- Cite This Article as
- Corresponding Author
The study compares two statistical methods:- Discriminant analysis and the Logistic regression model in predicting Drug Offenders, Drug peddlers and Non-drug peddlers. Of the 262 cases examined for Drug Offenders, Discriminant Analysis classified the Drug Peddlers correctly (56.3%) while it recorded (84.6%) success rate in classifying the Non-drug Peddlers. In the case of the Logistic regression, it recorded (92.4%) and (97.9%) success rate in classifying the Drug Peddlers and Non-drug Peddlers respectively. The overall predictive performance of the two models was high with the Logistic regression having the highest value (95.4%) and (71.8%) for Discriminant Analysis. Among the four characteristics examined, exhibit type and age were not significant variables for identifying Drug Offenders by both methods while exhibit weight is important identifying variable for both except gender which was significant in the Logistic model. The study shows that both techniques estimated almost the same statistical significant coefficient and that the overall classification rate for both was good while either can be helpful in selection of Drug Offenders. However, given the failure rate to meet the underlying assumptions of Discriminant Analysis, Logistic Regression is preferable.
[Balogun, O.S., Balogun, M.A., Abdulkadir, S.S. and Jibasen D. (2014); A COMPARISON OF THE PERFORMANCE OF DISCRIMINANT ANALYSIS AND THE LOGISTIC REGRESSION METHODS IN CLASSIFICATION OF DRUG OFFENDERS IN KWARA STATE Int. J. of Adv. Res. 2 (10). 0] (ISSN 2320-5407). www.journalijar.com
Share this article
This work is licensed under a Creative Commons Attribution 4.0 International License.