Information for Authors of Innovative Modules¶
The information on this page is explicitly aimed at the authors of innovative modules. In the following, innovative modules are abbreviated as “Inno-Modules”. As the author of an implemented Inno-Module, you will not only receive the normal Data Distribution File of the SOEP-IS but also a minimally edited version of the raw data of the year in which your module was surveyed. This is because the generation of the official Data Distribution File takes a certain amount of time. But by receiving the raw data you can work shortly after the survey has been completed and thus use the data from your module without having to wait for the official release of the data. However, there are a few aspects associated with the use of raw data that should be known and taken into account.
Structure of Raw-Data delivery¶
Important Note: The term “raw data” in this context does not refer to the same as in the SOEP-Core. While in SOEP-Core the processed cross-sectional data is referred to as “raw”, in SOEP-IS this is actually almost unprocessed raw data that we receive from our study institute. Although this is also cross-sectional data, it is only subjected to a data check and minimal editing. The datasets are only forwarded in this form to the authors of the innovative modules so that they can work with them at an early stage. For the official Data Distribution File, the datasets are integrated into the long datasets in a user-friendly way and published exclusively in this optimized format.
In the raw-data delivery we only provide the datasets of the current survey year. The data of the official Data Distribution File have to be ordered separately and will not be included. However, all the provided datasets in the raw-data delivery include the needed identifiers to correctly merge or append the data with the Data Distribution File.
On the first look, the structure of raw-data differs from the official data distribution file. Below you can find a screenshot showing an example of a raw-data delivery (top-level folder):
The included readme provides a short overview of the included files and may contain further important information. The dataset(s) in the top-level folder contain the respective variables from the innovative module of the author(s). The datasets of innovative modules from other authors are not included. The “SOEP data” folder contains the rest of the survey data as well as the questionnaire PDFs (English and German):
The German questionnaire is the one that was actually used for the survey. The English translation is only created for documentation purposes. The following table provides a short overview of the datasets in this “SOEP data” folder:
Dataset Name |
Explanation |
SUF equivalent (long data) |
|---|---|---|
soep-is-20$$-hbrutto |
Gross Dataset on household level |
hbrutto |
soep-is-20$$-pbrutto |
Gross Dataset on individual level |
pbrutto |
soep-is-20$$-hh2 |
Dataset from Household Questionnaire |
hl |
soep-is-20$$-ll2 |
Dataset from Biography Questionnaire |
biol |
soep-is-20$$-pe2 |
Dataset from Individual Questionnaire |
pl |
Provided the data from the official Data Distribution File is already available, the raw datasets can be appended or merged with the corresponding long datasets in order to add the new years to them and receive a long dataset.
It is also important to understand that the raw dataset of the biography questionnaire (“ll2”) only includes the individuals who answered this questionnaire in this wave. The data of individuals who have already completed the questionnaire in previous waves is only in the “biol” dataset of the Data Distribution File. For more information regarding the biography questionnaire see Individual Questionnaire.
Limitations of Raw-Data¶
Missing variables
Generally, we do not provide all the variables with the raw data that are available to us. This usually has to do with the fact that we want to keep the data as clear and easy to use as possible and do not integrate any variables that require additional data protection. Nevertheless, there is some information that we can provide on request or generate relatively fast. This includes, for example, basic regional data (such as federal states or east/west affiliation), durations of the items, open ended questions (string variables), date of the interviews, etc.
Cross-study variable names and labels
We aim to maintain the highest possible comparability with the SOEP Core in the SOEP-IS data. This is why we use identical variable names and the same variable labels wherever possible. In the case of versioned variables, this means that the version designation of the SOEP-Core is adopted, regardless of the actual version in the SOEP-IS. For more information an that please see “What do the suffixes “_v1”, “_v2”, or “_h” in variable names mean?”.
Generated data, weighting and official release
As already mentioned above, the raw-data delivery only provides the datasets of the current survey year without any generated data. This means that no weights or imputed variables are included in the raw-data. If it is necessary to merge the raw data with generated variables of the same survey year or to run weighted analyses, it is necessary to wait for the official release of the data. The generation process usually takes about a year. The data release is always scheduled approximately for the second quarter of the next year. However, due to the data embargo for the data of inno modules (see Inno Dataset), these variables will not be part of the official Data Distribution File. It is therefore necessary to merge the raw data with the Data Distribution File of the next year in order to use the weights of the same survey year.
Example of relevant dates from the survey year 2022:
2nd quarter 2022: Start of data collection
4th quarter 2022: End of data collection
1st quarter 2023: transmission of the raw data to the authors of the inno modules
2nd quarter 2024: publication of the official Data Distribution File of syear 2022 without the inno module data of 2022, but with generated variables, imputation and weights
2nd quarter 2025: publication of the official Data Distribution File of syear 2023 together with the inno module data of 2022 (one-year embargo)
Please note that the time schedule for the survey may change. This example is only intended as a general orientation. For detailed information on the survey years please see the respective wave reports.