The Limits to Bureaucratic Observation: On the Role of Judgment in Comparative Political Measurement

30 Pages Posted: 19 Jul 2010 Last revised: 24 Aug 2010

See all articles by Andreas Schedler

Andreas Schedler

Central European University (CEU) - Democracy Institute

Date Written: 2010

Abstract

Standard methodological advice in the social sciences tells us to base our measurement decisions on observations, not judgments. It presupposes that we divide the process of measurement into two phases. In a first judgmental stage, we reach all the judgments necessary to build our theories, form our concepts, delineate our objects of observation, and devise our coding rules. In a second observational stage, by contrast, we suspend our judgmental faculties in order to apply our formal rules of measurement in a quasi-bureaucratic fashion. However, does the full bureaucratic regulation of observation and measurement represent a feasible ideal in the comparative study of politics? In the present paper, I explore the limits to “pure” observation in three steps. In the first section, I describe the “bureaucratic ideal” of social measurement. In the second section, I review four conditions that may compel scholars to incorporate elements of judgment into their measurement processes: unobservable realities, unobserved realities, unexpected realities, and complex concepts. In the third section, I offer some thoughts about methodological standards that may guide judgmental modes of comparative measurement.

Keywords: measurement, cross-national data, comparative politics, observation, judgment

Suggested Citation

Schedler, Andreas, The Limits to Bureaucratic Observation: On the Role of Judgment in Comparative Political Measurement (2010). APSA 2010 Annual Meeting Paper, Available at SSRN: https://ssrn.com/abstract=1642035

Andreas Schedler (Contact Author)

Central European University (CEU) - Democracy Institute ( email )

Hungary

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
37
Abstract Views
1,035
PlumX Metrics