During the POLPAN Seminar on Tuesday, January 20th, Dr. Macin Zielinski (Institute for Social Studies, University of Warsaw) presented a paper titled ‘Problems with harmonization of the data on the household composition in POLPAN 1988-2013′.
The discussion will focus on two dimensions of the measurement of household composition in the POLPAN 1988-2013 survey-projects. The first focuses on data inconsistencies; the second on possible presentation of information on household members.
1. This part covers discrepancies, errors and other logical inconsistencies in the data. I will focus on comparability of the data itself more than on the way of the data presentation in the data set. Possible corrections and problems will also be discussed.
2. This part covers the comparability of measurement of the household composition between POLPAN waves. Measures used in subsequent waves of the study, problems with their comparability, and possible solutions will be discussed in the context of the between-waves data harmonization. In particular, two approaches will be presented: (a) Dis-aggregation of the data on household members, which allows to keep the whole existing information but leads to difficulties in data interpretation and usage; (b) aggregation of the data on household members, which makes the data easier to use but leads to the loss of some part of existing information.