Validating the CG-STYLE Measure in Family Caregivers of Individuals Living With Dementia

TL;DR

This study validates a new tool called CG-STYLE to measure how family caregivers of people with dementia provide care, showing it is reliable and useful for research and practice.

Contribution

The paper introduces and validates a new caregiving style measure (CG-STYLE) with five domains for dementia caregivers.

Findings

The CG-STYLE measure showed no significant floor or ceiling effects except for maladaptive behavioral management strategies.

The measure demonstrated convergent and discriminant validity through appropriate correlations with related and unrelated constructs.

Significant group differences were found across all CG-STYLE domains, supporting its reliability and validity.

Abstract

Family caregivers for individuals with dementia provide care based on unique caregiving styles. Using mixed-methods approaches, we developed a measure representing five caregiving style domains: understanding, adaptability, emotional expression (positive, negative), orientation to self-or-other, and behavioral management strategies (adaptive, maladaptive). We present data supporting the reliability and validity of this new Caregiving Style (CG-STYLE) measure. The sample included 209 family/friend caregivers for a community-dwelling individual living with dementia. Participants completed the CG-STYLE measure and comparator measures. We explored floor and ceiling effects (criterion ≤20%), convergent validity (correlations between similar domains should be moderate: 0.36 - 0.67 or high: 0.68 - 0.89), discriminant validity (correlations between dissimilar domains should be small: ≤ 0.3), and known groups validity (there should be significant differences between groups known to be different; determined by high versus low scores [median split] on caregiver readiness, competence, criticism, and active management). The CG-STYLE domains were devoid of floor or ceiling effects apart from a slight ceiling effect (30.3%) for maladaptive behavioral management strategies. There were moderate to high correlations between the CG-STYLE domains and known comparators supporting convergent validity, and negligible to small correlations between the CG-STYLE domains and an unrelated construct (i.e., pain intensity) supporting discriminant validity. There were significant group differences for all CG-STYLE domains (Cohen’s d’s range from 0.38 to 1.68). Findings support the preliminary reliability and validity of the CG-STYLE measure suggesting its utility in research, community, and clinical practice to target and tailor supports to caregiver’s unique caregiving styles.