Validation of the death reflection scale among older people

Human beings possess the capacity to cognize their own mortality, which compels them to process death awareness. The present study seeks to validate the Death Reflection Scale (DRS) among older individuals, which measures growth-oriented cognitions and prosocial behavior following confrontation with death awareness.

DRS was validated using a cross-sectional online survey of older adults (50+). To assess configural, metric, and scalar measurement equivalence across age groups of older adults, a multi-group confirmatory factor analysis was conducted in conjunction with dynamic fit index cutoffs. Construct validity was evaluated using Pearson's correlation and analysis of variance (ANOVA). The data were obtained from an online survey. The survey was conducted in May 2023 and people aged 50 and over were recruited from an online panel using quota sampling (by age, gender and federal state). A total of 1,806 individuals completed the survey.

Confirmatory factor analysis showed a good fit of the originally proposed 5-factor (CFI 0.949, SRMR 0.058, RMSEA 0.070) and bi-factor model (CFI 0.956, SRMR 0.067, RMSEA 0.067) with a general factor and five subscales. In light of several considerations, it is recommended that the DRS should be considered as a five-factor model, as originally proposed. Alpha ranges from 0.807 to 0.875 and Omega from 0.811 to 0.875, indicating good reliability. Partial scalar invariance was obtained, therefore mean comparisons can be made between groups of older people. Testing the construct validity showed only a partial confirmation. The exploratory analysis of the DRS with the Big Five personality traits revealed a correlation structure that can be plausibly explained by considering the facets of personality traits.

The value of the DRS lies in its perspective that death awareness should not only be viewed as a threat, but rather as a potential for a positive and growth-oriented perspective on death awareness and has been validated for older adults. The present study demonstrated that mean comparisons could be conducted between groups of older adults. Tests of construct validity yielded inconclusive results, indicating the necessity for further analysis.