Modeling of the UCLA Loneliness Scale According to the Multidimensional Item Response Theory

Auné, Sofía Esmeralda; Abal, Facundo Juan Pablo; Attorresi, Horacio Félix

doi:10.1007/s12144-020-00646-y

Cited by 12 publications

(17 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The internal consistency, test-retest reliability and discriminant and convergent validity of the UCLA-LS3 has been supported by several studies (e.g. Auné, Abal, & Attorresi, 2020;López-Ramos, Navarro-Pardo, Fernández Muñoz, & Da Silva Pocinho, 2017). Further confirming its utility over the two previous versions, the UCLA-LS3 has been adapted and validated in various cultural contexts and countries (e.g.…”

mentioning

confidence: 55%

The applicability of the UCLA loneliness scale in South Africa: Factor structure and dimensionality

Pretorius

2022

AJOPA

View full text Add to dashboard Cite

This study examines the generalisability of the University of California Los Angeles Loneliness Scale Version 3 (UCLA-LS3) in a South African sample of young adults. In particular, it examined the normative data, reliability, and factor structure of this scale. The participants were young adults (N = 337) who were randomly sampled from a university population and they responded to the UCLA Loneliness Scale. It was found that the sample had higher loneliness scores than those reported in the literature, potentially suggesting that loneliness may be a significant mental health concern amongst this group. Women reported higher levels of loneliness than men. Reliability analysis (Cronbach’s alpha) and analysis of the influence of individual items on the mean, variance, and alpha demonstrated that UCLA-LS3 had highly satisfactory internal consistency in the sample. Confirmatory factor analysis (CFA) was used to test four conceptualisations of the factor structure of UCLA-LS3: a one-factor model, a correlated three-factor model, a bifactor model with two subscales, and a bifactor model with three subscales. Notably, CFA demonstrated that the two bifactor models are a better fit than the one-factor and correlated three-factor models and that the bifactor model with three subscales is marginally a better fit than the bifactor model with two subscales. Ancillary bifactor analysis confirmed the dimensionality of the scale as sufficient variance was accounted for by the three subscales, after the variance attributable to the total scale was partitioned out. Therefore, UCLA-LS3 is best conceptualised as comprising of three subscales (isolation, relational connectedness, collective connectedness), in addition to a total scale.

show abstract

mentioning

confidence: 55%

The applicability of the UCLA loneliness scale in South Africa: Factor structure and dimensionality

Pretorius

2022

AJOPA

View full text Add to dashboard Cite

show abstract

“…Es un test breve, unidimensional, conformado por siete ítems de respuesta politómica, los cuales fueron redactados en base a entrevistas grupales a población general adulta y de la tercera edad residentes en el Área Metropolitana de Buenos Aires (AMBA). Aunque existía una escala adaptada al país para medir la soledad, se realizó un estudio de la misma que evidenció que las respuestas a los ítems se veían influenciadas por la dirección de respuesta (Auné, Abal & Attorresi, 2020).…”

Section: La Escala De Soledad De Buenos Airesunclassified

Análisis psicométrico mediante la Teoría de la Respuesta al Ítem: modelización paso a paso de una Escala de Soledad

2020

Self Cite

View full text Add to dashboard Cite

La teoría de la respuesta al ítem (TRI) es un conjunto de modelos psicométricos utilizados en el desarrollo, evaluación, mejora y puntuación de escalas de evaluación. Este artículo pedagógico provee un recorrido inicial acerca de cómo conducir, interpretar y exponer resultados de aplicación de modelos de la TRI aptos para opciones de respuesta politómica ordenada. Los datos utilizados como ejemplo para la modelización con TRI corresponden a la administración de la Escala de Soledad de Buenos Aires, un nuevo instrumento para evaluar la autopercepción de la soledad, a una muestra no probabilística e incidental de 509 participantes residentes en el Área Metropolitana de Buenos Aires (53% mujeres). El objetivo de este artículo es mostrar un recorrido por los pasos generales y los componentes necesarios para realizar un análisis con TRI para incrementar el acceso a esta poderosa técnica psicométrica.

show abstract

“…This structure has been supported using confirmatory factor analyses in relation to the third version of the UCLA [22] in two Turkish samples [50], and in a sample of 300 healthy Iranian adults [51]. Similarly, a sample of over 500 respondents from Argentina [52] supported this model using the second version of the UCLA [21]. Given the range of studies supporting the Russell model [22] model, we attempt to fit this model to our data, below.…”

Section: Loneliness Equivalencementioning

confidence: 80%

“…Among these packages, we used "lavaan" [79] to perform confirmatory factor analysis, following the one-factor solution proposed by [22], the two-factor solution proposed by [29], and the three-factor solution proposed by [54] (see Table 1 and [55]). We also attempted Russell's [22] bifactor model (as supported by [50][51][52] -see Introduction), where all items load on to a general loneliness factor, and in addition each item is allocated to a "positive items" or a "negative items" factor, but this did not give rise to a reliable solution, and is not discussed further in this paper. Next, we examined measurement invariance [66, [80][81][82].…”

Section: Discussionmentioning

confidence: 99%

Are we measuring loneliness in the same way in men and women in the general population and in the older population? Two studies of measurement equivalence.

Pollet¹,

Thompson²,

Malcolm³

et al. 2021

Preprint

View full text Add to dashboard Cite

Background: High levels of loneliness are associated with negative health outcomes and there are several different types of intervention targeted at reducing feelings of loneliness. It is therefore important to accurately measure loneliness. A key unresolved debate in the conceptualisation and measurement of loneliness is whether it has a unidimensional or multidimensional structure. One of the most widely used scales to measure loneliness is the UCLA Loneliness Scale. The aim of this study was to examine the dimensional structure of this scale and establish whether this factorial structure is equivalent in men and women. Methods and Sample: Two online UK-based samples were recruited using Prolific. The participants in Study 1 were 492 adults, selected to be nationally representative by age and gender, whilst the participants in Study 2 were 290 older adults aged over 64. In both studies, participants completed the UCLA Loneliness Scale (version 3) as part of a larger project. Results: In both studies, the best fitting model was one with three factors corresponding to Isolation, Relational Connectedness and Collective Connectedness. A unidimensional single factor model was a substantially worse fit in both studies. In both studies, there were no meaningful differences between men and women in any of the three factors, suggesting measurement invariance across genders. Conclusion: The findings of this study are consistent with previous research in supporting a multidimensional, three factor structure to the UCLA scale, rather than a unidimensional structure. Further, the measurement invariance across genders suggests that the UCLA scale can be used to compare levels of loneliness across men and women. Overall the results suggest that loneliness has different facets and that people can be satisfied with some aspects of their social relationships (e.g. a wide circle of friends), whilst dissatisfied with others (e.g. lack of emotionally close relationships). Thus future research on loneliness should consider treating the UCLA loneliness scale as a multidimensional scale, or using other scales which are designed to measure the different aspects of loneliness.

show abstract

Modeling of the UCLA Loneliness Scale According to the Multidimensional Item Response Theory

Cited by 12 publications

References 35 publications

The applicability of the UCLA loneliness scale in South Africa: Factor structure and dimensionality

The applicability of the UCLA loneliness scale in South Africa: Factor structure and dimensionality

Análisis psicométrico mediante la Teoría de la Respuesta al Ítem: modelización paso a paso de una Escala de Soledad

Are we measuring loneliness in the same way in men and women in the general population and in the older population? Two studies of measurement equivalence.

Contact Info

Product

Resources

About