Hello Lior,
1) Is there a number of years of education variable?
Ans: Not number of years of education, but highest educational qualification: w_hiqual_dv
2) Are there variables for the age of each child/the year each child was born, and not only for the birth year of the first child (ch1by4)?
Ans: The file w_indall includes age, sex, marital status information for everyone in an enumerated household. This file also includes variables which represent the identifier of the natural/step/adopted father and mother (fnspid mnspid). You can use those variables to reverse-identify children in that household. As the DOB for everyone is in this file you can also identify the DOB of the children. This only applies to co-resident children. In Wave 1 (and in Wave 6 for IEMBS) there are files a_natchild and a_adopt (f_natchild, f_adopt) which has information on all children ever born or adopted. After that the files w_newborn record the DOB of all children born since the last interview (if they are co-resident then you can also find out their DOB from w_indall file as describe before).
3) Are there variables for the gender of the children?
Same answer as 2)
4) Is there a comprehensive variable for the date/year of current marriage? I found variables such as lcmar1y and lmary4, but they seem to be limited to only few waves.
This is asked in the first intervew, after that changes since the last interview and their dates are recorded. So, you can use previous waves to identify when the current marriage was first reported and use the start date. Alternatively, you can use the partnership history file (https://beta.ukdataservice.ac.uk/datacatalogue/studies/study?id=8473#!/details) which lists the different partnership spells along with start and stop dates and partner id. You can match that file using PIDP and then using the interview dates to identify the current marriage and then the start date for that marriage.
Hope this helps.
Best wishes,
Understanding Society User Support Team
Updated by 
Updated by