Psychologists and sociologists usually interpret happiness scores as cardinal and comparable across respondents, and thus run OLS regressions on happiness and changes in happiness. Economists usually assume only ordinality and have mainly used ordered latent response models, thereby not taking...