I have a question regarding childl.dta and youthl.dta. The years of birth in these files are limited to 2000 and 2002, respectively. I wanted to ask whether youthl and childl data are also available for birth years before 2000? I need a data with birth years starting from 1991 at least.
Child data was first collected in 2003.
Youth data for 17-year olds is available earlier.
Younger children: no data before 2003 to best of my knowledge. Check the old ‘bioagel’ codebook for more information about data collection
Dear Miriam, thank you for your answer. However, it does not resolve the problem completely. If I look at the variable birthy in youthl.dta, I see that only individuals with birth years from 2002 to 2010 are available. The question is where the individuals who were born earlier are. There are also many observations with the values Frage in diesem Jahr nicht Teil de, in Fragebogenversion nicht enthalt, or unplausibler Wert.
Hey Victor,
yeah, I see you problem with the youth now, and I think I can offer you a solution - to the best of my knowledge.
I suppose the youth in youthl with birthy < 0 (any missing) have been observed before 2002. Before the child survey (age 0-14) started in 2003, they already had a youth survey in which they asked youth at age 17 only. The data from this youth survey is contained in jugendl.
So, what you need to do to get the birth year of those with birthy < 0 is the following.
Merge youthl and jugendl alongside pid and syear.
Then ‘tab birthy jl0233, m’
You will see the birth years that are missing in birthy contained in jl0233 from jugendl and you can use it from there.
Good luck.
Let me know if it works!
And read the documentations of the datasets that you want to use closely, as well as the documentations of the earlier version of this data (bioagel, biopupil, jugendl, in your case). It will help you understand this complex missing structure better and might make you go a little less crazy. Cause the new youthl and childl is harmonized but I think it is far from perfect now.
A question back to you: Did you see any variable containing info about child age / youth age?!
In earlier versions (bioagel, biopupil), ‘bioage’ contained age info. This variable has gone now. So I think we need to generate age from survey year and birth year?! Thanks! 
Hey Miriam,
Yes, it works! Thank you! Regarding the age variable, I also do not see it in the data, so I think you should generate age at the time of the survey using the survey year and birth year.
Dear Victor and Dear Miriam,
Thank you for working with the SOEP data and especially working with the information in Childl and Youthl. Both datasets are new and have been completely revised, so we are always grateful for any feedback on working with them.
Information to the childl-Dataset you find e.g. here. The childl dataset replaces the previously existing bioagel dataset from version 39 onwards. It contains the information from parent-child questionnaires collected from children aged 0 to 11 years since 2003, as correctly mentioned. The youthl Dataset exist since version 40 and contains certainly current information from the formerly jugendl-dataset and biopupil-dataset. Both are no longer updated.
Information on age you find in the ppathl-dataset. We recommend using the ppathl dataset as a basis and then merging the information from the individual datasets 1:1 via pid and syear.
In the dataset biol you find also some information concerning children of the respondents. Perhaps you’ll find what you’re looking for there.
Please feel free for any questions and any feedback.
Jana Nebelin (part of the SOEP Team)
1 „Gefällt mir“