Results
Surveys were analysed from a total of 842 participants over four communities. The mean (sd) age of the sample was 25.2 (5.3) years and 352 (41.9%) of the sample were male. The mean (sd) of the primary outcome was 8.1 (1.9), kurtosis 2.4, and skewness -0.7. The model had good fit statistics, χ/df = 2.69, CFI = 0.95 and RMSEA = 0.045.
Table
3 presents the means, standard deviations and reliability estimates for the items that comprise each latent concept. All construct reliabilities were acceptably high. Cronbach’s alpha is high for ‘importance of culture’ (0.98) indicating that some latent items could be redundant. However given the paucity of validated cultural indicators in this context and our exploration of these in subsequent analyses (in separate reports), we retained this measure in order to develop and refine cultural indicators. While this does not change the results, the high correlation between latent items confirms how closely the communities view these attributes as reflecting culture (Fig.
3).
Table 3
Means, standard deviations and Cronbach alphas for each latent concept
Importance of culture | 2.88 | 1.51 | 0.98 |
Practice culture | 2.59 | 1.03 | 0.82 |
Culture in School | 2.60 | 0.93 | 0.76 |
Aboriginal literacy | 2.45 | 1.56 | 0.96 |
Empowerment | 2.39 | 0.58 | 0.84 |
Community | 2.96 | 0.96 | 0.85 |
Motivations for education | 0.30 | 0.34 | 0.92 |
English literacy and numeracy | 2.23 | 0.73 | 0.91 |
Work | 1.33 | 1.00 | 0.80 |
General health | 3.23 | 0.67 | 0.84 |
Social and emotional wellbeing | 2.93 | 0.94 | 0.82 |
Substances | 1.97 | 0.87 | 0.82 |
Table
4 displays the bivariate correlations between the latent traits. Most were highly significant. Covariances between variables that were non-significant (corresponding to correlation whose absolute value is below 0.1) were removed in the final statistical model. All cultural variables correlated highly with each other, and with empowerment and community (Table
4). While ‘importance of culture’, ‘practice culture’ and ‘culture in school’ all correlated negatively with ‘English literacy and numeracy’, they all correlated positively with ‘Aboriginal literacy’ which in turn correlated positively with ‘English literacy and numeracy’, suggesting cultural factors have indirect impacts on education through Aboriginal literacy. In depth analysis of these mediated interrelationships are reported separately. Empowerment correlated highly with all education and work variables. There is more variation between health variables suggesting more complex pathways. ‘Substances’ (measured positively with higher scores indicating less use of tobacco and alcohol) was linked with better outcomes across culture, education and work.
Table 4
Correlationsa between variables
Practice culture | 0.64*** | - | | | | | | | | | |
Culture in school | 0.56*** | 0.68*** | - | | | | | | | | |
Aboriginal literacy | 0.18*** | 0.31*** | 0.26*** | - | | | | | | | |
Empowerment | 0.11** | 0.21*** | 0.21*** | 0.17*** | - | | | | | | |
Community | 0.32*** | 0.37*** | 0.46*** | 0.21*** | 0.39*** | - | | | | | |
Motivations for Education | 0.13*** | 0.12** | 0.17*** | 0.09* | 0.21*** | 0.07 | - | | | | |
English literacy and numeracy | -0.21*** | -0.14*** | -0.11* | 0.36*** | 0.30*** | -0.02 | 0.15*** | - | | | |
Work | 0.22*** | 0.15*** | 0.22*** | 0.16*** | 0.32*** | 0.04 | 0.26*** | 0.38*** | - | | |
General health | -0.07 | 0.01 | -0.11** | -0.05 | 0.07 | -0.04 | -0.10* | 0.02 | -0.14*** | - | |
Social and emotional wellbeing | -0.18*** | -0.10* | -0.15*** | 0.00 | -0.02 | 0.01 | -0.19*** | -0.03 | -0.31*** | 0.21*** | - |
Substances | 0.11** | 0.04 | 0.22*** | 0.19*** | 0.06 | 0.05 | 0.16*** | 0.15*** | 0.23*** | -0.05 | -0.01 |
The statistical model and the estimates of the standardized structural path coefficients are presented in Table
5. Several constructs load directly onto wellbeing. Constructs with the most direct positive impact on wellbeing include – in order of impact – ‘social and emotional wellbeing’, ‘English literacy and numeracy’, ‘Aboriginal literacy’, ‘substances’ (lack thereof), ‘work’ and ‘community’. Other constructs may influence wellbeing through mediated pathways and these will be explored in separate analyses.
Table 5
Standardized path coefficients for the statistical model
Importance of culture | -0.06 | 0.24 |
Practice culture | 0.08 | 0.24 |
Culture in school |
-0.16
|
0.03
|
Aboriginal literacy |
0.14
|
<0.001
|
Empowerment | 0.07 | 0.13 |
Community |
0.08
|
0.05
|
Motivations for education |
-0.11
|
0.004
|
English literacy and numeracy |
0.15
|
<0.001
|
Work |
0.12
|
0.02
|
General health | -0.07 | 0.06 |
Social and emotional wellbeing |
0.23
|
<0.001
|
Substances |
0.13
|
0.003
|
Although configural invariance was established as per the fit statistics above, metric (weak) invariance was not, CMIN = 83.4, df = 38, p <0.001. Factor loading in each construct were constrained in separate models to examine the degree of heterogeneity between remote and very remote communities and the following constructs were found to be statistically different: empowerment (p = 0.006), community (p = 0.016), culture in school (p = 0.003), Aboriginal literacy (p = 0.001), and general health (p = 0.02). Although the results are presented with pooled data, further investigation with broader datasets is required to investigate the diversity between groups. Further analysis with a larger dataset is required to explore these differences.
Discussion
These analyses confirm statistically the holistic nature of wellbeing for Aboriginal people in remote Australia, and the importance of culture, empowerment and community to government priority areas of education, work, health and wellbeing. The evidence reported here demonstrates that all of these factors interplay through both direct and indirect relationships. They all interrelate - they all influence one another and exist as one entity. This suggests that governments can ‘close the gap’ on their priority areas of education, employment, health and wellbeing through policy and programs that build from the Aboriginal priority areas of culture, empowerment and community, such as the Empowered Communities program [
41]. As such, the methodology and framework presented here provide a culturally and scientifically valid approach for different groups to work together to develop a shared system of knowledge, values and goals.
International studies confirm the influence of social determinants on health outcomes, particularly noticeable in socially disadvantaged groups such as Aboriginal and Torres Strait Islanders [
42,
43]. Our findings confirm that social factors directly influence wellbeing in this group, particularly social and emotional wellbeing, education, work, community and substance use.
Our findings emphasize a key role for culture and empowerment in the sphere of wellbeing. For example, Aboriginal literacy has a direct positive impact on wellbeing and other cultural factors appear to have equally important but less direct impacts on wellbeing. These include learning about culture at school, strong links between the community and school, learning in one’s first language at school (bilingual education), practicing culture through ‘caring for country’ and hunting for food sources, together with the importance of law and ceremony in one’s life. Validation of these indicators confirms the foundational role they play in the wellbeing of Aboriginal people in remote Australia.
Of relevance to remote communities - where people are more likely to live on or near their ancestral lands - is the prominence amongst these validated cultural indicators of measures relating to connection to language, land, law and ceremony. The interplay and role of these factors may play out differently in more urban communities where these connections can be strained. Our analysis suggests some indicators behave differently in relation to remoteness amongst our cohort. Further investigation is therefore required to better understand the impacts of remoteness, and cultural and geographic diversity to the holistic system of wellbeing.
The use of cultural indicators is growing in research such as in linking land management roles (‘caring for country’) and health [
38,
39] - and in government monitoring [
8,
14,
30,
33]. Data extracted from national surveys have also been used to show the importance of cultural attachment for wellbeing based on cultural activities, and use of land and language [
44,
45]. Dockery also showed differences in a derived indicator of cultural attachment between Indigenous people living in remote compared with urban Australia [
44]. However national datasets are limited in addressing cultural and geographic diversity and further work is therefore required to develop more sophisticated knowledge and indicators to better understand diversity and its impacts [
2]. For example, communities in tropical and desert regions can both be considered remote but differ vastly in climate and population density, but also in terms of cultural practices [
2,
25,
46].
Similar to the current research, localised approaches have used qualitative research to inform the development of quantitative measures to understand Aboriginal wellbeing [
47,
48]. As this body of research grows, it becomes important to corroborate findings across studies, such that localized research providing a ‘depth’ of understanding informs cultural indicators in national studies that address the ‘breadth’ of understanding, to better inform policy and progress.
Importantly the domains of the Interplay Wellbeing Framework that represent priorities identified by Aboriginal people in remote communities - including culture, empowerment and community - all correlated strongly with each other. Their relevance to more government prioritized areas of education, work and health was not only confirmed by the entire model goodness-of-fit statistics, but also through subsequent analysis showing strong relationships between Aboriginal literacy and both English literacy and numeracy, and wellbeing. This suggests that Aboriginal literacy developed through learning about culture in school, and learning in one’s first language in school, is a key stepping stone to achieve success in English literacy and numeracy, and improve wellbeing overall. These interpretations are consistent with the correlation analyses presented here and we will explore them in more depth through mediation analyses as part of structural equation modeling to be reported subsequently.
Strong correlations shown between empowerment with all education and work variables signify a key role for building empowerment to improve outcomes across government priority areas of education and employment, and wellbeing overall. Our findings support an existing model demonstrating the importance of building empowerment based on cultural and spiritual beliefs, to improve wellbeing for Indigenous Australians [
49,
50]. Our measures of empowerment are based on self-reported resilience, self-efficacy and identity. The strong correlations reported here between culture, empowerment and community indicators suggest building empowerment is closely linked to strengthening community and culture. Based on our data presented here, strengthening community means building safety, connectedness, trust and respect. Building cultural strengths means fostering connections with language, land, law and ceremony. Analysis also suggests that reduced use of alcohol and tobacco leads to stronger representation in culture, education and work. These interrelationships will be explored in more detail through subsequent mediation analyses.
Aboriginal communities are heterogeneous and our approach addresses this dependence on context through a mixed methods design. However, limitations of this paper include that participants were surveyed from four remote communities only and further data is required to understand how generalizable these findings are nationally. Framework development was guided by our research questions and may have evolved differently for a different purpose. Further limitations include the reliance on self-report and that many items were not normal in distribution. However, we tested the robustness of our results by rerunning the analyses in Stata with robust standard errors and found the same results in terms of statistical significance. Further, objective measures are also limited in addressing the non-tangible and subjective aspects of wellbeing that are recorded here.
Although there are some positive developments in this area [
41], Aboriginal wellbeing policy strategies have historically been criticized as ‘top down’, ‘one-size-fits-all’, focusing on disadvantage rather than strengths, and looking at education, employment and health separately rather than at the whole of system level [
2]. A lack of genuinely involving Aboriginal and Torres Strait Islander peoples in defining and measuring concepts that shape their lives through government policy has received increasing attention [
2,
13]. Further, the cultural diversity represented by Aboriginal and Torres Strait Islander peoples across Australia, and the vastly different life experiences of those living in remote compared with urban places has challenged national policy approaches.
Here we present an integrated model that is innovative in terms of its three primary characteristics: (1) the ‘shared space’ approach to working collaboratively across languages and cultural worldviews with the key stakeholder groups; (2) considering these challenges at a whole-of-system level, to understand how different components Interplay or work together as part of an interconnected system and (3) ‘bringing together stories and numbers’ to represent Aboriginal values in western monitoring systems to inform policy. An integrated research approach such as this is extremely rare in the literature and innovative in its approach and application in the proposed context. The Interplay Wellbeing Framework has since been considered for similar applications in substance use and identity research, and for evaluation of the wellbeing impacts of policies and programs.
The framework and data presented here form a baseline against which longitudinal data are being collected and will be reported in the future as part of a prospective cohort design. The Interplay Wellbeing Framework provides a statistical tool to measure and strengthen wellbeing. It has a web-based representation using tailored data visualisation software, showcasing integrated ‘stories and numbers’ to optimise accessibility to general audiences [
51]. The visualization represents statistics such as those presented here together with video stories recorded from Aboriginal community members voicing their perspectives on how areas represented in the framework play out in their lives. Data derived from this integrated approach can illuminate the relative balance of investment necessary to have the most significant wellbeing benefits for Aboriginal people.
Implications for policy and community
These analyses confirm that wellbeing is holistic and solutions must therefore be considered at the ‘whole of system’ level, meaning they must address all of the different areas that interplay to impact ones wellbeing. It further suggests that Australian Governments can best meet their ‘Closing the Gap’ objectives through policies and programs that are developed with the active participation of Aboriginal people and strengthen culture, empowerment and community.
As part of the shared space approach to knowledge translation, this research has been represented in a variety of output including the interactive Interplay Wellbeing Framework that includes statistics (numbers) and 30+ short video documentaries (stories) [
51], a talk at TEDx StKilda [
52], conferences, academic publications (in development) and a series of posters to report preliminary outcomes [
53].
These outputs were officially launched in November 2016 through a number of events: with policy makers at the Department of Prime Minister and Cabinet; and in each participating community. In addition to informing policy, participating communities have benefited from the shared space approach through their active participation and capacity development through the research, and report confidence of using both the networks and resources built in their future work. Detailed qualitative reports on this process will follow.
Based on subsequent collection of qualitative and quantitative data using a community-level survey, development work is underway to validate the Interplay Wellbeing Framework to evaluate the wellbeing impacts of programs, policies and service delivery.
Acknowledgements
This research is supported by the CRC-REP, hosted by Ninti One Limited. We acknowledge the support and involvement of our key organisational stakeholders: Centre for Remote Health (Flinders University and Charles Darwin University), Department of Prime Minister and Cabinet (PMC), Northern Star Resources, Yalu Marŋgithinyaraw Aboriginal Corporation, Marthakal Homelands Resource Centre, Central Desert Native Title Services, Poche Centre of Indigenous Health, Miwatj Health Aboriginal Corporation, Australian Bureau of Statistics (ABS), Muntjiltjarra Wurrgumu Group (MWG), Kalano Community Association, Wurli-Wurlinjang Health Service, StrongBala Men's Health Program, Flinders NT-Katherine, Katherine Stolen Generations Group, Banatjarl Strongbala Wumin Grup, Martu Rangers (Wiluna), Ngangganawili Aboriginal Health Service Community (NAHS).