Family Economy of 19th-Century Industrial Workers

George Alter, ICPSR, University of Michigan
Author Profile
This material is replicated on a number of sites as part of the SERC Pedagogic Service Project


This set of assignments leads students through the steps of identifying a research question and designing an empirical analysis. Students use a large collection of late 19th-century family budgets from the U.S. and Europe, which lend themselves to questions about consumer decisions and the labor-force participation of women and children.

Learning Goals

The activity is designed to teach students how to translate a research question into a hypothesis for empirical analysis. Students should learn the importance of descriptive statistics for understanding data, and they should begin to think about controlling variables that are not part of their hypothesis.

Context for Use

This assignment was developed for a course in quantitative methods for graduate students in history. It was also used in a 4-week course at the ICPSR Summer Program in Quantitative Methods. The assignment works best for students who will be able to use multiple regression by the end of the course.

Description and Teaching Materials

The activity consists of a sequence of projects
Project Reports for the Cost of Living Survey (Acrobat (PDF) 22kB Sep28 09)

Teaching Notes and Tips

There were several aspects of this assignment that helped to make it successful. First, I gave them a dataset (the 1888-90 US Cost of Living Survey) which already had a substantial literature. Students had to come up with their own hypotheses, but I provided a bibliography to get them started. Second, they wrote a sequence of "project reports," but they did not write a "term paper." This was intended to get them to think through the steps of defining their project before they actually looked at the data. It also broke up the project across the semester and allowed more time for me to work with them. Third, I emphasized that any reasonable hypothesis was acceptable, and that their grade would not be affected if the data did not support their hypothesis. This takes away the pressure to find the "right" answer. Fourth, this is a assignment works very well as a group project.

Instructors should also plan on time to teach students how to use statistical software.


The assignment was used with frequent in-class discussions and meetings with students. Since assignments were distributed acros the semester, there were more opportunities to monitor students' progress and their understanding of the research process and the substantive material in the course.

References and Resources