Full Citation

Title: Intuition to Recognize and Address Sample-Selection Bias in Historical Sources

Citation Type: Miscellaneous

Publication Year: 2017

ISBN:

ISSN:

DOI:

NSFID:

PMCID:

PMID:

Abstract: The idiosyncratic availability and survival of historical records often forces quantitative social science historians to rely on sources describing only a subset of the population that they wish to study. Whether and how such sources can be used to study this population is a key concern. In this paper, I illustrate the intuition of an approach that enables researchers to recognize the presence of sample-selection bias—a systematic difference between the sample and the population of interest that can render conclusions drawn from the sample uninformative about the population of interest—and to draw conclusions about the broader population without ignoring this potential pitfall. A key part of this discussion concerns the role of a variable that affects whether an individual is observed but not the outcome of interest. I focus specifically on differences in average stature between the Northeast and the Midwest in the antebellum United States as observed in historical military data. Using voting patterns in the Presidential Election of 1860 as this variable, I show what simple patterns in the military data can indicate the presence of sample-selection bias, and outline how the variable enables the researcher to understand the potential impact of this bias.

Url: http://www.ariellzimran.com/zimran_selection_ssh.pdf

User Submitted?: No

Authors: Zimran, Ariell

Publisher: Vanderbilt University

Data Collections: IPUMS NHGIS

Topics: Methodology and Data Collection, Other

Countries: