Introduction

Qualitative methods are increasingly being used to study a variety of health topics; however, these methodologies are just beginning to be used to explore injury prevention. In the field of injury prevention, qualitative methods have been used primarily in programme evaluation—for example, evaluating programmes designed to decrease occupational injuries, educational programmes to increase children’s and adolescent’s knowledge of a variety of risky behaviours, assessing a self-help programme aimed at preventing injuries to preschoolers, establishing effectiveness of hospital-based injury surveillance systems and evaluating an injury prevention programme in Australia.12345 Other published qualitative studies have addressed the reasons for drinking and driving in adolescents.6 Qualitative methods have also been used extensively by Rothe (first author of this paper) in studies of traffic safety issues such as seatbelt wearing, impaired driving, traffic law enforcement and risky driving.7891011121314 In these studies, interview findings were used to capture the experiences of both injury victims and traffic law violators, in order to aid in the design and implementation of interventions aimed at improving the safety of vulnerable populations. The objective of this paper is to discuss the purposes, formats and methods of interviewing and to introduce an innovative and culturally appropriate method of interviewing indigenous peoples, an adaptation of these peoples’ traditional “Sharing Circle”. This strategy is illustrated by a case study.

Qualitative interviewing as a research methodology should be used only when appropriate to the research question. Quantitative research questions are usually best suited for questions such as “how many?”, “what proportion?” and “what characteristics?”. To evaluate the effectiveness of treatment, a randomised controlled trial may be best; to identify the incidence of an injury, a survey might be suitable.15 Should the questions pertain to the meaning of a topic, such as “why and how do parents worry so much about their teen becoming involved in drinking and driving?”, then the focus shifts to psycho-socio-cultural factors, and qualitative methodology should be used, with interviews being a core strategy to answer these particular questions. The aim of qualitative interviews in injury research is to disentangle affected individuals’ motivations and perceptions and describe how they impact on injury in an everyday context.

Qualitative interviews are typically used to uncover normative patterns of behaviour, as they reflect psychological intents and motives, social action displayed in modes of interaction, and types of power relationships, cultural beliefs and shared knowledge. To accomplish this, researchers analyse social settings/situations, formal and/or informal organisations and cultural events. Within these settings, qualitative researchers may focus on exploring individuals’ beliefs and self-perceptions, their personal histories/biographies, their ways of interacting, the “taken-for-granted” features of everyday life such as unquestioned habits and routines, and the way in which people deal with their roles and functions within the context of their daily lives.16 The emphasis is on naturally occurring language that people use to describe features of their lives.16 This is in accordance with the basic premise that people’s social actions arise from their consciousness of themselves in relationship to others and their social and physical environment.17

Qualitative research is ideal for developing intervention approaches that consider such issues as people’s reaction to stigma, the importance of role modelling, the special needs of a vulnerable population, and the reasoning and belief systems behind risk-taking behaviour. For example, qualitative researchers examine risk-taking behaviour not so much as an individual characteristic, but more as a human action that is influenced by the actor’s motives, social relationships, cultural norms, traditions, roles and values, and the social context that supports (or at least tolerates) such behaviour. The phenomenon of “injury” (and therefore its prevention) may be quite different for those groups of people to whom injuries are rare than it is for those groups for whom injuries are commonplace and thus considered normal. These differences affect the type of intervention strategies that are most likely to be both effective and accepted by different communities. An epidemiological study that reports the incidence of suicide and identifies the subgroups at highest risk of committing suicide provides useful information about which communities to target for special intervention programmes. On the other hand, qualitative researchers can use interviews and other methods to carry out in-depth explorations of how individuals and groups define, react to and “make sense” of phenomena such as suicide. Thus, qualitative studies can provide information on how suicide is viewed and acted upon by communities, and how family members, friends, colleagues and other suicide survivors respond to it in terms of their own lives. Together this information can lead to more effective and socially/culturally sensitive prevention programmes. Another relevant example for the use of qualitative research in injury prevention is senior drivers and interventions to improve safe driving for this population. Whereas a randomised controlled trial can determine the efficacy of a senior driver course to improve safety, a qualitative study can address the issue of why seniors do or don’t take these courses. According to Rothe’s findings from a qualitative study, 60-year-old drivers typically do not consider themselves to be “senior” drivers, and believe that senior driving support should target those over 70 years old. Those drivers in their 70s consider “senior drivers in need of driver support” to be those drivers over 80 years old. Regardless of how traffic safety experts define senior drivers, in the “everyday world” of the senior, “senior drivers” are those over 80 years old. Anyone younger is likely to not participate in senior driver skills improvement courses because of their self image as a non-senior person.18

There is growing recognition, particularly within injury research, that risk-taking behaviour reflects a complex interaction between individuals’ beliefs and behaviours and their social and environmental context. Thus, the dominant beliefs and attitudes of the community—that is, the social milieu in which the individual lives and operates—and the cultural and physical environment need to be understood before successful prevention strategies are possible.19 The usual qualitative method of arriving at these issues is through directly engaging individuals of interest in person-to-person or small group interviews, or to interview persons associated with that individual, such as parents, healthcare workers, friends or community leaders. Interviews are sometimes used in health and injury research as stand-alone methods to generate insights about an issue or a reconstruction of experience, but are more often used as complementary methods which can be adapted along a continuum, creating an overlap between quantitative and qualitative methods.20 For example, interviews may serve as the starting point for quantitative research methods, establishing themes that are explored further by surveys, or, alternatively, may follow a statistical analysis to determine the social reasoning behind the quantitative findings.

A holistic view includes perspectives offered by various groups within communities such as ethnic minorities, women, children, non-English speakers and occupational groups. This broad inclusion offers a variety of opinions and perspectives, providing greater likelihood of capturing the essence of an issue within the community context. Another method to achieve a holistic description is through convenience sampling—for example, soliciting content experts or knowledgeable community members to speak on behalf of an issue. This increases the likelihood of obtaining information that may not be uncovered by a random selection of participants in particular communities.

In this paper, we describe conventional, commonly used qualitative interviewing methods. However, these interviewing methods may not be optimal for researching non-mainstream cultures—for example, indigenous peoples of North America (such as First Nations peoples, Métis and Inuit). Conventional interview methods may provide glimpses of how we deal with injury, but seldom offer insight into the feelings, learning and conditions that engender injury in an indigenous community. To arrive at a better understanding of indigenous people engaging in risky behaviours, we propose an alternative interview approach that more fully portrays cultural values, norms, interactive processes and argot. This alternative approach integrates culture-specific communication styles into qualitative interviewing techniques. An example of such an approach that penetrates into the cultural lives of some of Canada’s First Nations communities is the Sharing Circle (discussed in greater detail below).

Interview formats

Interview formats may be structured, semi-structured or unstructured. Structured interviews pose and seek detail on predetermined questions.21 All participants are asked the same series of questions with forced-choice response categories. This facilitates statistical analysis, but excludes answers that participants may believe are more relevant or valid, and does not permit the participants to expand or qualify their responses. This interview format is best used when the subject matter is already well researched and understood.16

The semi-structured interview also consists of questions developed in advance; however, each question may have a secondary “probe” and allow for “tags” and “prompts” that broaden the meaning of the answers, which allows varied information to be obtained. A probe may consist of a “why” question, whereas a tag is the introduction of such phrases as, “Do you mean that…” and “Give me an example of…” to attain a more embedded sense of meaning. A prompt asks the participant to “Tell me more about…”.

For the unstructured, or open-ended, interview, there are a few guiding questions, such as “what led up to your son’s death in a car crash?” The interviews are more akin to normal conversation whereby questions are formulated as talk progresses. This format is more likely to be used to start research projects when researchers do not know in advance which questions may be relevant,222324 or they are intended to be portrayed as autobiographies or storytelling.25 The conversation can be aimed in any direction, not necessarily in a natural progression. The interviewer becomes an attentive listener, shaping the process into a familiar and comfortable form of social engagement—a conversation—where the quality of information obtained is largely dependent on the interviewer’s skills and personality.26 Thus, unstructured interviews are arguably the most difficult kind of research interview to undertake. They usually constitute the preferred form of interviewing with indigenous people, whereby a few guiding questions are asked and the respondents fill in the rest through storytelling and anecdotes.1327

To acquire the best information using this approach, interviewers should be trained in interviewing (eg, have good listening skills) and be sensitive, empathetic and able to establish a non-threatening environment. Interviewers are often selected according to personal characteristics, such as age, sex, profession, cultural and or group membership, aimed at helping the interviewee to feel more comfortable and open. Matching of interviewer to interviewee usually enhances the quantity and quality of the information obtained. It is also important that interviewers be familiar with the project, its goals and the social context within which the research project is undertaken. Furthermore, the interviewers should be properly trained in the social interaction typically used by the target group, especially when dealing with a marginalised group such as street people or indigenous peoples. Poor interviewing skills, poor phrasing of questions, or inadequate knowledge of the participants’ culture or frame of reference may result in collecting irrelevant data.

When using unstructured interviews, researchers need to define and adapt questions to the study participants’ “everyday realities”—that is, their own use of language and world views—to avoid misunderstanding and to be consistent with their argot.2829 Interviews must be flexible. Participants often present information that is inconsistent with the researcher’s previous knowledge or which does not fit the study’s original design. In unstructured interviewing, such unexpected developments are welcomed, as they allow the development of novel interpretations of data and breakthroughs of understanding. Using unstructured techniques enables a focus on participants’ deep structured knowledge that may not become evident with other strategies. However, unstructured approaches are not suitable for every situation—for example, they are less conducive for changing or reworking a weak original hypothesis.29

Interview methods: person-to-person, focus group and Sharing Circle interviews

Interview methods can be conceptualised as being on a continuum. At one end is the person-to-person interview while at the other is the focus group interview that typically consists of six to 12 participants. However, there are also those occasions when an interviewer converses with a respondent, and a second person (or more) joins in and contributes valuable information. This could be a hindrance or an asset, depending on the arrangement, the topic of the interview, and the relationship between the participants.

A focus group consists of a small group of people who offer their opinions and experiences on injury-related phenomena, such as pain, attitude towards medical services (such as ambulances) or the implementation of traffic safety-related equipment (such as photo radar cameras), or their perspective on drinking and driving advertisement programmes. Questions are asked in an interactive group setting in which participants are free to talk according to a turn-taking process defined by the facilitator. The participants are usually offered the opportunity to “explain further” or “give an example”.

In addition to these more conventional interview methods, Rothe (first author of this paper) developed a novel type of interview methodology that is particularly suitable for studying injury in First Nations peoples. This method is an adaptation of the traditional First Nations communication tool, the Sharing Circle. It is a unique type of focus group that involves interviewing in a completely natural setting, in the local language and according to local norms and customs. The unique features of this methodology are described in more detail below, but involve greater involvement of the interviewer as a participant in the interview process than is usually the case, and involvement of participants in data analysis and interpretation. This is consistent with the equality practiced in traditional Sharing Circles.

Sharing Circles as a traditional method of communication

Sharing Circles themselves are traditional communication tools used by indigenous people (and other cultural groups who have a strong tradition of oral history) to discuss issues and topics in an egalitarian, supportive and non-confrontational manner. They reflect the indigenous people’s values of sharing, supporting each other and respecting life experiences through the use of personal interaction, and group consensus to identify problems and derive solutions.30 Sharing Circles are especially important in First Nations cultures.13 They incorporate oral traditions and styles of interaction entrenched in the culture. They are typically used to identify problems and determine solutions by providing culturally enriching environments that offer participants “balance” and “support”.1430 The Sharing Circle is used to teach culture and tradition,31 promote health,32 and provide spiritual counselling and healing to members of indigenous communities.

Using a Sharing Circle methodology as an interview strategy with indigenous populations can provide a greater richness of information than the more usual interview techniques, while also providing a highly culturally sensitive research environment.

Similarities and differences among interview methods

One-on-one interviews can draw out people’s experiences through storytelling. Focus group interviews are equally valuable. As above, Sharing Circle interviews are an adaptation of focus groups, using methods that are specifically designed for cultural sensitivity. Although both person-to-person and focus group interviews vary in their format and content (eg, either could be highly or loosely structured, depending on the study topic and what kind of information is sought), we have outlined general similarities and differences among these interview types in table 1.

To illustrate the strength of using Sharing Circle methodology in qualitative research with First Nations peoples, we will describe recently completed studies performed by the first author of this paper and reported in detail elsewhere.1333 These studies were approved by both the health research ethics board (health panel) at the University of Alberta and the Saddle Lake Community Ethics Committee (the First Nations ethics committee relevant to this project).

A study of impaired driving in First Nations communities

Background

Injuries occur disproportionately in First Nations members. Injury is the leading cause of death among 1–64-year-olds in this group, and they suffer six times more injury fatalities than non-First Nations people in Canada.3435 Motor vehicle collisions account for 40% of unintentional injury fatalities among Canada’s First Nations people,36 and more than 80% of fatal crashes involving young First Nations men were alcohol-related.36 The overall focus of the following research project was to understand the high rates of impaired driving in the First Nations communities by focusing on the experience of a First Nations community using Sharing Circle interview methodology to collect data. We use this example to illustrate how Sharing Circle interviews were developed, conducted and analysed, and how these findings were used.

Method and design

In partnership with the First Nations people of Saddle Lake, Alberta, 10 Sharing Circles were used to collect data in studies of young people and impaired driving. These groups served as the cultural platform for examining local drug and alcohol misuse practices, community norms and systemic influences underlying impaired driving.131427 The Sharing Circles were designed to capture these life issues in a way that supported traditional norms and values. The approach invited meaningful partnerships between the researchers and communities, balance amongst community participants, and offered opportunity for developing vision for change, a feature usually missed in First Nations research projects.

Participants

The studies described here involved four Northern Alberta Cree communities with four to nine participants per community. These communities were chosen because of their geographical proximity and their shared cultural background of Woodland Cree. Participants were recruited by convenience sampling, whereby the local co-investigators recruited volunteers by word of mouth. Consistent with First Nations communications in communities, the interviewers asked personal, professional and/or collegial acquaintances with vested interests in the social/health issues in the communities to nominate individuals aged 18–29. Thirty young men and women volunteered. After participants granted consent, the Sharing Circles were audio-recorded and transcribed. Participants were reimbursed for their time and expenses, and gifts such as tobacco pouches were provided as per the Woodland Cree First Nations peoples traditions.

Interview questions

Consistent with unstructured interview techniques, the questions were open-ended, with a series of questions, probes and tags, which provided ample opportunity for participants to describe any facet of their responses. In the first of two sets of Sharing Circles, five questions were asked: (1) Who are you and what is your relationship to drinking and driving? (2) How does your experience connect to what’s going on with your friends or others in your age group? (3) Tell us about drinking and driving as it relates to your family. (4) How does your experience differ from or relate to what happens in your community? (5) What do you think could make a difference in your community regarding drinking and driving? (eg, recommendations). The second set of Sharing Circles was part of the data analysis (described below).

The interviewer and the interviewed

In Sharing Circles, all participants are considered equal, which encourages support and suspends criticism. Those involved share, support and respect each other in an emotionally sensitive manner. Likewise, when Sharing Circles are adapted for research, they must be designed to ensure that there is no clear distinction between the interviewer and participants. Explicit in the Sharing Circle focus group research design is that the First Nations person facilitating the proceedings embodies a genuine and shared sense of history, culture, values, beliefs, action patterns and pragmatic issues embedded in participants’ lives. This leads to the “facilitator” becoming “part of the process” rather than an “objective observer”. For example, whenever there was a sense of redundancy in the interviews, the interviewer attempted to offer a counter perspective to gain an alternative to what was being said by participants. The discussions arising from these counter perspectives became part of the information gathered.

Sharing Circles: procedural tenets

In the Sharing Circle adaptation of focus group methodology, five basic culturally traditional tenets support the proceedings. Turn-taking is ensured by use of a talking stick, which is given to the next person wishing to speak. Once a participant has the symbol in hand, he or she is expected to “speak from the heart”—to express their feelings. Furthermore, participants are expected to “listen from the heart”—to listen without judgement, with open minds, always respectful of the person with the talking stick. Participants are to “speak spontaneously” and not rehearse what to say when it is their turn. Participants are to “speak leanly” meaning speaking without embellishment.

Data analysis

The key to the analysis in culturally sensitive qualitative research is to resist ethnocentric views. Akin to participatory action research, the participants in Sharing Circles interviews are involved in analysing and interpreting the data and in considering practical implications for community development and change. They become part of the research team. In this example, the data from the first set of Sharing Circles were initially reviewed by the lead researcher in consultation with the First Nations interviewers. This led to identification of themes. The Sharing Circle participants were then provided with copies of the transcripts and preliminary themes, and another set of Sharing Circles were held with the goal of verifying (or modifying) the analysis framework and the themes that arose from the data. These Sharing Circles also identified additional themes and interpreted all themes according to their own local experience and culture. This participatory technique of “verification” helps to reduce the concept of deviance versus normalcy, enhances the linkage between history and present times, and erases the social distance between insiders (indigenous people) and outsiders (researchers).

The Sharing Circle data were analysed by a process called constant comparison, in which each item is checked or compared with the rest of the data to establish analytical categories. The constant comparison strategy requires a coherent and systematic approach. Key points are: (a) data are inclusive—that is, not selected on the basis of whether they “fit” with the original intent of the research; and (b) categories are added to reflect as many nuances as possible rather than reducing the data to a few numerical codes. In the present study, sections of the data, such as discrete incidents, typically included multiple themes.

General comments on Sharing Circle methodology

In using the Sharing Circle methodology, it is crucial that the interviewers be indigenous (and well trained to conduct informative interviews), and that data analysis and interpretation closely involve indigenous persons, in order to document the social reality that is lived by members, capture heterogeneous perspectives, and attain data in a culturally sensitive way. Actively involving interviewers in the actual interviews (as described above) helps in the analytical stages in that participants are better positioned to trace how social life is organised and realised and why particular types of behaviour occur. This, in turn, is crucially important information for the qualitative researcher in integrating this information and understanding/interpreting the overall findings.

Examples of important findings from the Sharing Circles

Given that impaired driving is a major cause of injuries in First Nations communities, understanding the context of drinking and driving in these communities is crucially important in prevention. The social context that embedded impaired driving in these four study communities was a potent mixture of poverty, high unemployment, illiteracy, hopelessness, alcohol/drug misuse and violence. There was a strong pattern of “generational compression”—that is, where older people (50 years of age and older) interact with, party with and take risks with teenagers as if they were of a similar age.13 Older people in these communities had lost their historical status and traditional leadership role, no longer serving as positive role models. An effect of this role redefinition was intergenerational participation in high-risk, deviant acts such as drug and alcohol misuse, sexual abuse and impaired driving:

I did a lot of driving for my Grandfather so when I would drive for him he would always buy liquor for me so that when we get back we could drink together cause everybody always liked partying at his house so I would go and get drunk there and we would all have a good time.12

Although the research team initially focused on community norms, alcohol misuse and impaired driving, community members offered the point of view that the misuse of illicit drugs and driving is the “real” issue in today’s First Nations communities. They described how drug misuse has also become an intergenerational activity where young people are introduced and socialised into the use of drugs and alcohol by parents and grandparents. Older family members professed that they used drugs with their children; their children have now become users. Witness a man who helped construct a world of drug misuse for his children:

I’ve smoked dope for years. I’ve smoked pot but I’ve never done crack. I ate mushrooms which I don’t think I’ll ever do again and I’ve tried acid when I was younger…. I smoke dope and have beer. It all started with me, and my kids are all into the drugs now, but they’re all adults now. They’re all into crack.

There were also detailed descriptions of the police. The police always seemed to be in people’s consciousness. Although questions were about people’s relationship to impaired driving, participants spoke of the police as “sign posts” for conceptualising impaired driving. The police were described as having chased, investigated, harassed, stereotyped, charged or beaten them. Hence, there was a tacit understanding that community “insider” information could not be relayed to the police. The use of back roads and trails was an example of such community information. These roads were typically seen as a resource for drinking and driving, and thus for denying the police access to community information. The local Northern Alberta landscape, comprising primary, secondary, private (oil industry well-site roads) and emergency access roads (eg, forest fire) dotted by rough trails, becomes a link between individual drinking and driving behaviour and a community sense of practical living. Selective access information became a common preference rule or cornerstone for other insider information.24 Police were not told about drivers who had crashed and injured pedestrians. It is part of “the code”, best explained by a participant:

It’s just that if you go blab, you’re a rat. In all communities like this you get beat up for that. It’s mostly for protection - protecting the driver as much as you’re protecting yourself so. Take precaution if you can.

A significant piece of knowledge that engenders a sense of community is the shared attitude toward drug dealers. First Nations people may not support the use of crack cocaine because of its destructive impact on lives. Disapproval of selling drugs emerged as a common value. However, if the dealer is a member of the extended family and the participant profits from family members’ dealings, a counter “code of action” was quickly expressed that invoked a “code of silence” and supported dealing. For example:

It’s kind of hard for me to say get rid of crack because, that’s how my cousin makes his money, so I don’t want to be a rat… He wrecks a lot of lives. Well it’s not so much him but it’s the people that go buy it too. If I say, ok, he’s selling crack you know, I can stop it but I don’t because I have that love for my cousin and I wouldn’t want him to get in trouble. I’m basically being a hypocrite because I don’t want to let him get busted but yet I don’t like people doing crack. So that’s being a hypocrit so it’s hard for me to say anything about it. How can I say for him to stop selling crack when he gives me money… I’m spending the drug money too so. But I hate crack. I don’t smoke it. I hate what it does to people. I’ve seen people tweaking out from it like for them twitching and wanting another hoot. That 10 minute hoot!

These and other findings were presented to community chiefs and their respective councils, First Nations health centres, schools and police, as well as at First Nations conferences in Canada. Currently, these findings are informing the development of a proposed local addiction facility and the implementation of a Health Canada-supported heath promotion programme. The Alberta provincial transportation department has used the findings and incorporated them into a “content” document for their Traffic Safety Initiative. Furthermore, there is likely to be further research on the concepts of community coping and resilience as it relates to impaired driving and injuries.

Conclusions

Qualitative research has an important role in understanding injuries. Like many other events and situations, injuries take on the meaning people attribute to them, and this meaning needs to be considered in planning interventions and anticipating consequences of those interventions. Rather than being an isolated, individual phenomenon, injury relates to all sociocultural aspects of people’s lives. Injury research is especially important in indigenous populations because of their high injury rates. Accurate descriptions of sociocultural features of the phenomenon are required for successful interventions. This paper discusses different modalities of qualitative research interviewing, and introduces the reader to a new, culturally sensitive method of designing and conducting focus groups, the Sharing Circle. This methodology encompasses a culturally meaningful collaboration between the researcher and the researched, and could be adapted to study other indigenous cultures.