Loneliness in later life: Constructing and testing a model of loneliness
Abstract
Freedom from loneliness has been considered an important component of well-being in later life. Although loneliness research has been flourishing in the last two decades, there is a severe shortage of theoretical studies. The purpose of this investigation is to construct and empirically test, via path analysis, a model of loneliness that includes background variables, a personality variable, objective social-network variables, and cognitive/evaluative variables. The present study included 126 subjects (53 males, 73 females), whose age ranged from 60 to 96. Path analysis was used to test the model, and the revised UCLA Loneliness scale was employed as the primary dependent measure. The findings indicate that marital status, income, self-esteem, perceived social support, and cognitive comparison and perceived discrepancy directly predict feelings of loneliness. Social-network size and frequency of interactions, however, had no significant effects on loneliness. The data supported the ordering of variables predicted by the model. The hypothesized model explained 52.1% of the variance in the data set, and made a valuable contribution to the understanding of loneliness. The current model attempted to combine the "needs" and "cognitive" theories of loneliness, and the results of the study provided tentative support for this integration. The results of the current study bring into question the appropriateness of using the increasingly popular Social Provisions scale to study loneliness, particularly when loneliness is measured by the UCLA Loneliness Scale. Finally, directions for future research were discussed, and strategies for treating loneliness were suggested.
Degree
Ph.D.
Advisors
Swensen, Purdue University.
Subject Area
Psychotherapy|Gerontology|Psychological tests
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.