Measuring resilience using Strong Souls in the Longitudinal Study of Indigenous Children: Evaluating psychometric properties using a Rasch measurement approach

Abstract

Purpose Rigorously evaluated psychometric instruments are necessary to measure constructs relevant to wellbeing, such as resilience. The availability of high-quality instruments for use with Aboriginal young people in Australia is growing but remains limited. This study used data from the Longitudinal Study of Indigenous Children (LSIC) to psychometrically evaluate the Resilience subscale from the social and emotional wellbeing (SEWB) instrument, Strong Souls, for a nationwide sample of Aboriginal adolescents, using alternative psychometric methods to previous research.

Methods Using a Rasch measurement approach, cross-sectional data from Wave 9 of LSIC were used to ascertain the psychometric properties of the Resilience subscale from Strong Souls. Using the responses from 516 Aboriginal young people (age 11.5 to 13 years) to the 12-item scale, Rasch techniques were applied to determine item independence, response category adequacy, differential item functioning (DIF), person and item reliability, item fit and unidimensionality. Two versions of the instrument were evaluated: the full 12-item version, as completed by participants, and an 8-item version, as recommended for use by previous research based on different psychometric methods.

Main findings Both versions of the instrument met several Rasch model requirements for reliable measurement, including demonstrations of unidimensionality (first off factor construct < 2), item independence (all items Q3∗ < 0.30), and item fit statistics within an acceptable range (0.60 < X < 1.40). However, both instrument versions displayed less adequate person separation (PSI) and reliability (PRI) statistics (12-item scale: PSI = 1.18, PRI = 0.58; 8-item scale: PSI = 0.71, PRI = 0.33). Principal conclusions Using a Rasch measurement approach to psychometrically evaluate the Strong Souls Resilience subscale in a sample of Aboriginal young people from LSIC, this study provided novel evidence of the functioning of this popular instrument from an alternative psychometric perspective. With mixed results regarding meeting Rasch recommendations, these findings provide a strong evidence base for psychometric strengths as well as opportunities to improve the robustness of this instrument, and ultimately offer a tool that can more accurately inform services, policy and practice to effectively support resilience and wellbeing in Aboriginal young people.