Strategies for specification search as a cause of bias and inaccuracy of parameter estimates An important problem in model development, illustrated by MonteCarlo simulation and Bootstrap

As we all know, model predictions and estimation results are erroneous. Accuracy thus is a key issue of modelling – not only to obtain, but to properly describe.

The typical tool we would use for description of model accuracy is standard errors of obtained parameter estimates. However, standard errors are only designed to illustrate a smaller part of those model errors that may arise from the complex process of developing a transport model. This paper discusses and investigates some such limitations in relation to the errors that may arise from specification search. Initial analyses, based on a combination of a real data set and simulation tools, shows that there may be considerable inaccuracy and bias caused by systematic factors that are outside the scope of standard error.

Despite the fact that the concrete example, and implicitly much of the discussion, relates to models for discrete choice applied to transport demand, the general conclusions would apply also to a vast range of other types of modelling.

Karin Brundell-Freij, Lunds Universitet, Inst för Trafik och samhälle

Danske Keywords:
Modellering, Modellval, Specifikation, MonteCarlo simulering, Bootstrap, Bias, Standardfel

Engelske Keywords:
Modelling, Model selection, Specification search, Monte Carlo simulation, Bootstrap, Bias, Standard error


Link: Hent paperet som PDF-fil