Background: The study of well-being is becoming a priority in social sciences. The Day Reconstruction Method (DRM) was developed to assess affective states. The aim of the present study was to validate an abbreviated version of the DRM designed for administration in population studies, and to assess its test-retest properties.

Principal Findings: 1560 adults from Jodhpur (India) were interviewed using an abbreviated version of the DRM, and a week later they were re-interviewed using the original long version of the DRM, after which the abbreviated version of the DRM was compared with the original version. A regression model considering interaction terms was employed to analyse the impact of sociodemographic characteristics on net affect. Test-retest reliability was assessed, and found to be moderate. Positive affect showed more test-retest reliability than negative affect, while net affect had more temporal stability than U-index. The affect of sets A, B, and C, taken together, had a moderate predictive ability compared with the affect obtained using the full version of the DRM: AUC = 0.67 for positive affect; 0.66 for net affect; 0.61 for negative affect; and 0.60 for the U-index. Household income, gender, and setting all had a significant impact on net affect.

Conclusions: Net affect and positive affect showed moderate temporal stability, whereas negative affect and the U-index showed fair temporal stability. Evaluating the affective state using the abbreviated version of the DRM provides a profile of the population similar to that of the full version. The results provide considerable support for using the short version of the DRM as an instrument to measure subjective well-being in large population surveys.