The validity of well-being measures: A multiple-indicator–multiple-rater model.

Abstract

In the subjective indicators tradition, well-being is defined as a match between an individual's actual life and his or her ideal life. Common well-being indicators are life-satisfaction judgments, domain satisfaction judgments, and measures of positive and negative affect (hedonic balance). These well-being indicators are routinely used to study well-being, but a formal measurement model of well-being is lacking. This article introduces a measurement model of well-being and examines the validity of self-ratings and informant ratings of well-being. Participants were 335 families (1 student with 2 parents, N = 1,005). The main findings were that (a) self-ratings and informant ratings are equally valid, (b) global life-satisfaction judgments and averaged domain satisfaction judgments are about equally valid, and (c) about 1/3 of the variance in a single indicator is valid. The main implication is that researchers should demonstrate convergent validity across multiple indicators by multiple raters.