Intensive Family Observations: A Methodological Guide

Recruiting Families for Observations

Unlike many other settings, families often have a thick “membrane” which is difficult to pass through. Therefore, in both studies, we took a number of steps to gain access to the families. Most importantly, we first carried out in-depth interviews with parents. Conducting interviews provided valuable background information and was a way to build rapport before requesting families to allow researchers to conduct observations in their homes.

Ongoing Consent, Multiple Family Members, and the Risk of Coercion

The issue of consent was ongoing in both studies, and we often revisited it during the course of collecting observational data as well. To start with, when explaining the observational component of their studies, we both made it clear to the families that if they participated in this aspect of the studies, they would be consenting to visits, often daily, of two to three hours or more for about three weeks. Neither Lareau nor Rao provided extensive details about the family dynamics of interest because we worried that such details would make the participants even more self-conscious than they already were. Our institutional review board (IRB) did not require more than a general, accurate, summary of the goals of the study. ¹¹

Authors must, however, describe the study to the family members as part of the consent process. Here, researchers should think about, and possibly communicate to families, specific, routine activities which might be helpful for them to observe. Lareau generally told families, “Family life is a lot of work, and I am interested in learning more about all you do to help get kids through the day.” Lareau also listed a set of activities she wanted to observe (e.g., haircut, grocery shopping, family event, homework, before school, and so forth) “to get a sense of daily life.” But Lareau stressed, “it is up to you. We can always reschedule. You are in control here.” In the childhood study, one father asked what the research team would do and was relieved to learn that he would not have to talk to the researchers when they visited for observations. Rao had explained the consent process and detailed what the observations entailed in a one-page memo she shared with each family that she asked to participate in the observational component. This broad document explained that Rao would “hang out” with the family, visiting daily for anywhere from two to six hours, and complete at least 50 hours with each family over a two-week period. It explained that the family could opt out of the study whenever they wished. She also informed each family that if they decided they did not want Rao to note down an event that occurred during her observations, they could do so. None of the families in the unemployment study, however, requested to have anything “off the record.” Families in both studies thought the plan for observations was bizarre, but they generally had few questions. ¹²

Scheduling Visits

In both studies, if a family agreed to the participant observations, the fieldwork began whenever it was convenient for the family, usually immediately. We cleared our schedules in order to be available at the convenience of the families. In the childhood study, Lareau ritualized the entry and exit to the family. For the entry, Lareau and the research assistants would bring a dessert over to the family in the evening and talk to them all together, answer questions about the study, and make a schedule of visits. The schedule included the name of the fieldworker, time of arrival, and activity to be observed. The details looked as follows: Monday “before school and walk to school arrival at 7:00 a.m.,” Tuesday “homework and dinner, arrival at 3:30,” Wednesday “dinner and bedtime, arrival at 5:00,” and so forth. Lareau gave the family a magnet with the name of the university to post this schedule on the refrigerator. The calendar had the arrival time but not the departure time; the departure times were “played by ear” depending on how things were flowing but usually were after two or three hours. The parents in the childhood study signed the consent form. Furthermore, in the childhood study, each family had multiple fieldworkers, with one being a “lead fieldworker” who visited the family three to four days per week, kept track of the field notes, and managed any issues that surfaced in consultation with Lareau. A second fieldworker visited once or twice per week; Lareau visited once or twice per week unless she was the lead fieldworker (as she was in two families). Each family only had one fieldworker visiting at any given time. Lareau stressed that the schedule could be changed at any time if something came up. The scheduling visit helped build rapport with all family members.

Rao had a less formalized entry, but she too explained what the consent entailed on this first day with each family, answered any questions, and went over what the visits would look like on the weekdays and weekends. Rather than making a schedule of visits for the period of family observations, Rao more loosely discussed “coming over each evening.” The exact times were determined on a daily basis, and exact timing was confirmed by texting on the go regarding when Rao would reach the family’s house each day. (Texting was not a mode of communication during the childhood study.) The schedule of visit was thus more flexible for Rao.

Both authors also did separate, formalized “exit interviews” with family members at the end of the visits to probe more issues and to ask how the visits had altered family life. In the childhood study, the exit interviews often took place with the researcher and the child in a park while the parents were interviewed at home. In the unemployment study, exit interviews (done separately with the husband and the wife) were typically conducted in a secluded space in the home (such as the family den, closed off). Children were not included in the exit interviews.

To ensure that each visit went smoothly, we sought to ritualize our participation with the families in an unobtrusive way. Specifically, we used small gifts as a way to gain and maintain entry and rapport. We found it helpful to provide small gifts during the study such as brownies, fruit, wine, or something that suits the family’s preferences. Because Rao was often present for family dinners, ¹⁷ she made it a point to bring a small gift—usually a store-bought dessert for the family—at every third visit. Lareau ritualized the exit. She had a party with a special cake (with “thank-you” written on it), special paper plates the children usually chose, and pizza. This party was meaningful for the working-class and poor children in the study. Since the middle-class children had pizza and cake frequently, they did not see it as special and they did not appear to feel thanked by it.

The Visits

Since the family has a “thick membrane,” researchers should be prepared for initial awkwardness (see Elliot et al. 2016). Indeed, we found the first few visits to be extremely awkward, although things gradually improved, particularly after the third day. During the first few visits, family members frequently glanced toward us, and their eyes often tracked us as we walked up the stairs or crossed the room. Family members treated us as an honored guest in the beginning, offering choice seats in the house (which we declined), and being solicitous about our welfare. By the third or fourth visit, things started to normalize. Lareau noticed on the 10th visit that things were more relaxed. Interactions at the beginning of a relationship are different than interactions later. Information collected is usually more valuable after the relationship has deepened. We found follow-up visits, even a month or two later, important to sustain the level of rapport established earlier. The design implications are straightforward. If a researcher can only allot 30 days to visit a family, it is usually better to observe daily for two weeks and then conduct follow-up visits such as once a month for a while. (This schedule is more likely to yield rich data more efficiently than spread-out visits such as visiting once per week for nine months.) In both studies, we found that a higher density of visits in a short time helped everyone get more comfortable. Indeed, originally Lareau began the study by visiting three or four times per week, but she noticed that things flowed better after the schedule happened to include daily visits for a few days since researchers could keep pace with changes in the family’s schedule, families were less likely to forget about the visits, and everyone had more of a chance to adjust more rapidly. She then switched to daily visits (although it varied across families). An alternative approach is to determine how many hours of observations each researcher can spare (including travel and writing field notes). This figure, along with what the researcher deems necessary for their project and to which families agree, can guide the design.

The research questions help to guide the visitation schedule. For example, given Rao’s goals of gaining insight into marital interactions during one spouse’s unemployment, both spouses had to be home for this. As a result, Rao visited the family she was observing in the evenings (when the whole family would be at home) or just as the spouses were arriving home from work. All family members were at home for the majority of the time she spent observing families. On the weekends, Rao visited as per the family’s convenience, typically around 10 a.m. and into the evening.

In sum, in both studies, we spent time explaining to family members what we were planning, and we arranged for intensive, daily visits. The frequent visits helped us understand the flow of family life, reduced cancellations, helped to build rapport, and generally helped everyone adjust to the strangeness of the endeavor in a more rapid fashion.

Where Should Fieldworkers Sit or Stand? The Issue of Spatial Positionality

Fieldworkers should carefully consider where they physically position themselves in the home as it is a constrained site. In the childhood study, many of the working-class and poor families had living rooms that were very small. Fieldworkers’ actions are closely observed. Small decisions about where to sit and when to talk were all heightened. In middle-class homes, where the houses were large, fieldworkers had to make constant choices about which family members they should follow around such as when children abruptly got up from watching television and went upstairs. (Generally, once the parents knew that fieldworkers were in the house, fieldworkers did not check with the parents each time they moved from room to room. Fieldworkers did, however, check in with the children to see whether they were comfortable with them tagging along.)

As children became more comfortable with the fieldworker, they would sometimes not update the fieldworker on their every move. This means that fieldworkers constantly have to figure out what to do. Compounding the difficulty, some family members were more supportive of the project than others. For example, in the Black low-income McAllister family, an aunt (who had a serious drug problem) was staying there. The children—Harold, and his sister (Alexis, 9)—were comfortable with the fieldworkers, but they deeply disliked having their aunt around. (Harold angrily said, “She’s stupid.”) The aunt, in turn, was annoyed by the fieldworkers. One Friday late afternoon, Greg (a Black male fieldworker) in his 20s has been playing basketball when Harold went back to the house (possibly to use the bathroom) without updating the fieldworker. Greg wrote in his notes:

Harold disappears on me. I wait 10 minutes before I go get him. The basketball crew dwindles down to four. As I walk to the house, I see Harold’s mom. She waves. She says “Harold is upstairs.” “Go ‘head up; he’s up there.”

Furthermore, family members often have squabbles, and fieldworkers need to decide where to position themselves and how much to trail after family members when a conflict is unfolding. A field note from the unemployment study reveals the dilemma in the Jansson family where Robert, the father, had been unemployed for many months:

After dinner this evening, we were all sitting in the living room on the first floor. The kids, two-year old Taylor and four-year old Tessa were playing on the carpeted floor in front of the TV. Robert’s wife, Laura, was sitting on the couch opposite the TV (Rao was sitting next to her), scrolling through Facebook on her phone and intermittently looking up at the TV which was on a news channel. Robert was reclining on the couch perpendicular to Laura.

In the home, some rooms may be more “public” than others. The living room, kitchen, or den where family members congregate together is often a space where the researcher is less conspicuous as opposed to the master bedroom or bathroom. Different rooms within homes carry different expectations of privacy, and fieldworkers must remain attuned to this. At times, we waited for participants to invite us in, and at other times, we checked that it was all right that we followed them into the more private rooms (“is it okay to come along or would you prefer to skip it?”).

In addition, the presence of young children can complicate researchers’ mobility within the home. Researchers around children often need to acquiesce to children’s demands—smiling at them and playing with them. Yet, this can sometimes obstruct the researchers’ goals of observing. In a summary of field notes from the second visit to the Jansson family, Rao noted that:

Robert was in the kitchen, trying to find the Kung Pao chicken that he was supposed to heat up for dinner. Tessa, between the dining room and the living room with me, and was keen to play with me and show me her toys. She wanted to play catch with [a] large plastic ball, so we did that in the living room for a while. Taylor busied himself with jumping up and down on the couch in the living room. Then we played “princess dominos,” followed by a Dora the Explorer puzzle. I was a little frustrated at this point because…I wanted to be free to move around. Yet, I didn’t want to say no to Tessa because I want her to like me. So, my solution was to tell Tessa to move to the dining room, from where I could see Robert in the kitchen, and keep an eye on the living room and entrance, and play there.

Are Fieldworkers Like “Family Dogs”?

Hochschild (1989) famously told her families that they should treat her like “the family dog.” Yet, in both studies, we found that the “family dog” analogy was not quite apt. Social norms require researchers to greet family members, be introduced to others outside the home, and accept or refuse offers of food and drink during meals. The team-based approach of the childhood study and the solo fieldworker approach of the unemployment study both raised a few key points when it came to how families, and different members of each family, saw and responded to fieldworkers. In the unemployment study, Rao had met both spouses by the time of the first visit for family observation, but this was usually her first meeting with the children. Three of the four families she observed had children under the age of 6, who were typically excited and curious to meet her, meaning that she could not be as unobtrusive as a “family dog” might be. The fourth family had a teenage son, who, although initially awkward, warmed to her over time. The small sizes of these families in the professional middle-class meant that Rao was a visible presence in their homes. ¹⁸ The size of the families in the childhood study varied, but the researchers always stood out.

In both studies, we strove to collect observational data without disrupting the normal routine of each family excessively. We had a goal of getting a sense of the daily routines and interactions of the families, which we triangulated with our interview data. As such, while we were not invisible, we did seek to ensure that conversations did not center on our experiences or opinions and that families continued about their day to the extent possible. An example of this approach in practice is that we typically conversed when family members initiated conversations. In both studies, fieldworkers tried to be helpful and folded laundry, picked up toys with the kids, and helped kids with their chores. But we also had occasions where fieldworkers in both studies chatted with parents as part of the flow (while being cautious of not steering conversations) and asked “probing questions.” These chats were more informative when only the researcher and one person were together and had privacy. For us, this occurred while driving with a family member or walking in the neighborhood. A common space like the living room, where other family members may listen in, hindered more in-depth conversations.

While families know how to deal and interact with a “family dog,” they were usually unsure of how to treat fieldworkers—to ignore them, treat them as a guest, or behave in a familiar fashion. Hence, it is helpful for ethnographers to give families guidance; “Oh, don’t worry about me, I will just hang out and watch television with the kids” was something that Lareau would tell families. In the childhood study, the fieldworkers would often hang out with the children and play games with them. Still, this lack of surety from research participants about how to treat the researcher almost always alters dynamics within families. In the childhood study, children reported that the families tidied up before the arrival of the researchers and, in the middle-class white family, the eldest boy reported that “my mom talked more.” We speculate that families also might have changed the meals, dressed differently (wearing more company-appropriate clothing); in the case of spouses, they may have been less physically affectionate with one another or fought less. Despite these potential changes, we do think that observations capture relationship dynamics relatively deeply. Furthermore, when participants perform for the benefit of the researcher, it is still as useful as other common data collection methods such as in-depth interviews since performances can be helpful in revealing what behaviors participants value and idealize—or at least find socially valued (Pugh 2013). But, more importantly, it is difficult to constantly self-censor, especially over the course of numerous, daily visits. While family members may perform more, or present more of a front stage (Goffman 1959), we are reasonably sure that the family members became more relaxed and revealed more “backstage” moments of family interactions as the study unfolded. Of course, family members also perform for each other—including at the dinner table or in other family gatherings.

We also found that fieldworkers were not like “family dogs” because social norms required more strategic interaction with family members than the “family dog” analogy suggests. For example, when fieldworkers arrived and departed, family members asked about our trip there or about our plans for the weekend when we left. As an ethical issue and instrumental issue to increase integration in the field, we both sought to be helpful to family members; a form of social interaction obfuscated by the “family dog” analogy. For example, Lareau and the research assistants held babies, chased after toddlers, played basketball, talked about television programs with the children, and generally sought to be helpful with household chores. Rao similarly helped set and clear away tables for mealtimes and helped clear away children’s toys at the end of the day. This was tricky for Rao as she was interested in understanding the marital dynamics during unemployment including around housework. Typically, she would wait to see who would take the lead, for example, in setting the table and then help that person. Cleaning up toys was almost always the children’s responsibility in the families she observed, so she would quickly check with one of the parents if “I can help pick up a couple of things?” They would usually nod, and she would proceed to help. This way she was able to attempt a balance between not creating awkward boundaries between herself and the families but also not interfering too much in the running of things. Hence, ethnographers ideally will find a way to be helpful which is not too disruptive of the central research goals of the study.

Our interactions with family members in both studies were delimited by our conviction that fieldworkers should not significantly transform family life, particularly when an issue was core to the focus of the investigation. Consequently, Lareau did not give rides to family members (even though it would have been very helpful) since the focus was on how the families managed daily life. When the subject came up, she pled that the insurance regulations at the university did not permit it. In addition, as we discuss below, even if children’s actions made the researchers uncomfortable, if the parent was present, Lareau told the research assistant not to intervene. In both studies, fieldworkers sought to minimize how much we talked to family members. Lareau would nod sympathetically and look interested, but she said little as she did not want to disrupt the flow of daily life. Most of what she said was praise (“that is really cute”), curiosity (“how does that work?”), or self-deprecating comments (“when I was a kid, my sister and I fought like cats and dogs. We were wild.”). She also sat and looked interested, smiled, and nodded and helped with household chores, but said little. In retrospect, Lareau wished that she had probed more, for example, by asking an open-ended, relatively neutral, “How do you think that went?” after a visit to the dentist or doctor. Lareau also followed the technique of “holding up a mirror” by repeating back (word for word) what a family member was saying. For example, when a mother expressed deep dislike for a teacher (“I hate the school”) or anxiety about the future (“I am worried”), Lareau would repeat what she said either directly or in a slightly different fashion (e.g., “you hate the school” or “it is easy to be worried”). This technique is very helpful in interviews and can make people feel “heard.” If it could be enacted in a casual way—and if other family members were not in ear shot—it possibly might have led family members to say more.

When family members asked Rao questions about her opinions or her life, she usually responded succinctly and without detail. Rao’s responses were more concise than they normally would be in other social settings. However, both authors also offered a sympathetic ear when the situation warranted. For example, Rao was alone several times with Rebecca Mason, an unemployed mother. Rebecca and her own mother, who drops by the house frequently, have a fraught relationship which Rao observed. On reaching the Mason’s house, Rao would often see that Rebecca was on the phone arguing with her mother or upset because her mother had visited earlier. In these times, Rao would listen sympathetically to Rebecca, nod in agreement as Rebecca vented in frustration, and offered brief sentences of assurance like “that sounds like a difficult situation” or “I am sorry you are going through this—it must be difficult.” Through their communication and their (hopefully helpful) actions, researchers are able build rapport and find their place in the family.

How Often Should You Visit and for How Long?

There is no perfect amount of time to visit. Lareau and her assistants usually visited only two to three hours at a time, but some visits (e.g., when the family went to a funeral in a city two hours away) lasted many hours.

Budget and time constraints also shaped the extent of fieldwork possible: Lareau had research assistants who had to be paid for their work (e.g., about 10 hours per visit including travel time, observation time, and field note writing time). These budget constraints led her to settle on daily visits for three weeks per family. Lareau had an ambitious design with three social classes (e.g., middle class, working class, and poor) and two racial groups (e.g., Black and white families). The research was tiring and difficult to complete, and the number of families was, in the end, excessive. Because Rao was the sole researcher on her study, she was guided more by the constraints of her schedule. Although we both wished we could have followed each family for a longer period of time, limited resources made that impossible. Ideally, researchers should leave a site when they have reached theoretical saturation and are no longer learning new things (Lofland et al. 2006; Small 2009). In both the childhood and unemployment studies, we found that reaching theoretical saturation was realized for some, but not all, aspects of family life. Nevertheless, in both studies, we collected a great deal of data.

The duration of the visits was guided by a blend of the researchers’ interests and the family’s preferences. It is imperative to get to know families, be there long enough to see the topic of interest, and adjust to the specific nature of the events being studied. Lareau was studying parent–child interaction. This interaction was an ongoing feature of daily life; thus, in two hours, there would be dozens of interactions relevant to the study. Rao, however, was studying job searching and the impact of unemployment on families. In briefer durations, for example, two hours, she could observe how family members went about daily activities like doing chores. But it took observations lasting for longer, such as four hours, to reveal other aspects of their life, for instance, how spouses discussed the unemployed partner’s job search. The research question, broadly conceived, structured the visits.

When ending intensive observation, it is helpful to openly ask the family whether it is okay to stay in touch. Rao also stressed that research moves slowly, and it will be a long time before she would be done and able to share any “findings.” At times, families face new and unexpected developments that offer extremely valuable data for the study. Clarifying these expectations when the intensive observations are wrapping up can be helpful. We both continue sending annual holiday cards to each family. But some families from the childhood study felt hurt and angry after seeing the published book and severed ties with Lareau (for more details, see Lareau 2011, Appendix). With the remaining families, Lareau continues to send holiday gifts, and she sends more significant gifts on the birth of a child. In one case, she covered the cost of a tombstone after the death of a parent. Other researchers have discussed the importance of reciprocity in the field (Desmond 2016); further, in the childhood study, these gifts occurred several years after the last round of interviews. No future rounds of data collection are planned in either study.

Working in a Team Compared to Working Alone

Family observational studies are rare. Most studies have been done by one researcher working alone (e.g., Cooper 2014; Hochschild 1989; Paik 2016; Reich 2005; Stacey 1990; Stack 1974), but team ethnographic studies do exist (e.g., Lareau 2011; Bowman et al. 2019). As noted above, family observations often require having numerous visits (ideally in a short time period) combined with the need to spend many hours per day writing field notes. Clearly, family observational studies can be time-demanding for ethnographers; the time required to travel, carry out fieldwork, recover from doing fieldwork, and write field notes can consume almost all waking hours. Hence, using research assistants can help reduce the workload and increase the sample size. Research assistants also can have valuable insights; ask probing questions; and provide valuable racial, class, and gender diversity in a team which can help in building rapport with participants (as well as increasing the legitimacy of the study). Team studies can help the researcher study more families and visit the families longer thereby increasing the strength of the project in the eyes of some readers.

Researchers differ in their approach, but, in the childhood study, Lareau was the person to negotiate almost all aspects of recruitment and consent for the families, design the schedule of visits, decide what research assistants should focus on during the visits, and work closely with team members to ensure highly consistent data collection across sites. The research assistants were responsible for arranging transportation to and from the families (and were reimbursed for transportation and travel time), carrying out the observations, talking to Lareau on the phone after each visit, and writing field notes. ¹⁹ They attended team meetings, and research assistants were also invited to use the data for their dissertation projects. Interestingly, in the childhood study, fieldworkers had favorite families. Families also had favorite fieldworkers. In Black families, the boys showed their preference for the Black male fieldworker over other fieldworkers, thereby reinforcing the value of a racially and gender-diverse team of researchers in building rapport.

Still, there are multiple challenges with teamwork. First, teams are costly, requiring a major grant. Second, teams create challenges with ensuring a consistent pattern of data collection. Research assistants also can have missteps (just as the principal investigator can) and damage rapport. Third, there are hurdles for researchers working in a team as they navigate the complex process of moving from field notes to a research question, considering disconfirming evidence, and honing a theoretically informed thesis (Luker 2008). After all, ethnographic scholars often begin with “sensitizing concepts” and a general set of research questions, which are honed by the data collection process through an emergent process. In a team, this means that the principal investigator writing up the results of the study must have a deep understanding of all the data. Yet, this depth of knowledge can be difficult to gain unless the principal investigator collects the data. Fourth, since there are multiple possible lines of inquiry, team members may disagree about the most important questions to pursue; the principal investigator needs the input of research assistants, but the principal investigator must also provide leadership and manage differing interests. Fifth, since intensive observation is a challenge, research assistants can benefit from professional development opportunities. It is critical for the principal investigator to provide emotional support to team members, show gratitude, and be a good listener in times of difficulty. In addition, team members need help with their own professional development. But these mentorship relationships are time-consuming, and they likely come at a time when the principal investigator has many demands. Finally, as in any group, team members will cohere and, at times, fracture. Team members may also compete for the attention and mentorship of the principal investigator.

A lone ethnographer, such as Rao, has more control over the relationships forged with family members which is invaluable. Rather than relying on the field notes of others, the lone ethnographer can carry out participant observation and have a much deeper understanding of the setting. There is more flexibility to develop the focus of the study and more time to work on the study. However, the lone ethnographer usually has a smaller sample, which provides limited opportunities to introduce variation into the subject matter being studied. In short, the lone ethnographer is the traditional approach to carrying out ethnography. There are cases of team ethnographic research projects, but the team approach also has numerous challenges.

Writing Up Field Notes From Observations

In the long run, high-quality field notes are the lifeblood of an intensive observational ethnographic project. Rao usually wrote a brief one-page outline of the day’s visit each night, followed up by detailed field notes the following morning and afternoon, before returning to the family in the evening. In the childhood study, Lareau wrote field notes immediately after every field visit; other fieldworkers in the childhood study would telephone Lareau for about 20–35 minutes, recounting the events of the visit. These phone calls were crucial to provide Lareau with a vivid sense of families’ activities and give the fieldworker detailed directions on what should be written up in the field notes. Given the labor-intensive nature of writing field notes, only selective aspects of the visit could be highlighted. The team of three fieldworkers also reviewed each other’s notes and had a weekly meeting to discuss emerging analytic themes and challenges in fieldwork. A major constraint on the duration of the visit is the time needed to write field notes. Fieldworkers should never schedule visits unless they have time to write field notes (see, among others, Emerson et al. 2011; Lareau 2000). Normally, there is a ratio of at least two-to-one in terms of time spent writing vis-a-vis time spent observing to allow for sufficient detail in field notes. A two-hour visit can mean time allocation for five hours of writing field notes which is considered a reasonable “ballpark,” but it depends on the research focus and how much happens during that visit. Timely and detailed field notes that capture interactions including facial expressions, tones, gestures, and other such tactile experiences are the key to this form of data collection (Emerson, Fretz, and Shaw 2011).

Fieldwork is a complex task, and researchers vary in the skills they bring to the project. Some gregarious researchers can easily build rapport but then struggle to write detailed, vivid, and accurate field notes. ²⁰ Lareau found that the field notes of research assistants varied in quality, so she eventually sent them small tape recorders to carry into the homes. The audio recordings were then used to write up the field notes. ²¹ Fieldworkers always mentioned the tape recorders and displayed them openly (e.g., by putting it down on the dining room table or holding it in their hands). Hence, any taping should be done openly and with the permission of family members and normally be mentioned on every single visit. The tape recorders, did, however, create strain in some families as members would repeatedly glance at them, talk about them, and seemed conscious of them at times. Other family members, especially children, ignored them and did not seem particularly concerned about them. This pattern surprised us, but, as F. Erickson (July 1991, personal communication.) once said, “If they get used to me, then it doesn’t matter what I bring—tape recorders or video recorders.” Despite the benefit of the recordings, recording people can alter the dynamic. Of course, with the widespread use of cell phones, it is more common for scenes to be recorded or filmed. Still, recording makes some family members very uneasy. Lareau would not generally recommend it. Nonetheless, a recording dramatically improved field note quality for some fieldworkers. For discussions of the use of video see Aarsand and Forsberg 2010; Erickson 2006; and Golann et al. 2019.

Rao did not take any notes or jottings while she observed the families (nor did the researchers in the childhood study). In these small-sized families, three to four members, an additional person—Rao — stood out in the family. Jotting notes, however discreetly, for example, during dinner table conversations where each participant was entirely visible, would have been unfeasible. The families she observed were interested in Rao’s note-taking. At least one adult in each family she observed asked her about her lack of note-taking—they expected her to take notes. She explained that she made jottings based on her observations on the commute back to her home; then, she would develop those jottings into full-fledged field notes within the next 24 hours. ²²

In sum, ethnographers planning to conduct family observations face a number of logistical challenges including scheduling of visits, finding a physical space to position oneself in the home, carving out a way to interact with family members, and managing the writing up of field notes. We found it helpful to create regular, repeated, and routine interactions with family members as part of the observations; by making our actions more predictable, it helped everyone adjust.

Intrusion, Experiencing Powerlessness, and Deploying Deference

Fieldworkers straddle a delicate boundary between having access to private and quintessentially familial moments and maintaining a sense of discretion so that family members do not acutely feel that they are being intruded upon. Fieldworkers can at once be in the position to exploit their subjects since it is the researchers who ultimately represent the subjects even as fieldworkers depend on research participants for continued access to the field site (Duneier et al. 2014; Punch 1986; Stacey 1988). We found that fieldworkers’ worries about intruding upon families, while maintaining access to the family, leads to a sense of powerlessness for the fieldworker, which has not sufficiently been documented in previous accounts. Once inside the door of the family, researchers are deeply constrained and usually do whatever family members do including watching television shows they would never otherwise watch, listening to political views they may believe are abhorrent, and playing games with children long after the pleasure was subsided. Yet, the point is to learn about the routines of the family. To leverage this powerlessness for data collection, we suggest that cultivating a deferential stance can be a useful strategy. By deference we mean prioritizing family members as experts on their family, allowing family members to guide the researcher into the routines of family life—for example, where to sit, when to eat, what to eat. Using deference means shrinking one’s presence in the family, embracing one’s powerlessness for data collection purposes so that the researcher doesn’t become any more of a focal point than the researcher already is.

In the first few visits in particular, families fell into the role of hosts and worried themselves with taking care of us. In the childhood study, a fieldworker described one of the early visits to watch Billy’s baseball game with the Yanelli family.

At about 7:30 pm, I told Linda, the mother in the family, that I should get going. She said, “I’m gonna walk you out. I want to know that you get the right bus and all.” I told her I didn’t want her to miss any of Billy’s game, and so I could walk out myself. She repeated that she wanted to know that I got on the bus.

In addition, in someone’s home, fieldworkers have less control over when they can exit, spaces they can enter into within the home, or even when and what they can eat. For example, in the unemployment study, Rao had a challenging day when she was hungry observing the Smith family on a Saturday, arriving at one in the afternoon.

I had eaten an early lunch [around 11 a.m.] before making the 90-minute journey to the Smiths’ home, and I wasn’t totally surprised when a couple of hours into the fieldwork [around 3:30] I started getting hungry: my stomach was growling. Usually I take a couple of bananas or granola bars with me (easy to carry and filling), but today I’d been in a rush and forgotten to put them in my bag.

Deference also surfaces when the fieldworker accompanies the family to unfamiliar settings and needs to be able to engage in conversation topics that matter to the family. In this example, Lareau drove her car to meet Harold, a low-income Black boy, and his Dad (who lives separately from the mom) to go on a shopping trip to buy clothes for camp. The three of them walked to the bus stop to wait on a blazing hot summer day:

We wait for the bus. Harold’s Dad stands against the building in the shade where it is cooler. Harold walks around on the sidewalk but, after a few minutes, stands to his father’s left. They talk about the basketball game. I ask if they watched it. Harold tells me that the Knicks won. I ask, “Are you rooting for anyone?” He says, “Houston.” His Dad tells him that it is unlikely that they will win. He says, “They have to win back at home and then they have to win when they don’t have the home court.” Harold interrupts him, “Home court advantage.” Dad says, “Don’t bet the rent.”

Furthermore, the need for the fieldworker to defer to the family becomes complex when moving out in public space. As many others have noted (Steinbugler 2012), whites and Blacks are not interpreted as being “together” in public space. That same day, Lareau was sitting at a table in the fast food restaurant while Harold’s father was not there yet. Harold went to join her at the table. A stranger, an older Black woman, motioned to Harold that he should not try to sit at the table with Lareau since it was occupied. Lareau spoke up to her, smiled, and said that they were together. In this instance, the stranger chastising the child in the study is likely linked to the oft-discussed patterns of race relations in public spaces (McIntosh 1989; Steinbugler 2012). Lareau would have liked to have been more deferential to Harold in this setting, but it was complicated by other social norms.

The Urge to Intervene

The issue of researchers’ intervention during fieldwork is fraught, and we both generally sought to minimize their own presence in the field sites. ²³ This “low profile” raised issues of how, and whether, to intervene if fieldworkers observed troubling events and interactions as well as the impact of our presence on the experiences of family members. All of these decisions are difficult, and there are no easy answers.

In deciding when to intervene, researchers are always guided by their own ethical and moral beliefs. ²⁴ For example, Lareau felt it was important to respect the families and the ways that they lived their lives. Lareau believed that the parents were doing their best, and she was convinced that, with very few exceptions, there are not universally established ideal methods of child-rearing which have prevailed across historical periods. She did not believe that she held superior information of how the parents should behave, and she did not want to try to change them. She held on to this view even as there were emotionally taxing experiences connected to having a close-up look at family poverty. For the childhood study, Lareau established the rule that team members would only intervene if there was blood or if it was a life-threatening situation. (As we explain below, neither Lareau nor Rao was a mandated reporter.)

Most family life is mundane. Still, one difficult moment did occur when a toddler took Lareau’s keys and headed to an unprotected electric outlet where he might have experienced a severe electric shock. In this instance, Lareau leapt to her feet to intervene, but an aunt grabbed the keys from the child first. Lareau certainly learned of illegal activity and observed challenging moments, but the childhood study had a rigid rule to avoid interventions. Although extremely unlikely that she could have come to the attention of government officials, Lareau had planned to face any consequences with her employer or the courts (including a contempt of court citation and jail time) than follow a path of either revealing confidential information about families with authorities or destroying field notes (Khan 2019; Lubet 2019). Rao, on the other hand, informed the participants that the field notes would be kept securely on a password-protected computer, and any printouts would always be kept in a locked drawer when not in Rao’s possession. She also informed the families that in the highly unlikely event that the legal system should request her to turn over the field notes from her observations, she would follow suit in so doing. As a non-U.S. citizen, Rao was particularly wary of any complications of being caught up in the legal system. The approaches of Lareau and Rao present two of the various alternatives in terms of how to deal with the tricky, and not easily resolvable, issue of continuing to protect participants during and after fieldwork.

As noted above, our role in the field did, however, include helping out or lending an empathetic ear when family members discussed issues of importance to them. For the most part, we both strove to maintain a level of distance from the goings-on in each family. Other scholars have questioned whether “objectivity” is a desirable, or even feasible, way of collecting observational data (Stacey 1988). But we both decided that since we were interested in understanding family routines, we would try to minimize our presence and maintain an uninvolved stance to the extent possible so as not to disrupt routines even more. In the childhood study, this meant that none of the fieldworkers would intervene in family dynamics unless there was an immediate and grave threat to someone’s life.

But this decision did not shield us from difficult predicaments. Although most of the fieldworkers’ interactions with family members were comfortable, and even fun, at times it could be painful to watch family interactions. In one family in the childhood study, the Brindles, difficult moments were nearly continuous. One fieldworker found it painful to watch Katie hit herself on the forehead in full view of her mother and auntie while they all watched television. Relatedly, although Lareau would have intervened immediately if the child was in a life-threatening situation, she did not want to chastise mothers for not following historically specific standards of child-rearing since she thought it would be fundamentally disrespectful to the mothers. Nonetheless, despite this commitment, Lareau fretted when she saw Ms. Brindle leaving her 18-month-old son alone in the bathtub for several minutes, feeding him whole grapes (which were a choking hazard), and allowing him to carry around a washcloth with Comet cleanser all over it as she wrote in her journal:

My reaction is profound, deep, and negative…. She is letting him down. It makes me feel angry and powerless…. I hate to go. I hate to go back…I have done eight visits and even though it is getting better six more visits seems like absolutely forever.

Depending on fieldworkers’ own class positions and sensibilities, different issues troubled them. A working-class-origin, male, Black fieldworker, Greg, in the childhood study was deeply troubled by the rudeness he perceived middle-class children as showing to their parents; patterns that some of the middle-class-origin fieldworkers hardly noticed. He wrote about a white middle-class family:

Sammy (4) begins to cry and whine. Then he calls her mother in a whiny voice at least five times and then shouts MOM! before she responds. She seems rather annoyed, “yes Sam.” Sam is whiny and loud, ‘Mom can the people see me when I was in the way back [of the car]?” She responds rather calmly as she read the directions to the baseball field, “Well Sammy you know that the windows are tinted and the people might not be able to see you when you wave.” Sam begins to whine again, “I want them to see me!” Sam whines for at least 15 minutes of the 20-minute trip….

Mandatory Reporting

Some researchers are clear that they do not plan to intervene in families. Complicating this situation, however, is the reality that dominant institutions have (historically specific) standards for how family members should act, and, in some instances, ethnographers can be “mandated reporters” which means that they have a legal obligation to report specific behaviors to state agencies. Mandating reporting was not enacted in the United States until 1963 (McTavish et al. 2017), and the precise obligations vary across states, but most require mandated reporters to contact child protective agencies, for example, with evidence of “serious physical neglect,” “bodily harm,” or “intentionally, knowingly, and recklessly…create a likelihood of sexual abuse” (Keep Kids Safe 2019, see also Pennsylvania Family Support Alliance 2020). Although there is not much research on the effectiveness of mandated reporting, one meta-analysis found that negative experiences were reported in 73 percent of the articles in the meta-analysis, including complaints “the child was removed from harm, but the foster care environment was worse than the family-of-origin environment” (McTavish et al. 2017).

While we did not encounter the issue of mandated reporting firsthand, it is part of a more general set of moral choices which ethnographers routinely face (Anspach and Mizrachi 2006; Bosk and De Vries 2004; Emerson 2001; and Heimer and Staffen 1998). Furthermore, researchers’ decisions regarding how to respond to mandated reporting are also, realistically, part of more general decisions regarding how closely citizens comply with the law (e.g., speeding, reporting all forms of income on their taxes). So too, there will likely be variations in how closely researchers hue to the letter of the law on mandated reporting. One option is for researchers to gain a “Certificate of Confidentiality” from the National Institutes of Health (2020) before beginning the research (since it is not valid after the research has begun). Although there is ambiguity, many researchers believe that this certificate protects the researcher from being mandated to share information or report families (Paquette and Ross 2020). But, if researchers are going to engage in mandated reporting, we believe they must tell the family ahead of time and, ideally, as part of the consent process. This follows the best practice of other professionals (e.g., see Langford 2000). As much as possible, there should be transparency about this issue between researchers and research participants.

That said, given the complexity of family life and the nature of ethnographic research, we believe that it is naive to think that all ethical strains can be avoided. Indeed, ethical quagmires have been discussed at length in ethnographic research from classical studies (Whyte 1981) to more recent works (Duneier 2004; 2011; Duneier et al. 2014).