Psychometric properties of the 10-item Connor-Davidson resilience scale among Finnish older adults.

Objectives: Resilience, the ability to bounce back after adverse events may be an important factor in active aging. The 10-item version of the Connor-Davidson Resilience Scale (CD-RISC10) seems suitable for aging research owing to its low participant burden; however, its psychometric properties have not been comprehensively reported for older people. For example, no estimate exists of the test-retest reliability of the scale when used with older adults. Hence, this study evaluated the psychometric properties of the CD-RISC10 in a large population-based sample of community-dwelling older people.

Method: The scale's internal consistency, factor structure, construct validity, test-retest reliability, and user experience were analyzed among 1018 Finnish older adults (57% women) aged 75 (45%), 80 (33%) and 85 years (22%).

Results: The internal consistency of the CD-RISC10 was good (Cronbach's α = 0.85), test-retest reliability moderate (ICC = 0.61), and the scale was unidimensional. However, a two-factor solution was found among the 75-year-olds, where the ability to deal positively with change (e.g. using humor) explained an additional part of resilience. While no differences by gender, age, or education were observed in total scores, very good perceived financial situation was associated with higher resilience. The scale showed good construct validity, and user feedback indicated that administering the scale in research is quick and easy.

Conclusion: In general, the CD-RISC10 is a valid method to study older adults' psychological resilience. However, the present test-retest reliability values suggest caution in using the scale as an outcome measure in intervention studies.