The purpose of this study was to develop and psychometrically test the Korean Geriatric Loneliness Scale (KGLS).
METHODS:
The initial items were based on in-depth interviews with 10 older adults. Psychometric testing was then conducted with 322 community-dwelling older adultsaged 65 or older. Content, construct, and criterion-related validity, classification in cutoff point, internal consistency reliability, and test-retest reliability were used for the analysis.
RESULTS:
Exploratory factor analysis showed three factors, including 15 items explaining 91.6% of the total variance. The three distinct factors were loneliness associated with family relationships (34.3%), social loneliness (32.4%), and a lack of belonging (24.9%). As a result of confirmatory factor analysis, 14 items in the three-factor structure were validated. Receiver operating characteristicanalysis demonstrated that the KGLS' cutoff point of 32 was associated with a sensitivity of 71.0%, specificity of 80.2%, and area under the curve of .83. Reliability, as verified by the test-retest intraclass correlation coefficient, was .89, and Cronbach's α was .90.