« Previous | Contents | Next »
Listen
Effective Interventions Unit
Examining the injecting practices of injecting drug users in Scotland
Chapter 2: Methodology
Data collection
The primary means of data collection was direct observation using a video camera to record injecting events. The rationale behind this methodological approach was to give an insight into the injecting process that would deepen the understanding of how HCV transmission occurs in IDU environments. Thus, an exploration was sought of how IDUs prepare and administer their injections and what happens immediately afterwards.
During recorded observations researchers asked participants to identify the quantities and types of items being used and from where these items were obtained. In some instances conversation would then develop into a more detailed description of the process as it happened.
The research team sought to record events in accordance with the generally accepted principles of this methodology. This involved the recording of "whole events" using long single takes from the widest possible angle and with the least interference to the scene being recorded (DuFon, 2002, Erickson, 1992). The "whole event" was defined to respondents as the preparation and administration and post-administration stages of injecting. Conditions permitting, and with participant consent, other observational footage was also recorded. This style of ethnographic video recording encourages the researcher to use the camera as a "passive recording device" which allows events to "speak for themselves" (DuFon, 2002). However, on occasion, interjections were made by researchers to clarify certain points.
Recorded data were augmented with researchers' field notes. These provided a record of the circumstances surrounding each injecting event. Additionally, taped interviews were conducted with the majority of participants at their discretion. Interviews took place either immediately after an injecting event or during a separately arranged appointment. Most interviews were preceded by at least two recorded observations. All interviews took the form of an informal conversation loosely based around the issues raised in the study objectives.
An estimated total of 720 hours were spent in recruiting, establishing relationships with participants and collecting data. Twenty-one hours of video footage was recorded. Time spent with individual IDUs or groups of IDUs ranged from 3 hours to 40 hours. This time spent with participants did not involve the observation session alone. To develop rapport and a trusting relationship, researchers watched television, provided transport for participants, accompanied them to collect their methadone prescription and, on one occasion, the research team helped a participant to move house. Spending significant amounts of time in the company of participants reassured most that the researchers were interested in investigating injectors' own perspective on their injecting practices. It also ensured that the collected data reflected the process of injecting in the context of the IDU's everyday life.
Introducing a camera into observational research creates the possibility that those being recorded may embellish their behaviour with an element of performance. To reduce this possibility as much as possible, participants were filmed in their own environments and were repeatedly reminded that any footage filmed was not intended for public consumption. The focus was on how injectors "make up and have a hit" rather than "making a movie". Furthermore, injecting episodes were recorded with a small hand held digital camcorder (JVC GR-DVL 160 ER: 17cm x 8.5cm x 7.5cm) to maintain the element of unobtrusiveness (Carruthers, 2003).
Pre-recruitmen
t
Prior to data collection the study design included a preliminary phase in which members of the research team sought to acclimatise themselves to the lifestyles and behaviours of the city's IDUs. This involved canvassing opinion from as many relevant sources as possible. Local knowledge was identified through a variety of sources representing different perspectives. Current and former IDUs were asked for their opinions and advice. At the service provider level, professionals working in drug treatment, social services, needle exchange and outreach workers were asked for their input. Researchers also attended a drop-in clinics run by the city's homeless addiction team and worked for several weeks on a voluntary basis with a city-centre based soup kitchen. This consultation exercise helped researchers familiarise themselves with the concerns of IDUs and to identify current themes relevant to everyday lives of Glasgow IDUs.
Recruitment and sampling strategy
Recruitment and field work took place over a seventeen-week period from the first week of January 2003 to the first week of May 2003. A total of 30 respondents were recruited to the study. This sample was consistent with our initial aim to recruit between 30 and 50 current IDUs. The limited sample size allowed the establishment of trust to develop between researchers and participants and which allowed the researchers to gain access to injectors' own homes and other environments and to observe injecting practices in as natural a setting as possible.
Respondents were recruited using a targeted snowballing technique (Beirnacki & Waldorf, 1981). Snowballing provided a range of injectors at different stages of their injecting careers. Due to the limited time available in the field, participants were recruited from multiple venues to maximise recruitment opportunities. Potential participants were approached in two distinct settings: needle exchanges and in the street.
Agency based recruitment took place in four fixed site needle exchanges across the city. The most successful of these recruiting venues was a city centre based agency providing multiple services for drug users. Within this busy venue, recruitment occurred during daytime office hours. Potential participants waiting for a methadone prescription or new needles and syringes were approached in the agency's waiting room. In addition, a poster advertising basic details about the study and the research team's mobile telephone number was left on the waiting room walls. Recruiting at this agency produced 14 study participants; six initial contacts were met in the clinic, five directly and one who contacted the research team via the poster. Through these six, researchers were introduced to a further eight study participants.
The remaining agencies were needle exchanges run in local health centres during the evening in the Gorbals, Easterhouse and Parkhead areas of the city. In this context, participants were approached about the study after they had exchanged their needles and syringes. Recruitment at these agencies eventually led to a total of six study participants; three initial contacts introduced the research team to a further three respondents.
Street based recruitment involved opportunistic sampling. Individuals begging or selling the Big Issue magazine around a prominent open air, city centre car park were approached and engaged in a general conversation. This dialogue began with some general questions about the city centre drug scene. Once established, the conversation was steered towards the issue of injecting drug use. Subject to individuals identifying themselves as a current IDU, conversation would then evolve into a more detailed explanation of the research. This approach resulted in meeting seven participants; four initial contacts were made through whom the team were introduced to a further three respondents. The remaining three participants were approached in a similar manner within two separate housing schemes within the city.
Once researchers were satisfied that potential participants were current IDUs, the background and aims of the study were explained to interested individuals. If the individual was still willing to participate, an explanation of our intention to film injecting practices and conduct informal interviews was provided. At this stage they were given an information sheet further explaining the study aims and objectives and given an opportunity to ask questions about the study. They were also informed about their rights regarding confidentiality and anonymity and asked to sign a consent form. Consent to involvement in the research was presented as a product of an ongoing negotiation between the researchers and the participant. In other words, the signing of a consent form did not signal the end of respondents' right to withdraw consent; this point was made forcefully and explicitly.
Following this process of information and consent, arrangements were made to meet the participant at an appointed date, location and time. These arrangements were entirely the choice of the respondent. Respondents were then informed that they would be reimbursed 8 for each observation.
Data analysis
The process of reviewing recorded footage began at the early stages of fieldwork. Early reviewing sessions ensured that recordings were accurate and clear.
Erickson's "ethnographic microanalysis of interactions" provided a framework of analysis best suited to the exploration of how injectors inject (Erickson, 1992). This analytical method concerns itself with the minutiae of habitual or routine human behaviour. Through a process of repeated viewing of the multiple examples of the same type of event, that event can be broken down into component parts and then reassembled to explain the process to those unfamiliar with this routine behaviour. Additionally, this method also seeks to establish a range of recurring themes inherent in the process from common events to less typical ones.
All recorded events were analysed in accordance with the 5-step model that Erickson proposes:
Two researchers first watched each of the injecting events in their entirety.
From these viewings, three constituent parts of the process were identified: the start of that activity, the main focus of that activity and its conclusion.
The third step involved an investigation of how these three major sequences were organised and how they combined to form the whole event.
Each of the tapes was then transcribed onto a tape log sheet by two researchers reviewing the footage simultaneously. Tape logs consisted of a time code followed by some descriptive text. Each separate action, as well as dialogue, was noted, coded and entered into a Nudist software package. Coding was developed using the aims and objectives and from recurring themes which emerged from the footage.
Coded tape logs were searched for the frequency of certain events. Moreover, searches explored possible relationships between certain injecting practices and wider factors such as environment. These searches enabled researchers to develop a representative model of injecting.
Recorded observations were augmented by informal taped interviews and researcher fieldnotes.
Ethical approval
The Greater Glasgow Primary Care Trust Research Ethics Committee granted ethical approval for the study. Achieving ethical clearance required the committee to be satisfied that IDUs were in a competent state to provide informed consent, that anonymity and confidentiality would be ensured, and that effective safety measures were put in place for researchers' safety. In addition to ethical approval, advice was also sought form the Crown Office and Procurator Fiscals office in Glasgow prior to commencement of fieldwork. The Glasgow City Social Work Department also granted ethical permission.
Maintaining anonymity
To ensure anonymity, video recordings focused on the participant's hand movements and their chosen injection sites rather any face shots. Further, prior to any recorded observation, participants were encouraged to cover any identifiers such as tattoos, birthmarks, jewellery or scars that may have identified them and to remove any unique household items from the camera's view. Crucial to achieving both participants' initial consent and in maintaining their trust was the constant and continuing emphasis on the participants' right to view the recorded footage and withdraw consent (and any footage involving them) at any time. All of those who reviewed their footage were satisfied that the footage captured their actions as opposed to their identities. Seven participants declined to review their footage, most of these preferring not to see themselves injecting.
Safety issues
Two researchers recruited the sample and collected all the data with the support of a part-time interviewer. Researchers always worked in pairs and informed the study manager each day of their whereabouts. Each researcher was provided with a mobile phone and personal alarm. At the end of each day's fieldwork they informed the study manager of their safety.
« Previous | Contents | Next »