Determining profiles of Turkish families by number of children: an application of data mining technique
Sebnem Bese Canpolat, Turkish Statistical Institute (TurkStat)
Kadir Korhan Babadag, Turkish Statistical Institute (TurkStat)
In Turkey, one of the most widely debated issues in recent times is a continuous decrease in fertility rate and economic and social outcomes that will occur with the ageing of the population. In 2008, Turkey’s Prime Minister commented that to ensure that the Turkish population remains young, every family would need to have at least three children. He has repeated this statement on numerous occasions. In this context, main purpose of this paper to examine the link between profile of Turkish family and the number of children they are raising. The profiles of families having 1, 2 and 3 children have been worked to find out the profiles specific to families with 3 children and more and to determine geographical, socio-economical and environmental factors that impact on the number of children families have. Another objective of this study is to use the data mining technique which is a new method employed in analysing demographical data. Data mining allows users to analyze data from many different aspects, categorize it, and discover potentially interesting and useful patterns of information embedded in large databases. I used data mining techniques to discover which attributes have the highest impact on number of children. The primary data used in this study are the 2006, 2007, 2008, and 2009 Income and Living Conditions Survey’s data. Since 2006, TurkStat has started to carry out “Income and Living Conditions Survey”, where the panel survey method is used, within the scope of the studies compliance with European Union (EU). According to initial findings, geographical regions is the most important factor that affect the number of children. Another factors are dwelling tenure status, receiveing some regular payments from members of other private households, dwelling type, total income, and social assistance payment for children under 16 years old, respectively.
Presented in Poster Session 2