Kerbside Recruiting: Issues of rigour in using an innovative technique for recruiting qualitative research participants

Jan Browne

School of Health
University of New England
Armidale New South Wales
Contact: 03 9731 0775
Email: mailme@primus.com.au

 

Introduction

The main aim of this paper is to stimulate creative thinking and questioning about the possibilities for using innovative techniques for sample recruitment, when conducting research into highly sensitive issues and invisible or difficult to identify and access populations. It explores the value of the kerbside recruiting technique as a process through which a more diverse qualitative sample can be obtained, thus enhancing rigour through the potential for making multiple contrasts and comparisons.

Study title

Sexual Health Discourse: Interactions and negotiations between practitioners and patients.

Doctoral thesis submitted to the University of New England, May 1999.

Purpose of the research

To examine the social conditions of clinical consultations about sexually transmissible infections (STIs), pursue the meaning of discourses in sexual health, and explore the role of the clinical interview in the prevention of STIs.

The general question of the research was: "What does it take to do good sexual health care that results in STI prevention?"

Overview of the literature related to STI care

It is widely believed that health care practitioners play a vital role in the prevention of STIs, including HIV/AIDS. Health care workers are accepted as legitimate and trusted sources of sexual health information . The clinician/patient encounter may provide a fertile resource for safe sex promotion and education, counselling on sexual issues, and for achieving preventive sexual behaviour in the community. However, research specifically examining the experiences & perspectives of participants in settings related to the screening & treatment of STIs, or examining how the work of medicine is carried out at the coal face of sexual health, is rare. We have minimal knowledge about how discussion of sexual matters is managed in the clinic; what constitutes an adequate sexual history for STI prevention; and how an emergent sexual health history is used to guide a management plan that is individually tailored to the patient’s sexual experiences.

In addition, we have not explored how the changing face of medicine is impacting on the interaction about STIs between clinicians and their patients. Such changes include: new medical curriculums being developed to redirect medicine towards a preventive, community oriented model for practice ; increasing disclosure of sexuality by gay clinicians ; increased numbers of women practitioners ; altered funding schemes and introduction of sexual health centres ; changes in perceptions of health from a ‘sick role’ to individual responsibility ; and the adoption of multi-disciplinary team approaches in sexual health care . New insights into human sexuality and sexual behaviour have been generated throughout the HIV/AIDS era , yet a thorough examination of the applicability and usefulness of that knowledge for directing approaches to STI management has not been conducted. In addition, we have insufficiently examined how stigma is managed within the clinic & how disease entity impacts on interactions about STIs, although evidence suggests that health professionals retain prejudicial attitudes towards people with sexually related conditions .

Research Process

A qualitative research program, using the in-depth interview method of data collection and a rigourous grounded theory analysis was adopted. Informants were recruited from Sydney and Melbourne, Australia. A multi-stage data collection design, which combined selective, snowball and opportunistic recruiting processes, was used to select people whose experiences contrasted along the theoretical sampling categories identified as important to the research question. The overall aim of sampling was to obtain a sample of practitioners and patients that provided a wide diversity of demographic profiles, experiences in sexual health consultations, and type of health care setting. A sample of 63 people were interviewed. The study group included 34 people who have been tested/treated for STIs and 29 clinicians who have provided care for patients attending for STI screening or treatment in the previous 12 months. This paper focuses on the recruiting processes used to obtain the patient sample group.

The sample included practitioners (sexual health experts, general practitioners, a HIV counsellor, sexual health nurses) who work in diverse types of practice and health care facility, (general practice, community health, sexual health, general practice with HIV and/or gay men’s specialty, community health with lesbian women’s focus). The practitioner sample were of varied: ages and years of experience; country of origin; sexuality; with various practice experiences, including providing care to a predominantly middle class client base, providing medical care to fringe populations, HIV medicine, and women’s health; and STI caseload burden from 2% to 100% of patients with STI related attendance. The STIs included in the study were: herpes, chlamydia, genital warts, positive pap smear results, and negative test results.

The patient sample provided contrasts along the lines of age, sex, education, income, self description of sexuality, STI diagnosed, time since diagnosis, number of times attended for screening/treatment, type of facility/ies attended, age/sex/sexuality/previous attendance of clinician/s attended, IV drug use, sex work. Due to limitations of the study, it was not possible to include contrasts in terms of race, ethnicity, people with a disability, or younger people (<20).

Sampling process background

One of the primary processes by which we achieve rigour in qualitative research is through the process of theoretical sampling. Theoretical sampling is a reflective model of data collection, analysis and theory building, and is one of the central processes in implementing grounded theory method . As Patton (1990) notes, the qualitative research design in effect unfolds and emerges as the study occurs. This sampling method does not seek to obtain a ‘statistically representative’ sample. Rather, the aim is to procure the full range of possibilities and variations in experience identified as relevant to the study in terms of preliminary assumptions and provisional findings . Ongoing data analysis drives the selection of further participants. Cases are purposefully selected in a search for contrasts that clarify the analysis, help to identify, broaden and shed further light on emergent concepts or themes, and assist the researcher to extend, modify, develop and verify theory . The researcher is involved in searching for people who will provide ‘negative cases’ (people who provide contrasts in experiences to other informants). This is a systematic process that increases the rigour of the study and enables the researcher to compare and test the propositions being posed as the data evolves .

Another central way in which rigour is achieved in qualitative research is within data analysis, through the process of making multiple comparisons . Comparisons are required between informants of different demographics, but it is also necessary to make multiple comparisons along as many experiential lines as we possibly can in order to achieve theoretical density. Therefore, we need to look at what we know about our data, and to identify the full range of possibilities which are theoretically relevant to the evolving data.

In the sexual health discourses study, it was important to be able to look at both professional and patient stories about the STI clinical interview and to be able to compare those experiences. The two informant groups were:

1. Professionals who provide care for people attending for screening or treatment of STIs. (Sampling strategies for this group are not discussed in this paper)

2. People who have attended for screening or treatment of STIs.

Within the patient group, it was essential to obtain as much diversity of experience about STI testing as was feasible given the time and resource constraints of the study. In order to obtain theoretical density, the study sought for people who had (for example)

In qualitative research, there is a tendency to rely on the snowball method of sampling for obtaining informants. Snowball sampling is particularly useful when the research explores sensitive matters that require knowledge of ‘insiders’ to locate people. Thus the process is best used in situations where the phenomenon being studied itself leads to open and available social networks. It is an approach that comprises making contact with an initial group of interviewees known to the researcher to meet the criteria for the research. These informants are asked to connect the researcher with other potential informants in their networks, who are then asked to provide further connections to people who meet the criteria for the research. In this manner, the researcher casts a net into an ever-widening chain of people .

However, people who have been tested for STIs are a study sample who are especially 'invisible'. Achieve some of the categories of experience which were necessary to achieving theoretical complexity was difficult. Accessing this group required the adoption of special techniques. In order to enhance participation and honesty, the sample was not cross matched and patients were not recruited from medical clinics or facilities. It was believed that obtaining a clinic based sample might decrease the likelihood of agreement to participate, increase the problems of anonymity and confidentiality, and decrease the truthfulness of responses. The sample would also have been strongly biased in terms of reporting on the activities in specific places rather than allowing comparisons across varied types of facilities. A number of alternative recruiting methods were adopted.

Sampling methods used for recruiting a patient sample

Given the marked difference in obtaining access to a professional compared to a patient sample, a flexible approach to sampling recruitment procedures was needed in selecting people who had been tested/treated for an STI. Although searching for the patient sample among the general population rather than through clinics and practitioners was a difficult process, it reaped rewards in terms of the richness and diversity of experiences found. The researcher aimed to establish a rich data base that shed light on the many variations of patients’ experiences in consultations about STIs. When the study was originally proposed, a complex combination of methods for recruiting participants was devised, and it was thought that this group of methods would be appropriate to secure a reasonable diversity of experiences related to STI care. Both direct contact methods and indirect methods were used.

The methods adopted included:

Indirect Contact:
Forms of advertising;

Direct contact

table 1 Patient sample: Success of recruiting strategies

Strategy Advertising         Direct contact   Total
  Brochure Self Help Group Local Papers Internet Flier Kerbside Recruiting Snowball  
Recruited 1 6 3 0 0 18 6 34

Table 1 shows the variation in success of recruiting strategies. Despite the use of these multiple strategies, some of which required a great deal of effort but were entirely unsuccessful, it was difficult to obtain a sample that provided the sampling contrasts required for the study. It became evident that people who have been tested for STIs are an especially ‘invisible’ and difficult to access population. Problems included, for instance, that the informants obtained frequently do not know anyone else that has been tested. Alternatively, the one person that they do know who has been tested is often their x-partner, and relationships with that person may be especially strained. They were usually not keen to phone and ask previous partners to participate in a research study. Therefore, snowball techniques, although usually very useful in recruiting qualitative samples, were not the most useful strategy for this study. Although a reasonable size sample had been obtained using the other processes, there was a need for testing out the propositions which were being developed in the study to see whether the theory that was being developed held across more diverse cases. Therefore, in order to attain a higher level of rigour, it was necessary to obtain further informants, and for those informants to provide certain types of diversity in experience. For instance, people who use IV drugs, people with very low/high incomes, people who have only positive experiences of STI consultations (Browne 1999).

Having adopted the processes of kerbside recruiting in a previous study which examined the safe sex interactions of male sex workers and their clients , this strategy was adopted for the current study. (Note: Ethics committee approval had been obtained for the recruitment of participants in public places.)

What is kerbside recruiting?

Kerbside Recruiting is a technique for recruiting informants into a qualitative research study in which the researcher directly approaches people on the street and invites them to participate in the research project.

Approach adopted

Kerb-side recruiting was by far the most successful recruiting process used. The process involved the researcher and a friend (who is a clinical psychologist) spending two evenings approaching people on the street in a popular ‘café society’ area in Melbourne, and inviting them to participate in the study. In this setting, people were told the interview would be by telephone, completely confidential and anonymous, and fairly brief. Once a person agreed to speak with the researcher, the topic was briefly described. The person was then asked whether or not they would fit the criteria for the study (they have been tested for an STI, and could participate whether the test result was positive or negative), and be willing/able to assist with a brief, confidential and anonymous phone interview. Willing participants provided their phone number and agreed on a time that was suitable for the researcher to call them for the interview.

Over two evenings spaced one month apart, 26 people agreed to be interviewed. Participants were subsequently phoned at the agreed time. Efforts were made to space the interview appointments out so that there would be time for transcribing and analysis between each interview, although this was not always possible. The researchers were pleased (and a little surprised) at the open, friendly and generous manner in which the public spoke to them and the willingness with which people agreed to participate. Of the 26 who agreed to an interview, a total of 18 interviews were completed. Eight were not interviewed due to various difficulties with follow-up and contacting. The major problem faced in follow-up was that people were not home at designated times, and after several repeated phone calls leaving messages on answering machines or with house-mates, the researcher left it up to the informant to make the contact as it was felt that being too ‘pushy’ would annoy people. Some informants did, however, return the calls and contacted the researcher.

In selecting passers-by to invite to participate in the study, the researchers: approached only people walking on the street, not those inside of premises; avoided ‘couples’ and people ‘deep in conversation’; did not persist when people gave negative signals, - politely accepted first refusals and walked away. The selection process was to attempt to make contact with every person or group walking in the street without bias. Once the researcher had finished talking with one person and said good-bye, they then approached the next person/group walking along. Using this approach it was felt that to some extent personal preferences in the type of person invited to contribute were minimised.

As can be seen in Table 1, using kerbside recruiting, it was possible to obtain a good size, diverse sample in a very short period of time. Most importantly, the richness of the data increased in terms of:

Issues, advantages and problems

Safety issues must be taken into consideration, both for researchers and passers-by. In some instances, identification badges may be a good idea to provide evidence of the legitimacy of the researcher’s presence on the street. We would not recommend researchers adopt this approach alone, as there are some risks in terms of being alone on the street at night. Adopting a team approach provided not only visual authenticity to the project, but also a sense of greater security. There was one situation in which a woman appeared to feel threatened by the approach of a stranger (even though female), which required an immediate ‘back down’ and reassurance in order to alleviate her concerns. An up-front, extroverted approach is essential to recruiting people on the street, therefore the method is not suitable to shy or nervous researchers.

The kerbside recruiting method is not suitable for all studies, but is very useful for highly sensitive and invisible topics - when the sample cannot be identified in other ways, or samples obtained would be biased in some way, such as those only readily available through singular venues for instance, clinics.

Careful attention was paid to the selection of an appropriate assistant in order to ensure professionalism, courtesy and efficiency. It was necessary to select a person who was not only familiar with the study and research methodology, but also experienced in recruiting research participants and managing any difficulties that might arise given the unique situation of the recruiting process.

One advantage of the process was the collection of some interesting, fortuitous data. For instance, by using this recruiting method and incidentally conversing with people on the street about STI testing, the researcher discovered a number of people who were unable to participate because they are not having, and may not have had, sexual relationships. Such comments suggested that popular images of a highly sexually active population may need further research as to how accurately such images reflect the experiences of the community. The researchers were also surprised at the number of people who said that they ‘should’ get tested because they have been at risk of contracting STIs, but despite those risks, have never been tested. Such comments indicated the importance of research studies to explore the value of widespread public education about the need for STI testing. It also indicates the need for informing the public about where testing is available anonymously and confidentially. Additionally, the importance of research that examines the processes involved in making the decision to have STI screening, as well as the issues which people who have been at risk, but not tested, face in their ongoing sexual relationships, was highlighted.

Summary

Kerbside recruiting is an effective strategy for obtaining a sample of informants for qualitative research, and is especially suited for finding ‘invisible’ research populations. Using this approach requires not only an upfront attitude, but also careful attention to safety issues, adopting approaches that show respect for the public, and their needs for confidentiality and anonymity. The process facilitated greater theoretical complexity and verification of propositions developed in the study through a sample that was both rapidly collected and provided significant experiential contrasts and comparisons. The study indicated that personal, face to face contact remains the most powerful and efficient method for recruiting research participants.

Acknowledgements

The assistance of the following are acknowledged. I wish to express my sincere gratitude to the informants, who readily gave their time and generously shared some of their most intimate and personal stories. Sallee McLaren of La Trobe University, who assisted with the recruiting process, and my mentor, Associate Professor Victor Minichiello.

This research was support by the National Health and Medical Research Council (NHMRC).

References

Bermingham, S. and Kippax, S. (1998), HIV-related discrimination: a survey of New South Wales general practitioners, Australian and New Zealand Journal of Public Health, 22, 1, 92-97.

Bolton, R. (1995), Rethinking anthropology: The study of AIDS, in Brummelhuis, H.T. and Herdt, G. (eds.), Culture and Sexual Risk: Anthropological Perspectives on AIDS, Australia, Gordon and Breach.

Browne, J. and Minichiello, V. (1995), The social meanings behind male sex work: Implications for sexual interactions, The British Journal of Sociology, 46, 4, 598-622.

Browne, J. and Minichiello, V. (1996), The social and work context of commercial sex between men: A research note, Australian and New Zealand Journal of Sociology, 32, 1, 85-92.

Browne, J. and Sullivan, G. (1999), Analysing in-depth interview data using grounded theory, in Minichiello, V., Sullivan, G., Greenwood, K. and Axford, A. (eds.), Handbook for Research Methods in Health Sciences (1st Edition), Sydney, Pearson Educator.

Cockerham, W.C., Rutten, A. and Abel, T. (1997), Conceptualizing contemporary health lifestyles: Moving beyond Weber, The Sociological Quarterly, 38, 2, 321-342.

Commonwealth Department of Health and Family Services (1996), General Practice in Australia: 1996, Canberra, Author.

Dawson, J., Fitzpatrick, R., McLean, J., Hart, G. and Boulton, M. (1991), Gay men's views and experiences of the HIV test, in Aggleton, P., Davies, H. and Hart, P. (eds.), AIDS: Responses, Interventions and Care, London, Falmer Press.

Featherston, W.E. (1990), Sexual identity and practices relating to the spread of sexually transmitted diseases, Primary Care, 17, 1, Mar, 29-45.

Herdman, E. and Kippax, S. (1995), Institutional Discrimination: Critical Ethnography of HIV/AIDS Related Discrimination in a Hospital Setting, Sydney, National Centre in HIV Social Research.

Jones, G., Bligh, J. and Valentine, C. (1996), Venus and Freud: An educational opportunity?, Genitourinary Medicine, 72, 4, Aug, 290-294.

Labonte, R. (1996), Health promotion: Ten years on from the Ottawa Charter, Australian Journal of Primary Health-Interchange, 2, 4, 7-24.

Leenaars, P.E.M., Rombouts, R. and Kok, G. (1994), Service attributes and the choice for STD health services in persons seeking a medical examination for an STI, Social Science and Medicine, 38, 2, 363-371.

Mindel, A. and Tenant-Flowers, M. (1998), Services for sexual health: Where should they be provided, Medical Journal of Australia, 168, 8, April, 373-374.

Minichiello, V., Sullivan, G., Greenwood, K. and Axford, R. (1999), Handbook for Research Methods in Health Sciences, Sydney, Pearson Educator.

Mulhall, B., Anderson, B., Venables, S. and Donovan, B. (1995), Venereology as a specialty in Australia, Annals of Academic Medicine, Singapore, 24, 4, July, 644-647.

Plummer, D. (1992), The Medical Establishment, in Timewell, E., Minichiello, V. and Plummer, D. (eds.), AIDS in Australia, New York, Prentice Hall.

Rosenthal, D. and Reichler, H. (1994), Young heterosexuals, HIV/AIDS and STDs, Canberra, Australian Government Publishing Service.

Steinberg, J.K., Wellman, J. and Melrod, J. (1991), A proposal for strengthening medical school training in STD prevention techniques, Public Health Representative, 106, 2, Mar-Apr, 196-202.

Strauss, A. and Corbin, J. (1994), Grounded theory methodology: An overview, in Denzin, N.K. and Lincoln, Y.S. (eds.), Handbook of Qualitative Research, Thousand Oaks, Sage.