De Jong Gierveld Loneliness Scale: validity, reliability and fairness in Peruvian adults

Abstract:

Background: Loneliness, transient or long-lasting, constitutes one of the mental health problems of great public impact with peculiar characteristics. It can be defined as a subjective and unpleasant experience, in turn associated with other serious psychological symptoms. This variable in question has not been addressed in a timely manner, among other reasons due to the scarcity of instruments for specific populations. Although the De Jong Gierveld Loneliness Scale (DJGLS), based on Weiss' multidimensional model, has been adapted and validated in different contexts, it is still insufficient in Peru. Precisely, the objective was to determine the psychometric properties of the DJGLS, its internal structure and factorial invariance.

Methods: An online survey of 1248 Peruvians between 18 and 70 years of age (M= 27.37, SD= 11.29) from all regions was used. The validation of the DJGLS was analyzed with Exploratory Factor Analysis (EFA), Confirmatory Factor Analysis (CFA), convergent validity, measurement invariance and internal consistency reliability.

Results: Psychometric properties were found with adequate values in its internal structure by means of the CFA, where it was found that the components of the scale are interrelated and the data matrix is factorizable. Here we present a model of two specific factors and one general factor, which is consistent with theory and has practical utility, revealing acceptable reliability values and invariance between sexes.

Conclusions: Adequate psychometric properties, which allow for a better data collection process in further related research, are revealed.

Keywords: loneliness; validity; reliability; fairness; adults.