Study design
In seeking to develop an appropriate tool and in taking the perspective of Indigenous stakeholders with an interest in the food system, we planned our process over five steps that involved : (i) test of clarity of construct meaning; (ii) inductive derivation of community capacity constructs; (iii) application of community-derived constructs to the capacity development of multi-sector food-interest groups; (iv) a cross-check of community-derived constructs and their meanings to literature-derived constructs, and; (v) achieving consensus on tool constructs.
Data specific to this paper were collected over a three-year period (2010–2012) that involved two on-site visits to one community, and two urban-based workshops that brought together stakeholders from each of four remote Indigenous communities who had various roles in the food system. These data were augmented by meeting minutes and reports recorded from the multi-sector food-interest group meetings and workshops that occurred in each of the four communities as part of the GFS Project.
Ethics approval for all aspects of the GFS Project was attained from the Human Research Ethics Committee (HREC) of the Northern Territory Department of Health and Menzies School of Health Research (ref. HREC 09/07), Cairns and Hinterland Health Service District Ethics Committee (ref. HREC/10/QCH/71-678) and the Central Australian HREC (ref 2009.02.02). Signed partnership agreements were negotiated with participating organisations and communities.
Although here we are documenting a tool development process with community participants rather than a straightforward qualitative study involving community, we in the research team have endeavoured to adhere to the RATS guidelines [23] where applicable, in clearly documenting the relevance of the study questions, the appropriateness of the method employed, the transparency of procedures undertaken and the soundness of our interpretive approach.
Study setting
Two of the remote Indigenous Australian communities that participated in the GFS project were situated on the North Australian coast, another was in escarpment country in Northern Australia and a fourth was in the Central Australian desert. The communities varied in size from 250 to over 2000 residents, and shared a relatively recent history of European colonisation. Multiple languages were spoken in each community with English rarely the first language. The four communities were of Aboriginal heritage and diverse in cultural practices. All were accessible only by long-distance road travel, small aircraft and/or boat. They were all serviced by limited public facilities – for example schools and health clinics – and private resources such as banking, telephone and a community store.
Participants
In each community, stakeholders with an interest in the food system were invited to meet as part of the GFS project. In three of the four communities, an average of 10 (8–12) stakeholder meetings (including annual planning workshops and review meetings) occurred over the period of the GFS project. In the fourth community only one annual workshop and two review meetings took place. Participation in these workshops/meetings was voluntary and open to any community member (both Indigenous and non-Indigenous) or external stakeholder who had an interest in, or who might impact on, the local food system. Participants included community elders and other community members, gardeners, store managers, owners and operators (including store and take-away outlet staff), public health nutritionists, other health professionals, school, crèche, and aged-care staff, and government officials involved in food related policy and service delivery. A member of the research team (project facilitator) and a local resident Aboriginal community co-ordinator (employed by the project) facilitated these meetings.
One community was conveniently selected as the field site to test the community capacity constructs due both to its relatively close proximity to Darwin (a 700 km round trip via road) and also because the food-interest group in this community had met more frequently over the entire study period compared to the other three communities and had a relatively stable group of people attending meetings.
Tool development process
Step 1. Test of clarity of construct meaning (March 2010, Year 2 of the GFS project).
This first step aimed to determine the relevance and meaning ascribed to each of the literature-derived capacity development constructs (as listed in the introduction and shown in Additional file 1: Table S1) from the perspective of members of the food-interest group in the field site. Eight people representing community elders, the health centre, local government, the community market garden and the store participated in the community-based food-interest group meeting. Seven of these people were Aboriginal. Members of the food-interest group were asked as a collective to firstly review each construct and its definition, to comment on its meaning and then to appraise their capacity using each of the constructs. Each construct was discussed in depth with specific and practical examples given by the facilitator to assist participants to consider the application of the construct to their situation. All constructs were considered by participants as important to community capacity development and through the appraisal process specific areas of capacity development were identified as in need of strengthening. The appraisal process took 50 minutes in total. Results from this appraisal process are shown in Additional file 1: Table S1.
In appraising the group’s capacity, the eight constructs were displayed on an A3 sized paper around an image of a coloured shooting target or bull’s eye. The purpose of the shooting target was to provide a visual scoring system for members to consider how well the food-interest group was gaining or utilising their capacity to address community level food security, and to point out where on the circles they might place themselves. In contrast to that intended, this visual aid of a shooting target was viewed by participants in the opposite way where the central circle of the target represented the construct strength as ‘just developing’ and the outer circles represented progressive steps to ‘fully developed/very strong’ , just as a heavy stone causes ripples to spread more distantly. From this the concept of a ripple with five concentric circles emerged. In subsequent activities in step 3 this concept was successfully used with a large drawing of a ripple placed in the middle of each of the participating community groups. Each of the measures were then appraised by each group in relation to their relative position from ‘just developing’ in the centre to ‘fully developed/very strong’ on the outer circle (see Additional file 2: Table S2).
Step 2. Inductive derivation of community capacity constructs (November 2011, Year 3 of the GFS project).
This step aimed to determine the most pertinent set of attributes perceived by participants as important to community capacity. Fourteen participants, from all four project communities participated in an urban-based workshop with 11 of these being Aboriginal. Two to six people represented each community. In community groups, participants were asked to identify and describe the characteristics of a ‘really strong’ food-interest group. The term ‘strong’ is a colloquial term used in remote Indigenous Australia generally used to denote the strength or resilience of an object, person or structure in relation to its purpose. These characteristics were then categorised by two of the authors (JB & CB) according to common themes and each category was given a label. These categories were then checked with the participants. The eight capacity building constructs that emerged from this exercise are presented in Additional file 2: Table S2.
Step 3. Application of community-derived constructs to the capacity development of multi-sector food-interest groups (November 2011, Year 3 of the GFS project).
To test if step 2 workshop participants could then use the community capacity constructs and “ripple tool” to appraise their community’s capacity, the four-interest groups discussed each construct amongst themselves and marked on the ripple circle where they viewed the development of each construct. Similar to the method described by Gibbon et al. [17], discussion points and reasons given by the group for the scoring of each construct were recorded. Additional file 2: Table S2 shows that the areas identified as requiring strengthening varied across communities. In the three communities where the food-interest group had met more than twice, the constructs “knowledge and skills” and “good planning” were scored relatively higher than the constructs “leadership/ right people” and “commitment to action”, which were identified as needing the most strengthening.
Later in November 2011, this process was repeated in the field with a community food-interest group. Ten people participated of whom five were Aboriginal. The process for completing the appraisal was modified from that used by McDonald et al. [24] with the role of the facilitator being to:
-
Provide a verbal description of each construct based on the characteristics of the construct identified previously by workshop participants
-
Ask the group to reflect and comment on how strongly they saw the construct performing in their community;
-
After listening to the group’s discussion, place a card with the construct label on a circle of the ripple tool that seemed to best represent people’s comments and ask if this matched with their perception; and then,
-
Manage the discussion until consensus was reached and write agreed discussion points on a sticky note and also display on the ripple;
-
Take a photo for future comparison with the next community capacity appraisal.
Participants demonstrated an understanding of each construct and were able to relate them to their community context. Most participants engaged in the discussion and different views were expressed and taken into consideration by the group in reaching consensus on the construct score. Food-interest group members scored the constructs “knowledge and skills” and “commitment to action” as very strong/fully developed, and “community ownership and support”, and “communication” as just developing and in need of strengthening (see Figure 1). The group perceived the process of planning to be well developed, but were concerned that prioritised actions were not being implemented due to the already heavy workload of participants. Overall, participants identified that more investment was needed in gaining greater support from the wider community and encouraging participation from sectors and community groups not represented.
Step 4. A cross-check of community-derived constructs and their meanings to literature-derived constructs (February 2012, year 4 of the GFS project).
The aim of this final step was to match community-derived constructs of community capacity and their meanings to literature-derived constructs and to lastly propose a capacity appraisal model that could be used by community level multi-sector groups to consider their capacity. This step involved a review by the authors (JB, CB and SL) of all data collected from steps 1 to 3 augmented with food-interest group meeting minutes and workshop reports, and the comparing and contrasting of the capacity building characteristics and associated meanings captured in these data with the literature-derived constructs until all relevant data had been matched. Data were also checked for any new emerging constructs and/or divergence in meaning within constructs. Through this process, no new constructs to the literature-derived constructs were identified and all the literature-derived constructs were found to be represented by participant-derived constructs, except for ‘assets-based approach’ and ‘resource mobilisation’ , although elements of these were identified. Different terminologies to describe constructs were also used.
Step 5. Achieving consensus on tool measure terms (October 2012).The literature-derived constructs were refined in a workshop with three community coordinators from two study communities and consensus was reached on the structure of the appraisal tool shown in Figure 2.