Analysis of attrition biases and trip reporting errors for panel data

摘要: Abstract The objective of the study is to assess nature and magnitude reporting errors attrition biases in first two waves a household panel survey data set that consists weekly trip diaries. This paper investigates relationship between first-wave diary subsequent decision participate second-wave same for those households which chose participate. An econometric procedure developed accounts bias relates from waves. model system offers mechanism by relationships among mobility, errors, behavior can be statistically evaluated. It three elements: wave-one equation, probability model, wave-two equation. conditional distribution error terms these components evaluated assuming chronological dependency errors. result used develop correction term error. Results empirical application are reported together with use an efficient sample weighting factor fully utilizes information results.