Discriminant Analysis of Unemployment and Literacy Rate by State in Nigeria

Abstract

This paper used Fisher's linear discriminant functions to classified and determined if unemployment and literacy rate in Nigeria by State differs. The multivariate data covariates investigated are unemployment and literacy rate in Nigeria states. The data were classified using low and high rates. Fisher's linear discriminant functions, Standardized Canonical Discriminant (SCD) Function, and Group Centroids (GC) Function were obtained. The result of Wilks' Lambda statistics show that the equality of group means between low and high unemployment rate is significant at 5% but the literacy rate is not significant. This result indicates that the unemployment rate in Nigeria by state differs significantly while the literacy rates are similar by state. The classification results show that 18 states were correctly classified with a high level of the unemployment rate, while 18 states are accurate with a low unemployment rate. However, only one state was misclassified, using discriminant Score functions p-values.