Please use this identifier to cite or link to this item:
Title: Comparison of Efficiency of OLS Regression,Logistic Regression and Discriminant Analysisfor Classification Binary Outcomes
Authors: Hatchavanich, Doungporn 
Issue Date: 2008
Publisher: Chulalongkorn University Printing House
University of the Thai Chamber of Commerce
Source: Doungporn Hatchavanich (2008) Comparison of Efficiency of OLS Regression,Logistic Regression and Discriminant Analysisfor Classification Binary Outcomes. University of the Thai Chamber of Commerce Journal Vol.28 No.1.
Journal: University of the Thai Chamber of Commerce Journal 
Abstract: The most widely used statistical methods for analyzing categorical outcome variablesare Discriminant Analysis and Logistic Regression. If a dependent variable is a binaryoutcome, an analyst can choose among Discriminant Analysis, Logistic and OLSRegression. The statistical assumptions required for Discriminant Analysis are essentiallythe same as for OLS Regression. Logistic Regression makes no assumption aboutthe distribution of the independent variables. They do not have to be normally distributed,linearly related or of equal variance within each group. The purpose of this study was tocompare the predication accuracy of the three statistical methods mentioned above inorder to highlight the most appropriate method. The results showed that the percentageof predicted value resulting from these three methods was highly correlated to thecoefficient of determination. However, the normal distribution of the predicted variableswas not examined. For every level of VIF and the percentage of qualitative variables, theDiscriminant Analysis method gave the most accurate result. Nevertheless, for data havingcoefficient determination of more than 0.5, all methods gave the same performance in theaverage of predicted value.
ISSN: 0125-2437
Rights: This work is protected by copyright. Reproduction or distribution of the work in any format is prohibited without written permission of the copyright owner.
Appears in Collections:JEO: Journal Articles

Files in This Item:
File Description SizeFormat 
328fulltext.pdf571.3 kBAdobe PDFThumbnail
Show full item record Recommend this item

Page view(s) 50

checked on Jul 11, 2019

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.