About this resource
The purpose of the ESS Multilevel Data is to make it easier for the research community to analyse the ESS-respondents with reference to the context they live in. The resource contains data about:
- individuals (the ESS respondents)
- regions (mainly data collected from EUROSTAT)
- countries (data collected from different sources)
EU-, EFTA- and candidate-countries have a regional ESS variable that is possible to map with the NUTS nomenclature. But the countries do not use the same NUTS-level as their regional level in the ESS survey, and thus data must be collected from different files. Moreover, some countries are not part of the NUTS nomenclature. Because of the complexity at the regional level, we have found it neccessary to offer two gateways into the data: One service for Online Browsing, and one service for Multilevel Download.
In the service for Online Browsing, we offer a hierarchical dataset for each ESS round. The hierarchical datasets contain three datasets or levels:
- Individual level: The ESS-survey data, for example ESS4
- Regional level: ESS region is created by combining data from different NUTS-levels
- Country level
Data from these three levels can be included in the same analysis. If you add a variable from the country level dataset and a variable from the ESS-survey data, the respondents from the same country will get the same value. The N of this analysis will equal the number of respondents.
You can also analyse each level separately, for example country. N in such an analysis will equal the number of countries included.
If you would like to add data from a particular NUTS-level to the ESS-data, you must use the Multilevel Download.
The Multilevel Download makes it possible to add information about countries and several regional levels to the respondents in the ESS data. Please note that the country specific region variable decides how it is possible to merge contextual data on regions to the individual level ESS data: If a country uses NUTS 1 as its regional variable, it is not possible to merge information from NUTS 2 or 3 to the respondents from this country. The region variables are of the following types:
|Type of ESS region variable||Possible to merge with NUTS 1||Possible to merge with NUTS 2||Possible to merge with NUTS 3|
|Local administrative unit||Yes||Yes||Yes|
|Combined NUTS levels|
|Countries not covered by the NUTS||No||No||No|
There are less data available at the NUTS 3 than at the NUTS 1 level. To avoid missing data, it could be an alternative to add data about the NUTS 1 or NUTS 2 levels.
In the Multilevel Download you can generate and download a customised subset of the ESS data and the contextual data by selecting rounds, countries and variables. The data file can be downloaded in SPSS, Stata, SAS and other formats. A Study Description and documentation of the selected variables is generated for each customised data file and is included in the download.
How to generate a customised subset in Multilevel Download (click to open or close)
- Go to ESS Multilevel Download
- Log in, if not already logged in
- Select ESS-round
- Select countries and click "Next"
- Select variables. You can select all variables in a variable group, for example "Media use and social trust", or you can select individual variables in this selection process, you will also choose which regional level you would like to add data from:
- ESS regions
- NUTS 1
- NUTS 2
- NUTS 3
- Select data format and click "Download dataset"
These variables are by default included in the download:
- IDNO "Respondent's identification number"
- CNTRY "Country"
- Region "ESS region"
- NUTS 1 "NUTS 1 region"
- NUTS 2 "NUTS 2 region"
- NUTS 3 "NUTS 3 region"
- DWEIGHT "Design weight"
- PWEIGHT "Population size weight (must be combined with dweight)".
There are numerous sources with contextual data. The ESS Multilevel Data does not contain everything available. The selection of variables is based one the recommendations in the report of the ESS Upgrade and Planning Group on Contextual Data: 'Conceptual and Statistical Issues in Multilevel Analysis'. We have also tried to find variables that could be relevant for the rotating modules in the ESS.
|The regional variables are downloaded every second year. The variables downloaded in 2012, have been merged with the ESS4 and ESS5 survey data. The variables downloaded in 2014, have been merged with the ESS6. The variables downloaded in 2016, have been merged with the ESS7.
We will not update the regional variables already added to an ESS round.
If the datasources have corrected data, it is possible that the same variable will have different values in the data downloaded in 2012 compared to the data downloaded in 2014 or 2016.
The downloads may include different variables.
The contextual variables are collected to shed light upon themes like
- demography and geography
- political institutions
- composite measures
Most variables are collected for several years. More information about the contextual variables, can be found on this page.
In case of mismatch between data found in this resource and data presented by one of the sources, you should always rely on the original source. It is also important to credit the original source in written publications.
You will find a list of our sources on this page.