From Research Questions to Data Acquisition

Moving from research questions to data acquisition is a creative act; there is no formula. But here are some tips.

1) Don't embed your assumptions in your research questions, especially not as binaries:

Arias: "How does the increasing participation of electrical and computing engineers in public-oriented ICT organizations... help stabilize or destabilize capitalist discourses of the information society"--as it turned out, neither alternative was true: rather he found "the emergence of a type of engineering oriented to serve social needs that both the market and the State leave unattended."

Williams: "Do Nepalese scientists subscribe to the ideology of Big Science (and the liberal norms of scientific practice)"?--even western scientists don't always subscribe. Better: "what role do western scientific norms play in the way Nepalese scientists conceptualize their own scientific institutions and practices."

2) Use hierarchy to make increasingly fine-grained questions. See example from Richard Arias' proposal.

3) You cannot ask your informant to do your work for you. Arias never asks a question using the phrase "capitalist discourse." Williams never asks using the phrase "ideology of big science."

4) As an aid to developing questions, brainstorm various causal mechanisms: For example if you study scientific beliefs, you might imagine various causes: "Scientists think this because" (their data suggested it, they read a science fiction book, they are influenced by religious belief, corporations pay them, etc.). Now think about data acquisition (such as interview questions) that would expose such causal mechanisms (from both directions: "how did you think of this" and "what science fiction have you read"). Note that some mechanisms will not be revealed by interview, and even strongly denied (e.g. corporate influence). Be strategic in data acquisition: choose the right tool for the right job:

A. Various Research Experiments

a. Survey: for example we used the Bath County Computer Attitudes Survey to determine if there is a change in student increase in attitudes towards computers. Survey was carried out pre and post intervention. Applied to two classrooms, one control (using normal math website) and one experimental (using our ethnomath website). This is a "quasi-experimental" evaluation (not "true" randomized experiment). Results here.

b. Intervention in a scientific field: examples such as Brian Martin, "Sticking a needle into Science" and Deborah Heath, "A Modest Intervention."

c. Collaboration with laypeople: examples such as DiSalvo et al, participatory action research (eg "bucket brigade"), participatory design, etc. The laypeople can be positioned on a spectrum from dependance to independance.

d. Collaboration with/as scientists: examples such as Phil Agre's "critical technical practice" and Pheobe Senger's "Recoding Affect" project.


B. Various Ethnographic Approaches

a. Cultural probes: various means to elicit information, typically used in design research.

b. Interview: this can include either structured or semi-structured questions; or a biographical approach (eg "life history", topical history, "describe a typical day", etc.).

c. Participant observation -- as Genevive Bell calls it, "deep hanging out."


Back to Ron Eglash home page