Response Shift and Measurement Invariance of Chronic Stressor Scale

Abstract

Introduction:

Troxel’s Chronic Stressor Scale (CSS) is widely used in survey-based comparative research to investigate racial or gender differences in stress-health relationships, but its measurement invariance remains untested. This study aimed to examine the measurement invariance of CSS across age, gender, race, ethnicity, and temporality.

Methods:

This analysis used the longitudinal data from the Health and Retirement Study (N = 2,263). Model fit was evaluated using infit/outfit statistics, person/item separation, and reliability indices. Differential Item Functioning (DIF) was assessed using Rasch’s Rating Scale Model.

Results:

CSS demonstrated strong item reliability but poor person reliability. Significant DIF was identified in housing stress between Black and White Americans, in self-health stress and occupational stress between younger and older groups, and financial stress between 2012 and 2020.

Discussion:

Despite strong item reliability, CSS’s significant DIF across race, age, and temporality suggests caution when directly comparing composite or item scores across diverse populations and time points.