National Epidemiology Centre and CIBERNED, Carlos III Health Institute, Madrid, Spain.
University Carlos III of Madrid, Madrid, Spain.
Australas J Ageing. 2021 Jun;40(2):e100-e108. doi: 10.1111/ajag.12893. Epub 2021 Feb 10.
To analyse the psychometric properties of the De Jong Gierveld Loneliness Scale, 6-item version (DJGLS-6), in a multiethnic sample of older Chilean adults.
A cross-sectional study was designed using a sample of 800 older people, and 71% were self-declared indigenous. Data quality and acceptability, internal consistency, and construct and structural validity of DJGLS-6 were analysed. The scale was also analysed according to the Rasch model.
The DJGLS-6 scored higher for the non-indigenous group than for Aymara and Mapuche (P < .001). KR-20 coefficients (internal consistency) for emotional loneliness subscale reached 0.44 for Aymara and Mapuche and 0.62 for non-indigenous, and from 0.73 (non-indigenous) to 0.86 (Aymara) for social loneliness subscale. The confirmatory factor analysis identified one dimension. None of the two subscales fit the Rasch model.
The results indicate that DJGLS-6 is a reliable and adequate scale for non-indigenous Chilean older people. However, it should be culturally adapted to indigenous ethnic minorities.
分析多民族的智利老年人群体中，6 项德文孤独量表（DJGLS-6）的心理测量特性。
采用横断面研究设计，样本量为 800 名老年人，其中 71%为自报的土著人。分析 DJGLS-6 的数据质量和可接受性、内部一致性以及结构和结构有效性。还根据 Rasch 模型分析了该量表。
DJGLS-6 非土著组的得分高于艾玛拉和马普切组（P<.001）。艾玛拉和马普切组情感孤独分量表的 KR-20 系数（内部一致性）为 0.44，非土著组为 0.62，社会孤独分量表的范围从 0.73（非土著组）到 0.86（艾玛拉组）。验证性因子分析确定了一个维度。两个分量表均不符合 Rasch 模型。
结果表明，DJGLS-6 是适用于非土著智利老年人的可靠和充分的量表。然而，它应该针对土著少数民族进行文化适应性调整。
