Last modified: 2018-05-28

#### Abstract

In standard quantile regression (QR), quantiles are estimated one at the time. An alternative approach, which is referred to as *quantile regression coefficients modeling* (QRCM), is to describe the functional form of the regression coefficients parametrically. This approach facilitates estimation and inference, simplifies the interpretation of the results, and generates more efficient estimators. Moreover, thanks to the imposed parametric structure, it makes it easier to estimate quantiles in situations involving latent variables, missing or partially observed data, and other complications arising in survival analysis, longitudinal data analysis, and causal inference, where applying standard QR proves difficult and computationally inefficient. We describe the method, discuss applications, and illustrate the R package *qrcm*.

#### References

Frumento P, Bottai M (2016). Parametric modeling of quantile regression coefficient functions. *Biometrics*, 72 (1), 74-84, doi: 10.1111/biom.12410.

Frumento P, Bottai M (2017). Parametric modeling of quantile regression coefficient functions with censored and truncated data. *Biometrics*,, 73(4), 1179-1188, doi: 10.1111/biom.12675.

Frumento P (2017). qrcm: Quantile Regression Coefficients Modeling. R package version 2.1. https://cran.r-project.org/package=qrcm