Correlation, variance inflation and multicollinearity in regression model

Tay, R 2017, 'Correlation, variance inflation and multicollinearity in regression model', Journal of the Eastern Asia Society for Transportation Studies, vol. 12, pp. 2006-2015.


Document type: Journal Article
Collection: Journal Articles

Title Correlation, variance inflation and multicollinearity in regression model
Author(s) Tay, R
Year 2017
Journal name Journal of the Eastern Asia Society for Transportation Studies
Volume number 12
Start page 2006
End page 2015
Total pages 10
Publisher Eastern Asia Society for Transportation Studies
Abstract Since observational data are often used and variables in real life are often correlated, correlations among the variables are common in transportation research. In practice, this problem is often addressed by examining the correlations among the explanatory variables and estimating the variance inflation factors. More importantly, it is a common practice to exclude variables that are highly correlated from the regression model. This study provides some simulated cases to demonstrate that including highly correlated variables, as measured by correlation coefficients and the variance inflation factor in the estimation models, will not necessarily create significant estimation problems, at least in terms of coefficient estimates. Therefore, depending on the purpose of the research, researchers and practitioners should not automatically exclude any variables from their regression models simply because they are highly correlated with other explanatory variables. Some discussions on possible strategies to deal with multicollinearity are also provided.
Subject Road Transportation and Freight Services
Keyword(s) Correlation
Variance Inflation
Multicollinearity
Regression Model
DOI - identifier 10.11175/easts.12.2006
Copyright notice ©
ISSN 1881-1124
Versions
Version Filter Type
Altmetric details:
Access Statistics: 10 Abstract Views  -  Detailed Statistics
Created: Thu, 31 Jan 2019, 11:26:00 EST by Catalyst Administrator
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us