How can I download the micro data?
Before you follow the download instructions, please make sure having a zip programme installed on your computer. Please download the open source zip programme 7zip if necessary.
I have troubles to open the zip file. What can I do?
Why is a variable from the GGS Core Questionnaire missing in a specific country survey?
Although we encourage the countries to strictly follow the GGS Core Questionnaire, countries might not have implemented the full questionnaire. The absence of a variable indicates that this question has not been asked in the specific national survey. Please refer to an overview on the availability and country specificities which is available on GGS Data Description
Why do some variables contain a value that is not listed in the codebook?
Although we encourage the countries to strictly follow the GGS Core Questionnaire, countries might implement a questions with a country specific notion. These country specific values of a variable are coded as a four to five digit long value starting with the country code (see variable acountry). Please refer to an overview on the availability and country specificities which is available on GGS Data Description
Why does some variable have an extension that is not listed in the codebook?
Although we encourage the countries to strictly follow the GGS Core Questionnaire, countries might implement a question that differs to a considerable extent from the Core Questionnaire. In that case we add a four digit long extension to the variable that starting with the country code (see variable acountry). Please refer to an overview on the availability and country specificities which is available on GGS Data Description
Why are there a small number of inconsistencies in some people’s sex and age across waves?
The sources of inconsistent data can occur at the moment of data collection as well as during data processing. First, different respondents might have been interviewed across waves. It could be that the interviewer does not come across the same respondent as in Wave 1 and unwittingly interviews another person, for instance, another household member of the respondent in Wave 1. Or it could be that there are issues with inaccurate recall about birth dates amongst respondents, particularly the elderly. Second, errors can occur in the data entry phase when paper and pencil filled questionnaires are entered into a statistical software computer package. The person digitizing the questionnaires might make typing errors in the data base. Third, respondents might be erroneously linked across waves. This is a realistic scenario in contexts where paper and pencil modes of data collection are applied. A questionnaire containing a wrong or unreadable identification number could lead to the situation that the linking across panel waves might be based on socio-demographic characteristics of the respondent. When more than one respondent in the sample reports the same birth month and year they might be incorrectly linked across waves.
We therefore recommend you consider such factors, particularly when analysing data from countries that used a paper and pencil questionnaire. After pooling the data of both waves, we recommend creating a variable that identifies for each respondent who participated in both Wave 1 and Wave 2, whether birth month and birth year were reported consistently. This variable takes two values: The value 0 indicating that both variables contain identical information in both waves. The value 1 indicates that one or both of the variables show an inconsistency across waves. For respondents with a score 0, we can recommend the use of the GGS as a panel without restrictions. For respondents with a score 1, we find at least one inconsistency on at least one of the variables. We recommend using these respondents for panel analyses with caution, advising data users to run robustness checks including and excluding those respondents.
The easiest solution is to use the using option when opening a stata file. For example:
- use [Data filepath], using(a301-a401a_a a601-a701)
This would only open sections 3 & 6 of the dataset but you can adjust it so it includes the sections you want to include or exclude.
The variables arid and brid represent the unique ID for individuals in the wave 1 & 2 data and can be used to match respondents across the waves.