APPENDIX A

Methodology

Given that scholars studying white racial subjects have often found whites to experience and discuss race in ways that are contradictory and elusive, ethnography is a particularly useful method for exploring how race is discussed and lived.1 From January 2011 to December 2012, I conducted ethnographic research with 30 affluent families in a midwestern metropolitan area who self-identified as white. The ethnographic approach I utilized allowed me to access what Amanda Lewis describes as the “ ‘everydayness’ of whiteness” within the institution of the white family.2

Because this research included work with kids, child-centered approaches to data production were used. Critical youth studies call for methods that are innovative and lead to accurate depictions of children’s viewpoints rather than adult memories of what it was like to be a child.3 Additionally, child-centered research recognizes the power dynamics between children and adults and seeks to reduce this power as much as possible.

Overall, my research is based on data gathered through (1) ethnographic observations of 30 white families in their everyday lives in Petersfield and (2) semistructured, in-depth, child-centered interviews with 36 white kids between the ages of 10 and 13 and their parents.

INCLUSION CRITERIA

While many scholars debate how to best measure class, in this study, families were characterized as affluent if at least one parent (though oftentimes both parents) (1) holds a graduate or professional degree, (2) has a career as a lawyer, medical doctor, engineer, university professor, business CEO/manager, scientific researcher, or similar occupation, and (3) owns a home. While occupation and education have often been used to measure class in sociological research, I also included home ownership to account for measures of wealth.4

The ages of the child participants in this study fall in the second half of middle childhood, or ages 10 to 13. Middle childhood is a developmental stage in which moral principles such as “justice, fairness, compassionate caring, and feelings of responsibility for one’s fellow human beings” emerge.5 This developmental stage is also when children are in the midst of developing a social and ideological perspective of the world.6 Thus, middle childhood is a crucial developmental stage for the formation of racial ideologies.

SELECTING A RESEARCH SITE AND IDENTIFYING NEIGHBORHOODS

I selected Petersfield as my research site because this community is large enough for variation across groups of affluent whites but small enough that understandings about the community are locally shared. Suburbs surrounding Petersfield are almost exclusively white, and a sizable affluent population lives in the Petersfield area due to various industries, a major university, and government-related activities. Additionally, members of this metropolitan area occupy political positions at extreme ends of the political spectrum, though I did not know about the extent of the political segregation of this community or the impact that politics would have on this project initially.

In order to describe and interpret the everyday meanings of race in the lives of white, affluent kids and their parents, I immersed myself as an actor into the Petersfield community. From the outset, I appreciated that it would be my interactions with members of this community that would help me generate my data. After moving to this city, I spent approximately three months figuring out the lay of the land. Through an inductive process in which I spent most of my time in public places talking to strangers and building relationships, I was able to identify different neighborhoods within the metropolitan area that were literally and symbolically distinct. Rather than entering the field with preconceived notions about what I would find, I selected the neighborhoods of Sheridan, Evergreen, and Wheaton Hills after learning more about what was actually happening in this particular environment. Drawing on grounded theory, I made theoretical sampling decisions after learning more about the community under study.7 Of note, before entering the field, I considered sampling from a different suburb than the one I ultimately included in my study. However, as I learned more about how people in Petersfield constructed local meanings around place and race, the suburb I ended up including emerged as a better comparison group during the middle stages of my data-collection process. This choice was also shaped by the difficulty I experienced gaining access to families in this alternative suburb. This is evidence of the utility of grounded theory, or the idea that “simultaneous involvement in data collection and analysis … [leads] the researcher subsequently to collect more data around emerging themes and questions,” as well as of practical issues of access that present themselves in ethnographic work.8

RECRUITMENT OF PARTICIPANTS AND THE IRB

After selecting the neighborhoods, I used a structured snowball sampling method to recruit participants. My snowball sample began with multiple nodes—individuals I met through seeking child-care positions, people to whom I was introduced by friends, and people I met through spending time in coffee shops, yoga studios, and gyms. Through building personal relationships in this community, I was able to recruit families to participate. A few of the mothers in different parts of the city helped with recruitment efforts in substantial ways through drawing on their own social networks and connections to private school communities and youth athletics. Almost every family helped me find additional families, including some parents sending emails to their friends introducing them to me. To these parents, I presented myself as both an “insider” (white, educated at prestigious institutions, grew up in a similar family and community as theirs) and an outsider (knowing “nothing” about the community, asking them for their explanations of various dynamics and observations). I always talked about my work in this initial recruitment stage as “a project on how parents talk to their kids about social issues” or a project on “families with middle school kids” rather than using the word “race” up front for fear that they may not be willing to participate because of the controversial nature of race relations in the United States. However, my consent and assent materials were more straightforward. Only two families declined to participate after initially learning of my project, and both indicated it was due to their busy schedules, one of the biggest challenges in this study overall.

Researchers studying children often cite the IRB process as a major roadblock to gathering information directly from children about their everyday lives, including how they think about and make sense of the world around them. I actively sought to construct an IRB application that paid particular attention to the power dynamics implicated in an adult-research and child-participant interview model. I wrote the child-assent form in language that was child friendly, asking a child within the age range to look over an early draft of the document. In my IRB proposal, I included a discussion of why these elements of the child-assent form were so important to safeguarding against the possibility of a child feeling unwanted pressure to participate in my project but also to establish rapport with the kids from the onset of the interview process. I also discussed the importance of child-centered research methods and implemented best practices for these methods in my research design.

PARTICIPANT OBSERVATION

In order to facilitate the collection of my ethnographic data, and especially during the first year of data collection, I spent significant amounts of time in public places observing the interactions, behaviors, and language of white families. Spending an extended period of time in public places allowed me to acquire a sense of the larger social geography of race and the cultural milieu of the Petersfield metropolitan area.

Public places were not the only places in which I gathered my data. I was also able to successfully form relationships with families that provided me access to the private places where affluent kids spend a lot of time. Through these relationships, I was also able to access spaces that are designated for children and authorized adults. Because affluent children spend so much time in private spaces, the observations I made in these places were especially important for my research. I also spent informal time in the private homes or out in public with most of the families in the study. In some families, I spent multiple afternoons a week with the kids for long stretches of the data-collection period. In other families, I spent only a few hours with the family. Most of these more personal in-home experiences were with children with whom I formed relationships, as a friend of the family, as their babysitter, or through the process of asking if I could observe a home for a few hours following an interview session. Because children talk about race sometimes in very spontaneous, unpredictable moments, the observational component to this project was key.

INTERVIEWS WITH KIDS

In an attempt to include children in the research process from the very beginning, I sought the help of kids within the appropriate age range in developing an activity in which I compiled photos of popular celebrities of different races that I could use as part of my interview. Drawing on this activity as well as techniques outlined in various methodological texts, I established rapport with the children by talking to them about topics that mattered to them, by making them feel as if they were the experts rather than I, by reinforcing the point that I was not testing them or that there was a “right” answer to my questions, and by connecting with them in ways that made the interview process fun and interesting to them.9 I asked kids questions in terms that I thought they would understand, I encouraged them to ask me questions throughout the process, and I repeated their language back to them rather than insisting on using sociological jargon. I also encouraged the children to laugh, as well as to be serious, and I was always careful not to push them to answer questions that appeared to make them feel unreasonably uncomfortable. A few children were very shy, but the majority of them appeared to be very comfortable and reported enjoying the experience and telling me that they thought it was important that adults hear their voices.10

Child interviews generally lasted between 30 and 60 minutes. I usually conducted these interviews at the child’s home in the living or dining room. Occasionally, I interviewed children at a coffee shop or restaurant. In total, 36 children were interviewed. Gender representation of the children was a relatively equal split between girls (20) and boys (16).

INTERVIEWS WITH PARENTS

I also conducted more traditional, semistructured interviews with the parents of the children in the sample. Parents were interviewed separately from kids in all but one case. In the other cases, parents were frequently nearby, sometimes drifting in and out of the room in which the interview was being conducted. Most parent interviews lasted 75–90 minutes. In seven families, mothers and fathers were interviewed separately. In three families, both parents were interviewed together at the same time. In only one family, the father was the only parent interviewed, while the other 29 families included interviews with mothers. Parents generally shared similar views, which helped reduce concerns that this gender imbalance would distort my findings. All of the children in this study have heterosexual parents, most of whom were married at the time of data collection. These interviews often took place in the family home, though occasionally parents invited me to their work office or a restaurant to conduct the interview.

DATA ANALYSIS

After transcribing interview recordings and typing up field notes as well as changing all identifying details to protect the confidentiality of my participants, I used MAXQDA to code my data in multiple ways. First, I coded “categories that make sense in terms of … relevant interests, commitments, literatures, and/or perspectives,” otherwise known as “initial coding” or “open” coding.11 Drawing on “qualitative grounded theory coding” as described by sociologist Kathy Charmaz, data were then recoded.12 Here, the objective was to develop theoretical categories from the data on the basis of what participants told me and what I observed about how they constructed ideas about race, racism, inequality, privilege, and so forth, to generate theory about racial socialization processes.13 I coded my data while still collecting data, which allowed me to conduct an inductive analysis of my data and develop further questions to ask participants or next directions for my ethnographic data collection on the basis of emergent themes.

RESEARCHER STANDPOINT

My own social position as a white woman in my mid- to late 20s shaped not only the data I produced but also my interpretations of it in more ways than I can possible state here. In short, I was able to access private white, family-centered spaces given my own whiteness, gender, and age without anyone questioning who I was or why I was there. Numerous times, parent and child respondents told me that they only felt comfortable talking to me or that they only answered my questions the way they did because I was white. Parents trusted me with their children and even asked me for advice about college and other topics given my own educational background, which I offered when appropriate. Parents also assumed due to my age and gender that one day I would want to have kids and as such were happy to talk to me at length about parenting. Finally, I was familiar with many of the aspects of these families’ lives given my own experiences growing up in a similar community, and I drew on this knowledge to gain insider access to this particular social world.