I am a Research Scientist working the IPUMS projects, primarily responsible for historical data processing. I have led many historical census data construction projects over the past 25 years at the University of Minnesota, beginning with helping to create the public use sample of the 1880 U.S. Census while working toward my doctorate in History. Since 2000, I have served as the Data Services Manager at the Minnesota Population Center. In this position I have managed and supervised all data entry and processing of historical census data at the MPC. Typically, this work involved the construction of samples (one or five-percent density); more recently the emphasis has shifted to the processing of complete-count databases for the United States decennial censuses from 1790 through 1940. The scale of the complete-count databases has required considerable collaboration with the team of software developers (Science and Technical Core). Much of this work utilizes machine learning techniques to (1) assign numeric codes to the original alpha strings, (2) automate logical edits designed to correct transcription errors, and (3) construct machine-generated links between individuals enumerated in different censuses.
My substantive research has drawn upon historical census samples to study such topics as changing prevalence and characteristics of boarders and lodgers in U.S. households. I have published on methods of linking individuals between census samples, coding complex census variables, and played a leading role in creating documentation for historical U.S. census samples.