Social Capital in First Nations Communities: Conceptual Development and Instrument Validation Javier Mignone, Janet Longclaws, John O’Neil, Cam Mustard Centre for Aboriginal Health Research University of Manitoba We wish to acknowledge the primary support of the Canadian Population Health Initiative. Additional support from the Canadian Institutes for Health Research is also acknowledged with appreciation. Community A on the Bridging scale. For construct validation purposes results were expected to correspond with these predictions. 97% of items from the Bonding scale, 95% of items from the Bridging scale, and 84% from the Linkage scale were in the predicted rank order. These results were deemed to provide evidence of the construct validity of the scales. To examine whether empirical support could be found to justify the multi-component conceptualization of each dimension of social capital, factor analyses were run for each scale. Results justified the multi-component conceptualization of each dimension of social capital, but only to a relative extent as to what was predicted in the framework. The final two steps of the analysis sought to determine if demographic characteristics of respondents accounted for the variance in social capital mean scores and in factor mean scores combining all three communities. Sub-group differences within communities were also examined. For these goals, stepwise multiple regression analyses were conducted. The findings validated the idea that social capital scores may vary over and above sub-group differences within communities. Despite various imperfections and areas that require further assessment and development, the study presented a final version of the social capital instrument. A comparison between scales suggested that the Bonding scale performed the best in the assessment, with lower “don’t know” percentages, and better reliability and validity, followed by the Linkage scale and finally the Bridging scale. Implications of Findings The study provided a rigorous effort in developing and operationalizing the construct of social capital, and of creating a measurement tool. This was achieved, but with limitations. Study results offered both a framework and a tool that can be used in social epidemiological studies (which does not exclude its potential use in other research fields), and at the same time supplied the basis from where to proceed for further conceptual and instrument refinement. Two research implications emerge. The line of inquiry that leads to the theoretical development and empirical testing of population health determinants pathway models that incorporate ecological level factors require precise conceptual formulations of social environmental variables and the use of valid measures. The present study has taken an important step in fulfilling these requirements for First Nations health research. Thus, the first implication, that we now have an initial tool with which to advance along this line of research. In fact, the 2002 wave of the Manitoba First Nations Longitudinal Health Survey has already incorporated a significant segment of the Bonding scale in their survey. Nonetheless, construct and measurement validation are part of an ongoing process, which brings us to the second implication. The use of study findings to continue developing the construct of social capital (and maybe other constructs) and improving the tools for its measurement. Ongoing work is required via an effective participatory research process that continues to combine conceptual analysis, grounded theory development, and quantitative evidence. Policy implications of the study need to be considered from three points of view. One is the innate political nature of the concept of social capital, the other is the political utilization of the concept, and third the potential of policy to impact social capital. Inherent to the way social capital was conceptualized in our study is the notion of community as entity of empirical inquiry and policy. The idea of higher or lower levels of community social capital is not value free, given that it presupposes the good of the community as a whole as base criterion. The implication is that empirical inquiries that incorporate the construct of social capital need to make this fact explicit in interpreting their findings. The latter relates directly with the second area for consideration, the political use of the concept. The bottom line is the implication that research findings in this area must be subject to First Nations community and organizations representatives’ interpretation, as would be policy decisions that may derive from such findings. Lastly, the construct here developed suggests that policies, intentionally or unintentionally, may impact community social capital stocks for better or for worse. That social capital can be a source of inquiry highlights the fact that the effects of policy at this level could and should be monitored, if not considered from the start. Conclusions Despite limitations, the study achieved its goals of developing a conceptual framework of social capital for First Nations communities and of developing a culturally appropriate measurement instrument, with good discriminatory power, internally consistent scales, and adequate construct validity. More so, these objectives were achieved via a research process of true partnership between First Nations organizations and communities, and university based researchers. Abstract The two contributions of the study were to articulate a conceptual framework for social capital in First Nations communities and to derive culturally-appropriate measures of the dimensions of social capital. The conceptual framework expressed dimensions of socially invested resources, culture and networks. The instrument developed to measure social capital was tested for important psychometric properties. This study took place in partnership with the Assembly of Manitoba Chiefs (AMC) via its Manitoba First Nations Health Information and Research (HIR) Committee, and three First Nations communities from Manitoba. Objectives and Questions The main objectives of the study were to: (1) Present a conceptual framework of social capital for First Nations communities; (2) Develop instruments, culturally appropriate for First Nations communities, for the measurement of social capital. The research questions were: (1) What are the dimensions of social capital in First Nations communities? (2) What are the estimates of the psychometric properties of an instrument developed to measure social capital in First Nations communities? Methods Based on differential criteria of size, geographic regions, economic development and cultural representation, the HIR committee of the AMC chose three Manitoban First Nations communities to be part of the study from seven that had volunteered to participate. The first phase of the study used ethnographic methodology with two aims, to contribute to the development of the conceptual framework and to generate an initial list of instrument items. Over a period of approximately three weeks in each community, primary data collection techniques involved a combination of in-depth interviews, informal focus groups, participant observation, archival research, and unobtrusive observations. The total number of interviewees reached 85 individuals. Based on the concept analysis and on the results of the ethnographic study, dimensions of social capital were identified for measurement and a list of questionnaire items was composed. After extensive feedback and seven drafts, a final version of the questionnaire was pilot-tested. A total sample of 462 respondents from the three communities was achieved (Community A - 204, Community B - 135, Community C -123). Primary analysis goals were to produce a measurement device that had good discriminatory power among First Nations communities, was made up of internally consistent scales, and had good construct validity. Results As a result of the iteration between theory and qualitative evidence, social capital was formulated as constituted by three dimensions: Bonding social capital, Bridging social capital, and Linkage social capital. Each dimension was postulated as including three mutually dependent components: Socially Invested Resources (SIR), Culture, and Networks. The descriptors for each component were the following: SIR - Physical, Symbolic, Financial, Human, Natural; Culture – Trust, Norms of Reciprocity, Collective Action, Participation; Networks – Inclusive, Flexible, Diverse. This model considers social capital as a feature of communities. The following is the operational definition of Social Capital that resulted from the first phase of the study: Social capital characterizes a First Nation community based on the degree that its resources are socially invested, that it presents a culture of trust, norms of reciprocity, collective action, and participation, and that it possesses inclusive, flexible and diverse networks. Social capital of a community is assessed through a combination of its bonding (within group relations), bridging (inter-community ties), and linkage (interactions with formal institutions) dimensions. Results of the psychometric analyses that assessed the reliability and validity of the survey instrument were the following. The issue of high “don’t know” percentages in item responses was studied. Analytical options were suggested to deal with this matter. The first analysis goal was to find those items that formed an internally consistent scale and eliminate those that did not. The difficult balance between excluding unreliable items while not sacrificing too many potentially valid items was pursued. The internal consistency of each scale presented coefficient alphas of 0.84 (Bonding), 0.73 (Bridging), and 0.81 (Linkage). The qualitative phase of the study had hypothesized that Community B would be expected to perform better on the Bonding and Linkage scales, and