In this section
@inbook{Graham:2021:10.1016/B978-0-08-102671-7.10055-7,
author = {Graham, DJ},
booktitle = {International Encyclopedia of Transportation Volume 1 7},
doi = {10.1016/B978-0-08-102671-7.10055-7},
pages = {283--290},
title = {Causal Inference for Ex Post Evaluation of Transport Interventions},
url = {http://dx.doi.org/10.1016/B978-0-08-102671-7.10055-7},
year = {2021}
}
TY - CHAP
AB - This chapter reviews statistical methods for ex post evaluation that can be used to quantify the causal effects of transport interventions. It introduces the potential outcome model for causal inference as a framework within which to conceptualize ex post evaluation, defines its key elements, and explains the theoretical principles underpinning techniques for treatment effect estimation. The aim of these techniques is to quantify changes that have occurred due to explicit intervention (or “treatment”) net of other effects, in nonexperimental settings with nonrandomly assigned treatments. The chapter argues that these conditions tend to characterize ex post analyses of transport interventions and that the key issue for successful evaluation involve identifying and adjusting for consequent sources of bias in estimation. Contemporary causal inference approaches achieve this either through model-based adjustment or by exploiting sources of exogenous variance that are orthogonal to the treatment. The chapter reviews a number of such approaches that have featured in the recent literature and that hold potential for practical application.
AU - Graham,DJ
DO - 10.1016/B978-0-08-102671-7.10055-7
EP - 290
PY - 2021///
SP - 283
TI - Causal Inference for Ex Post Evaluation of Transport Interventions
T1 - International Encyclopedia of Transportation Volume 1 7
UR - http://dx.doi.org/10.1016/B978-0-08-102671-7.10055-7
ER -
Transport Strategy Centre
South Kensington Campus
天美传媒
London SW7 2AZ - UK
enquiries.tsc@imperial.ac.uk or +44 (0)20 7594 5995