Construct Validity of the Social Provisions Scale: A Bifactor Exploratory Structural Equation Modeling Approach

Area of Science:

• Psychological assessment
• Social psychology
• Psychometrics

Background:

• Existing theory suggests perceived social support is multidimensional, measurable by the Social Provisions Scale (SPS).
• Previous research shows inconsistent evidence for SPS multidimensionality, with high factor correlations.
• A novel bifactor exploratory structural equation modeling (ESEM) framework is proposed to address measurement complexities.

Purpose of the Study:

• To investigate the internal structure, gender invariance, and predictive validity of the Social Provisions Scale (SPS).
• To apply a bifactor-ESEM framework to account for general and specific factors in SPS data.
• To assess the utility of bifactor-ESEM in psychological measurement.

Main Methods:

• Utilized a bifactor exploratory structural equation modeling (ESEM) framework.
• Analyzed item responses from 376 participants.
• Examined internal structure, gender invariance, and predictive validity of SPS scores.

Main Results:

• A bifactor-ESEM representation of the SPS data was supported.
• Item thresholds and latent mean invariance for six of seven SPS factors were supported.
• Social Integration levels differed by gender (men scored higher).
• SPS scores demonstrated predictive validity for loneliness and psychological well-being.

Conclusions:

• The study provides evidence validating the Social Provisions Scale (SPS).
• The bifactor-ESEM approach is effective for analyzing complex psychometric data.
• Findings support the SPS's utility in assessing social support and its impact on well-being.