Report on Family Law Research in Nunavut

APPENDIX ONE: Methodology and Recommendations for Future Researchers

This research project employed a variety of methodologies. Results varied for the four components. A more detailed discussion of these sources follows, along with some discussion of pitfalls and possible improvements for future researchers, particularly those planning to do quantitative work in the North.

1.   Review of Existing Statistics: Sources

Our review included 1996 Census data on families and households, the NWT Drug and Alcohol Survey (1998) and the 1999 Nunavut Labour Force Survey. We also relied on earlier compilations of data, including Justice Canada's Report on Selected Crime Statistics 2000 (which provides a community-level view of charging and clearance data from the Royal Canadian Mounted Police); the Government of Nunavut's Report on Housing in Nunavut 2000; and other data compiled by Nunavut Statistics and the NWT Bureau of Statistics prior to the division of the northern territory into two jurisdictions.

We were unable to take advantage of several important national studies that touch most directly on family law issues. In particular, the General Social Survey, including Cycle 10 on Families and Friends (1995) and the National Longitudinal Study on Children and Youth (a biennial survey since 1994) do not cover Canada's territories. High cost and other logistical difficulties involved in conducting quantitative research in the North have meant that there is a significant shortage of data concerning family life in the territories. For this reason, the Government of Nunavut determined that a quantitative study focussing on details of family relationships and family breakdown would be a useful planning tool and could contribute to bridging the gap between what is known in southern Canada and what is known in the North.

2.   Household Survey

The Sample

The survey was administered through in-person interviews with 342 individuals in five Nunavut communities.

Based on the GSS methodology, only one individual, older than 15 years of age, was interviewed in each household. In each household, the interview subject was chosen on the basis of whose birthday was next in the year relative to the interview date. Overall, there were 193 female respondents and 149 male respondents; the eldest respondent was 82 years of age.

The five communities were chosen to be as representative as possible of geographical differences in Nunavut. Each of Nunavut's regions was represented. The survey was conducted in two communities on Baffin Island, two communities in the Kivalliq (Keewatin) Region and one community in the Kitikmeot. The communities ranged from Nunavut's largest community, Iqaluit (the capital, pop. 4,627),[129] to one of the smaller communities, Chesterfield Inlet (pop. 363). As most of Nunavut's population live in communities of between 1,000 and 1,500 people, we chose two communities in two different regions with populations in that range: Pond Inlet (pop. 1,276), in the Baffin Region and Cambridge Bay (pop. 1,387) in the Kitikmeot Region. Finally, Coral Harbour in the Kivalliq Region with a population of 822, was chosen to represent the small-to-mid-size communities in the territory.

The sample size in each community was roughly proportional to the percentage of Nunavummiut living in a community of comparable size. The initial intention was to have 500 completed surveys. Out of the total population, we estimated that approximately 20 percent of respondents should be from Iqaluit, because approximately 5,000 people live there. Twenty percent of the respondents should be from Cambridge Bay, as approximately 5,000 people live in the two regional centres. Just under 10,000 Nunavummiut live in the eight large communities that are not regional centres, each with populations between 1,000 and 1,500 residents. Consequently, approximately 30 percent of respondents should be from Pond Inlet. The seven mid-sized communities with populations between 500 and 1,000 make up another 20 percent of the territorial population, so twenty percent of our sample was taken in Coral Harbour. Finally, small communities with populations less than 500 people make up about 10 percent of the population of Nunavut, and Chesterfield Inlet constituted ten percent of the intended sample.

For a number of reasons, the final breakdown of responses did not precisely match the original structure of the survey. In a few communities, particularly Cambridge Bay, we had considerably lower response rates than expected. That Cambridge Bay was one of the two last communities visited may explain the lower response rates. However, it may also reflect problems with recruiting surveyors. In Iqaluit, data collection was terminated after a serious incident when a surveyor was locked in a house and threatened by an intended respondent. We were delighted that the surveyors in Pond Inlet and in Chesterfield Inlet fully met the target numbers of surveys.[130] For the purposes of this overview report, data have not been broken down by community. The results, while not sufficiently broad-based to be representative of Nunavut as a whole, still provide a meaningful sample on many important questions relating to family life.

Survey Instrument

The survey was based, to a large degree, on the General Social Survey (the GSS), Cycle 10 "Family and Friends" conducted by Statistics Canada. The 1995 GSS surveyed a wide variety of family issues, including household composition, marriages, common-law relationships, family history and changing family roles. It examined attitudes and opinions on a range of questions relevant to family life, and sought detailed financial information on family members. Certain adaptations were required in order to administer a survey similar to the GSS in the North. We chose not to pursue some of the lines of research relating to attitudes, perceptions and finances of the respondents. We also added more questions than were directly relevant to family law issues per se. We asked questions about contact between children and their parents, financial and material support, and public legal education. As much as possible, we attempted to identify areas in which it was important to incorporate culturally different norms. For example, questions about adoption and guardianship were refined to differentiate between court-ordered adoption, custom adoption and informal guardianship arrangements.

The final significant change to the survey was to adapt it from a Computer-Assisted Telephone Interview instrument to a paper questionnaire in order to facilitate in-person interviews.

The survey instrument was translated into Inuktitut, but not into Inuinaqtun, the dialect of western Nunavut. Surveyors were provided with a copy of the Inuktitut translation for reference.

The Interviews and Surveyors

Bilingual Inuit surveyors were selected from within the communities where we carried out the survey. In most cases, we relied on recommendations from the Hamlet Office as to the individuals to hire. While this may have had some disadvantages in terms of people having to share private information with familiar people, those disadvantages were, in our view, outweighed by the combined advantages of significantly reduced cost and increased confidence on the part of interviewees. We were extremely pleased that the surveyors outside of Iqaluit reported no difficulties in convincing respondents to participate in the survey. In our view, the use of local surveyors facilitated the exchange of information and the perception that this survey would be useful for Nunavummiut.

Survey designers travelled to each community to conduct a one-day workshop with the surveyors to familiarize them with the survey instrument. Surveyors were then asked to conduct interviews on a random basis. Depending on the community, we achieved this random basis in a variety of ways. In Iqaluit, we simply used a randomized dwellings list provided by the Research and Statistics Division of the Government of Nunavut. In the smaller communities, where the sample represented a larger proportion of the households, we used a more approximate form of randomization. For example, in Pond Inlet, 150 surveys were to be conducted in a community of 325 dwellings. Accordingly, we asked the surveyors to visit every second household in an area of the map and divided the map between three surveyors. A comparable approach was used in all the other communities: every fourth dwelling in Chesterfield Inlet, and so forth. Surveyors were paid for each completed survey.

Data analysis

A contractor entered the data from the completed surveys into an SPSS database of her own design. Considerable work was required, thereafter, to evaluate the aggregate data and to manipulate them in order for them to be responsive to our key issue concerns.

3.   Service Inventory

The service inventory was conducted by telephone during a short period of time before the household survey was undertaken. Researchers developed the questionnaires based on their knowledge of services in a number of different communities. The questionnaires were then given to municipal officials, either the Senior Administrative Officer (the SAO) or the Assistant SAO (the ASAO), and in some cases, both. Because in most cases, the SAO was a qallunaat (non-Inuit), and the ASAO an Inuk, we were interested in knowing whether there were significant differences between knowledge of services and perception of issues between the two groups. This did not materialize to an appreciable degree.

We completed interviews in 17 of Nunavut's 23 communities, not including the five communities where the household survey was conducted because in those communities we were able to get a more complete picture for ourselves. We also excluded the very small communities with populations of less than 150. Because of difficulties contacting the appropriate officials, we missed the Kivalliq regional centre of Rankin Inlet.

4.   Focus Groups and Individual Interviews

In each community that we visited for the purpose of conducting the survey, we also used a range of qualitative research strategies and outreach activities.

Perhaps most important, we conducted community meetings. The meetings were held in either community centres or the Hamlet Office. Before going to the community, we contacted a number of individuals there and advertised the meetings on the local radio. Attendance varied considerably, from a large group of 40 individuals in Pond Inlet to a small group of five in Cambridge Bay, where the meeting was rescheduled because of an elder's funeral. A wide variety of individuals actually attended the meetings: some community activists, certainly, but also an array of people of different ages (from babies to great grandparents) who heard about the meetings on radio or through announcements or word-of-mouth.

Meetings were loosely structured, with a short presentation on the basic tenets of Nunavut family law (child and spousal support, custody and access decision making, and the rights of common-laws) followed by an open session for questions and comments. Meetings were conducted mostly in Inuktitut, although translation to English was available for the lawyer and anyone else who needed it.

We also conducted one-on-one interviews with interested individuals, including social workers, CJC members, lawyers, the Senior Judge, the Director of Adoptions and Child Welfare, and church ministers. In the course of our outreach activities, we also met with a sizeable number of concerned individuals, including parents and grandparents, youth, teachers and others. Finally, we benefited from feedback and information provided by MQ and the Nunavut Family Law Working Group, which, as noted before, elsewhere, is a group that includes representatives of Health and Social Services and the Department of Justice, family law practitioners, the judiciary, the public trustee and public guardian.

These meetings provided important information for planning purposes. There were wide-ranging comments that shed light on crucial aspects of the development of family law in the territory.

5.   Recommendations for Future Research

In undertaking such a large task, inevitably mistakes were made, but also successes enjoyed and lessons learned.

In retrospect, we made a serious error in choosing such a complex survey instrument as the basis of the household survey. We chose this tool with the goal of comparability between Nunavut and the rest of Canada in mind. However, we underestimated the significant difficulties involved in adapting a CATI-based instrument to a paper survey for door-to-door interviewing and in coding and analyzing the data. As well, the survey was broader than it needed to be, resulting in a certain amount of wasted effort. More narrowly defined research goals would have prevented this problem. We also made a serious mistake in not designing the database contemporaneously with the survey to allow for more straightforward analysis. If a comparable project is undertaken in the future, it would be wise to have someone with quantitative research and programming expertise involved at a much earlier stage.

A notable success, from which the project benefited, was the involvement of Inuit in adapting the survey instrument, identifying communities to visit, organizing and facilitating community meetings, and administering the survey as well as commenting on the final report. At each stage, community-level knowledge about what to ask, how to ask it and when, was essential to identifying important information and creating conditions in which people were comfortable sharing it. In future, it would be desirable, if possible, to continue this involvement in writing the final report. Also, since adapting an existing survey produced unexpected difficulties, it would be worthwhile to attempt to build the survey instrument itself, using even more of a participatory approach.

Another success-and one of the difficulties of this project-was to tie the research project directly to the goal of program and policy design. One of the frequent objections of Inuit and other northerners to large-scale research undertakings is their sense that there is nothing for them to be gained from it. Because there have been tangible results from the research (in terms of the development and training of mediators and the creation of the Family Support Office), it is hoped that those who gave their time as participants will feel that their contribution led to a concrete result for community benefit.

In light of the community distrust for research endeavours in general-and the considerable needs identified in this paper-we do not recommend that further quantitative research be undertaken in the near future. Any quantitative research that does take place should focus on the subgroup of the population that has experienced separation and divorce in order to gain more information about those areas with which family law is primarily concerned; in doing so, however, factors such as the extended family structure or the role of custom adoptions should not be neglected.

The one area clearly requiring additional research is qualitative work with elders and mature interviewees to look at traditional and contemporary approaches and principles in family law. With such work, it might be possible to report more definitely on the different norms concerning separation, decisions about children, obligations to support family members, equality between spouses, problem solving, violence and healthy families. It would be invaluable to be able to look at how these norms interact with the formal legal system.

MQ will follow this research with a consultation process to evaluate our findings and proposals for law reform. However, when that consultation is wrapped up, the real challenge-requiring most of the family law energy in the territory-will be to put these findings into action in programs and policies that suit the North.

Date modified: