Measuring the effects of covariates on duration data through completely censored and lengith biased cps-like data.

This study shows the problems that arise when the estimation of the differences in the duration of unemployment experienced by workers with different characteristics is calculated using cross-sectional data and methods that do not take account of the longitudinal nature of the duration data. We propose an alternative method that avoids these problems using techniques of Statistical Survival Analysis.Length Bias; Censored Data;