Navigating the Use of Secondhand Data and Cross-Institution Collaboration

• In Event: The Necessity of Seeing Anew: Dreaming Up New Possibilities for Research on Nondominant Communities

Abstract

Purpose: Drawing on our (2016) understanding of historicity and Jones and Raymond's (2012) conception of third-party video, this paper explores the methodological risks and opportunities that are assumed by analyzing second-hand ethnographic data. Detailing the collaborative research process between scholars at The University of Colorado, Boulder and The University of California, Berkeley, this paper considers the historicity of data - recognizing the history that is unique to each piece of data and how paying attention not only to the history of the study participants but to the data itself, contributes to richer understandings of the study's collaborators.


Framework: A cultural historical activity theoretical approach to research focuses on the ways multiple domains of history impact the way we understand and experience the current moment (Cole, 1998). Historicity is essential to building a nuanced yet detailed understanding of the structural conditions that influence how people live and how people understand themselves now and in the future (Author, 2016). We bring this lens to how we look at third party video - data that we did not collect ourselves. The videos for this project can be understood as akin to third-party videos in that we had no influence over “where, when, or how data were collected”; instead, the data are "products of a localized interest in an encounter or series of events” (Jones and Raymond 2012, p. 112), in our case, the interests of the Boulder team at the time the data were collected. Looking at these data from a historical perspective allows us “to understand the tensions, constraints, and possibilities of [third party video data]” (Gutiérrez et al., 2017, p. 23).

Methods: The analysis reported here developed within a larger multi-sited ethnography (Marcus, 1995; Vossoughi & Author, 2015) that followed youth across everyday activities in their after-school program, homes, and community. A total of 14 families participated (12 Latino, 2 White); each self-identified as working class.

Data: We draw from qualitative data sources: 175+ hours of video-recorded observations, activity logs, surveys, interviews, family-documented photos/videos, and memos. Units of analysis included family home practices (Level 1) and instances of ingenuity (Level II). Subcodes (Level III) were developed inductively.

Findings: By engaging the Boulder team as thought partners, the Berkeley team was able to maintain the study's rigor while also ensuring that the data were not misrepresented. The Berkeley team also developed insights that may have eluded the Boulder team, as they gained access to new information and ideas, employing new analytical frames while also accounting for changes in the context in which the data were being analysed.

Significance: In an academic environment of increased collaboration and data sharing, this paper highlights the methodological and ethical implications of doing large research projects, particularly those across academic institutions.This paper also examines how a CHAT approach to data analysis positions the researcher to account for the historicity of people's everyday practices.

Authors

• Patrick Johnson, University of California - Berkeley

• Krista L. Cortes, University of California - Berkeley

• Karen Villegas, University of California - Berkeley

• Bryce Larkin Chessell Becker, University of California - Berkeley

• Kris Gutierrez, University of California - Berkeley