Rectypes 2000
1. VARIANZ
In addition to the household indicator this file contains the variables STRAT1, STRAT2, SAMPOINT and INTNR. Some software packages (such as STATA, SUDAAN) are able to use these to estimate variances. All four variables provide information on the respective subsample for the start of each first wave, i.e. they are saved at the case-level (variable HHNR).
STRAT1 identifies the levels, which were relevant for pulling the Primary Sampling Units for the respective sample. For subsample B, these were the five nationalities. Therefore, "artificial" levels were created for subsample B corresponding to the other subsamples and filed under STRAT2.
The variable SAMPOINT identifies the respective PSU (e.g. in subsample A voting constituencies, in Subsample D not present).
Due to data protection laws the various values of the variables STRAT1, STRAT2 and SAMPOINT were given transformed values, in order to prevent regional units from being identified.
The variable INTNR is a variable to which every interviewer assigns a number, so that clusters of households that were surveyed by the same interviewer can be identified.
2. HBRUTT00
Similarly to the collection of the supplementary sample 1998 (sample E), this file contains all Brutto information from all households in the Innovation Sample in the year 2000 that were recently surveyed using the Random-Route-Method. In this case, it doesn't matter if these households were successfully surveyed or not. Information such as this can be accessed for the use of methodical investigations through the participation of households in (SOEP) surveys.
3. QJUGEND
In the year 2000, a youth questionnaire was introduced to be used instead of the biography questionnaire. This was aimed at all "new" participants who had reached the minimum age of 16 and were therefore able to take part in the SOEP survey. The 232 data sets that exist as of now supplement the information collected from the likewise first-time answering of the person questionnaire, in order to gain retrospective details on education, as well as basis indicators on education success. A thorough revision, as well as a supplementation of the youth questionnaire indicators took place in 2001, in addition to the fact that the youth participants of sample F took answered this new questionnaire for the first time. As a result, the data set QJUGEND represents, so to speak, a type of pre-test for the recently prepared biography data set BIOYOUTH (available from 2001 onwards).
Reworking of labels
The VAR LABELS and VALUE LABELS have been be completely reworked for all previous years (up to and including 1999). Missing labels were included where applicable and the systematic was standardised (for instance for sub-items or variables with just one answer category). Furthermore, the labels were made consistent over time. At the same time the reworked label text was transferred to the English labels, so that these too were retrospectively fully identical to the German systematic.
$PGEN 2000
For the current data distribution, extensive revisions were made to the variables from earlier waves. For instance, note that there are far fewer missing values -1 (k.A.) for many variables related to the occupations. The education variables in all $PGEN were reworked and supplemented. New variables include a differentiated labour force status for all participants and education information generated on the basis of data first collected in the year 2000 which dealt with the highest level of education and employment achieved up till now. The existing generated education variables were retrospectively reworked, extrapolated, as well as supplemented: you will now be able to access data on the temporarily absent respondents, as well as information on current school attendance, apprenticeship or studies. Furthermore the variable BETR$$ in $PGEN was recoded (the data on the size of the firm and therefore the codes in SOEP have changed over time). We would like you to take this into account when updating programs.
Contact: Jürgen Schupp and Peter Krause
$PEQUIV 2000
The $PEQUIV files were updated. This affects:
Contact: Markus Grabka