Can the Short-Form UCLA Loneliness Scale Be Used to Measure Loneliness Among Chinese Older Adults? From Classical Test Theory to Rasch Analysis

Abstract

Background

Loneliness among older adults, is a subjective experience and a public health issue in aging societies. Psychometrically sound and culturally sensitive measures are needed for developing precisely targeted interventions in culturally distinct groups. This study tested the Short-Form UCLA Loneliness Scale (ULS-8) among Chinese older adults.

Method

Confirmatory factor analysis, internal consistency, and the correlation with the single question of loneliness were conducted with a sample of Chinese older adults. Rasch analyses assessed the unidimensionality, response category functioning, item difficulty, and targeting of the ULS-8 for older Chinese adults.

Results

Data from 347 Chinese older adults (mean age 71.36 ± 9.51 years) were analyzed; 74.64% of the participants were female. The ULS-8 showed acceptable internal consistency and criterion validity in Classical Test Theory. Confirmatory factor analysis and Rasch analysis indicated that the ULS-8 did not demonstrate a unidimensional structure. Additionally, Rasch analysis revealed (1) a misfit in item 3, indicating a problem with construct validity; (2) the need to combine response categories; and (3) that Chinese older adults are less likely to endorse a high level of loneliness when using the ULS-8.

Conclusions

To ascertain the adequacy of the loneliness measure, it is crucial to customize a new short version of the loneliness scale for Chinese older adults through Rasch analysis.