Ethnostatistics

Conceptual Overview and Discussion

The meaning of statistics needs to be understood so the distinctive domain of ethnostatistics is clear. Statistics involves theories and techniques to manipulate data. It is the occupational home of statisticians and quantitative professionals who teach, develop, improve, and use statistics. Second, statistics is a label for rule-governed calculations. Third, statistics refers to almost any numerical summary that is the outcome of rule-governed calculations. Statistics aims to produce better counting from a technically correct perspective.

Formal texts and training provide procedures and guidelines for creating statistics, yet little is known about how statistics are created and used by their makers and users. Ethnostatistics, outlined by Robert Gephart in 1988, is a new field of methodological scholarship that studies how statistics are actually constructed, interpreted, and used by professionals. Ethnostatistics emerged from ethnomethodological research into informal practices needed to enact measurement and to interpret measures in social and organizational research. Ethnomethodology demonstrated that formal accounts and practices of science fail to describe fully what scientists actually do—the content of scientific practice. Scientific practice relies in part on commonsense reasoning and practical choices. Ethnostatistics explores the commonsense reasoning and practical choices in statistics production and use.

To study tacit knowledge and informal practices, ethnostatistics investigates the activities in which statistics and measurement are locally produced and practically relevant. Ethnostatistics explores the immensely varied practices that terms like measurement and statistics gloss over. It produces highly detailed and vivid case descriptions of the locally organized production of statistical or quantitative knowledge. Ethnostatistics thus enriches our understanding of methodological aspects of quantitative social science and provides a qualitative perspective for understanding quantitative social research methods. Ethnostatistical [Page 354]case studies thus assess quantitative research practices, challenge methodological conventions, and suggest improvements in quantitative research.

Ethnostatistics explores three levels of analysis. The first level, producing a statistic, uses anthropological techniques including observations and ethnography to develop case descriptions of the work activities of quantitative professionals as they create statistics. Ethnostatistics then undertakes a second level of analysis by putting statistics to work in computer simulations that assess underlying practical and technical assumptions made in producing statistics. These studies explore how alternative assumptions would influence outcomes and interpretations of statistical measurement in specific research contexts. Third-level ethnostatistics uses literary analysis to explore the rhetoric of statistics—how textual practices persuade readers that the statistics reflect truths.

...