'How are you feeling?' Assessing reporting bias in a subjective measure of health by quantile regression

In this paper we investigate reporting heterogeneity in the Visual Analogue Scale (VAS) when it is used to measure current health status in cardiovascular patients. We provide a new framework to identify reporting heterogeneity using quantile regressions. EQ-5D responses are used as a proxy to control for objective health. The objectiveness of this generic measure is supported by other measures of actual health. The data comes from a Norwegian, health-related quality of life study. We find substantial evidence of reporting bias in VAS related to gender and education. For some quantiles we observe reporting heterogeneity related to age and weight.