This five-day course covers the concepts and skills needed for analyzing and interpreting quantitative data collected as part of ethnographic field research. Researchers will learn how to: (1) develop quantitative measures of behaviors, attitudes, and material objects; (2) provide group-level summaries of quantitative data; (3) frame expectations about group differences and relationships between variables; (4) test those expectations with quantitative data; and (5) justify why a specific test is appropriate for a given kind of data.
In addition to lectures, the course involves class activities, visualizations, and analysis of real data, to illustrate the main concepts and skills and to walk participants through the steps of quantitative data collection and analysis. A supplemental web site contains primary course materials—lecture powerpoints, readings, activity modules, and datasets analyzed in the course.
A key goal of the course is to familiarize participants with techniques for analyzing the kinds of quantitative data commonly collected as part of ethnographic field research. These include:
- Scale construction: creating a quantitative estimate from several measurements
- Summarizing and graphing quantitative data
- Data exploration: multi-dimensional scaling and cluster analysis
- Testing hypotheses about group differences and statistical relationships
- Application of the cultural consensus model
- Basic linear regression, including identifying statistical confounding
- Assessing the assumptions underlying a given test
We will begin with an overview of common approaches to quantifying perceptions, attitudes, and behaviors, focusing on specific issues to address in developing a quantitative measure. We will then turn to basic approaches to summarizing quantitative measures, comparing measures across individuals and populations, and characterizing relationships between measures. These initial sessions will serve as a foundation for later sessions focusing on scale construction, visual data exploration, testing for group differences, applying the cultural consensus model, and fitting a basic linear regression. In addition to these basic skills, participants will be given a roadmap to more complex models and tests which they may encounter in their research, including logistic regression, repeated measures, and statistical interactions.
During the course, participants will practice analyzing data using R Commander, a user-friendly (and free) graphical user interface for the open-source R statistical programming language. This will give beginners the opportunity to learn basic statistical procedures in a user-friendly “point-and-click” environment while also giving participants some exposure to the written commands used to call those procedures in a widely used (and free) statistical package.
The course readings are linked on the schedule.