Digital Research Learning Curve: Practical Lessons from a Seven-Year Historical Census Database Project

AHA Session 170
Saturday, January 7, 2012: 2:30 PM-4:30 PM
Chicago Ballroom A (Chicago Marriott Downtown)
Sarah L. Franklin, University of North Alabama
The Future Is Here: Digital Methods in Research and Teaching in History
Setting up Shop: Things You Need to Do Before You Begin
Rodney D. Anderson, Florida State University
Getting It Right: Error Detection and Verification Procedures
Claudia P. Rivas Jiménez, Universidad de Guadalajara
So What Good Is It? Quantitative Data as Qualitative Evidence
Tam Spike, North Georgia College and State University
The Audience

Session Abstract

Our panel explores the benefits and challenges of historical database construction, offering practical lessons from The Guadalajara Census Project, 1791-1930 (GCP), funded by the U.S. National Endowment for the Humanities, Division of Preservation and Access, from 1999 through 2006.  The project datasets derive from manuscript census returns from three centuries, making it one of the largest urban census databases currently available--nearly 150,000 cases, mainly individuals, their families and households (thirteen million pieces of data).  The point here is not the quantity, but the extensive experience the GCP gained in database design and implementation. Forging this project in the midst of changing computer technology, from floppy disks to DVD-ROM, the GCP staff constantly revisited basic procedures and policies.  It is that practical experience which will be the focus of this practicum. 

Our session has little to do with Guadalajara but rather deals with issues common to all historical database construction.  In particular, this session is aimed at a new generation of young historians who come to the craft already comfortable in computer and digital technologies, but who might think that database construction requires a background in quantitative methods.  In fact, windows technology has led to a vast simplification of analytical procedures, resulting in a “deskilling” of quantitative methods.  Now scholars with little or no quantitative training can easily incorporate data sets into broader qualitative objectives. 

Each panelist will cover a different aspect of the project: planning, variable construction, data coding and entry, verification, and preliminary research findings.  In covering these topics, we will focus on what we regard as a set of “best practices” in historical database design, and point out the pitfalls inherent in the process.  The papers will illustrate hands-on procedures through on-the-screen examples, emphasizing how software tools work, and why certain procedures were adopted over others.  We will highlight the interdisciplinary nature of database design, including how Geographic Informational Systems (GIS) function in a historical database format.  We will also present tentative findings relevant to many historiographic problems and to a variety of historical approaches.  We will show, for example, that census manuscripts may be deconstructed as cultural text.  Among social and cultural patterns relevant to current historiographic debates, we will illustrate how tracking mechanisms provide clues to kinship and social networks, and how aggregate data analysis disguises important variations in the real-time experience of individuals.

In brief, historians pursue research in a variety of venues, and from many different methodological approaches.  In this practicum, our objective is to show that database construction can be a useful, easily learned, tool in much of that research.

See more of: AHA Sessions