Name: Sibling data (Syskondata)
Period(s): - (SMS) Code book: V

Information
This section presents sibling data on the cohort members, which was collected from three sources; 1) Delivery Records (information about older siblings born before 1954), 2) Population Registration Records (mantalsskrivning) and 3) The 1970 Census.
The data set contains variables on the number of siblings, birth date (year and month) for each sibling, sex and the number of older and younger siblings. The variable IDMAN is used to determine if the sibling is older or younger than the cohort member when they have the same birth year.

The definition of siblings is rather broad in this data set. The information on the elder siblings have been gathered from the delivery records (which contain information as regards to whether or not the mother has given birth to children prior to the cohort member(s)), while information gathered on the younger siblings have been excerpted from the Register of population and income 1964 and the Census register data. The latter registers do not differentiate between biological siblings and other children who live within the household, i.e. children who have biological relations to only one of the parents.

Marja Walldén, professor emerita at Stockholm University, has written numerous reports using the Metropolitan data on the importance of sibling positions. Her definition of siblings is indeed more restrictive in nature than the available data. Her studies are delimited to 11,200 of the 15,117 cohort members since siblings have to be related to the same biological mother (measured indirectly through estimations). Walldén compared the data from the register of income and population (Mantals- och inkomstlängden) from 1964 with the Census data from 1970. If any “siblings” to the cohort members (born prior to 1964) were found in the latter data set but not the former, the assumption was made that the “sibling” did not share the same biological mother as the cohort member in question and that they cohabited with one another for other reasons. Walldén further chose to exclude cohort members with very large number of “siblings”, especially those who were born during the same year. The data set that she has used for her studies is available although not presented herein since clear operational definitions to the various variables are not available at this moment in time. Further information on these studies can be found in Walldén (1982; 1990 and 1992).

The data set was originally comprised of 15,091 observations. Since the cohort originally includes 15,117 observations, 26 observations must have been removed for some unknown reason. Additionally, two observations have been excluded from this presentation due to the lack of identification number. One of these observations only has missing values in the variables (except for the variable TVILL, where the observation was given an invalid value).The other observation, which was deleted had values in the other variables, was given the value of 23 in the variable ANT_SYSK (the number of siblings), which is an unrealistic value (especially when many of the siblings were born during the same year).