 |
Saturday, Nov 22 |
|
| ||||
Study Name:
Panel Socio-économique / Liewen zu Lëtzebuerg (PSELL I)
Panel Socio-économique /
Liewen zu Lëtzebuerg (PSELL II)
see: http://www.ceps.lu/paco/pacolupa.htm
Study Director: Pierre Hausman
Principal Investigators: Gaston Schaber, Frédéric Berger, Georges Heinrich, Blandine Lejealle, Mireille Zanardelli, Jean-Yves Bienvenue
Host Organization: CEPS
/ INSTEAD
Centre
d'Études de Population, de Pauvreté et de
Politiques
Socio-Économiques
International Networks
for Studies in Technology, Differdange/Luxembourg
http://www.ceps.lu
Year Initiated: PSELL I: 1985
PSELL II: 1994
- Funding Sources
The Luxembourg Government through a contract with CEPS/INSTEAD primarily funds the study. Additional funding is provided through EUROSTAT in correspondence with cloning the ECHP (European Community Household Panel) sample for Luxembourg.
- Host organization
The study is conducted by CEPS / INSTEAD.
- Governing Body and role of external researchers
The study is steered by members of CEPS/INSTEAD with strong co-operation with the Central Statistical Office in Luxembourg (STATEC). Because the newer waves of PSELL II (from 1997 on) are used for the production of the ECHP, requirements of EUROSTAT are partly integrated into the design of PSELL II.
- Sample selection
The PSELL I and II take the resident population as the
target population. Anyone living in the Grand Duchy
who is concerned by social welfare or social protection. Excluded are :
1) foreign residents who have no links with the country's social security system or who do not live in a household, where at least one of the members has such links 2) elderly persons living in a collective household such as an old people's home. The PSELL I and II survey is randomly sampled out of the National Social Security Register of Luxembourg. First individuals are selected and then the whole household, where the individual is living is selected and interviewed. There are approx. 8500 persons and 3000 households in the gross sample. The basic sample represents 97% of the population living in the country.
The PSELL I (1985-1994) was stopped because the sample was no longer representative for the population in Luxembourg, in a country which experienced a relative high immigration.
- Follow rules
Each member of the family is followed in principle. The follow up will stop if the person or family migrated into another country. The exception from the last rule will be done for households, which leaves the country, but move only near to the border of Luxembourg.
- Oversampling
There is no oversampling done.
- Weights and attrition bias
Estimations of population totals and related figures must be weighted in order to produce unbiased estimates. The weights comprise adjustment for the sampling procedure in wave one, and adjustments to account for attrition in later waves. Therefore the sample is re-weighted every wave (cross-sectional weights).
- Sample “refreshing”
The current PSELL II is refreshed all two years. The former PSELL I was refreshed by integrating additional young families with children.
- Demography
- Household and family
- Living conditions
- Labour market
- Education
- Income
- Health
- Housing
- Mode
The PSELL I and II are collected in face-to-face interviews. using paper and pencil questionnaires. A minority of interviews is conducted also over the telephone. The questionnaires are entered into the computer using an advanced data entry program, which checks for completeness and consistency of the data.
- Instrument design (paper and pencil, CATI, EHC)
Paper and pencil questionnaires are used.
- Dependent interviewing
none
- Calendar year, survey year and point of survey measures
All three methods are used.
- Family composition editing
Some editing is done in the process of data entry, but most work is done after using numerous SPSS syntax programs has entered the data.
- Production of specialized files
A subset of PSELL I and II have been entered or will be entered into four international comparable databases:
PSELL I:
PACO: Panel Comparability Project (http://www.ceps.lu/paco/pacopres.htm)
PSELL II:
ECHP: European Community
Household Panel (http://www-rcade.dur.ac.uk/echp/)
PSELL I +
II:
LIS: Luxembourg Income Study (http://lis.ceps.lu/access.htm)
CHER: Consortium of Household Panels for European Socio-economic
Research
(work under progress)
- Data and documentation standards
The data of PSELL are stored as SPSS system files. The comparable subsets of variables of PSELL I and II are available in different other formats.
Currently detailed documentation about the original PSELL data is rather conventional organized. Summary information is also available within the survey documentation system set up by EUROSTAT. For the future we consider to document PSELL data within the NESSTAR (Network Social Science Tools and Resources) system, which is using the DDI documentation standard.
- The original data of PSELL are only distributed on request. A comparable subset of data from PSELL I is available within the PACO project which distributes micro data via the PACO-CD.
- Virtual data centers and functionalities
The Luxembourg data in LIS can be analysed via the LIS information system over the Internet. The user submits syntax programs (SPSS/SAS/STATA) via e-mail to the LIS server and receives the results as listing files back via e-mail. A download of micro data is not possible within the LIS system.
The new CHER project is considering the possibility to load data into the NESSTAR system, which allow to do the user interactive tabulations.
- Improving contractual use of data (confidentiality and encryption)
The micro data of PSELL I and II can be used within research visits the large-scale facility IRIIS-CI, which is run by CEPS/INSTEAD.
- Comparability of formats (evolution of data standards)
The PSELL panel supports following standard classifications:
ISCO 1988 (3 digits)
NACE
ISCED 1997
NUTS1
Currently a subset of variables from PSELL I and II is available in three international comparable formats:
Currently work is done to enter the PSELL data into the CHER (Consortium of Household Panels for European Socio-economic Research database) database.
- Value of joint multinational analysis projects
Co-operation within multinational research projects were very important for:
(a) integrating PSELL data into comparable databases
(b) analysing PSELL data within cross-national analyses
Institute for Social Research | University of Michigan | Privacy | Conditions of Use
