subject: The Snowball Sampling Method [print this page] I conducted fieldwork in four locations and two countries: Framingham, Massachusetts, and Marietta, Georgia, United States; and, Governador Valadares, in the state of Minas Gerais, and Piracanjuba, in the state of Goias, Brazil. I utilized ethnographic techniques including participant observation; and results from a 20-question survey instrument conducted with a total of 100 interviewees. In addition, for this paper I selected excerpts from a total of 273 interviews conducted in both countries: 100 formal in-depth interviews and 173 informal unstructured interviews. I draw from Vibram Five Fingers only a few selected interview excerpts (from both formal and informal) in order to highlight particular topics at the micro-level, and introduce them throughout this paper in the form of vignettes.
The ages of persons interviewed ranged from 18 to 74 years and the average age of all interviewees was 42. The average length of stay in the United States for all Brazilian migrants interviewed was seven years. Immigrants who remained in the United States had been in the country an average of 10 years, while returnees' average length of stay was five years. In light of the acknowledged US Census undercount (Margolis 1995) and overall enumeration discrepancies, it is virtually impossible to draw an accurate statistical sampling framework for all Brazilian immigrants in the United States. Therefore, to circumvent this methodological challenge, I used a "snowball sampling"1 technique for generating contacts for all interviews. Ethnographic techniques consisted of asking interviewees open ended questions on Five Fingers Shoes multiple occasions and using participant observation. For example, I attended mass at various churches (Brazilian immigrant Pentecostal, Presbyterian, and Catholic churches); attended breakfasts, lunches, dinners, family gatherings, outings, business events, and barbeques in both countries; talked to locals, immigrants, returnees, and established contacts; and made in-depth observations and took notes. A grounded theory approach enabled me to identify themes and patterns as they emerged at the same time as I collected data (hence from the 'ground up') (Glaser and Strauss 1967; Charmaz 2006). The Yellow Pages in Brazil provided useful information about business services in both countries. For example, US-based stores had Brazilian names or were advertised in the specific Brazilian-US Yellow pages. I contacted respective business owners whom I later interviewed, and with their permission, also took photographs of their establishments.
I selected Governador Valadares and Framingham because of their status, well established in the literature as traditional sending/receiving communities. In contrast, Piracanjuba and Marietta constitute a new migration corridor unstudied until now. Governador Valadares is one of the oldest sending communities in Brazil, with migrants leaving mainly to destinations in Massachusetts, New York, Florida, California, and Connecticut. I conducted exploratory research in Piracanjuba and found out that most interviewees were either returnees themselves, or they had family members and/or loved ones residing in the United States, mainly in Marietta or the Greater Atlanta area. Hence my snowball sample in Marietta and Piracanjuba opened up valuable research opportunities, and I selected both as research sites since 1 also wanted to evaluate spatial and socioeconomic interrelationships in two destination regions in the United States.
This study is not intended to be a representation of all Brazilian immigrants in the United States or of all returnees in Brazil. The aim of this study is to evaluate at the micro-level, how migration processes between two 'migration corridors' - an older one between Governador Valadares and Framingham, and a more recent one between Piracanjuba and Marietta - shape migrant experiences, livelihoods, and notions of place. That is, I evaluate transnational ways in which migrants (re)create places and spaces within the migration process. To better comprehend the context and symbolic nature of Brazilian migration processes, a discussion of some of the conditions and caveats of immigrant geographical origins and US ethnic identity categories is in order.
