Public Selves,Inequality,and Interruptions: The Creation of Meaning in Focus Groups with Teens

Abstract

Focus groups have received substantial attention over the past few decades, particularly as they are considered to provide rich, interactive data, yet only occasionally do researchers discuss the process of conducting focus groups with young people. This paper contributes to wider debates on focus groups through engagement with three interrelated topics, each with unique reflection on focus groups with teenagers: the advantages of focus group interactions, particularly in relation to hierarchies of age and the research relationship, how focus groups shape self-representation and “truth-telling,” and, finally, the challenge of “unruly” data. The author addresses these topics through drawing on several sources of data: 18 focus groups with secondary students on the topic of school rules, exit questionnaires collected from focus group participants, and in-depth interviews with the primary investigator and three research assistants.

Introduction

Focus groups are “group discussions organized to explore a specific set of issues” in which the group interaction is considered part of the data (Kitzinger, 1994, p. 103) and have been used in consciousness-raising as well as market, community, and academic research (Kamberelis & Dimitriadis, 2005). Writing on focus groups covers many practical, theoretical, and political issues related to organizing and conducting them (e.g., see Bloor, Frankland, Thomas, & Robson, 2001; Kitzinger & Barbour, 1999; Smithson, 2000; Vaughn, Schumm, & Sinagub, 1996), yet only a minority of commentators have specifically discussed focus group research with young people. In this paper I contribute to this literature on focus groups by concentrating on dynamics of interaction, self-representation and “truth-telling,” and unruly data, all with some consideration of focus groups with adolescents based on research conducted with 18 focus groups of secondary students, including exit questionnaires, on the topic of school rules in southern Ontario. I also draw on reflective, in-depth interviews with the primary investigator and three research assistants involved in this project to ultimately argue in support of focus groups as a method to be used with teenagers.

Literature

Dynamics of interaction

Focus groups are valued for the rich interaction they provide, fitting well with a theoretical orientation that privileges the social (Bloor et al., 2001; Kitzinger & Barbour, 1999; Smithson, 2000). Although focus groups are staged social events, they nonetheless are an attempt to capture the way meaning is made and negotiated: Participants explain and defend their positions (Warr, 2005), challenge others’ comments (Hyde, Hewlett, Brady & Drennan, 2005) and share excitement (Vaughn et al., 1996), teasing, and joking (Kitzinger, 1994). Kitzinger particularly values “turning points,” when participants change their minds through group interaction. Kitzinger and Warr both thus described focus group data as “unruly” because focus groups provide less studied personal “account-making” than that emerging through interviews and life histories. Kitzinger has argued that these unruly data are instructive “because people's knowledge and attitudes are not entirely encapsulated in reasoned responses to direct questions” (p. 108). The consequent partial, fragmented, and contradictory material illustrates how meaning is made and challenged within groups (Warr, 2005). Focus groups can also bring to light common assumptions that are frequently otherwise left unarticulated (Bloor et al., 2001), making this technique especially useful for research with children and youth as it can provide access to shared culture and “indigenous” terms/categories unique to peer groups (Bloor et al., 2001; Kitzinger, 1994; Morgan, Gibbs, Maxwell, & Britten, 2002).

Market researchers usually recommend focus groups with participants who are strangers to each other (Kitzinger & Barbour, 1999), yet if one of the key advantages of focus groups is that they allow us to see people interact within a somewhat normative social context, then focus groups with people who already know each other in preexisting groups might be particularly valuable (Hofmeyer & Scott, 2007; Kitzinger, 1994; Warr, 2005). Kitzinger also found that people in such groups were more likely than others to challenge each other on what they were saying. Practically speaking, they are also easier to organize. Others have raised concerns with such an approach, however. Focus group researchers such as Hofmeyer and Scott (2007) and Hollander (2004) worried that preexisting hierarchies, inequalities, and subgroups can silence participants. Vaughn et al. (1996) argued that focus groups with people who know each other might be particularly unwise with younger participants as they might be less able to keep confidentiality and must later interact within mandatory environments such as schools. Hofmeyer and Scott have contended that in groups of any age, facilitators must highlight the risks and benefits, note that confidentiality cannot be assured, carefully observe power dynamics, and ensure that participants can speak confidentially to them afterward if needed. The research topic of investigation and the context in which participants know each other are both important to consider as it might also be with friends that young people are most comfortable.

Another advantage of focus groups dynamics is that they can shift the power relationship from researcher to participants (Kitzinger & Barbour 1999, Wilkinson, 1998). Through numbers, focus group “participants have more control over the interaction than does the researcher” (Wilkinson, 1998, p. 114); participants can thus shape the nature of the discussion and provide mutual support (Warr, 2005). Kitzinger and Barbour (1999) contended that focus groups can even politicize through problem-solving and the creation of shared identities. As power shifts toward participants, the age-based power imbalance between facilitator and younger participants can be partially diluted (Wilkinson, 1998). Finally, such groups might provide participants with security in that each individual does not have to answer every question (Vaughn et al., 1996).

Self-representation and truth-telling

A recurring criticism of focus groups reflects the flip side to the making of meaning in groups, however: Does the group distort individual opinions? Does it affect participants' genuine self-presentation? These questions emerge from a position that privileges the individual's solitary voice, thus locating the individual outside of social relations and social context (Wilkinson, 1998). Yet there is, as Hollander (2004) has put it, a greater diversity of contexts to consider in focus groups than in interviews, which can silence and/or shape what is said in particularly complex ways. She has suggested that focus groups are complicated by what drew the group together, participant identities and inequalities, and the conversation topic.

Hyde et al. (2005) agree that focus groups with youth are particularly valuable for presenting elements of participants' subcultures. Their focus groups addressed the question of adolescent sexual health, and they found that male participants were “caught up in a roll of showing off and exaggerating in front of each other” (p. 2593). This valuable information on the dynamics of peer masculinity raises some similar concerns, however. Hyde et al. asked how a researcher is to determine between the validity of the presentation of subculture and the validity of people's lived experiences being presented: “Thus, although focus groups offer, on the one hand, opportunities to witness group dynamics of the sub-culture at work, they frustrate the analyst trying to distinguish when reports should be taken as truthful or untruthful” (p. 2592).

Individual self-representations might be quite different from subcultural self-representations. Mitchell (1999) found that in her research on girls' peer groups, focus groups prevented her from hearing about the private lives of the lowest status girls and cautioned against relying solely on focus groups where “members cannot leave the research encounter behind for their separate lives” (p. 37). In their study on boys and masculinity Frosh, Phoenix, and Pattman (2002) similarly found that boys would talk about peer bullying in individual but not group interviews.

Such concerns about normative censure are shared by a number of commentators on focus groups, including Warr (2005), Hollander (2004), Mitchell (1999), and Kitzinger (1994). Yet Kitzinger contended that although minority voices can be muted within focus groups, sometimes this process is informative in itself because “knowing what is (and is not) expressed in a group context may be as important as knowing what is expressed in a confidential, one-to-one interview” (p. 112). In this way Kitzinger has challenged researchers who would give individual interviews the status of accuracy over focus groups. She countered that people generate meaning through a variety of forms of talk and interaction; one is not more valid than another. This position is underscored by Wilkinson (1998), who has contested the individualistic position that focus group interactions “contaminate” individual views, arguing that what is produced in an individual interview is also produced within a social context that can involve censure. Finally, focus groups can provide support for nonnormative positions, with shyer people sometimes more likely to speak in groups than in one-on-one interviews.

In part because of this question of group dynamics and with an eye to what kind of data researchers are seeking, most researchers have argued that focus groups work best when used in conjunction with other data collection techniques (Bloor et al., 2001; Hofmeyer & Scott, 2007; Kitzinger & Barbour, 1999; Mitchell, 1999). For example, Smithson (2000) contended that focus groups are useful but only “as a forum for generating public discourses about a topic, and not as a way of uncovering participants' ‘real’ views” (p. 114), particularly as dominant voices sometimes drown out minority ones. She argued that this problem can be addressed through group homogeneity across gender and age, strong facilitation, and acceptance of silence. Mitchell (1999) argued that focus groups should be used only in conjunction with other research techniques, notably interviews. Bloor et al. (2001) have recommended pregroup questionnaires to collect demographic information and to access initial differences of opinion. Kitzinger (1994) also feels that pregroup surveys encourage people to argue their position. In sum, most researchers seem to agree that focus groups should not stand alone in cases where researchers are primarily interested in individual views.

Focus groups with young people

A small number of focus group researchers have drawn attention to several other issues that arise from specifically conducting focus groups with young people. For example, Morgan et al. (2002) have argued that in research with children, parents' roles as gatekeepers must be negotiated. In their work on children with asthma they found that some parents did not want their children to participate and others were unable to commit to ensuring that their child would attend. Although such gatekeeping is less significant when working with teenagers, the issue of parental consent still complicates the ease of conducting focus groups with them, and other gatekeepers can emerge, as I discuss below.

Morgan et al. (2002) have found it beneficial to keep focus group sessions short and to have activities rather than simply discussion, particularly for shyer children, and to allow participants to fiddle with a toy while the group progresses. They recommended two facilitators, although they suggest that in smaller groups having two facilitators might tip the power dynamics in favor of the adults.

Commonly, focus groups are recommended to have between 8 and 12 participants, although Kitzinger and Barbour (1999) have argued that this number is usually taken from marketing focus groups and that sociological researchers often opt for smaller groups, such as 5 or 6. Vaughn et al. (1996) recommended 8 to 10 participants for adult groups and 5 to 6 for groups with children or teenagers. Morgan et al. (2002) also found that smaller groups work better with young people, although they also encountered problems with too few participants, creating more of a serial interview. Hyde et al. (2005), while conducting focus groups with adolescents, found that larger groups (up to 12) were more relaxed although they also generated more cross-talk, which made transcription difficult later.

Data Collection

This paper is a reflection on a series of focus groups conducted with secondary school students on the topic of school rules over the summers of 2004, 2005, and 2006. Nine groups were conducted in a semirural region of southern Ontario (“Whitton”) and nine in a large city in Ontario (“Big City”). Each group was composed of participants who already knew each other in some capacity. The characteristics of the focus groups are described in Table 1.

Groups were created either out of friendship circles accessed through word of mouth or, more commonly, through a service organization, such as a club. Participants were asked what they knew and thought of the rules and their enforcement, what they would change, how they appealed unfair accusations, and whether they had ever participated in creating their school rules. All participants shared pizza during the group, and each received an honorarium of $10 at the end. After each focus group participants were asked to fill out a short exit questionnaire asking for some demographic information and if there was anything else that they had wanted to say but did not in the wider group. Unfortunately, exit questionnaires yielded a very limited number of new comments. These postgroup questionnaires were also sometimes left nearly blank, producing problematic gaps in our demographic data. ¹ Overall, the exit questionnaires were a somewhat useful check on the focus group material but perhaps not as useful as interviews would have been. ² Finally, data reports on each of the Whitton and Big City regions were sent to all participants who supplied contact information, along with an optimistic invitation for comment. I was disappointed that no comments were received, and among the Big City participants a third of the reports were sent back as undeliverable.

Reflection on these focus groups is bolstered through four interviews that were conducted in the early fall of 2007, all of which considered the above data collection process. The project involved a research team composed of me and three research assistants. I conducted interviews with each of these assistants, asking them to contemplate the focus groups as well as the transcription process. Participants each received an honorarium of $20. “Lori,” my most senior assistant and a senior undergraduate, had done the onsite and post-focus group transcription for almost all the Whitton focus groups and had coded and written a report on the Whitton and Big City focus groups. “Simone” (a senior undergraduate) and “Agatha” (a masters of arts student) alternatively facilitated and conducted onsite transcription for most of the Big City focus groups. They also transcribed most of these focus groups afterward. Finally, a colleague interviewed me. All dimensions of data collection were approved by my university's ethics board. All participants and regions are referred to with pseudonyms.

Table 1.

Focus groups

Description of Focus Group	Participants
Whitton
1 Youth center in mall	15–17 years; 5 females and 2 males.
2 Street youth drop-in shelter	16–21 years; 4 females and 10 males; 7 youth were out of school, 3 due to age
3 Political youth group	16–18 years; 3 male
4 Performing arts group	1 13-year-old, the remaining members 16–17 years; 3 females, 1 male
5 French school group	15–18 years; 6 female
6 Catholic school group	17–18 years; 2 female, 2 male
7 Boys and Girls Club	13–16 years; 4 female, 2 male
8 New immigrant youth group	15–18 years; 2 female, 5 male
9 Lesbian, gay, bisexual, transsexual, queer and questioning youth group	15–19 years; 5 male, 4 female
Big City
1 Inner city youth drop-in	14–18 years; 2 female and 7 male
2 Informal group	14–15 years; 2 female and 1 male
3 Youth leadership group	15–17 years; 3 female
4 New immigrant youth group	14–18 years; 4 female, three male
5 Alternative hobby group	14–15 years; 3 female, one male
6 Youth homeless shelter	17–19 years; 3 female, 2 male
7 Boys and Girls Club	14–16 years; 3 female, 4 male
8 Catholic youth group	15–18 years; 3 female
9 Native group	18–19 years; 1 female, 4 male

Interviews were transcribed and coded on paper for open and abstract codes. Coded segments were subsequently organized into common themes and points of difference across the interviews. In this paper I am concentrating on the common, dominant, emergent themes that arose through this process: dynamics of interaction, self-representation and truth-telling, unruly data, and other related practical challenges such as facilitation.

Emergent themes

Dynamics of interaction

As the literature suggests, we found this interaction to be rich. My research assistants noted how familiarity between respondents led to playful teasing between them, finishing each other's sentences, and calling each other out. A nice example of such process was in Whitton Focus Group 6 when Ang's friends suggested that her position on respect was undermined by a past incident:

Ang: Like I don't think that you should swear at a teacher. I don't think that/

Kierstyn: What about Mr. Hendricks?

[Ang and Kierstyn giggling]

I: What? What?

Tetrad: What about Mr. Hendricks?

[…]

[laughing]

Ang: No… No! I've never ever, I've never ever sworn at a teacher.

Kierstyn: Ok then [inaudible]. Like swearing or respecting?

I: So you're saying that Ang, there's a teacher she does not respect?

Kierstyn: Well that she hasn't respected before. [chuckles]

I: Ahhh

Kierstyn: But like he was rude to her so she was rude back him. […]

Ang: I think that if you … give out respect, then you should receive respect [chuckling]. ³

We noted that participants bounced ideas off each other, debated, and had to justify their positions as well, as we see here in Whitton Focus Group 9:

Jason: I think it's perfectly fine to wear a hat in the hallways but why would you need to wear one if you're sitting at a desk?

Allison: Because it's a cool thing, I mean it completes your outfit. Why not?

Jason: If you're sitting down, they're only seeing from here up anyways [indicates towards his middle]. So you don't have an outfit … […]

Allison: If you don't have an outfit, maybe [the hat] makes the outfit.

Jason: I just don't see the point of wearing a hat in class

Betty: What if you have a really, really bad hair day, and you put a toque on?

Jason: Then you live with it. You put a hairnet on or something.

Although group discussions in the end tended to produce some consensus, the process illustrated a diversity of opinions and loosely reflected an everyday dynamic through which young people negotiate meaning.

Other focus group researchers have recommended some kind of game or project to stimulate interaction, especially among younger participants (Kitzinger, 1994; Morgan et al., 2002; Peterson & Barron, 2007). Along these lines, we included an activity involving the distribution of rules on cue cards into three piles: good rules, bad rules, and debated rules. This cue card activity helped generate interaction and was a useful tool for shifting the focus away from the facilitator, especially when one of the participants would be responsible for reading through the cards. As my research assistant Lori said,

[With] the cards, it kind of happens a little more organically, especially if they happen to be in disagreement and they'd be like [changing voice] “oh no, I don't think so” and then they kind of go off on each other.

Lori also noted that the cue cards provided for a degree of spontaneity.

Like you wouldn't have to specifically ask questions so much, which can be difficult, right, because like you don't even know what you want to ask. [laughs]

The cards also provided a useful cross-group comparison at the end (Kitzinger, 1994).

As noted already, a key advantage of focus groups is that they can shift the power relationship from researcher to participants (Kitzinger & Barbour 1999; Wilkinson, 1998), particularly in research with young people. This value was raised by two of my research assistants and in my interview. As Simone noted,

I think unfortunately whenever you get people, I don't want to say older but I think they saw us as older, and younger youth, you have this automatic hierarchy, right? So I think even though there was this hierarchy between Agatha and I and these young people, they [inaudible] in greater numbers … none of them had any problem disagreeing with the things that we said. None of the groups… . Whereas I wonder if it has been one-on-one if they would be less likely to have corrected the assumptions we made in the questions, maybe?

Despite there being two adult researchers at the interview, Simone felt that the participants were willing to challenge the facilitators' assumptions. This response might also have been helped by the fact that all but two groups were held on the participants' “turf.” Smithson (2000) has noted that an advantage of focus groups is that the group has “access to shared knowledge of which the moderator is ignorant” (p. 112). We certainly experienced this when participants enthusiastically focused on discussing specific incidents, teachers, and/or administrative staff.

Within our groups, some participants remained silent. This could be seen as an active disengagement to prevent disagreements from disrupting group interaction (Jordan et al., 2007) or a problematic silence (Hollander, 2004) resulting from the discomfort with the group dynamics and context (Mitchell, 1999). ⁴ For instance, we noted that frequently female participants in large, mostly male groups were likely to contribute less, a concern discussed below. On the other hand, sometimes silence can be seen as a valuable ethical option where, in contrast to an interview, participants are less likely to feel obligated to answer questions. The lack of private dissenting comments made on our postinterview questionnaires corroborates this possibility. Overall, the vibrant group dynamics in which participants negotiated their views suggest that the focus group is a particularly effective tool for gaining insight into young people's public views. The issue of normative censure is of concern, however.

Self-representation and truth-telling

As discussed above, Hollander (2004) and others have raised questions regarding the accuracy of focus group comments, a concern that commonly reflects critics' arguments that the group contaminates individual opinions. Many focus group advocates have countered that individual views are not isolated and that they shift and change through social interaction, interaction well captured through a focus group (Kitzinger, 1994; Warr, 2005). Nonetheless, we have seen that many working with focus groups have stressed the need to complement this method with another out of concern that participants can be silenced or might adjust their answers within a group (Hofmeyer & Scott, 2007; Hollander, 2004; Mitchell, 1999). My research assistants also raised such questions of accuracy. Simone and Lori both observed problematic silences (Hollander, 2004) among the girls in larger mixed-gender groups. ⁵ Also, Simone felt that the young people were giving very scripted answers in two of the groups she conducted, and Lori felt that sometimes participants exaggerated their rule-breaking as a way to show off to their peers, whereas in others they lied when they said that they did not break the rules at all.

These worries about truth-telling relate to identity, context, and power. First, the social categories of gender and age are both referenced here. Hollander (2004) and Vaughn et al. (1996) have found that gender dynamics can silence certain members, especially girls in groups with large numbers of boys. My assistants join these researchers in suggesting that sex segregation might counter such silences as well as disruptive behavior, although our smaller focus groups, with equal numbers of female and male participants, seemed to garner more equal participation. Age is another category relevant to conducting focus groups with young people. For instance, Vaughn et al. (1996) suggested that in focus groups with young people, participants should be within 2 years of one another in age. We had two groups (Whitton Focus Groups 7 and 9) with significant diversity in ages, and in each the older participants tended to dominate the conversation, supporting this recommendation. Again, however, this was not a consistent pattern across mixed age groupings, and such groupings also provided some fruitful inter-age interactions. As Tina (Whitton Focus Group 4) said to a younger participant on the pressures to dress in revealing clothing, “I think it kinda changes when you get to high school like, generally people don't really [stammers] respect people who wear like skimpy, like it kinda changes in high school.” Homogeneity of either age or gender might bring comfort to some participants, although insight through interaction across difference can be lost along with insight into the ongoing salience of inequality in the public construction of meaning.

Beyond the status composition of groups, focus group context is relevant to these concerns about self-representation and truth-telling in other ways. Recall Hyde et al.'s (2005) observations that peer group subcultural norms shape what is said in focus groups and Lori's similar concerns about participants showing off. As Hyde et al. noted, it is quite difficult to untangle what parts of focus group participation reflect peer cultural norms and what parts reflect more individual positions independent of the group. This is particularly difficult if we recognize that both the performance of self in a group and performance of self individually are “true” and interlinked (Frosh et al., 2002). After conducting both interviews and focus groups with boys, Frosh et al. argued that different contexts provide “different ways of ‘doing boy’” (p. 32) but that one is not more valid than another. They conducted both focus groups and individual interviews with their participants, providing a nice contrast in presentation of self, although they also interviewed each boy twice and found that presentation of self changed between interviews. Conducting a sufficient number of focus groups can also allow researchers to compare contexts. For example, Whitton Focus Group 2 and Big City Focus Group 1, both with disenfranchised youth, stood out in terms of participants' experience with rule-breaking and evidence of potential posturing or bravado about that rule-breaking.

The context in which the groups took place was also relevant in our research for a reason quite unique to groups with young people, which ties in to the question of inequality. Unfortunately, in two focus groups, Whitton Focus Group 9 and Big City Focus Group 4, adults in a position of authority over some of the participants were present, a police officer in the former and several parents wandering in and out of the latter. As Simone observed, “If anyone's answers were scripted, um, I, I would assume it would, it would have been those kids‘.’ … Their families were sitting virtually on top of them.” ⁶

A second group that concerned Simone in relation to the question of context was a group of homeless youth living within a shelter in Big City. Quite unlike our other sessions with disenfranchised youth, this group spoke in favor of virtually all school rules. Simone felt that their responses seemed incongruent with their lives as a number of them had been expelled at some point. She hypothesized that within this focus group the facilitators were seen as authority figures like those that structured their lives in the shelter, and they therefore moderated their responses, or the participants had fully internalized expectations of obedience. Despite her earlier comment on participants' willingness to challenge the facilitators, Simone suggested that these participants “were just so indoctrinated … like they were just so used to giving the right answers.” Later, when I asked if this desire to repeat back what one is told undermines the data, Simone said,

No, I don't know if it undermines the data per se but I think it says a lot about how young people think … cause I think, I think even though maybe those are scripted answers for them, I think that a lot of them actually believe them.

Peer relations, the presence of adults in several focus groups, Simone's concerns about participant compliance, and Lori's similar concerns about some participants wanting to “say the right thing” all raise the important issue of power. Reflections on focus groups with young people could emphasize inherent features of children or adolescents in order to explain group dynamics; for example, that young people are more likely than adults to lie, to internalize rules, or to impress peers. I do not adopt such disparaging or homogenizing conceptualization of young people, however (Raby, 2007). In fact, the diversity of dynamics across our focus groups belies such generalizations, as do other researchers' commentaries on self-representation and truth-telling among adult participants (Hollander, 2004). Instead, my research assistants' concerns seem much better explained through the previous attention to context but also to the lenses of power and inequality directly related to age.

Despite the promise of focus groups to shift power relations, ongoing effects of age and research-based power imbalances shaped this project. First, because participants were accessed through service organizations that they were already involved in, they had little choice about who would be in a focus group with them and where it would take place. There were even occasions where, despite having signed letters of consent, participants did not seem to have an idea in advance of what the research was about. Such organizational issues can also be present for some adults involved in focus groups, of course, although young people's dependent status might particularly limit their sense of autonomy within service organizations. Second, many of these organizations were also governed by rules created by adults, perhaps seeping into the research relationship, as Simone suggested. Third, as mentioned above, in two groups participants could not control the presence of adults with authority over their lives. Finally, within the research process, the participants' comments are interpreted by adult researchers. In this last instance, discourses of young people, as compliant, rebellious, or peer influenced, for example, can all shape how their comments are interpreted and whether they are considered legitimate (Raby, 2002), adding a further power imbalance to the one already embedded in any research relationship.

What can we make of all these concerns? Context affects the responses of participants in any kind of data collection. This can be addressed in part through triangulation, saturation of the field, attention to research design and a thoughtful consideration of the many ways in which inequalities are manifest in research. It is also important to recall that some of the inequalities between adults and youth are, in fact, countered through focus groups, as discussed above. Of course, the relevance of these concerns also depends on the topic under investigation (Kitzinger, 1994). For example, for us the public nature of focus group discussions seemed to well represent the public nature of meaning-making around school rules. In fact, the repetition of views that quickly emerged across the focus groups we conducted suggested a commonality of public opinion that focus groups were successful at capturing. As Lori noted, “They all say the same thing [and] at the same time they all have their unique twists on it.” Focus group discussions undoubtedly include omissions and partial fabrications in self-representation, as do other forms of data collection, yet consistency of comments across groups, authenticity of exchanges such as the one between Ang and her friends above and the relevance of public presentations of self all point to the value of these data.

Unruly data and other practical challenges

In this final section I address a more practical challenge that arose in our focus groups, unruly data and the related issues of access and facilitation. In focus groups, people's account-making is frequently partial, fragmented, and contradictory, making it unruly yet uniquely valuable (Kitzinger, 1994). Yet, in my research this unruliness occasionally shifted into chaos, undermining transcription and contributing to debates between my research assistants and me about facilitation, specifically between allowing natural, spontaneous interaction and attempting teacherly control.

The greatest challenge of conducting our focus groups arose in large groups or groups with more than 4 or 5 participants. The dynamics of the groups in general involved young people who knew each other, and consequently they sometimes had very excited exchanges. Interruptions, cross-talk, loose conversational style, and slang are all potentially linked to such dynamics. In such groups, participants would be more likely to develop side conversations, sometimes creating a flurry of cross-talk that was impossible to transcribe. There were several groups where almost half of the data was lost through what one research assistant described as “pandemonium.” Participants were hard to identify and were more likely to provide short, undeveloped answers to questions when in larger groups. Smaller groups, in contrast, seemed to produce more conversation, with a greater percentage of participants involved. Both Lori and I suggested in our interviews that groups of 3 to 4 participants were ideal, particularly if they were all friends (even if they were of different ages or backgrounds).

Despite our consequent desire to keep groups small, sometimes this was not possible. Service organization personnel were invaluable to this research. Their organizations provided a structured context through which to coordinate and conduct the focus groups, and participants were drawn from a wide range of social circles and class backgrounds. Yet this strategy posed some quite serious challenges. Often we could not control group size, particularly when focus groups were conducted during a drop-in time. Also, we could not control when or where the groups would take place, nor what to expect in terms of setting. For instance, some of the spaces had challenging distractions (big windows onto the street where friends would be walking by or a television on in the background) or unfavorable acoustics. In one venue the focus group was held at the end of the day, and the participants were restless and impatient. Further, we had little control over how the project was presented to potential participants. Service organization personnel acted as gatekeepers in terms of whether the organizations would participate at all, who was encouraged to participate, and how information about the project was communicated. Finally, the preestablished culture of organizations was something that we were unfamiliar with. In contrast, the groups we conducted with friends who had been located through word of mouth were frequently smaller, engaged, and more easily transcribed, although they also tended to be groups of young people who were relatively invested in school (and school rules). Two groups were even conducted at the university, allowing us full control over the space in which the group would occur.

Despite these challenges, I argue that if the only way to access some groups of young people is through such large groups (e.g., through service organizations), then it is worth conducting such groups. For example, it is only in our two largest, most challenging groups (Whitton Focus Group 2, Big City Focus Group 1) that we heard participants critically assess the rule against fighting, providing important data on student engagements with school rules. The tradeoff was in the transcript.

All of our focus groups were conducted by a facilitator and also had a second person who would sit in the corner and take detailed notes, with particular attention to recording who was speaking. Later transcription was also facilitated by having two tape recorders, one digital and one conventional. Ideally, the same person who conducted the onsite transcription would be responsible for the later transcription. Despite these precautions, we found the transcription frequently challenging. ⁷ In sum, the richness of the data produced in focus groups with young people who are friends might be compromised through later challenges with transcription, especially in large groups, but these groups can nonetheless provide valuable data.

In training Simone and Agatha on how to conduct focus groups, I facilitated the first Big City focus group with them both present. Despite my emphasis that participants should speak one at a time, this group turned out to be large and fairly chaotic, with participants frequently talking at the same time. Afterward, Simone and I debated the extent to which such a group should be managed by the facilitator. Specifically, Simone felt that I should have been more teacherly, although she debated between structure and spontaneous communication:

I think that that was such a, they're so not used to [a loose structure], especially given like where … where a lot of those kids come from, you know underprivileged and they're probably kids that everybody's trying to control and get in, in order. So that it was probably a really bizarre place for them to have so much … freedom in a way, that they could just like talk whenever they wanted … It's hard, right, because at the same time we don't want to set them, you don't want to set up a hierarchy between you and the kids but at the same time that's all they really know and so there was so much distraction in the room, there was a lot of data I think we just logistically didn't have, logistically that we didn't catch, which is unfortunate for the research but if maybe we had kind of set more ground rules we would have caught more data, but at the same time, would that have compromised the type of data that they would have given?

This middle ground proved difficult to negotiate. By leaving participants to talk among themselves the focus group best captures the participants' natural, spontaneous interaction. Sometimes in our focus groups participants even seemed to forget they were in a focus group, and these moments often proved very rich. At one point in the first Big City focus group, for example, one of the younger, quieter boys talked into his shirt about the need to be able to defend yourself in fights and to carry weapons because otherwise people will disrespect you. For Agatha, this was the most powerful moment in her entire time with this focus group project.

For me to hear, you know, children at that age talking about things like this … the fact that they had, you know, policemen arresting them even at that age, and, and that whole thing becoming normal for them was something that was just like ‘wow’. It, it was a very, it was a very strong moment for me… . So it was real to me. And now every time I hear anything about gun violence in Big City, I just remember that I, I so vividly remember that.

The kind of interaction that yielded such important material might not have arisen if we had been more strict about structure but also proved very difficult to capture later when it came to transcription, and consequently, as Simone noted, we lost a lot of valuable information. This points to the importance of a skilled facilitator and limiting group size but also to a possible tradeoff between quality of focus group data produced and ability to actually record the data.

The experience, identity, and perceived identity of facilitators are also important in focus groups, as they are in other forms of data collection. Smithson (2000) has suggested that it might help to have a facilitator who has a similar background to the participants. As she noted, however, shared identifications are quickly complicated by the relevance of diverse, intersecting identifications. Age is only one of many salient features of our identity that can affect the insider or outsider status of a facilitator working with young people (Raby, 2007). One attraction to having research assistants facilitating focus groups with teenagers is that they are frequently younger, although most advice on conducting focus groups emphasizes the need for someone who is experienced and skilled at facilitating (Bloor et al., 2001; Morgan et al., 2002), and other features of identity such as gender or race might be equally or more important. It is also quite difficult to determine the effects one's perceived identity might have on a group.

All three of my research assistants commented on their identity and how they thought it might be perceived, but in the end such comments did not consistently support an ideal of shared identity. For example, Simone at one point suggested that she felt rapport with one group she facilitated due to her closeness in age (mid 20s), yet she in part attributed this to a shared class background. Also, this importance of shared age was countered by Lori, who positioned me as young although she is almost 15 years my junior:

Lori: I think they would be more intimidated than if they were to do it with you. Rebecca: Even though I'm the prof?

Lori: … I'm kind of scary looking [laughing] … My presence is more kind of intimidating.

Agatha also noted that some participants might have held back because they saw us as being like teachers but that overall she has found that here in Canada (as opposed to her home country, on Africa's west coast) “a lot of people here, young people, are outspoken nonetheless, whether or not, you know, they are speaking with a grown up or not,” suggesting that age was relatively unimportant. My research assistants discussed rapport in terms of race and status as well as age. Simone assumed that perhaps Agatha would have been able to better establish rapport with a racially diverse group she facilitated (Big City Focus Group 6) because Agatha is Black. Yet, when discussing a group she found particularly challenging (Big City Focus Group 7), Agatha noted that the racially diverse participants might have been less respectful of her because of her age and might have adjusted their behavior had I (a White professor) been there:

I just felt that I had no control whatsoever over, you know, the whole process… . I keep wondering if something would have been different if you were with me. I keep thinking that something may have been different in that [pause] and, and those moments sometimes when I wished I looked a little older. Or even, you know, like I possessed a little more authority. Because then I keep thinking if you were with me and I had introduced you as a professor, then maybe they would sit down a little and, but I don't know whether they, you know, I don't know. It, it's just, it was just so [pause] all over the place. (Agatha)

In this case shared racial background and similarities in age were not seen to be useful for the facilitator. In fact, within this focus group several quite troublingly racist comments were made by participants about a teacher of African heritage, adding to a chilly climate for Agatha's facilitation and providing an example of how the balance of power in focus groups can undermine a facilitator (Kitzinger, 1994). Ultimately for this research into public talk about school rules, I was not convinced that a facilitator's similar age or racial background necessarily drew more information from participants or led participants to feel more comfortable. Overall, facilitator skills seemed more important than identity, although this quite likely depends on the research topic.

Conclusion

This examination of focus groups has considered their dynamics, particularly in terms of the richness of the data they produce, inequality, and concerns about self-representation. I have also touched on some of the more practical issues we negotiated in conducting these focus groups with adolescents. I come away from this reflection with several conclusions. First, inequalities between young people and adults support the focus group method by tipping the balance of power away from the facilitator(s) toward participants and creating a comfortable, engaging environment for most participants. Yet focus groups are not a panacea. Young people share a lack of power in our society, inevitably shaping any research adults conduct with them, particularly in terms of organizational and parental gatekeeping, the context of the focus groups, and common discourses of childhood and adolescence providing an added interpretive filter (Raby, 2007). The one unifying feature in conducting research with young people is this relative lack of social power, although this is disproportionately experienced in the pronounced diversity across groups of young people which must also be considered in terms of participants' self-representation.

Second, I concur with Bloor et al. (2001), Kitzinger (1994), and Hollander (2004) that the effectiveness of focus groups as a data collection tool depends in large part on what kind of information one is trying to access. If a researcher's primary focus is on specific, individual experiences or views, then focus groups, on their own, might not be the ideal form of data collection with young people or adults. Participants might prefer to keep individual experiences private, for instance, particularly when they will continue to know other participants in other contexts and when they are subject to bullying (Frosh et al., 2002; Mitchell, 1999). Also, within some of the focus groups, certain answers might bestow a level of status on participants and therefore be preferred. However, where scholars are interested in public meaning-making and young people's interactions, the focus group is an excellent tool, particularly when supported by some form of triangulation, even if only to provide participants with a venue through which to privately disagree. As the primary focus of our research was on public meaning-making among students on the topic of school rules and as we found a high degree of both spirited debate and repetition across focus groups, the focus group seemed valuable in our case.