Dataset: Generations and Gender Survey France Wave 1 & Wave 2

Abstract

The Generations and Gender Survey (GGS) provides micro-level data with the aim of significantly improving the knowledge base for social science and policymaking in Europe and developed countries elsewhere.  
In Europe 2020, the European Union develops a strategy "to help us come out stronger from the crisis and turn the EU into a smart, sustainable and inclusive economy delivering high levels of employment, productivity and social cohesion". The economic crisis affects not only day-to-day decisions, but also fundamental choices at all stages of people's lives:  marriage and childbearing, the combination of employment and caring responsibilities for the young and the old, retirement, housing, and ageing well. The GGS has been developed to provide scientists with high-quality data to contribute scientifically grounded answers to these key policy questions. Survey content focuses on intergenerational and gender relations between people, expressed in care arrangements and the organization of paid and unpaid work. Key feature of the survey are:  
- Cross-national comparability. In each country data is collected on the basis of a common international questionnaire and guidelines about the methodology. Data processing includes central harmonization of national datasets.  
- A broad age range. It includes respondents between the ages of 18 and 80.
- A longitudinal design. It has a panel design, collecting information on the same persons at three-year intervals.  
- A large sample size. It has an average of 9,000 respondents per country at Wave 1.
- A theory-driven and multidisciplinary questionnaire. It provides data for policy relevant research by demographers, economists, sociologists, social policy researchers, social psychologists and epidemiologists. The questionnaire is inspired by the theory of planned behavior.
- Possibility to combine the survey data with macro data provided by the GGP Contextual Database. This combination enables analyses of individuals and families in their cultural, economic, political, social and policy contexts.

Variable Groups

Document Description

Full Title

Generations and Gender Survey France Wave 1 & Wave 2

Alternative Title

GGS France Wave 1 & Wave 2

Identification Number

GGS.W1.W2.15

Date of Distribution

2012-12-20

Version

Working Version: GGS Wave 1 Version 4.3 and GGS Wave 2 Version 1.3.

Update of variable catagories and documentation with the release of Poland Wave 2 Version 1.3.

Date: 2018-02-26

Guide To Codebook

In the field “Study Description”, users can find metadata about surveys. This includes the distributors, keywords, abstract, and guidelines on the bibliographic citation.  
Country specific metadata include information on survey producers, methodology and processing. This information was provided by the GGP-country teams, based on a metadata grid with pre-structured questions. Links to relevant references (e.g., working papers and questionnaires) are also provided.  

The field “Data Files Description” provides metadata about the data file, such as file contents, missing values, as well as changes across different GGS versions.

The field "Variable Description" provides information on each variable, such as question text, descriptions of country specific categories and variables, universe (i.e., subset of respondents to whom the question was asked), country specific deviations to GGS routing, descriptions of the ways in which consolidated and derived variables are calculated. Variables are ordered according to the sections of the GGS codebook.

PLEASE NOTICE THAT WE DOCUMENT ONLY VARIABLES HAVING VALID CASES.  
VARIABLES HAVING ALL SYSTEM MISSING CASES ARE NOT DOCUMENTED.  
This is the reason why the total no. of variables in the documentation is smaller than the total number of variables in the SPSS and STATA files.

Full Title

GGS_W1-V.4.3.&W2-V.1.3_France

Producer

Name Affiliation Abbreviation Role
Arianna Caporali Institut national d'études démographiques (INED) AC

Study Description

Full Title

Generations and Gender Survey France Wave 1 & Wave 2

Alternative Title

GGS France Wave 1 & Wave 2

Parallel Title

Étude des relations familiales et intergénérationnelles (ERFI)

Identification Number

GGS.W1.W2.15

Authoring Entity

Name Affiliation
Régnier-Loillier Arnaud INED

Other identifications and acknowledgments

Name Affiliation Role
INED's Surveys department INED Tracking, data cleaning, weighting.

Producer

Name Affiliation Abbreviation Role
Régnier-Loillier Arnaud INED Researcher

Funding Agency/Sponsor

Name Abbreviation Role Grant
Institut national d'études démographiques (National institute of demographic studies) INED Financing of Wave 1 and Wave 2
Institut national de la statistique et des études économiques (National institute of statistics and economic studies) INSEE Financing of Wave 1 and Wave 2
Agence nationale de la recherche (National agency of research) ANR Financing of Wave 1
Caisse nationale des allocations familiales (National fund of family allowances) CNAF Financing of Wave 1 and Wave 2
Direction de l'animation de la recherche, des études et des statistiques (Ministry of Employment) DARES Financing of Wave 1
Conseil d'orientation des retraites (Pensions Advisory Council) COR Financing of Wave 1
Direction de la recherche, des études, de l'évaluation et des statistiques (Ministry of Health and Solidarity) DRESS Financing of Wave 1 and Wave 2
Caisse nationale de l'assurance vieillesse (National fund of elderly allowances) CNAV Financing of Wave 1

Data Distributor

Name Affiliation Abbreviation
Institut national d'études démographiques - 133 boulevard Davout 75980 Paris Cedex 20, France. INED
Netherlands Interdisciplinary Demographic Institute - Lange Houtstraat 19, NL-2511 CV The Hague, The Netherlands NIDI

Depositor

Name Affiliation Abbreviation
INED (Institut national d'études démographiques)

Bibliographic Citation

United Nations 2005. Generations & Gender Programme: Survey Instruments. New York and Geneva: UN, 2005.

List of Keywords

Date of Collection

Start End Cycle
2005-09-26 2005-12-03 Wave 1
2008-10-20 2008-12-12 Wave 2

Country

FRANCE  (FRA)

Geographic Coverage

Whole metropolitan territory of France

Geographic Unit

Department (97 administrative departments in French metropolitan territory).

Unit of Analysis

Individuals

Universe

WAVE 1
French speaking persons aged 18 to 79 on December 31st, 2005 living in private household in France.

WAVE 2  
Persons who took part at Wave 1 and agreed to be re-contacted for Wave 2 (88% of the 10,079 persons interviewed at Wave 1)

Kind of Data

Survey data

Time Method

Panel

Data Collector

Sampling Procedure

WAVE 1 SAMPLING PROCEDURE
1. Sampling frame
1.1 Type of frame: Dwelling lists: master sample of 1999 National Census data + sampling frame of dwellings constructed after march 1999.
1.2  Frame coverage: Whole population of metropolitan territory of France.
1.3 Frame size: Master sample of 1999 census: 2,022,889 dwellings (7% of all 1999 dwellings) from 349 primary units (of 3435 in census data).
1.4 Level of units available: dwellings.

2. Sampling method
2.1 Sampling method type: A first random sample of 16009 households addresses from 1999 census master sample (14752) and sampling frame of dwellings constructed after March 1999 (1257). To complete this first sample: two random samples of 1000 households adresses from 1999 census master sample (2*838) and sampling frame of dwellings constructed after March 1999 (2*162). So a total of 18009 dwelling adresses.
2.2 Sampling stage definition
- PSU: Cities if more than 20,000 habitants, local groups of towns if less than 20,000 habitants.
- SSU: Dwellings addresses.
- TSU: NA.
2.3 Sampling stage size
- PSU: 349
- SSU: 18009
- TSU: NA.
2.4 Unit selection: All primary units of the master sample were selected.  
2.5 Final stage unit selection: Dwellings were selected using simple random sampling.
2.6 Within household unit selection: First-name method. Among those eligible in the household (age 18-79 years on December 31st, 2005), the person whose first name begins with the letter the closest to the beginning of the alphabet was selected.
2.7 Stratification: Explicit: Two stages stratification. First stage was done by dwelling categories with overrepresentation of main residences (5 categories: main residences at 1999 census, secondary residences at 1999 census, occasional residences at 1999 census, vacant residences at 1999 census and dwellings built after 1999 census). The second stage was by age with overrepresentation of households whom referent person at 1999 census was in the scope of the survey (younger than 79 years old in 2005). The second stage was not done for new dwellings as the data on the age of these household's occupiers were not available.
2.8 Sample size:
- Starting size sample: 18009
- Aimed total size at Wave 1: 10000
- Aimed total size at Wave 3: none
2.9 Estimated Non-response
- Yearly attrition: None: the yearly attrition was not estimated.
- Non response measures: Oversampling : to reach the objective of 10000 respondents, the sample size was 18009.
- Within household non-responses measures: None - the household was marked as nonresponse.

WAVE 2 SAMPLING PROCEDURE
Sampling: panel (wave 1 database of addresses). All the Wave 1 persons who had agreed to be recontacted for a second interview (i.e., 8,341 respondents which is equal to 88% of Wave 1 respondents).

Mode of Data Collection

Method: Face-to-Face (personal interview)
Technique: Computer-Assisted (CAPI)

Type of Research Instrument

Structured questionnaire in French.

Characteristics of Data Collection Situation

WAVE 1 DATA COLLECTION
1. Interviewers
1.1 Total number of interviewers: 552 interviewers of INSEE (data collector)
1.2 Number of interviewers in the field: Each interviewer had a fixed number of interviews to carry out. Some of them can finish the work earlier than others.
1.3 Network organization: At least two field coordinators in each of the 22 regional agencies of INSEE (data collector).
1.4 Working arrangement of interviewers: Fully employed survey administrators from INSEE.
1.5 Payment of interviewers:  Paid per interview (completed interview: one hour paid, refusal: fixed compensation) + compensation for one day of training + travel expenses.

2. Interviewer training:  
2.1 General interviewing: Interviewers were already trained in general interviewing techniques: CAPI, interviewing techniques, appointments taking and general knowledge on surveys carried out by INSEE.
2.2 Survey specific: In august 2005, professional regional trainers of INSEE were trained centrally at INED in GGS specific issues and then trained the 560 interviewers in each of the 22 French regions.
2.3 Length: General training: 6 days. Survey specific: 1 day.
2.4 Control of performance: The new interviewers were accompanied by the regional survey manager for their first interviews. Some of the experienced ones were accompanied too as part as their annual performance review.
2.5 Interviewer survey: None.

3. Contact protocols
3.1 Advance letter: Each household selected in the sample received an advanced letter with a leaflet introducing the survey and announcing the coming of an interviewer, the reason of his/her visit, the survey subject, the agencies involved in the survey, the future use of the data, etc. The household was also invited to visit the survey website (www-erfi.ined.fr) to find more information about the survey and the project.
3.2 Cold contacts: Face-to-Face
3.3 Scheduling / scattering: Yes. To get the highest response rate, contacts attempts were scattered over different days of the week and different parts of the day.
3.4 Contact history: Yes. For each contact attempt, the interviewer had to report the date, the time and the outcome in a grid.
3.5 Min number of contacts: 7.
3.6 Max number of contacts: 14.

4. Questionnaire localization
4.1 Validation: In March 2004, a test of the first version of the questionnaire was carried out by interviewing 111 persons (face to face paper) to improve the translation and evaluate the length of the interviewing.
4.2 Pre-test: Two pre-tests carried out using CAPI version of the questionnaire in November 2004 (134 respondents) and April 2005 (180 respondents).
4.3 Length of interview: Respondents were globally very cooperative. The average duration of interview was about 65 minutes, with a large variance according to the composition of the household and the complexity of the family structure and history of the respondent. A quarter of the interviews took place in the presence of another person. Some difficulties were encountered by respondents to answer fecundity questions. Two questions (religion appartenance and civil partnership) were considered as "sensitive" according to the French law and needed a signed authorization from the respondent. 6% of the respondents refused that their answers to both "sensitive" questions were recorded.

WAVE 2 DATA COLLECTION
1. Interviewers
1.1 Total number of interviewers: 454 interviewers of INSEE (data collector).
1.2 Number of interviewers in the field: Each interviewer had a fixed number of interviews to carry out. Some of them could finish the work earlier than others.
1.3 Network organization: At least two field coordinators in each of the 22 regional agencies of INSEE (data collector).
1.4 Working arrangement of interviewers: Fully employed survey administrators from INSEE.
1.5 Payment of interviewers: Paid per interview (completed interview: one hour paid, refusal: fixed compensation) + compensation for one day of training + travel expenses.

2. Interviewer training:  
2.1 General interviewing: Interviewers were already trained in general interviewing techniques: CAPI, interviewing techniques, appointments taking and general knowledge on surveys carried out by INSEE.
2.2 Survey specific: In September 2008, professional regional trainers of INSEE were trained centrally at INED in GGS specific issues and then trained the 454 interviewers in each of the 22 French regions.
2.3 Length: General training: 6 days. Survey specific: 1 day.
2.4 Control of performance: The new interviewers were accompanied by the regional survey manager for their first interviews. Some of the experienced ones were accompanied too as part of their annual performance review.
2.5 Interviewer survey: None

3. Contact protocols
3.1 Advance letter: Each household which agreed to participate to Wave 2 received an advanced letter recalling the survey subject and announcing the coming of an interviewer, the reason of his/her visit, the agencies involved in the survey, etc. The household was also sent a brochure concerning the survey and the invitation to visit its website.  
3.2 Cold contacts: Face-to-Face
3.3 Scheduling / scattering: If 2005 respondents were no longer present in the household, the interviewer made an appointment to meet them, and if they had moved to another zone, another interviewer was given the new address. When 2005 respondents had gone to live outside the regions covered by the survey (but within metropolitan France), the interviewers went beyond their usually allotted zone to meet them.
3.4 Contact history: Yes. For each contact attempt, the interviewer had to report the date, the time and the outcome in a grid.
3.5 Min number of contacts: Dk
3.6 Max number of contacts: DK

4. Questionnaire localization
4.1 Validation: No.
4.2 Pre-test: Two pre-test were carried out. The first was limited on the new module on occupational history introduced with Wave 2. It was carried out in november 2006 using Paper and Pencil on 20 persons of all ages. This test showed recall problems for older persons and persons with numerous occupational changes. There were also problems of data overlap which were then overcome thanks to CAPI automation. The second test was carried on the entire questionnaire from March 24 to April 25, 2008, in two different parts of France (Aquitaine and Limousin) on 180 persons. The test could not be run on people who had participated 3 years previously. In order to check “the retrieval chain of data collected in 2005” (i.e., the filter questions in Wave 2 aimed at saving time) the test questionnaire was preceded by a short module focusing on the respondents’ situation 3 years previously.
4.3 Length of interview: The average duration of interview was about 55 minutes, with a large variance depending on the complexity of the respondents’ occupational history and hence their age. A third party was present in 25% of the 2005 interviews, whereas this percentage dropped to 18% in the 2008 interviews. The interviews went very well on the whole. This is confirmed by the fact that 97% of respondents agreed to participate to Wave 3. The main problems concerned the occupational history grid, which turned to be not so easy to complete especially for older respondents. As to the items considered “sensitive” by the Quality Label Committee of the French National Council for Statistical Information (CNIS) at Wave 1, the question about religion membership was no longer asked; the other one on civil partnership was not regarded as sensitive anymore.

Actions to Minimize Losses

WAVE 1 ACTIONS
1.  Dealing with nonresponse
1.1 Screening: No
1.2 Refusal conversion: Usual techniques of refusal conversion i.e. specify that personal information will stay confidential and not divulged to a third party, insisting on the research purpose of the survey and on its international dimension.
1.3 Incentives: No

2. Tracking of sampled units
2.1 Respondent contact information: Respondent's address and telephone number (home, mobile, work) were collected.
2.2 Other contact information: The interviewers were instructed to ask for the addresses of two persons in the respondent's circle of acquaintances (close family, friends) to remain able  of resuming contact with the respondent via one of these persons in case of a move.
2.3 Cards: Various strategies were organized to keep in touch with a maximum number of persons. A thank you letter was sent to all the respondents, which varied in content according to 3 scenarios:
   - if the respondent had refused the follow-up, he was thanked simply;
   - if he/she had accepted the follow-up and given the address of at least one contact person, he was thanked by     enclosing a "change of address card" as well     as an e-mail and a phone number so that he can inform INED of any moves;
   - if he/she had accepted the follow-up but if we had no contact person, he was thanked the same way as in the previous case and we added to the letter a "     contact person card" and a prepaid envelope.  
Four other mails were sent to every respondent between Wave1 and Wave 2 (three mails with first results of the survey and a greetings card) .The aim of these regular mailshots is to keep in touch but also to update the address file (death, change of address...).
2.4 Additional surveys: None.
2.5 Administrative records: No.

WAVE 2 ACTIONS
1.  Dealing with nonresponse
1.1 Screening: No
1.2 Refusal conversion: As for Wave 1.
1.3 Incentives: No

2. Tracking of sampled units
2.1 Respondent contact information: Update of respondent's address and telephone number.
2.2 Other contact information: Record of respondent's circle of acquaintances addresses. In case no address was available, the 2005 addresses were reused.  
2.3 Cards: The respondents who agreed to participate to Wave 1 were contacted by mail five times between the two waves, i.e., approximately every six months (see above documentation about wave 1). The respondents who had agreed to participate to Wave 1 were contacted by mail five times between the two waves, i.e., approximately every six months (see above documentation about wave 1). The same basic follow-up procedure was used with the respondents between Wave 2 and Wave 3 (two letters each year: initial results and a New Year greetings card, etc.).
2.4 Additional surveys: None.
2.5 Administrative records: No.

Control operations

WAVE 1 CONTROLS
Controls (consistency checking) were done automatically during the interview thanks to the CAPI questionnaire.

WAVE 2 CONTROLS
CAPI allowed to retrieve information from answers to Wave 1 questionnaire. This had several advantages: it shortened the length of the interviews, prevented contradictory information from being collected between two waves and saved the respondents from having to answer sensitive questions (those referring to deaths, separations, religion, etc.) more than once. In addition, respondents were provided them with several chronological landmarks (the age at which various events occurred, their children’s year of birth, etc.) to help them remember sequences of events and dates. Respondents were never confronted with their previous responses to the 2005 questionnaire (no questions such as “In 2005, you stated that…” were asked).

Weighting

WAVE 1 WEIGHT VARIABLES
The country-specific weight variable (aweight_1501) was built at the respondent level to adjust the sample because of:
- an overrepresentation of women (56,6 %) in the sample compared with the structure of the French population (51,3 %),  
- an under-representation of men,  
- an over-representation of over-55s (in particular men),  
- and an under-representation of 18-29 year-olds (again also more marked for men).  
Non-response was corrected by fitting. Weighting variable was computed using the following variables: sex, age reached in 2005, citizenship, social and occupational group, type of household, number of household members, and size of urban unit.  
Because of a rather strong distortion of the educational level (over-representation of the higher degrees), this variable was not taken into account in the weighting calculation.
The other  weight variable (aweight) is a standardized weight based on the country specific population weight (recommended for use).

WAVE 2 WEIGHT VARIABLES
The country-specific weight variable (bweight_1501) was built to adjust the sample because of attrition between the two waves. This variable takes into account the attrition between the two waves as well as the weighting variable calculated with reference to Wave 1. Between the two waves, attrition was equal to approximately 35%  (some people refused to participate in wave 2, some had died, some had moved without leaving an address, and some had gone to live in institutions). The data can therefore only be used in a longitudinal approach, or a cross-sectional approach  representative of 2005, provided that a weighting variable is applied to correct for attrition and to fit the 2008 findings with those of 2005.  
The other weight variable (bweight) is a standardized weight..

Cleaning Operations

WAVE 1 CLEANING OPERATIONS  
The datafile was cleaned (anonymized, wildcode checking...) by INED's survey department before being sent to data harmonization center for harmonization.

WAVE 2 CLEANING OPERATIONS  
Data adjustment and cross matching between wave 1 and wave 2 were carried out by INED's survey department before being sent to data harmonization center for harmonization.  Severe underreporting of children in Wave 2 was brought to light. For more detailed information, refer to the documentation about Wave 2 methodology under "Study Description - Other Study Description Materials".

Response Rate

WAVE 1
Frequency of final disposition codes:
I = Complete interview: 10079
P = Partial interview: 48 (not included in the data base)
NE = Not eligible: 2430
NC = Non-contact: 1657
R = Refusal: 2242
O = Other non-response: 729
UC = Unknown eligibility, contacted: DK                
UN = Unknown eligibility, non-contact: DK                                                                      
eC = Estimated proportion of contacted cases of unknown eligibility that are eligible: DK
eN = Estimated proportion of non-contacted cases of unknown eligibility that are eligible: DK

WAVE 2
Frequency of final disposition codes:
I = Complete interview: 6,576 (however 42 were then removed from the dataset because the respondents were most certainly not those surveyed three years earlier).
P = Partial interview: 8
NE = Not eligible: 75 (respondents deceaced) + 17 (respondents in institution) + 52 (respondents abroad)
NC = Non-contact: 221 (could not be contacted) + 16 (no household member corresponds to the person interviewed in 2005) + 249 (moved without leaving address)
R = Refusal: 794
O = Other non-response: 190 (could not be interviewed)+ 77 (could not be interviewed before the end of the collection period)
UC = Unknown eligibility, contacted: DK
UN = Unknown eligibility, non-contact: 21 (address could not be reached) + 45 (address card not processed)
eC = Estimated proportion of contacted cases of unknown eligibility that are eligible: DK
eN = Estimated proportion of non-contacted cases of unknown eligibility that are eligible: DK

Completeness of Study Stored

The GGS questionnaire was adapted to fit with the French sociodemographic and legal context and because of time and budget constraints: some questions were added, dropped or modified. The added questions, French specific are not stored in the GGS harmonized data file.

Restrictions

In order to access micro data files, users have to sign and submit a Statement of affiliation, confidentiality and acceptable usage. They also have to submit a title and abstract of their research project. They can use the data for all their research projects, except for datasets from Australia and Norway. Users of these datasets need to submit a new application form if they want to use the data in a different research project. The access rights from Wave 1 data are transferred to the Wave 2 data.

Access Authority

Name Affiliation E-mail address Universal Resource Identifier
UNECE Population Unit - Palais des Nations - CH-1211 Geneva 10 - Switzerland. Tel: +41 22 917 24 77 - fax: +41 22 917 01 07 ggp@unece.org http://www.unece.org/pau/

Citation Requirement

In any work emanating from research based on the Generations and Gender Survey micro-data, I will acknowledge that these data were obtained from the GGP Data Archive and refer to the publication that describes the model survey instruments: United Nations 2005. Generations & Gender Programme: SurveyInstruments. New York and Geneva: UN, 2005

Deposit Requirement

Users of GGS micro-data are required to send any research papers based on the Generations and Gender Survey micro-data or aggregate tabulations to the Population Activities Unit of the UN Economic Commission for Europe, for inclusion in the GGP publications archive.

Conditions

In order to access, it is necessary to subscribe to the GGP Data User Space, and to follow the instructions available on the GGP data access webpage.

Disclaimer

The authors and producers bear no responsibility for the uses of the GGS data, or for interpretations or inferences based on these uses. The producers accept no liability for indirect, consequential or incidental damages or losses arising from use of the data collection, or from the unavailability of, or break in access to the service for whatever reason.

Notes

Metadata on the French GGS can also be retrieved from the documents available under "Other References Note".

Related Materials

Website of Étude des relations familiales et intergénérationnelles (ERFI)

Wave 1 French questionnaire (in French)

Wave 1 French questionnaire - Show Cards (in French)

Wave 2 French questionnaire (in French)

Wave 2 French questionnaire - Show Cards (in French)

Other References Note

French country presentations at the GGP International Working Group Meetings

Modifications to the Generations and Gender Surveys questionnaire in France (wave 1)

INED working paper by Pascal Sebille and Arnaud Régnier-Loilier (Documents de travail no. 144)

Presentation and Modifications to the GGS Questionnaire in France (Wave 2)

INED working Paper by Arnaud Régnier-Loilier, Leila Saboni, and Béatrice Valdes (Document de travail no. 173)

Inconsistencies in the Number of Children Reported in Successive Waves of the French Generations and Gender Survey

Article by Arnaud Régnier-Loilier (Population, English Edition, 69(2), pp. 167-190)

Data Files Description

File Name

GGS_Wave1_France_V.4.3..NSDstat

Contents of Files

GGS Wave 1

VARIABLES HAVING ALL SYSTEM MISSING CASES ARE DROPPED BEFORE PUBLICATION IN NESSTAR.
This is the reason why the total no. of variables in the Nesstar data file is smaller than the total number of variables in the SPSS and STATA files.

Variables are ordered according to the sections of the GGS codebook: Household, Children, Partnerships, Household Organisation and Partnership Quality, Parents and Parental Home, Fertility, Health and Well-Being, Respondent's Activity and Income, Partner's Activity and Income, Household Possessions, Income and Transfers, Value Orientations and Attitudes, Interviewers' report.
The variables begin with a letter designating the wave of data collection ("a" for the first wave likewise "b" for the second wave). We have attempted to keep the names of variables the same across the waves, and all the new variables would be identified as follows ["wave letter"]n e.g.  bn301.  
Although we encourage the countries to strictly follow the GGS Questionnaire, countries might implement a question that differs to a considerable extent from the GGS Questionnaire. In this case either we add country specific response values, or we introduce a country specific variable.  
Country specific values are added when the question follows the model questionnaire, but the answers are not at all or partly compatible. They are at least 4 digits long (F4 format) and begin with the country code: e.g., Australia 2401. Hence, the country code, as an example, for Australia is 24.  
A country specific variable is introduced when the question differs from the model questionnaire albeit measuring the same concept. This kind of variables is identified with a suffix given by the country code plus a number, e.g., Australia a119_2401.
In order to have an overview of GGS country code, please refer to the variable "acountry".

File Structure

Record Group

Overall Case Count

10079

Overall Variable Count

1372

Type of File

Nesstar 200801

Extent of Processing Checks

WAVE 1 DATA HARMONISATION
The data is submitted in an already pre-harmonised form. It is prepared and organised according to the GGS standards.  
Harmonisation aims at achieving a clear and comparable format of the GGS micro-data files that would be adequate for cross-country comparison.  The harmonisation procedure basically is composed of:
1. Label checks  
This step makes sure that all the variables are named the same across the countries and refer to a particular question in the GGS Questionnaire. Also the value labels are checked. They should be the same across GGS datasets.  
2. Dealing with grids
The GGS Questionnaire holds several grids of either event history information or members of the household. Such data needs to be harmonized with specific attention to order and logical consistency of grid-rows (be either household members or events such as births). In data sense each row of the grid is represented by variable name followed by a subscripted number ("_#"). Each subscript thus represents one household member or one event. Part of the grid harmonization is grid sorting. Grid rows are sorted according to pre-defined key. For example in the household grid, the household members are sorted according to their relationship to the respondent i.e. the relation to respondent variable (ahg3_# or bhg3_# ). Respondents would appear, first, followed by their partners and children if any and then followed by other household members. As there may be more then one child (or other relative) living in the household they also would need to be sorted. In the case of the household grid, age is used as the secondary sorting key (starting with the oldest person to the youngest).
3. Routing
Routing check ensures that the structure of underlying data set matches the structure of the GGS questionnaire. Its main goal is to code any given variable in the dataset to either a valid response, nonresponse or skip as indicated in the questionnaire. Consequently, the indicated skip in the quetionnaire is represented with a system missing code (. in STATA, sysmis in SPSS), while the missing information for other reasons is coded into non-applicable/no response (i.e. codes 7, 8, 9 in SPSS or .a, .b, .c in STATA).  
4. Consolidation  
The process consolidates the information scattered over several variables into a single one. The consolidation procedure is carried out in the Children Section, the Partnership Section and the Parents and Parental Home Section.
5. Imputation  
Due to its sensitive nature, the respondents are reluctant to share income information with the interviewer. In order to be able to use income information in a cross country comparative study and not to loose too many observations in the process it is necessary to impute the approximately correct distribution of the income variable in each country.  
6. Calculation of derived variables
We calculate derived variables out of the following variables:
- grid variables (i.e., household grid, children grid, and partnership history grid); the codebook starts with the constructed variables that sum the key socio-demographic characteristics of the respondent.
- month and year variables,  
- hours and minutes variables,
- frequency and unit variables.  
Occupation variables are recoded into ISCO-88 1 digit.
Explanations of the ways in which consolidated and derived variables are obtained, are available under the field "Note" of the "Variable Description" sections.
For a more detailed and technical procedure please refer to the Data Cleaning and Harmonisation Guidelines.

Missing Data

The following missing values have been assigned:
- 6, 96, 996, etc. = Unknown (only for consolidated variables in the group "administrative variables")
- 7, 97, 997, etc. = Don't know
- 8, 98, 998, etc. = Refusal
- 9, 99, 999, etc. = Not-applicable/no response

Version

Harmonized dataset, GGS Wave1, version 4.3.

Notes

IMPROVEMENTS INTRODUCED WITH V.4.3. (August 2016):
Variables corrected with Version 4.3.
- fertintent (no more ambiguous labelling)
- a1101 (corrected error in coding)
- aweight (now available also for NLD CZE SWE POL)
- aregion (now available also for HUN)
- aplace (now available also for HUN)
- a5112 (corrected routing error for ROU)
- a5113 (corrected routing error for ROU)
- a5114 (corrected routing error for ROU)
- a5115 (corrected routing error i for ROU)
- a211b_ (corrected error for POL & GEO)
- ankids (corrected error for POL & GEO)
- a1008mnth (corrected error for NGR & BEL)
- a108 (now available for SWE)
- a109_1 (now available for SWE)
- a109_2 (now available for SWE)
- a149 (now available for SWE)
- a309 (now available for SWE)
- aregion (now available for SWE)
- a620_ (corrected error for DEU & CZE)
- a402 (corrected error for POL)
- a149 (corrected error routing error in NOR)
- a344 (corrected error routing error in NOR)
- a256_ (corrected error for POL & GEO)

IMPROVEMENTS INTRODUCED WITH V.4.2. (February 2014):
The update from v4.1 to v4.2 does not include corrections of existing variables.  
The update only includes additional variables which are derived from the pre-existing datasets
- Variables derived from grid variables and variables which concern the respondents and his/her partner: numdissol numdivorce nummarriage numpartners livingwithpartner childprevp femage maleage femeduc maleeduc fertintent numbiol numres numnonres numstep numallchild ageyoungest ageoldest numrespleave numotherparentleave coreschild coresparen coresgrandp coressibl.
- Variables derived from month and year variable:  
a808Dur a822Dur a907Dur a911Dur a914Dur; a303cAgeP a315AgeP a316cAgeP a374cAgeP a608AgeP a610AgeP a617bAgeP a621AgeP a914AgeP a941AgeP; a107AgeR a121AgeR a150AgeR a239aAgeR a239bAgeR a240AgeR a301AgeR a302bAgeR; a311AgeR a314bAgeR a314dAgeR a371AgeR a372bAgeR a603AgeR a608AgeR a610AgeR a613AgeR a614AgeR a619AgeR a621AgeR a816AgeR a822AgeR a871AgeR a5116AgeR a5117bAgeR; a302bTdiff a314bTdiff a314dTdiff a372bTdiff.
- Variables derived from hours and minutes variables: a324_hour a520_hour a540_hour.
- Variables derived from frequency and unit variables: a205mnth,a241mnth,a325mnth,a355mnth,a359mnth,a363mnth,a367mnth,a521mnth,a541mnth,a1008mnth,a1102mnth; a203c_?w a204c_?w.
- Occupation variables recoded into ISCO-88 1 digit: a828_1dig a832_1dig a861_1dig a917_1dig a921_1dig a933_1dig a5112_1dig a5114_1dig.

IMPROVEMENTS INTRODUCED WITH V.4.1. (April 2012):
- Variables corrected: amarstat, anpartner, ankids, a148, a1008 (now correctly dropped)
- Value labels defined: a828 (user missings)

IMPROVEMENTS INTRODUCED WITH V.4.0 (March 2012):
- New constructed variables: amonth, asex, aage, abyear, aeduc, aactstat, aparstat, amarstat, anpartner, ankids, ahhsize, ahhtype
- New consolidated variables on respondents' current activity: a870, a871m, a871y, a873, a874, a875
- New consolidated variables on respondents' partners current activity: a940, a941m, a941y, a943, a945
- Variables previously unavailable, now available: a1002u, aweight (aweight standardized weight based on country specific population weight, recommended for use), aweight_1501 (country specific population weight)
- Variables corrected: a383 (age of current partner now rounded), a602 (corrected, with consequences on response rate to subsequent questions), a612, a622, a623, a624, a625, a626, a627*, a628*, a629*, a631
- Changes in the variable/value labels: a149, a309, a322, a380 (country specific list has been recoded into one digit ISCED 97 codes); a203b_* (Country specific codes have been recoded: a) "assistente maternelle","nourrice", "garde a domicile", "babysitter" coded into GGS code 1 "babysitter", b) "creche","ecole maternelle", "halte garderie" coded into GGS code 3 "nursery", c) "before/after school care" coded into GGS code 4 "after school care", d)"centre airee" oded into GGS 5 "Self-organized childcare group", e) "other" coded into GGS code 5 "other institutional arrangement". PLEASE NOTE THAT THIS CODING DEVIATES FROM WAVE 2 VERSION 1.0. In V.4.1. the group "Centre aéré" has been erroneously coded as "Self organized childcare group". It should be coded as  "4-After-school care-centre". In a future revision, wave 1 will be made consistent with Wave 2); a203c_* (when R mentions two or more original childcare alternatives that have been merged into one GGS code the frequency of that services used are summed up in a203c_*. By doing so we get closer to the true value of childcare use in France but we might however overestimate country differences between France and other GGS countries on this variables)

IMPROVEMENTS INTRODUCED WITH V.3.0 (August 2010):
- New consolidated variables:  a370, a383, a384
- Variables renamed: a203c_1u..._`i'u and a204c_1u..._`i'u renamed into respectively a204c_1u..._`i'u and a204cu_1...`i'
- Variables corrected: a211b, a212_*, a213_*, a214m_*, a214y_*, a215_*, a216a_*, a216m_*, a216y_*, a217y_*, a218_*, a219_*, a220m_*, a220y_*, a222_*, a223_*, a224_*, a407, a410, section 5 (parents), a5108_1, a5108_2, a612, a615, a616, a622, a624, a625, a627_*, a628_*, a629_*, a631, a914y, a917, a918, a919, a920, a925_*, a1003 and a1004 (previously coded as first, second, third... mentionned, now coded per item as a yes/no question)
- Variables previously unavailable, now available: a112a, a130, a225, a307a, a319, a926, a927, a928, a929, a623, a630, a831, a838, a914m, a610m, a610y
- Variables not available anymore: a516, a517_*, a536, a537_*, a607_*, a608m, a608y, a611
FIRST DATASET RELEASED: V. 1.8. (March 2010).

File Name

GGS_Wave2_France_V.1.3..NSDstat

Contents of Files

GGS Wave 2

VARIABLES HAVING ALL SYSTEM MISSING CASES ARE DROPPED BEFORE PUBLICATION IN NESSTAR.
This is the reason why the total no. of variables in the Nesstar data file is smaller than the total number of variables in the SPSS and STATA files.

Variables are ordered according to the sections of the GGS codebook: Household, Children, Partnerships, Household Organisation and Partnership Quality, Parents and Parental Home, Fertility, Health and Well-Being, Respondent's Activity and Income, Partner's Activity and Income, Household Possessions, Income and Transfers, Value Orientations and Attitudes, Interviewers' report.
The variables begin with a letter designating the wave of data collection ("a" for the first wave likewise "b" for the second wave). We have attempted to keep the names of variables the same across the waves, and all the new variables would be identified as follows ["wave letter"]n e.g.  bn301.  
Although we encourage the countries to strictly follow the GGS Questionnaire, countries might implement a question that differs to a considerable extent from the GGS Questionnaire. In this case either we add country specific response values, or we introduce a country specific variable.  
Country specific values are added when the question follows the model questionnaire, but the answers are not at all or partly compatible. They are at least 4 digits long (F4 format) and begin with the country code: e.g., Australia 2401. Hence, the country code, as an example, for Australia is 24.  
A country specific variable is introduced when the question differs from the model questionnaire albeit measuring the same concept. This kind of variables is identified with a suffix given by the country code plus a number, e.g., Australia a119_2401.
In order to have an overview of GGS country code, please refer to the variable "acountry".

File Structure

Record Group

Overall Case Count

6533

Overall Variable Count

1453

Type of File

Nesstar 200801

Extent of Processing Checks

WAVE 2 DATA HARMONISATION: see "Extent of Processing Checks" "WAVE 1 DATA HARMONISATION".

Missing Data

The following missing values have been assigned:
- 6, 96, 996, etc. = Unknown (only for consolidated variables in the group "administrative variables")
- 7, 97, 997, etc. = Don't know
- 8, 98, 998, etc. = Refusal
- 9, 99, 999, etc. = Not-applicable/no response

Version

Harmonized dataset, GGS Wave2, version 1.3.

Notes

IMPROVEMENTS INTRODUCED WITH GGS_Wave2_V.1.3 (August 2016)
Correction of the following variables that were previously  erronous: b343_*, bnnumdissol, bnumdissol, bnnumdivorce, bnumdivorce, bnnummarriage, bnummarriage.

IMPROVEMENTS INTRODUCED WITH GGS_Wave2_V.1.2 (April 2015)
The update from v1.1 to v1.2 does not include corrections of existing variables. The update only includes additional variables which are derived from the pre-existing datasets.  

- Variables derived from grid variables and variables which concern the respondents and his/her partner: bnumdissol, bnnumdissol, bnumdivorce, bnnumdivorce, bnnummarriage, bnummarriage, bnumpartnerships, bnnumpartnerships, bnrespartafterw1, blivingwithpartner, bchildprevp, bnchildprevp, bfemage, bmaleage, bfemeduc , , bmaleeduc, bfertintent, bnumbiol, bnumnonres, bnumres, bnumstep, bnumallchild, bageoldest, bageyoungest, bcoreschild, bcoresgrandp, bcoresparen, bcoressibl , bhhtype.
- Variables derived from month and year variable: b121AgeR, b150AgeR , bn152AgeR, b239aAgeR, b239bAgeR, b240AgeR, bn304Agb303cAgeP, b311AgeR, b315AgeP, b316cAgeP, b371AgeR, b372bAgeR, b372bTdiff, b374cAgeP, b5116AgeR, b5117bAgeR, b603AgeR, b608AgeP, b608AgeR, b610AgeP, b610AgeR, b621AgeP, b621AgeR, b871AgeR, b907Dur, b911Dur, b914AgeP, b914Dur, b941AgeP.
- Variables derived from hours and minutes variables: b324hour, b520hour, b540hour, b221hour_x.
- Variables derived from frequency and unit variables: b203c_xw, b204c_xw, b205mnth, b241mnth, b325mnth, b521mnth, b1008mnth.
- Occupation variables recoded into ISCO-88 1 digit: b828_1dig, b832_1dig, b861_1dig, b917_1dig, b921_1dig, b933_1dig.
- Three groups of variables derived from section no. 8 "Activity and Education History": 1) variables counting the total number of different activity and education situations Rs has had since age 16 (i.e., bnnumworkstatuses, bnnumstudentstatuses, bnnumemplstatuses, bnnumselfemplstatuses, bnnumhelpfamstatuses, bnnumunemplstatuses, bnnumretiredstatuses, bnnummilitarystatuses, bnnumhomestatuses, bnnummatleavestatuses, bnnumparleavestatuses, bnnumdisabilitystatuses, bnnumotherstatuses, bnnum1401, bnnum1501, bnnum1801, bnnum1301, bnnumparttime, bnnumfulltime, bnnumboth, bnnumparttime_1801, bnnumparttime_1802, bnnumpartfulltime_1803, bnnumfulltime_1804, 2) the total duration in month of each of the different situation (i.e., bndurstudentstatuses, bnduremplstatuses, bndurselfemplstatuses, bndurhelpfamstatuses, bndurunemplstatuses, bndurretiredstatuses, bndurmilitarystatuses, bndurhomestatuses, bndurmatleavestatuses, bndurparleavestatuses, bndurilldisabledstatuses, bndurotherstatusstatuses, bndur1501, bndur1401, bndur1301, bndurparttime, bndurlastparttime, bndurstudwhilework), 3) the age of R at the beginning and end of part-time employments (i.e., bn876_xAgeR, bn877_xAgeR, bn878xAgeR, bn879_xAgeR).

The availability of these variables in each different country-specific file depends on the availability of variables used for their calculation.

IMPROVEMENTS INTRODUCED WITH V.1.2. in Nesstar GGS micro data files (April 2015):
Publication of variables that were previously deleted before dataset release in Nesstar. The following variables are concerned: grid variables, month and year variables, hours and minutes variables, frequency and unit variables, and occupation variables.

IMPROVEMENTS INTRODUCED WITH V.1.1. (March 2013):
- Variables corrected: bage, b204*, b510, b530, b626, bn871* (only one country-specific value),  
- Changes in the variable/value labels: b906b, bn1401sqq, b203b (Country specific list of childcare arrangements has been recoded in GGS values as follows: a) "1-Babysitter (nanny)" comprises "Assistante maternelle", "Nourrice", "Garde à domicile", and "Babysitter", b) "2-Day care centre" does not correspond to any value in French GGS and it says empty, c) "3-Nursery or Preschool" comprises "Crèche", "Ecole maternelle", and "Halte-garderie", d) "4-After-school care-centre" comprises "Garderie avant ou après l'école", and "Centre aéré", e) "6-Other institutional arrangement" comprises "Autre organisation". PLEASE NOTE THAT THIS CODING DEVIATES FROM WAVE 1 VERSION 4.1. In V.4.1. the group "Centre aéré" has been erroneously coded as "Self organized childcare group". In a future revision, wave 1 will be made consistent with Wave 2.)
- Variables now available: b203*  
- Variables dropped because all system missing: b875 (not implemented in the French GGS)
IMPROVEMENTS INTRODUCED WITH V.1.1. in the derived variables calculated for Nesstar GGS micro data files (March 2013):
- Variables corrected: b203c_*w and b204c_*w (previously frequencies greater than 7 were put equal to system missing, now they are put as equal to seven).
- Variables previously dropped and now available: bint_st and bint_pr.

FIRST DATASET RELEASED: V.1.0. (November 2012).

Notes

WAVE 2 DATASETS - Main differences compared to WAVE 1 datasets
Wave 2 datasets include an additional new section that had not been implemented in the Wave 1 data collection. It is the section no. 8 "Activity and Education History". Respondents report comprehensively on their activity and education history since age 16. Two additional sections are also present at the end of wave 2 dataset: "Interviewer observations" and "Interviewer report" (respectively sections no. 13 and 14).
A set of constructed variables at the top of the data file increase the usability of the GGS data by summarizing key socio-demographic characteristics of the respondent (age, birth year, sex, level of educational attainment, activity status, partnership status, number of co-resident partners, number of children, household size, household type). An additional set of variables consolidates information on the current activity of the respondent and his/her partner that is otherwise spread over the questionnaire. Another set of consolidated variables concern respondents' parents and parental home.

WAVE 2 DATASETS - Variables names
Variables in the Wave 2 data sets that are consistent with variables implemented in the Wave 1 questionnaire are named identically. Wave 2 variable names start with the letter "b" compared to letter "a" in Wave 1. Variables that have not been implemented in Wave 1 but collected in Wave 2 begin with "bn".  
In Wave 2 datasets published in Nesstar, the variable "brid - R identification number" has been renamed into "arid" (same variable name than Wave 2). This allows the user to merge Wave 1 and Wave 2 datasets in Nesstar.
In Wave 2 datasets published in Nesstar, variable labels have the indication "(W2)". This allows the user to distinguish Wave 2 variables from Wave 1 variables, on the basis of the variable labels.

Download

Metadata Index

This is the Metadata Index for a Nesstar Server.
Nesstar is a tool used for analysing, visualising and downloading datasets.

Click the "Explore Dataset" button to open the dataset.