MEASUREMENT INVARIANCE OF THE SHORT UCLA LONELINESS SCALE IN SPANISH AND PERUVIAN OLD PEOPLE: LATENT MEAN DIFFERENCES AND EVIDENCE FOR DIFFERENTIAL EFFECTS ON PERCEIVED HEALTH

Abstract

The objectives of this study are to evaluate the measurement invariance of the University of California Los Angeles Loneliness Scale (UCLA-LS) three-item version (UCLA-LS-3) in older adults in Peru and Spain, to compare the latent means of loneliness, and to evaluate the psychometric properties of the scale with Item Response Theory (IRT) models, and evaluate the possible moderating effects of the country on loneliness-health relationships. Peruvian sample was composed of 235 old adults from the city of Lima. The Spanish sample was composed of 443 old adults. The three-factor structure of RUCLA-3 anchored to the health measure fitted the data reasonably well in Spain and Peru. The R-UCLA-3 may be considered invariant for these two samples. The latent means of loneliness are different, the Peruvian average of loneliness being greater than that of Spain. The R-UCLA-3 is an invariant measure in older adults in Peru and Spain, with adequate psychometric properties through IRT models.