Limited Dependent Variable Models

Course details

statistics seminar | level: intermediate | register now
for questions related to this event, contact ugent@flames-statistics.com
affiliation: Ghent University

Abstract

The aim of this seminar is to present and discuss models for continuous dependent variables that are restricted or have limited ranges. Central to our discussion will be the concepts of truncation and censoring.
Truncation occurs when the observed data in the sample are only drawn from a subset of a larger population we are interested in. The sampling of the subset is usually based on a value of the dependent variable. e.g A study of the determinants of incomes of those whose incomes are above a given poverty threshold. In this case, only households with income above that threshold will be part of the sample.
Censoring on the other hand occurs when the values of the dependent variable are restricted to a range of values. As in the case of truncation the dependent variable is only observed for a subsample. However, there is information about the whole sample. e.g. Household purchase of durable goods in which low expenditures for durable goods are censored to a zero value.
The models for truncated and censored data we will discuss include:
• Tobit model (using censored data)
• Truncated regression (using truncated data)
• Heckman Selection model (sample selection model)
A demonstration of how each model is fitted and interpreted will be done using the R software.

Prerequisites

Basic knowledge of linear regression models

Background readings

Cragg, J.G. (1971) Some Statistical Models for Limited Dependent Variables with Application to the Demand for Durable Goods. Econometrica, 829-844

Fee

Free

Venue

Ghent University

Instructor

Emmanuel Abatih