A psychometric evaluation of the Connor–Davidson Resilience Scale for Australian Aboriginal youth

ABSTRACT

Objective: There is a paucity of quantitative measures of resilience specifically validated for young Aboriginal people in Australia. We undertook the first investigation of validity and reliability of the Connor-Davidson Resilience Scale (CD-RISC) in a sample of Australian Aboriginal people, with a focus on youth.

Method: We conducted a cross-sectional study of resilience among a sample of 122 Aboriginal youth (15–25 years old) in New South Wales and Western Australia, featuring self-completes of the 10-item CD-RISC in online (N = 22) and face-to-face (N = 100) settings. A Rasch analysis using the 122 CD-RISC responses determined item independence, response category adequacy, differential item functioning, unidimensional measurement, person and item reliability, and item fit. Confirmatory factor analysis was also conducted, complementary to the Rasch analysis.

Results: Four problematic items from the original instrument were removed, due to item dependence (items 2, 6 and 9; Q_3,* > 0.30) and differential item functioning (item 4; > 0.43 logits between males and females). The final 6-item instrument exhibited improved item separation (ISI = 2.14) and reliability index values (IRI =.82) – suggesting an improved structure – however several limitations such as a prominent ceiling effect were evident (i.e., positive measure targeting coefficient of 0.99 logits).

Conclusion: Findings suggest the CD-RISC instrument should be applied in Aboriginal contexts with caution. Further psychometric examination of the CD-RISC with Aboriginal youth is warranted before it can be used with confidence by researchers and clinicians.

KEY POINTS

What is already known about this topic:

1. Resilience is generally thought of as positive outcomes despite exposure to risk. This can be a particularly important concept to consider in the period of youth and adolescence, when young people are experiencing wide ranging physical, emotional, and social changes and challenges.

2. Aboriginal youth face unique circumstances that are necessary to consider when considering resilience relevant to an Aboriginal perspective. There is currently a lack in the literature in regards to defining and conceptualising resilience from an Aboriginal perspective, particularly relevant to young people.

3. There is a dearth of instruments available that have been rigorously examined for their appropriateness and psychometric properties relating to resilience in Aboriginal youth.

What this paper adds:

1. This paper provides not only a first investigation of this version of the CD-RISC using Rasch analysis, but also the first time the CD-RISC’s validity and reliability has been assessed for an Aboriginal sample in Australia.

2. Analyses results illustrated a ceiling effect, highlighting the fact that for the instrument to be accessing the full range of resilience in the sample, new items need to be developed to tap into higher levels of resilience.

3. With the findings of this paper as a foundation, further investigation and adaptation of the CD-RISC could potentially lead to a useful screening tool for assessing resilience and identifying Aboriginal youth who may require further support.

KEYWORDS:

• Aboriginal
• measurement
• psychometrics
• Rasch analysis
• resilience
• youth

Acknowledgments

We gratefully acknowledge the investigators and staff of the Young Indigenous Peoples’ Resilience and Wellbeing research project (YIPRaW) for development of the survey design and dissemination, and the contribution of the Aboriginal and Torres Strait Islander youth who participated in the study. We also respectfully acknowledge the cultural guidance of the Elders and senior Aboriginal men and women in both Western Australia and New South Wales. The YIPRaW research team and staff who contributed are Kim Usher, Corinne Reid, Roz Walker, Katrina Hopkins, Debra Jackson, Petra Skeffington, Reakeeta Smallwood, Kevin McGregor, Cecilia Strutt, Grantley Winmar, Naomi Rumble, and Cindy Woods.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

^1. For the purpose of brevity, and as the work described in this paper was conducted on Aboriginal land with Aboriginal communities (as opposed to the Torres Strait Islands), the term “Aboriginal” will be used in reference to First Nations people of Australia, with no disrespect intended towards Torres Strait Islander people.

For the purpose of brevity, and as the work described in this paper was conducted on Aboriginal land with Aboriginal communities (as opposed to the Torres Strait Islands), the term “Aboriginal” will be used in reference to First Nations people of Australia, with no disrespect intended towards Torres Strait Islander people.

^2. Abbreviated versions of CD-RISC item wording from Campbell-Sills and Stein (2007) work are used throughout this paper due to original author copyright restrictions (Connor & Davidson, 2003).

Abbreviated versions of CD-RISC item wording from Campbell-Sills and Stein (2007) work are used throughout this paper due to original author copyright restrictions (Connor & Davidson, 2003).

Additional information

Funding

This research was supported by the Australian Government through the Australian Research Council’s Discovery Indigenous funding scheme (project number IN170100008). The first author is a recipient of an Australian Government Research Training Program (RTP) Scholarship.