School of Humanities and Social Science Division of Social Science 226 Reconstructing Historical Lives from Demographic Data Supervisor: GIETEL-BASTEN Stuart Arthur / SOSC Student: LIU Junde / QSA Course: UROP1000, Summer Data Linkage is a hefty operation that is both time consuming and, more importantly, drains a ton of computational power. It involves running a matching-algorithm over every single possible combinations of records and determining whether or not there is a match. The data linking operation detailed in this report is completed by an AWS cloud server with 96 CPU cores running fully-loaded for 10 hours. Althrough the price to pay is high, the potential rewards for data linkage is also high. For a start, a well-linked dataset give the researcher more variables to work with. In our case, a linked dataset would allow us to look at the birth side and the death side of the equation together, instead of separately. Also, linking the datasets is the ultimate test to see if there’s any error present in the datasets. If the match rate falls short to our expectation, the only explanation would be that there are significantly amount of human error in the transcribing process. With this knowledge, we can then discover what went wrong in the transcribing process and then devise strategies to address the issues. Reconstructing Historical Lives from Demographic Data Supervisor: GIETEL-BASTEN Stuart Arthur / SOSC Student: MAHBUBANI Esha / SBM Course: UROP1000, Summer As a group, we were tasked to input data from Wilmington, Delaware in the late 1800s to early 1900s. This data was taken from scanned records using FamilySearch’s services, where we looked at births and deaths in the region. Since the records were physical copies, we were tasked to interpret the records, then inputting the data into an excel sheet to make for easier analysis. The purpose of this data was to be able to identify trends based on different variables, like race, sex, location, and more, in relation to the births/deaths. These trends could then be compared to the norm for that time. For example, one area we could look at is the infant mortality rate, based on sex, to see whether female or male babies tend to survive longer. Similarly, we could look at the average death between the sexes and then conclude the reasoning behind it based on the era’s treatment of women compared to men. Reconstructing Historical Lives from Demographic Data Supervisor: GIETEL-BASTEN Stuart Arthur / SOSC Student: WANG Xiaohan / GCS Course: UROP1000, Summer As a group, we were tasked to input data from Wilmington, Delaware in the late 1800s to early 1900s. This data was taken from scanned records using FamilySearch’s services, where we looked at births and deaths in the region. Since the records were physical copies, we were tasked to interpret the records, then inputting the data into an excel sheet to make for easier analysis. The purpose of this data was to be able to identify trends based on different variablessuch as race, sex, and location in relation to the births/deaths. These trends could then be compared with the norm at that time for further analyses.
RkJQdWJsaXNoZXIy NDk5Njg=