Homelessness is a major problem in many global cities, and research on the issue recognizes that urban aspect. But homelessness is not just a problem for cities, as criminologist Michael Young describes in his case study for SAGE Research Methods Cases titled, “Homelessness, Addictions and Mental Health in a Northern, Rural Context.”
He was approached by the non-profit Inuvik Interagency Committee in 2010 to study the growing homeless problem in a remote city in Canada’s North West Territories, Inuvik, a city of roughly 3,600 people, is located about 125 miles north of the Arctic Circle on the vast delta of the Mackenzie River as it empties into the Arctic Ocean. Young was tasked with helping to identify gaps in services to that remote city’s homeless population and what role, if any, such gaps played in people being homeless. Using both qualitative and quantitative approaches – surveys and focus groups – Young learned how to familiarize himself with the community, and vice versa, as a first step toward getting a solid understanding of what was happening in Inuvik.
In this interview, Young describes his experiences conducting this research, details some of the nuts and bolts of his mixed methods approach, and explains how geography mediates culture in ways that wise researchers need to understand.
Young has been a professor in the School of Humanitarian Studies at Royal Roads University in Victoria, British Columbia since 2008. Before joining the faculty at Royal Roads, he taught at a number of institutions in British Columbia, including Camosun College, University of Victoria, his alma mater Simon Fraser University, Vancouver Island University, and Kwantlen Polytechnic University.
What drew you to study homelessness? And what in particular led you to Inuvik, which at 200 kilometers north of Arctic Circle is a little off the beaten path?
My interest in homelessness grew out of my studies in criminology. Those who study crime and criminal behaviour have long been exposed to the issues of socioeconomic class and participation in the justice system. The lack of housing is frequently noted as a cause of crime, but so is the relationship between release from prison and the lack of housing options. Both are matters of social justice, which caught my attention early on in my academic career.
I met a colleague from Inuvik at a meeting of criminologists in the late 1990s. Apparently, my approach to understanding and researching community reintegration caught the attention of at least one person, who later contacted me about preparing a proposal for the research in Inuvik. Homeless persons are frequently caught in a cycle of obtaining and losing housing because they cannot reintegrate well or have not even integrated in the first place.
What problems did you encounter in getting data? And what innovative ways did you use to address that? Was there an ethical component?
Starting with the last part of the question, there are ethical components to research. In this case, getting the data required approval of the research from my home university and from the Aurora Research Institute, the body responsible for research in the Northwest Territories.
Besides the bureaucratic aspect, on the ground I was confronted with two major ethical issues. The first has to do with participant expectations. There are often assumptions that researchers will advocate for individual participants, in this case to help them obtain housing. Clear guidelines about the objectives of the research and expectations were discussed with participants to help clarify this aspect of the research. The other issue was a moral issue. What personally would I do for participants? For example, after starting up a relationship with someone as a researcher, I was often asked for money. This was a difficult position to be in. One the one hand, giving someone a few dollars was not going to break me. On the other hand, word does get around when one starts handing out money. It would be incorrect to say I had a zero tolerance policy on this, but I tried.
What did you want your final dataset to look like? What methods proved most useful?
The final qualitative data set was a mix of digital interviews and focus group conversations that were transcribed into a written document. The data were coded using pseudonyms to identify participants. Given that there was not too much data to read, I used categories developed from the data to lump similar ideas together, look for counter examples, and identify trends. Lofland and Lofland’s work on analyzing social settings was instructive in this process.
The survey data, which may or may have not been that revealing, was easier to analyze because the instrument (QoLHHI) has some instructions and the researcher can use previous studies as a guide. The QoLHHI data were coded in Excel and analyzed with SPSS according to the scales provided by the developers of the instrument.
Of course, as researchers we use relevant literature relating to content analysis for the qualitative analysis in our research, and this research was no exception. However, it was important for the research that the voices of participants were clear and the use of direct quotes that best illustrated the emerging themes was our strategy for analysis.
How did the location of your research affect what you were able to do? What was unique to this study and locale, and what broader takeaways would be applicable to mixed methods researchers anywhere?
Starting with the last part of the question, the best takeaway for me was that regardless of the population of interest it is essential to get to know the people and place first before engaging in the “real” research. Social context is foundational to community-based research design and practice. Values, customs, traditions, history and such are not to be taken for granted so the familiarization process needs to be done well.
One of the more unique features of doing research in the Arctic, and probably most rural locations, is that geography impacts social conditions in more ways than physical coping. It defines social roles and values more than we think, much like indigeneity. A person’s connection with the community is mediated through their connection with the broader surroundings. This seems a bit esoteric, but it is something you experience when dealing with people in their own social realities. So yes, the location had an effect, but the effect was positive in that it opened up a different way of reaching out to and interacting with participants and others living in the community.
Once our presence was known and we were accepted, it was much easier to connect with who would later become participants and with others, like community elders, who could provide great insights into community dynamics, history, and a myriad of other aspects we had not considered in the period leading up to and including the research.
What advice would you offer anyone conducting similar research? What might you do differently?
First, as I note in the article, do the legwork as much up front as possible and be ready for change. In this case, the seasonality of homelessness affected the presence of people in the community, or their lack thereof. The more you know others in the community the more likely you will be able to locate your participants through informal networks. Persevere in your pursuit of accessing people and getting information, but be patient, considerate, and kind. Sometimes you have to listen to a lot of stories before your questions are answered, but there are usually nuggets of information in the stories that buttress your research. I say these things as much to myself as to others because community-based research is difficult work. The benefits, however, are obvious if you can get your results into an actionable format for dissemination.
As for doing things differently, given the chance I would spend more time in the community trying to get to know my potential participants beforehand. I would also attempt to use them more in the research process, although I would need to try a few different strategies to determine which is best. Could they conduct focus groups independently? Could they attend a conference to talk about results? At present I am not sure because most participants were not functioning well, but with time and the opportunity they might have risen to the challenge.
Referring to methods, I might consider alternatives such as adding a photo voice component so that participants could respond to visual, compared to written, stimuli. My thinking is that the results may be more revealing about certain aspects of homelessness concurrent disorders, but I would pretest this approach before developing a project.
Finally, I did think about running focus groups comprised of homeless persons, service providers, and service agency representatives. This could provide rich data that could be used to inform policy, although I assume it would be a difficult thing to do as it would be uncomfortable, particularly for homeless persons, and it could be potentially emotional and political for everyone involved.
The best piece of advice I can give, aside from what I have written above and in the case itself, is to demonstrate gratitude to research participants and the community in which the research takes place. Providing at least one venue for dissemination of the results to the community is an important consideration.