Measurement invariance of a household water insecurity metric in Greater Accra, Ghana: Implications for test-retest reliability.

Abstract

The mitigation of household water insecurity is recognized as an important component of global poverty alleviation, but until recently was difficult to measure. Several new metrics of household water insecurity have been proposed and validated, but few have been field-tested for reliability in diverse contexts. We used confirmatory factor analysis to test the psychometric equivalence of one such metric-the Household Water Insecurity Experiences (HWISE) scale-across two survey waves administered 18 months apart in similar climatic conditions among households in a peri-urban community outside of Accra, Ghana. The HWISE metric was not equivalent across survey waves, which may be attributable to the metric itself, sample size, subtle instrumentation changes, or other unobserved factors. Test-retest reliability may also be difficult to achieve given the dynamic nature of household water use, and we discuss the implications of using household water insecurity metrics as longitudinal measures of well-being in global anti-poverty programs.