Presentation on theme: "Introduction to ESDS Government surveys and services Vanessa Higgins ESDS Government Centre for Census and Survey Research (CCSR) University of Manchester."— Presentation transcript:
Introduction to ESDS Government surveys and services Vanessa Higgins ESDS Government Centre for Census and Survey Research (CCSR) University of Manchester
My mission… Intro to Economic and Social Data Service (ESDS) and how we can help you What data is available and what it is like Registration and access Key issues to consider when analysing the data
Intro to ESDS Government and how we can help you
ESDS Government ESDS began in Jan 03 Comprises 4 specialist services –ESDS Government –ESDS Longitudinal –ESDS Qualidata –ESDS International ESDS Government provides access and user support for key large-scale government surveys such as Labour Force Survey, Expenditure and Food Survey etc Access remains via the UKDA
ESDS Government - some of the things we do! Helpdesk (email@example.com or 0161 275 1980)firstname.lastname@example.org Survey pages incl. how to get started Online guides – SPSS, STATA, Weighting, Employment Research, Health Research, Social Capital, Data Sources for Scotland User Group seminars (data users and data creators) Publications Database Derived variables - consistent over time - consistent with Census Teaching datasets Training
Which surveys? General Household Survey/Continuous Household Survey Labour Force Survey/NI LFS Expenditure and Food Survey (previously the National Food Survey and Family Expenditure Survey) Family Resources Survey Time Use Survey British Social Attitudes/Scottish Social Attitudes/Northern Ireland Life & Times/Young Peoples Social Attitudes ONS Omnibus Survey Annual Population Survey National Travel Survey British Crime Survey/Scottish Crime Survey Health Survey for England/Wales/Scotland Survey of English Housing
Integrated Household Survey (Continuous Population Survey) Will bring together –Labour Force Survey –Annual Population Survey –General Household Survey –Expenditure and Food Survey –Omnibus Survey Expected to start c. 2008 Materials on our website IHS Consultation Meeting 23 rd January 2006 (pm)
QUALITY OF DATA Main data collectors: –Office for National Statistics (ONS) –Central Survey Unit of the Northern Ireland Research and Statistics Agency (NISRA) –National Centre for Social Research (NatCen) Very experienced in design, management and analysis of social surveys Permanent panels of highly trained field interviewers Management and Quality Checking (Relatively) high response rates – but falling Widespread use by secondary analysts
Registration All users can access study descriptions, online documentation, including questionnaires, free of charge without registering with ESDS In order to access the datasets you need to register with ESDS –Register online using your ATHENS username and password –Simple online form, takes about 10 minutes –You need to register a usage of the data as part of this process –Non-commercial users: free of charge –Commercial users: £500 Charge for per study and will need to apply for ATHENs username and password –You need to agree to the End User Licence when you register
The End User Licence This is not public data! We need to know who you are, how to contact you and what you are using the data for Agree not to attempt to identify individuals Only use the data for your stated purpose (you can re-register if you want to use the data for another use) Do not pass the data to unregistered parties (that includes deleting the data before passing on PCs!) Tell us if you publish using the data Some more sensitive or detailed datasets require more stringent licensing procedures –Special conditions: British Crime Survey –Special licences: Annual Population Survey and Labour Force Survey at the moment
Special Licences Agreement between ONS and ESDS Data where the disclosure risk is deemed too high for standard dissemination methods. E.g.: –Local authority level microdata –Month of birth Paper application procedure –Need to agree to stringent data stewardship conditions –Required to read the guide to good practice –Application has to be approved by ONS Annual Population Survey and Labour Force Survey at the moment but hope to roll out to other surveys
Registering the data for teaching If all students are registered ESDS users, you should register a new use of data, tick teaching only and add the name of each student to the usage details If students are not registered ESDS users a paper form can be used for teaching Lecturer takes responsibility for the class Students simply fill their details in on a paper form Return form to the UK Data Archive
Obtaining data once registered Choice of downloading entire files: –SPSS, Stata, tab delimited formats for most files –Comes zipped up with documentation Or explore the data and download subsets in Nesstar (c. 200 files in this format – including most of the Government surveys) –Allows you to explore metadata without being registered –Can do basic exploratory anlyses (including OLS) online without downloading files –Can define subsets to download in a wider range of formats including SAS
Key issues when analysing the data -Choosing the appropriate survey -Using the documentation -Using the appropriate file/dataset -Preparing the variables for analysis -Weighting -Sample size and sampling error
Choosing a dataset for research Is secondary analysis appropriate? –And has someone already done the analysis for you? –Check published reports. Is population data available? –If data is available for 100% of your population of interest this is preferable to sample data. Which datasets cover your main topic? Which other topics are you interested in? Measurement over time Geography Respondents – whole household, children? Sample size This information will be in the documentation To help with the decision –ESDS Getting Started guides –ESDS topic guides
Documentation VERY important to read the documentation before you start the analysis Documentation available online from UKDA/ESDS Dont need to be registered Should contain –Information about method/sampling/response –Codebooks of variables –Information about derived variables –Questionnaires Unless you can track the variable back to the question asked, and work out who it was asked of, you dont understand the data!
Using the appropriate files Once you have downloaded data you need to ensure you are using the file you need Each survey differs in format Formats can differ within surveys Some surveys/years have one file with all the data Some surveys have lots of different files –Separate files for topics –Separate files for unit of analysis Household Individual Carers –What is your unit of analysis?
Preparing variables for analysis These datasets contain hundreds of variables –Straightforward and complex questionnaire variables –Straightforward and complex derived variables –Often many variables on the same topic –Are you using the optimum variable available? Categories Population base –Do you need to derive your own variable from those available? –How are missing values dealt with? Unless you can track the variable back to the question asked, and work out who it was asked of, you dont understand the data!
Weighting Weighting is becoming increasingly important/common Weighting corrects for known bias in datasets Bias arises because of: –Sample design e.g. when one individual per household is selected for a sample or one group of the population e.g. children, are over-sampled to boost numbers –non-response i.e. unequally distributed non-response Third form of weighting often referred to as grossing which weights to produce population estimates Different packages handle weighting differently http://www.esds.ac.uk/government/resources/statguides.asp –are weights available? –how does the documentation suggest you use the weights? –what will your analysis package do with weights?
Sample size and sampling error Sample data is subject to sampling error The larger the sample the greater the accuracy with which the sample represents the population If the sample is small it may, by chance, be different from the population Govt surveys often publish a range of sampling errors for their main point estimates but you will probably need to calculate some yourself
Good practice: Some pointers Irresponsible use can threaten future access Keep data secure You wont understand the data unless you consult the documentation! Consider the impact of the sample design –Sample data has sample error; small cell counts are likely to be unreliable –Will the sampling design affect the standard errors? –Does weighting need to be applied to avoid bias Getting started guides, stats guides and topic guides are available on the ESDS site if you dont know where to start Contact helpdesk if you have any problems or find errors in the data User groups and news lists can help keep you informed Published results from surveys and published articles that have used the surveys Survey Link Scheme – can attend an interview yourself http://qb.soc.surrey.ac.uk/sls.htmhttp://qb.soc.surrey.ac.uk/sls.htm
Using the data in teaching Methods courses –Using the data in a hands on manner –Using substantive exemplars to demonstrate a methodological point –Using the surveys as methodological exemplars Substantive courses –Making your point using data –Integrating methods into substantive courses Teaching datasets –General Household Survey –Labour Force Survey –British Crime Survey –Health Survey for England
Some other useful resources Question Bank –Linked to from ESDS –Contains searchable questionnaires for major surveys –Contains articles about major topics/concepts –Runs the survey link scheme –http://qb.soc.surrey.ac.ukhttp://qb.soc.surrey.ac.uk PEAS (Practical Exemplars on the Analysis of Surveys) –Contains learning materials on the impact of survey design on analysis (e.g. weighting and complex sample designs) –Uses the government surveys as exemplars –http://www.napier.ac.uk/depts/fhls/peashttp://www.napier.ac.uk/depts/fhls/peas ONS website contains published reports, methodology etc. –http://www.statistics.gov.ukhttp://www.statistics.gov.uk
The important last slide… Helpdesk: email@example.com (0161)275 1980 Website: http://www.esds.ac.uk/government JISCmail: firstname.lastname@example.org