Links to website Homepage Get data Introduction The database POPLINK is one of the world’s most information-dense historical population databases: it contains life-course data for 340 000 individuals and 1,2 million records between 1900 -1950. By combining POPLINK with the Skellefteå regions in POPUM, the individual-level linkage between historical and modern data is made available and covers almost 400 years. Shortname DDB IDS compatible Yes Period 1900 - 1950 Territory Skellefteå region, Umeå region, Sweden Category Longitudinal Contact information Organisation Demographic Data Base, Umeå University Web Address http://www.ddb.umu.se/english/?languageId=1 Location Umeå, Sweden Postal Address Umeå University, SE 901 87 Umeå, Sweden Contact Persons Anders Brändström Anders.email@example.com Citation DDB (POPLINK) QuestionnaireDownload questionnaireThe questionnaire was submitted 10 March 2015 by Annika Westberg. Scope / Status Original goal The database POPLINK is one of the world’s most information-dense historical population databases: it contains life-course data for 340 000 individuals and 1,2 million records between 1900 and 1950. The Demographic Data Base has digitized the Skellefteå and Umeå regions up until 1950 to enable linkage to modern registries and biobanks. By combining POPLINK with the Skellefteå regions in POPUM, the individual-level linkage between historical and modern data is made available and covers almost 400 years. Main sources: Parish registers such as catechetical registers, birth and baptism registers, banns and marriage registers, migrations registers, and death registers. Current status The database is under construction. The completed parts were constructed between 2009 and 2015.There is no clear date for its completion. Sample definition Complete registration of parish registers where individuals are followed during their presence within the included parishes. When choosing the target population, an important feature for maximizing the number of Complete genealogies over generations is a low mobility. Homogeneity is vital in the medical sciences, as it will help minimize confounding by population stratification, reduce selection bias and facilitate correct imputations of genotypes. The coastal area of Västerbotten has until the early 20th century been characterized by a very low population turnover. There is also optimal possibility of linking large modern registries covering the whole area. 1. Which sources form the basis for the sample: Full count of parish registers for parishes selected by the research community. Parishes are grouped in two main regions. Individuals are followed during their presence within these regions. 2. Sampling units: Complete registration. 3. Variables used for selection: Complete registration. 4. Selection method: Full count. Geographic area under observation Two geographically connected regions in northern Sweden:Skellefteå region (seven parishes),1620-1950 andUmeå region (two parishes), c.1900-1950. Keywords demography, life course, parish registers, church records, genetics, occupations, migration, population history, epidemiology, literacy, fertility, mortality, family intergenerational Sources Sources From yearEnd yearSourceExplanationPDF 1630 1950 Baptisms Include births and baptisms. PDF 1700 1950 Marriages from church registers PDF 1620 1950 Burials Include deaths and burials. PDF 1720 1950 Population registers (maintained by church) PDF Collection procedure • Data collection period: The data from 1900-1950 was transcribed 2009-2015. • Data collection method: Transcription from scanned original sources. • The transcription was done: By individuals from scanned original sources. • Automatic checks when transcribing, random sample checked by proof reading at the time of registration. • Purpose of the transcription: Research. • Control methods by researcher: Consistencies are checked by logical controls and computer programmes. Observations Units of observation Unit of observation ExplanationNumber Individuals 340,000 Married couples Families Households It might be difficult to identify households. Farms Depends on how the population register was kept. Can differ from parish to parish and from time to time. Are there any related observations that are not included in the database? Explicit information on related persons not present in the parish is included in the database (for example “daughter of farmer Nils Olsson”, or “Farmers daughter”). How do the units of observation enter observation? Birth, start of registration, in-migration. How do the units of observation leave observation? Death, end of registration, out-migration. Dates estimated Sometimes only year is given. Are some entry or exit dates unknown? Only in rare cases. Mainly for older periods (i.e.18th century). Can observations be linked to geographic locations? Yes Are the dates and locations of movements within the observation area recorded? Yes Are all individuals who lived in the households of sample members recorded? Yes On individuals: Dates on birth, baptism, death, burial and marriage. Legitimacy, age, gender, marital status, occupations with HISCO-coding, cause of death (ICD-10-coded), migration, relationships, literacy, smallpox vaccination. On households: Households might be difficult to identify. Families, including number of children (both biological and non-biological), are identified by relation and place of residence. Linkage • Which sources and units of observation have been linked: Births/Baptisms-Y, Marriages-Y, Deaths/Burials-Y, Population registers-Y. • Documentation of linking: We use a combination of computerized and manual linkage and link in three steps. First within the closest geographical unit (parish), then we link relations with parents and children and finally within a bigger geographical unit. • Software: CoreLink, RelLink and RegLink for computerized linkage. ManLank and SirLink as computerized aid when linking was done manually. • What are the rules for linking: Several different rules are used during computerized linkage. Key variables are date of birth, sex, first name and last name. All links are logged to be traceable. • How each reconstructed person is traceable to the original sources /transcribed data: Volume, page and row in the original sources are recorded for each individual. • How is linkage represented in the database: Every individual has a unique identification number.Every record has a unique identification number and is linked to individuals through the unique person identification number. • Linkage percentage: 97-98%. • Quality of linkage (own evaluation): 100%. Variables Events Variables On individuals: Dates on birth, baptism, death, burial and marriage. Legitimacy, age, gender, marital status, occupations with HISCO-coding, cause of death (ICD-10-coded), migration, relationships, literacy, smallpox vaccination. On households: Households might be difficult to identify. Families, including number of children (both biological and non-biological), are identified by relation and place of residence. Coding / Reference systems Occupational titles: Own coding system and HISCO. Locations (including geo-referenced systems): Religion, civil status, etc. Cause of death according to ICD-10 Data representation INDIKO: Web tool for extracting and visualizing data (mainly visualizing) for the c 1650-1900: http://www.ddb.umu.se/english/service/indiko--parish-registers-on-the-web/?languageId=1. DDB library: a set of standardized java methods for analysis and data extraction. CoreLink: computerized record linkage software. PERSONA: a new open source software for digitizing longitudinal data will be ready for use in late 2015 (http://www.ddb.umu.se/tjanster/v42---utveckling-i-forskningens-tjanst/). Kinship relations Recording A specific table contains information about related individuals. Given relations are to parents, partners and children. From this table sib ship groups can be created and families followed over generations. Depth of information Up to fifteen generations. Publications 1. Main publications about the database itselfEdvinsson, S. (2000). The Demographic Data Base at Umeå University - a resource for historical studies. In P. H. Hall, R McCaa, & G. Thorvaldsen, Handbook of International Historical Microdata for Population Research. Minnesota Population Center.Johansson, E. (2003). Church Records - Part I: From Orality to Reading Tradition. Church Records - Part II: Baptism, Teaching to Observe, and the Demographic Data Base (DDB). Opening Reflections. In Interchange, 34(2&3).Nilsdotter Jeub, U. (1993). Parish Records. 19th Century Ecclesiastical Registers. Demografiska databasen, Umeå.Vikström, P, Edvinsson, S, & Brändström, A. (2002). Longitudinal databases – sources for analyzing the life course. Characteristics, difficulties and possibilities. History and Computing, 14Wisselgren, M., Edvinsson, S., Berggren, M., & Larsson, M. (2014). Testing Methods of Record Linkage on Swedish Censuses. Historical Methods, 47, 138-151.2. Main or exemplary publications on research based on the databaseEgerbladh, I., & Bittles, A. H. (2011). Socioeconomic, demographic and legal influences on consanguinity and kinship in northern coastal Sweden, 1780-1899. Journal of Biosocial Science, 22, 1-23.Edvinsson, S., Brändström, A., Rogers, J., & Broström, G. (2005). High Risk Families: the unequal distribution of infant mortality in nineteenth century Sweden. Population Studies, 59(3), 321-337.Engberg, E. (2004). Boarded out by auction: poor children and their families in nineteenth-century northern Sweden. Continuity and Change, 19(3), 431-457.Maas, I., & Leeuwen, van, M.H.D. (2002). Industrialization and Intergenerational Mobility in Sweden. Acta Sociologica, 45, 179-194.Vikström, L. (2010). Identifying dissonant and complementary data on women through the triangulation of historical sources. International Journal of Social Research Methodology, 13(3), 211-221.