Box-cox transformation in regression analysis: an Illustration with telecommunications data
No Thumbnail Available
Date
2007
Authors
Adeleke, I. A
Eno, E. E
Opara, I. A
Esan, E. O
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The aim of this study was to illustrate the application of Box-Cox transformation and other remedial measures on multiple regression analysis using telecommunication data. We had an initial regression with R2 of 25% inspite of the claim by the operators that the regressor variables usually affect the response variable. An application of Box-Cox transformation and removal of leverage points improved the regression to about 82%, which makes the model more useful. We therefore, suggest that Box-Cox should be applied to data with problem of non-normality.
Description
Being text of paper presented at the 2007 Annual Statistical Conference of the Nigerian Statistical Association at Asaba, Delta state
Keywords
Statistics , Regression analysis , Telecommunication , Data
Citation
Adeleke, I. A. et al (2007) ‘Box-cox transformation in regression analysis: an illustration with telecommunications data’ presented at the 2007 Annual Statistical Conference of the Nigerian Statistical Association at Asaba, Delta state