we had almost close to a million for across Europe the second positive thing is that you can attract them across days because you can you can see the individuals who logged in and did the the survey in one particular day might be different to the ones that did it a week later and so you can capture for example campaign effects but they have also their limitations and their limitations and of course not to be forgotten they are self-selected so they are people who are actually interested in finding out their own position they are people normally younger more comfortable with the internet with new technologies they are higher educated so the average population they are male or tend to be more male than anything else and of course they tend to be slightly had more left wing so in the individual level data the problems that we have is the low response rate in normal surveys and then you can see also in the complete failure of some of exit polls of the predictions of correct results of elections that things we didn't expect to happen so this is a failure a methodological failure that comes also from the non answering but not responding to surveys or people actually lying when they respond to surveys and then in the positive side that we have new types of data collections that can can capture some of these problems solve some of these problems but they create others so in the individual level there are some choices to be made on the other side the next type of data that we use are party data party data actually are there to give us two types of information or even the three three types of information the first one is how big are the political parties or how important so this is easy to collect it's percentages of vote shares and seats in parliaments or participation in governments that already has been shown in previous presentations that can come from the databases like palgov and so on then we have the positioning of political parties so political parties have specific position on different policy dimensions and this can be collected through either manifestos or speeches coded speeches or even expert surveys or even candidate surveys so you have very different data collection mechanisms to capture this and third is the salience question to what extent a specific issue is salient for a specific political party and that until now has come from either service from experts so expert service or party manifestos party data until now have been tested and tested again and again and again and we know that the various sources give very different information and that's why one has to be very careful what to choose but overall they they are accepted as problems so let me just give you one small problem that you might have for example using candidate surveys you know the response rate is very very low so you have occasionally 10 until 15 percent of response rates and there is a slight bias of people who actually did not get elected because they have more time after the elections the third type that is the easiest let's say has to do with context data and context data are easier in a way because they we're in the happy position that the government does all the work for us so for example euro stats collects interesting information about the economic activity but also the the type of population the categories of population that we have in a country there are other institutions like the world bank the OECD and so on that collects context data also CSES as a survey also collects context data to support the survey collection but they also other surveys that do the same the most problematic thing about context data if you want is that you cannot find them all in one place and also sometimes we have problems with the access and the comparative comparability between countries but especially if this comes from statistical offices we can't do much it's a government decision and as long as the European so you the euro stats or European community actually forces harmonization we are good when in topics that is not enforced we have a problem context data are important on country level but also they become more and more important on a smaller scale and these are where the problem starts smaller scales can be regions can be electoral districts can be also very very small areas defined by geo referencing I will get into that a bit more and there the problem that we always have is that electoral districts or regions even change over time and they're overlapping and so it's in data collections happen actually with a very different district in mind so there is no easy way of harmonizing or actually knowing that you're actually referring to the same district with all the data that you're using moving on the question that I'm asking myself as what is the most exciting development in the data world for our political behavior research as I said one of the even new things is geo referencing now within you technologies we are much more able to georeference our data and that allows us to link so link information between for example individual level surveys and context data and that can happen through simply geo referencing while we are collecting the information on the individual like a survey but also if we actually collect data on individuals through their social media or social media activity so this is also the new types of data the big data if you want that come out the good thing that is happening as well is that there is a huge amount of data that is created by individuals themselves so we don't need to force them go after them in the houses and knock the doors and call them to actually capture their opinions or specific questions but people actually want to share much more data out there at the same time though these people even though they're willing to give this data to big corporations like Facebook and Twitter and so on they're not as willing to give us this data for research purposes and that is that is an issue and that brings me to the last point that actually talks about the limitations and the kinds of data that I think are vital for us and here we talk about research data research data are slightly different to data in general because they are the kind of data that are collected or can be used for research and here we have the problem that data that that we we see out there being produced are actually behind copyright walls or behind big problems of data security and there we have the the question that arises of how can we overcome these problems in order to be able to use this data properly on the one hand the role of the data archives becomes more important because they can bring us this data through a secure environments and and then we as researchers can go in a secure environment either where the data archive is so in a room or in a virtual room and actually have access to the data and we hope that this type of data will be more accurate because people will be actually because the data will be actually capturing the behavior of the people directly and not capturing the opinions of the people that actually they are giving them to us when we ask them few months about after the elections and and then the data will become more accurate and we will not be asking the question to ourselves are these voters lying do they actually remember right what they voted do they recall the correct the correct questions do they understand what we mean no it will be much more accurate because we will be capturing their actual behavior this was also allow us to have a better other type of context data the media data as you know media studies are much more complicated because somebody has to sit down and code them for weeks on end and media data can through technology can become much easier be coded through algorithms and can be connected directly to other types of data so data linking is the future and through that we can also achieve shorter embargo time so you know sometimes where a colleague colleague collects data and they say okay I collected this data and I need them at least for two years just my own research and then promises that he would give it to us or she will give it to us but actually forget so like five years later we still don't have the data well this is not it will not be the case once we overcome this problem and have open data more and one last thing that we actually need is a better training because at least I was very well trained in analyzing for example survey data or like what we call a two dimensional matrix data where you have your variable and your respondent and and I'm not very well trained to analyze Twitter data for example so what we need is better training that corresponds to analyzing these new types of data so we live in exciting times technologies on our side even though service might be facing a little bit of a problem so I believe that we are all in a situation that we can definitely access more data and analyze more interesting data than before so that's all from me I pass on to Jen. Thank you thank you that was really great at this point so there's now time for questions just again a reminder that to ask a question you need to type it into a box which is in your control panel if you can't see the question you might need to maximize the control panel so look for a box with a red arrow in it and if you have a question specifically for Hannah or Alexia or myself just perhaps include our name in the text so I can work at who it's for and also as I mentioned earlier and there are some supporting materials for this webinar these are now available for you in the handout part of the control panel so you'll be able to find the slides and also two information sheets there and we'll also be making the materials available on the website after the webinar and so I'll just turn to the question so we've got a question first of all for Hannah and this is does the CSES have individual level data on ethnicity and or nationality? Yes thanks Alina for that question we do have data on ethnicity as well as race so those are two variables that we have been including since module one and all throughout two module four and also five. My experience is that not all countries provide these demographics some do most of them provide only one of the two but also as you might know in some countries it's not even legal to ask for ethnicity so so there will be some missing countries as well so I really recommend you to to go to our SACA tool or else to the code book or or the data set and look whether the countries that that you're interested in have these variables there is another variable that might be of interest to you that we've included recently in module four which is the country of birth and also the year arrived in current country for people who indicated that their country of birth was not the country they reside in so these two variables are available only from module four. That's great thanks Helen. I just had two questions for you that are linked as you were there you could answer them so one is whether there's or how many questions are repeated across the modules if any that would sort of allow overtime analysis what sort of themes could be could be done over time. Yeah I've mentioned I'm sorry go ahead. Then just the other question was about whether there was any other plans to repeat particular modules. Let me answer the second question first that's easier because no modules will not be repeated what we how we construct our modules is that there are always open calls to any researcher or anybody who's interested to hand in proposals for new modules and so it does happen that certain questions certain topics overlap and then the same or similar questions will be asked but generally every five years completely new modules in place. However there are certain items not only vote choice and turnout which are at the core but also certain more items that have been so far repeated in all modules so that would also enable longitudinal analysis and that I've mentioned on one of the slides let me just tell them again so satisfaction with democracy is another one this is next to turnout and vote choice and political efficacy party attachment evaluations of parties on the like dislike scale and ideological assessments of parties on the left right scale and then certain demographics. Great thank you okay well I can ask Alex here a question if possible I was wondering whether you could tell me the the tools you mentioned such as the EU box I was wondering whether there was any examples of these kind of sort of policy positioning data tools but the data from that had been archived and made available for researchers of all. Yes so basically the 2009 European elections the tool was called EU profiler and this has been archived by GASIS and the 2014 election tool there were two tools actually there one was 2000 sorry that was a EU and I that has been already archived again by GASIS and EU box is now the process of being archived by GASIS and we we're still missing the the data for political parties positioning so everything should be found in the GASIS archive. And I just do you have any examples of the sort of work that's been done using these data sources and this is a new type of literature that is called literature on the VAEA so the voting advice applications and it has two main components so two main components one is the methodological one analyzing the the the limitations and the the possibilities of the data and the other type deals with either party positioning so how political parties move and what dimensionalities they offer in the European space and the other questions is matching between individuals and parties so representation questions quality of representation so to what extent the party you voted for corresponds to your true policy position so are we in other words if you want is are we as individuals voting according to our policy preferences or is there any other reasons that lead us to vote for specific parties for example preparing the leader or we wanted to avoid another political party so tactical voting and so on.