Background Due to a history of colonization, Aboriginal and Torres Strait Islander became one of the most disadvantaged groups in Australia. The decades of assimilation policies, which culminated in the removal of Aboriginal children from their parents, disarranged the social support derived from their kinship system. In non-Indigenous populations, social support has been reported as a protective factor against stress (i.e. buffering hypothesis) and coronary heart disease; providing also better outcomes for breast cancer and type-2 diabetes. There are no psychological instruments validated to measure social support in Aboriginal Australians. The aim of the current study was to evaluate the validity and reliability of the Social Support Scale in an Aboriginal population.

Methods The Social Support Scale (SSS) is a 4-item measure developed to evaluate the emotional, appraisal, instrumental and informational domains of social support. Data was collected from to two different samples: participants of the (1) Teeth Talk Study (n=317), an oral-health randomized controlled trial (RCT) conducted among Aboriginal adults; and (2) the Baby Teeth Talk Study (n=367), an RCT conducted among pregnant Aboriginal women. The psychometric properties of the SSS were evaluated with Graphical Loglinear Rasch Models (GLLRM), which consist of recent methodological advances that extend the Rasch Model to incorporate local dependence (LD) and differential item functioning (DIF).

Results Overall fit to a GLLRM was found (X²(96)_sample1=52.7, p=0.057; X²(25)_sample2=22.2, p=0.625) after the inclusion of LD between items 3 and 4 (g_avg1=0.50; g_avg2=0.66). Item 2 displayed DIF by employment status (X²(4)=12.60, p=0.013, g=0.21) in Sample 1. There was no DIF by sex or educational attainment. Unidimensionality was confirmed in both samples (g_obs1=0.80; g_exp1=0.78, p=0.654; g_obs2=0.75, g_exp2=0.77, p=0.163). The SSS displayed good reliability (R_sample1=0.82, R_sample2=0.84) and probability of person separation (P_sample1=0.77, P_sample2=0.78). Targeting was poor (TTI_sample1=0.28; TTI_sample2=0.49).

Conclusion The same unidimensional structure was found in both samples, providing confidence in the robustness of results. Item 2 (‘There are people in my life who appreciate what I do’) displayed DIF by employment status, indicating that participants who were employed were more likely to feel appreciated than those unemployed given the same level of social support. The current study consisted of the first validation of a Western-developed psychological instrument to measure social support in Aboriginal Australians. The results show that the SSS is a culturally-valid and reliable instrument that can be applied in future health research among Aboriginal and Torres Strait Islanders.