Measuring relational wellbeing: construct validity in pre-COVID-Era UK; generalizability across COVID-lockdown-Era India, Greece, and UK

TL;DR

This paper introduces and validates a new measure of relational wellbeing, showing it works well across different countries and times, including during the COVID era.

Contribution

The paper presents a new eight-item relational wellbeing measure and validates its one-dimensional structure across multiple cultural and temporal contexts.

Findings

Exploratory and confirmatory factor analyses supported a one-dimensional structure for relational wellbeing after removing a problematic item.

The one-dimensional structure showed consistent fit across pre-COVID UK and post-COVID India, Greece, and UK samples.

UK participants reported significantly lower relational wellbeing compared to participants in India and Greece during the lockdown era.

Abstract

In the present studies, we examine the construct validity and criterion-related validity of a previously unpublished, eight-item measure of relational wellbeing.

First, in two pre-COVID-Era pilot studies within the UK (n’s = 207 and 146, respectively), results of exploratory factor analyses revealed that—with the possible exception of one item regarding close relationships—the items assessed individual differences along a single dimension (i.e., relational wellbeing), rather than two distinct dimensions (i.e., social connections and close relationships). Second, in an initial pre-COVID-Era main study within the UK (n = 192), results of confirmatory factor analyses provided support for the hypothesized one-dimensional factor pattern, although the same problematic item from the pilot studies continued to under-perform relative to the other seven items.

In a subsequent COVID-Lockdown-Era main study across India (n = 205), Greece (n = 354), and the UK (n = 390), results of confirmatory factor analyses established that—after omitting the same problematic item that had surfaced in the preceding studies—a one-dimensional factor pattern provided equally satisfactory fit for the three samples.

Although we had not set out to test a priori hypotheses regarding mean similarities or differences in relational wellbeing among our COVID-Lockdown-Era studies, results of an analysis of variance revealed that persons within the UK scored significantly lower in relational wellbeing than did persons in India or Greece.

As noted above, one particular item repeatedly performed poorly in factor analyses; this item ideally should be dropped from the relational wellbeing scale in future research.