 Good afternoon and welcome to our webinar, Education in the Long Run, which data are available. This is a webinar jointly organized by the UK Data Service in Celsius, the Center for Longitudinal Study, Information and User Support. My name is Bernd Lichtwald. I'm from the UK Data Service at the University of Essex and I'm very happy that today I will be joined by Wei Jun from Celsius at UCL. I will give the initial talk on data we hold at the UK Data Service on Education, Access and Support. And in the second part of the webinar, Wei will present on the ONS Longitudinal Study, Access and Research. But before we continue, I would like to make sure that you are all hearing okay. I will launch a poll to see whether everybody can hear okay. So you should now see on your screen. Can you hear us? Yes or no? Please can you vote? Okay, it's in progress. Okay. So I think there is no problem with the sound and we can continue. Thank you very much. Let's start with the first part of the webinar with introduction to data on education that the UK Data Service provides. I will give an overview of the data, access and support. But first of all, why a webinar on education? Education is a key factor in shaping an individual's life chances. It not only determines later employment chances, but also living and housing conditions, nutrition and health, participation in social life, friends and partner options. And finally, the educational opportunities and achievements of our children. Education is commonly regarded as the key to remaining competitive in a constantly changing globalised world, as it is the key to meet the demands of a knowledge-based economy. The UK Data Service holds a wide and rich range of data collections suitable to shed light on all sorts of education-related research questions from key longitudinal studies to international microdata, government data and qualitative data collections. The roadmap for today's talk will be as follows. First, I will give a brief overview of what the UK Data Service is. So who are we? What is the UK Data Service? Second, I will highlight what longitudinal data on education are available via the UK Data Service. And third, I will show you where to find data, longitudinal data on education, where to find resources and help. Now, what is the UK Data Service? The UK Data Service is a comprehensive resource funded by the ESRC. It is a single point of access to a wide range of secondary social science data. At the moment, we provide around 7,500 data sets. However, we do not only collect, preserve and disseminate data, but we also provide support training and guidance. Now, this is a website of the UK Data Service, and I would like to show you the location of today's materials. So once we are finished here, I will send the slides to the communications team and they will upload the slides to today's event page. And today's event page you find when you click on news and events, and then select on the left-hand side. I assume you will look at it no earlier than tomorrow and it would be then past events. And when you then click on the description for the webinar of today, you will find at the very bottom a link to the slides and also a link to the webinar recordings. Now, back to the UK Data Service. Who is the UK Data Service for? The UK Data Service is for academic researchers and students, for government analysts, for charities and foundations, for business consultants, independent research centers and think tanks. Our data sources are official agencies, and here mainly central governments or ONS. Also international statistical time series, we have data from individual academics, so from research grants. For example, if a researcher has an ESA grant three months after completion of the project, one of the obligations of the researcher is to deposit the data with us for secondary analysis and to make it available to others. We have data from market research agencies, we have public records, historical sources, and also we have access to international data via links with other data archives worldwide networks we are involved in, and I will say a little bit more about that later on. The types of data collections we hold as survey microdata, aggregate statistics, census data, and qualitative and mixed methods data. Key data we hold, a large scale government UK surveys, surveys following individuals over time, so that as a longitudinal major UK surveys. International microdata, qualitative and mixed methods data, census data, business microdata, administrative data, and also and that's fairly new controlled international microdata. I would like to start with a letter, and here I would like to announce that the UK data service is part of the International Data Access Network since 2018, short IDIN, and the IDIN project is a collaboration between six research data centers from France, Germany, the Netherlands, and the United Kingdom, to facilitate research use of controlled access data between these research data centers and countries via reciprocal provision of safe or remote desktop access. The network will enable researchers to work remotely on controlled access data provided by partner countries. And I'm very happy to say that the UK data service has set up a secure data access point within the UK data service safe room here at the University of Essex. To facilitate on-site direct remote access to the French secure access data center data from the UK. And I have provided a link here to the data so you can see the whole list. It would be too long to show you, but you can scroll down the list and also provide an address where to apply for data access. And just a very brief overview of the themes covered and the producers involved. But I just want to highlight here basically that lots of data sets have, of course, information on education training and skills. Another safe and remote desktop access point we have for the Institute for Employment Research, FTZ or Research Data Center. So basically another access point here at the University of Essex in our safe room is to the IAB, to the Institute for Employment Research in Germany. And again, I have provided a link here to the data sets that are available via the UK data service safe room. And I have also copied a small screenshot for you. And again, I just want to highlight one thing and that's the National Educational Panel Study, NEPS, focusing on education particularly and those data are available via our safe room here at the University of Essex at the UK data service. There will be more later on, you have seen there are more countries involved, but this is very much work in progress. Now back to UK surveys. I know that today we are focusing on longitudinal data. However, also the description of today's event contains information that we would touch on other than longitudinal data briefly. And that's what I'm doing now. So UK surveys containing lots of information on education as well are, for example, the labour force survey, the skills and employment survey series, the British social attitudes survey, the general lifestyle survey previously known as the general household survey and the integrated household survey. International data are available other than the controlled international data which are subject to secure access conditions. So these are not that I'm presenting here on this slide. For example, one could look into Africa development indicators and look at the primary school completion rate of different African countries, for example, or you can look into the OECD education statistics and, for example, try to find out what is the typical graduation age in tertiary education. And here, for example, the second stage of tertiary education is get six and compare, let's say, United Kingdom and Germany and you would find out that the typical graduation age in United Kingdom for that PhD level would be 24 whereas in Germany that would be 28. Now qualitative mixed methods data. Here I just mentioned three, but I would like to hide it particularly the school leaver study and you might look a little bit astonished at the date 1978. Well, why would that be relevant? I come to that in a minute. Basically, what has happened here is Ray-Pal has asked teachers at a comprehensive school on the island of Sheppey to ask their pupils at the end of their school time to provide an imaginary account of their life over the next 30 or 40 years and they ended up with 142 handwritten essays and why that is of interest in particular for the longitudinal aspect of today's webinar is this exercise has been repeated in 2009 and 2010 and again 110 essays were collected and then coded and compared to the data across time. So that's actually quite interesting and I'm not going into the details of what was found but actually it was a really interesting study especially in this combination with a follow-up 2009 and 2010 which wasn't planned originally. Now coming to longitudinal data. Longitudinal surveys involve repeated surveys of the same individuals at different points in time. The advantage of longitudinal data over cross-sectional data and I find its beauty is that they provide the chance to look at the histories not just point-in-time situations. Usually there are large samples, they are nationally representative and on a regular basis new respondents are added in order to keep the numbers up and address the issue of attrition which is a problem for all longitudinal surveys. Original data allow researchers to analyze change at an individual level and they are usually a little bit more complex to manage and analyze but they are definitely worth it. So I have prepared just in order to give you an overview of what sort of documentation one can ideally expect because I think that's particularly beautiful here for BHPS on top where you can easily see the different variables and in which waves you will find them so you can easily see not all variables are obviously in all waves. And at the bottom there is a bit of the user guide which is a learning format study and here I just wanted to highlight the module character of the survey so in the first one you have information from the birth records in the second one for example medical records in the third education records and in the fourth teachers edit as well so that's actually quite nice to illustrate how the composition is done. Some examples of longitudinal data this is not a comprehensive list but I will just try to cover a few very important longitudinal data sets particularly with focus on education. Now the first three are the famous birth cohort studies so the 1958 National Child Development Study NCDS 1970 British cohort studies and Lenin cohort study and then there's another one highlighted in green which you can see further down the list is Next Steps or LSYPE Longitudinal Study of Young People in England formerly known as that just recently joined the others at the CLS the Centre for Longitudinal Studies in London and it actually fits in quite nicely in between the rather long gap between the 1970s British cohort study and the millennium cohort study. Moving on there's also the British Household Penal Survey which is located here at the University of Essex it's instituted for social and economic research its successor is Understanding Society and also called the UK Household Longitudinal Study and the letter is four times as big and followed the BHPS who had 18 waves but the British Household Penal Survey is part of Understanding Society from Wave 2 and meanwhile we also hold harmonized data sets which are really beautiful to analyze. I will also talk a little bit more about all of these studies later but just to give you a first overview we have growing up in Scotland also very important for policy in Scotland and finally I've listed CLOSER the cohort and longitudinal studies enhancement resources here on this slide this is a website of the Centre for Longitudinal Studies and as you can see here are the four national longitudinal cohort studies and three of them follow 17,000 or 19,000 young people born in a particular week of a particular year in Great Britain and next steps follow the lives of 16,000 people in England born in 1989 and 90 and I've also provided a link here a URL where you can find podcasts related to research based on these data sets. This is an overview of the sweeps of all these British birth cohort studies and next steps cohort study and you can see here for NCDS we have 10 sweeps so far and you can easily see when the next one occurs so for NCDS it would be in 2020 at the age of 62 and again in 2020 for BCS 70 or the response then being 50, et cetera, et cetera so I think that's quite an easy overview also in case you hadn't noticed but I mentioned it before that next steps is now managed by CLOSER was formally managed by the Department of Education now this is an overview of the linked education administrative data sets key stage one and two and these are all secure access data sets I will explain later what that means but quite interesting here also for example study number 8226 is a millennium cohort study of state linked data 2006 to 2012 or on the right hand side all the different other linked education administrative data sets key stage one and two moving on to understanding society as I mentioned before it is a very big and large longitudinal study and it is now serving 100,000 individuals in 40,000 British households it has replaced the BHPS but it incorporates it and thus basically retains the longevity whilst adding to the sample size and adding to the scope of the study so there are new components for example the innovation panel and there's also greater detail available on ethnic minority groups and also qualitative and biomedical data collections are added and I didn't mention that before but also for the cohort studies biomedical data increasingly play a role and are added to the data sets and then linked with consent obviously to the data sets if possible and that enables fantastic new research also in relation to education so the understanding society data linkage covers three types of data linkage and that is linkage to geographical identifiers and administrative data linkage and organizations so here for example identifiers of the schools that children attend or recently attended now I have mentioned before that understanding society and BHPS data are harmonized and we have now data sets which have been harmonized and actually here the recent one is study number 6614 understanding society waves 1 to 8, 2009 to 17 and the harmonized BHPS data set in one so basically that covers now from 2009 after 2017 or basically including BHPS 1991 to 2017 so a fantastic longitudinal data set this is a slide containing the overview of the data linkage phases that understanding society plans and you can see regarding education there are quite a few administrative data sources which are planned to be linked to get linked to the data and you can also see the status and how the informed consent collection went or is planned or when it has to happen so that's actually quite a nice overview and I will not go into any more detail but if you want you can look at it later on you know where you will find the slides and then you can inform yourselves about the details closer the cohort and longitudinal studies enhancement resource aims to maximize their use, value and impact both at home and abroad so all these studies they include it brings together 8 leading studies and closer basically works to stimulate interdisciplinary research and develop shared resources provide training and share expertise and this way closer is helping to build the body of knowledge on how life in the UK is changing both across generations and in comparison to the rest of the world so the areas of work that CLOSER does include data harmonization, data linkage research impact and training and capacity building this is a very nice overview of the studies involved and also in a timeline so you can see what started when and you will recognize a couple of studies I have highlighted before for example the NCDS the BCS 1970 the MCS and understanding society but also you can see CLOSER and I want to come to CLOSER data sets so the UK data service is quite pleased to announce the first harmonized data sets created by CLOSER and they are now available to download so these data sets harmonize body size and body composition variables for example across five of the prestigious cohort studies of the UK the MRC National Survey of Health and Development NCDS 1970 BCS MCS and the even Longitudinal Study of Parents and Children and I would like to quote Professor Alisson Park here saying as an experienced researcher I know the huge challenges involved in creating harmonized data sets from different data sources and the importance of ensuring the methods used can be usefully exploited by other researchers Longitudinal harmonization provides fascinating insights as to how the UK is changing over time and I'm delighted CLOSER is helping make this work easier I can only agree with that and indeed it does so here we have the data sets and you can see CLOSER work package 1 harmonized height, weight and BMI and CLOSER work package 2 includes harmonized socioeconomic measures in four Longitudinal cohort studies and then we also have CLOSER training data set now I would also like to mention the growing up in Scotland data set it's a Longitudinal Research Study and this study tracks the lives of thousands of children and their families from the early years through childhood and beyond and the main aim as for most of the studies is to provide new information to support policymaking and in this case particularly policymaking in Scotland so first cohorts for 2002 to 3 and 2004 to 5 and 2010 to 11 are collected and again they are also focusing on education as well as other aspects and as other studies introduced before they also ask for consent from parents to link data collected from them and their children to administrative data helped by health and education authorities more on Longitudinal studies and their impact is available but I don't have time to go into impact and policy relevant research if you want to know more about that there is another webinar available in the slides and I have given the URL here where you can find all those now how would you go about finding Longitudinal data on education so one option is you search using key data so you click on get data on the left hand side you select key data and then Longitudinal studies and then you will present it with a list of Longitudinal studies you can scroll down and select another option is you search by theme for this you would go to get data data by theme and then select education or you use the data catalog and you type education in the box and then you specify on the left hand side cohort and Longitudinal studies and here you would get 410 results another option is instead of using studies you could use series so again you would on the left hand side give the topic and the data type so Longitudinal studies and this time we are looking at series and then you get 22 hits another option is you look at variable in question bank the option to look at variable in question bank is nice but it is not complete I would like to give you a little bit of a warning here it's very nice to try it but it's not comprehensive and for that reason it is not the first choice but interestingly enough you can here search as it suggests for variables and I have just typed highest education in here to demonstrate what happens then and I have specified on the left hand side as a country or as country selected from the list European Union countries and I was presented with 18 results and I then sorted it according to the earliest date so that would be 1995 and then I wanted to compare variables I found so I have added certain variables to my variable basket circled in green and finally I wanted to see what's in my basket in order to compare the question texts of different variables in different waves and also different studies to see whether I could actually compare those variables on highest education that's actually quite useful for your analysis and research anyway data access who can access the data basically all registered users however which data can be accessed and the particular access conditions vary according to the user type so whether you are from a UK higher education further education institution or not whether you are from the UK or not from the UK second it also depends on the usage or the project characteristics whether you request the data for a commercial or non-commercial project and third whether the specific data access conditions attached to the chosen data show end user license, special conditions, special license or secure lab data access conditions or even safe home access we provide web access to the data and for secure lab data also remote access to our data we have a few studies for safe home access only so that's on-site access only but that's not the majority so I'm not highlighting that too much here accessing data step by step so you need to register with us you then agree to an end user license you select the data you specify your project give in maybe 30 words minimum describing your project and then you can either download the data if it's an end user license data set or if it's a special license or other data set then you would need to complete forms which you then prompted the same here on our website so basically everything is explained on our website step by step and the data access conditions that we have end user license, special conditions special license, secure lab data access conditions and for secure lab data access you would need to also attend a safe researcher training following an approval as a researcher and also you would then be given a passport and a username the end user license basically specifies that you do not share the data with anyone who's not authorized that you use it just for the purpose that you have access for and that you don't share your login details obviously the special license access you have to give a little bit more detail you have to apply for it and obviously there's a higher risk of disclosure attached to these data sets so the institution with responsibility for the researcher also needs to ask explicit permission of the data owner to release the data to the researcher and secure lab data access conditions are slightly different so the data is not available for download access requires accreditation and also an agreement to the services user agreement and breaches penalty policy the applications are screened and the individual or institution having ownership of the data or the designated authority have then to screen the applications as well and access is only granted to those requiring the data for statistical research purposes and who can justify the need for the data however we can assist with those applications and help you to succeed this is how it works you register, you order the data you become an approved researcher you complete the agreement complete the training, receive your unique login then you work remotely in the secure lab area you will then ask for statistical disclosure control of your requested output we will check it and then hopefully if there's no problem release it to you for publication you can also explore the data online for example with Nesta which is an online data browsing and analysis system and this is how you find it you go to get data, explore online and click on Nesta and here's one example I have provided for you what you could for example do in this case I have used general household service time series data set 1972 2004 and I have looked at the highest qualification and here just select the degree or higher and you can see that you have the total numbers on the bigger in the bigger graph but in the right hand side bottom corner you find percentages and you can see hopefully that in 1972 we had roughly 1% of the respondents having a degree or higher whereas in 2004 almost 6.5% held the degree or higher so that's actually quite positive here you find a couple of links or useful links to get started with Nesta and see how to go about using Nesta for your browsing finally support and resources which support and resources can you find and expect from us so we have bigger tutorials and webinars we have student resources for example and in line with what we are talking about today we provide a longitudinal data module you can see it in the little box provided and also the link to it we have case studies available online guides, themes, advice on managing and sharing data we provide teaching data sets and resources particularly for teaching and we want to help this and individual user support video tutorials you can find at use data, video tutorials and then here thematic guides thematic guides can be found when you click on get data and education so then you find thematic guides on education if you want to search all case studies you would go to use data data in use, if you want case studies based on longitudinal data you would go to get data, key data and then longitudinal studies further down you find case studies relating to longitudinal studies and if you want case studies using data on education you would go to get data, education and here you find case studies using data on education and in the box highlighted in green I have just provided a few case studies on education for example, titled factors associated with academic achievement exploring the middle on GCSE attainment or our ethnic inequalities in education changing teaching with data I had mentioned that we have lots and lots of resources on teaching so if you go to use data teaching with data you find lots of resources at your fingertips there longitudinal teaching data sets particularly you can go to our data catalog type teaching data sets or teaching data set specifies a data type cohort in longitudinal studies and you will be prompted to list data, data by theme education and again here you find teaching data sets on the theme of education. Finally if all this was too fast or not clear enough you can go to frequently asked questions or to help or to contact and then if nothing else contact us we have a telephone number but we would prefer you to go to ukdataservice.ac.uk help get in touch and then select which of these online forms are appropriate for your purpose. We are also on Twitter, Facebook and YouTube and you can subscribe and I would recommend that to the ukdataservice email subscription list justmail.ac.uk etc etc and you will be updated about the newest developments now questions I don't think we do now I would like to now hand over to my co-presenter Wei so I am handing over to Wei Hi my name is Wei Xun thanks for the invitation from organizers to speak today. I am from the Centre for Long-Tunnel Studies Information and User Support Celsius for short we are based at UCL funded by the ESRC as well so today I would like to talk to you about the Office of National Statistics Long-Tunnel Study and so I will go through quite briefly the structure of the Long-Tunnel Study I will refer to that as the LS data for short and then what sort of educational variables we have in LS and then also a little bit about my own research into education how it relates to economic activity in terms of working or not through the life course so we start sorry the slides are working now so firstly just really general introduction into the LS so the LS is in fact what looks like a panel study that includes individual microdata from 1% of the UK population and this was established in 1974 and that includes the census since 1971 so every census since 1971 we have had five waves of this with each wave capturing over 500,000 people so the dumpling of the LS was based on birthdays which means that is a random selection it's four birthdays out of 365 days a year so this just about totals just over 1% so because the dumpling is based on the census so all the information in each of the census are available there are also life events data that's linked to the members in the LS study which I will talk about later on so one of the very big advantages of using the LS data is very large sample size which enables people to do all sorts of comparisons of subgroup analysis or geographical based analysis in fairly small units which we can talk about also later so here is a schema of the LS data so if we start from the central spine where the white boxes are so this tells you the number the numbers for 1971, 81, 91, 01 and 2011 each census captures between 530 and 590,000 people the dumpling has grown a little bit just from census over time because we're actually getting more people from the UK so since 1971 the sample members are linked back in at each subsequent census via the NHS central register so each time an existing LS member appears as subsequent census they are linked back into the data set using a unique identifier that we give them so hence the sampling is fairly stable and we also have a dynamic sampling in LS in which people can actually enter by being born on one of the four birthdays in the year so from 1971 to 2015 we've had 325,000 new births and that's at the rate since 2015 about 7,000 per year you can also enter the LS through immigration which is a technical term in the LS and that's actually registering at the NHS register so that's been 242,000 since 1971 to 2015 and that happens at about 9,000 per year and with exit events you can exit the study by dying we've had 280,000 since 71 and that's about 6,000 per year and you can also embark and that's actually living in the country so the sampling actually stays really stable we've had births and deaths and entries and exits so we've been able to maintain a stable 50,000 members per wave and then we also have linked life events so I already talked about birth, death and embarkation we also follow up example members if they have given birth so that's for women whether the infant lived so hence we have infant death we also have widowerhoods or widowerhoods as well as cancer registration so all of this would have been derived by either health or administrative records again linked in via the NHS register so next slide okay so one of the very big advantage of the LS data is slow attrition the census has been mandatory to fill in so we have very few people who are not linked compared to most longitudinal studies that possibly that have been mentioned before so here is a diagram infographic that shows I think since 1971 each time so the members in 1970 are original example that each time they appear census exactly what happened to them so in 1981 you see from 530,000 we have 418,000 that are present 91 352 291 to 90 and 2011 to 40 you see in the pink circles where most people have gone and actually they have died this is again something quite unique about the LS is that it includes because it's four birthdays are randomized over the year so it's calendar day and month so the LS includes all ages and it's completely representative of the UK population as census each time so that we actually if you look at 2011 at the end almost 50% of example in 1971 have died for example says immigration and that means deregistration from the NHS Central Register and the true attrition in the orange circles right at the end it's true loss to follow up it's only around 10% after 40 years and I think that's actually a very good number so what sort of information does LS contain so basically all information from the census are available in LS so that includes demographics ethnicity things like marital status so social status information family composition so where they were living the geographies what they were doing occupationally whether they were traveling to work so more topics were added in successive censuses now we're also recording religion national identity and passport language intention to stay now we also have limiting long-term illness self-rated health and also caregiving since 2001 I have already shown that the life events that are linked in to the LS include death and that includes up to 8 courses of death birth to LS members maintenance events to birth members and also birth weight and I think birth weight possibly was recorded up to 2015 immigration and immigration and again that's to do with NHS registration with oral health and cancer registration so now I will talk to you about what sort of education information is present so how do I go back in the LS so as a rule of thumb the best way of ascertaining what is available in the LS is to go to the census forms which is what I've screen grabbed for you here 1971 so we have two questions relating to education so firstly question 13 have you obtained any of the following so it talks about GC A level higher school certificate and higher grade of Scottish highest also ordinary national certificates ordinary national diplomas and question 14 if you read the question it says have you obtained any of the following qualifications since reaching the age of 18 so these are the answer would relate to basically a professional degree level qualification and here it gives you a list up to 6 of these are recorded and unfortunately that's what's available in the dataset is up to 6 post 18 qualifications in 1981 the question asks have you obtained any qualifications after the age of 18 so it's fairly similar to 1971 question 14 so again we would only know what up to 6 post 18 qualifications these people have had 1991 fairly similar again it asks about after reaching the age of 18 so we again would have post 18 age 18 qualifications but now there's more categories looking at degree and also professional qualifications up to here you also know the institution for each of the qualifications so in 2001 we have moved to a question that asks which of these qualifications do you have and take all of the ones that apply so now you can actually have different levels below age 18 so you can see here all levels GCSE A levels a degree and also then higher degrees and recues it asks a several questions about professional qualifications in question 17 again it says take all the boxes apply the one caveat of this is that the question asks if you're aged 16 to 74 so you lose the people who are aged over 74 but the good thing about having this information now is that for people who have answered the questions in 71 to 91 you're able to retrospectively gain back input back the qualification levels it would have had in certain ages and that's what I've done in my study so in 2011 the qualification question has been combined into one and now is asking those aged 16 and over all the qualification levels that's possible so again it's possible to use this information to retrospectively impute education level below age 18 earlier censuses provided that they appear so we have at the LS website a derived variable for codes for standardizing education for degree non-degree up to 2001 and that just really needs to be updated to include 2011 so that has been done by some researchers that we've worked with in the past so basically with this information that it's possible to compare cohorts of people over time so you have five censuses worth and you also have all ages so it's possible to construct many cohorts so if you want to do time series comparisons it's also possible to follow up an individual over time I guess most useful thing would be following up them over before the age of 35 before education actually gets solidified and also what's possible is to sorry I didn't mention this before but you also have information about those who are co-resident in the household at census so that includes parents as well as spouses so if you wanted to compare those people with education levels versus the LS member so you can have intergenerational mobility in education or you can look at differences between spouses and then you can actually ask some really interesting questions over time in generations and over people in the household how that evolves so now I will explain briefly how to access the ONS LS data so we at Celsius handle basically all the front end interactions between us and users who are can be anyone in the UK so based in the UK if you're academic, charity government individual researchers as long as you qualify through the approved projects and approved research process you're able to access the data in a secure setting so we handle the process between inquiry and also all the way through to data outputs so what we really encourage users to do is to contact us at Celsius at UCL.ac.uk our email address if you have a potential research idea that you think you could explore using LS data we also have a wealth of information in training materials on the Celsius website which is this link highlighted in blue here in the middle as I already mentioned we provide support from inquiry to application process to the research approval to constructing your data set so each project will get a tailor made data set just to suit your needs we'll prepare this for you we will also provide on-site some limited on-site support while you are accessing the data we also do remote access support so this is for people because the safe setting there are three at the moment one in London, in Pimlico in OMS office and there's another one in Southampton in Titchfield in another OMS office and there's another one in Wales so understanding that not all research can actually get to these safe settings so we provide a remote service by which the users will send us their analysis scripts and we'll run them generate output and clear those output to an acceptable disclosure level and then return those to you so all of this process will be supported by us so now I will talk to you briefly about my own research so I realize I don't have a lot of time left so I was looking at how basically how employment status lingers over the life course education was one of the variables I used so very quickly my example I picked from 21 and all of those age 16 to 24 and I've tracked them forward 40 years so each 10 year gap from say age 16 to 24 wanted to know whether their education had any effect on them being employment at 10 years later and actually my main variable was economic activity how that changes over time and so each time the education would be contemporary to the outcome so here if we see men and age 26 to 34 how influential was education level at predicting at putting people in unemployed status versus employed status so any old ratios lower than one shows the protection of having higher education level versus a reference no education or missing which is a really big proportion of my example in 1971 you know all the cohorts historical so this is showing that having higher education is protective against not working basically age 26 to 34 and also 36 to 44 so I have two non-working categories one for unemployed and one for other inactive which is everything else that's not inactive and you can see that the O.R.s kind of creep upwards for GCCO level as people age and so if we look further down the life course it's still protective for those who are unemployed and inactive age 46 to 54 but it's no longer so at the very the oldest age group that I have in the 56 to 64 the oldest transition should I say but it is protective at the older ages for having higher education for not working because they're ill and again the reference category was to be working whereas in women again O.R.s symptoms lower than one shows protection from being work less so so this time we see the whole entire life course and between 26 to 64 and whether education was actually protective against being unemployed or other inactive in these ages and it seems to actually strengthen in older age in women so that means women with higher education are more likely to be working between the age of 46 to 54 64 I would think possibly that's because that's when caring load is lower in a woman's life course and if you look at the other work less category which is looking at the home family it's really very similar picture so the protection is kind of very strange pattern like a z-shaped pattern so protection is the weakest in middle life between 36 to 44 and it strengthens again at the older ages 46 to 64 again we think that's because it's less caring load in the household so women can actually work so thank you for listening. So if you're interested we would really like to hear from you please get in touch with us by email at celsus at ucel.ac.uk and also I forgot to mention that the portal offer is free. Okay thank you very much so we've had a question how you controlled for other factors influence employment so I really just had to fly over that so this is the adjustment that I've made so we come activity at the last follow-up age group, education level whether they're married, whether they report to self-reported sickness and that applies only in the last two follow-ups whether there's children in the household whether there's illness in household members whether there's spouse working and also there's a costed operation at the world level. Okay Are there any more questions? Okay it seems that we have provided all the information that was wished Thank you very much, that's it from us today and if you have any other questions please get in touch we are very happy to answer Thank you