Does it matter how happiness is measured? Evidence from a randomized controlled experiment

Abstract

A continuous and a discrete rating scale were implemented for a single item happiness question in a representative online survey. A randomized controlled experiment enables unique analyses on data quality and distributions of happiness scores. The continuous measure is found to provide more information by overcoming the endpoint aversion present in discrete measurements. Furthermore, findings raise doubts about inferences drawn on correlates of discrete happiness scores. Evidence is provided that the gender happiness inequality depends upon gender specific question design effects.