PRINCIPAL COMPONENTS ANALYSIS OF NIGERIAN ECONOMIC VARIABLES
Eze-Emmanuel, Peace & Ette Harrison Etuk
Department of Statistic
Rivers State University, Port Harcourt
Email: adelepeace678@gmail.com and ettetuk@yahoo.com
ABSTRACT
Principal Components Analysis (PCA) of Nigeria economic variables was done to decide on the most important variables to be considered in determining those variables that have positive effect on Nigerian economy. It is important to determine the significant proportion of those variables that contributed to the Nigerian Gross Domestic Product because it helps reveal especially with regards to revenue generation by the government. This research work examined the performance of these variables using quarterly Nigeria Gross Domestic data from 1981Q1-2013Q4. The methods used are Principal Components Analysis (PCA) and factor analysis (FA) multivariate technique. Using Minitab 17 statistical software, the data were examined; summary statistics are as follows: covariance matrix, correlation matrix, standard deviation, Eigenvalues, Eigenvectors, transformation of the sample onto new subsample (Sorted Unrotated and Rotated Factors), computing the principal components and finally plotting the graphs. Our results showed that 24 economic variables from 31 variables have almost perfect (positive) effect on the economy and identify the sorted rotated factor loading as the better (appropriate) method, when using economic data variables.
Keywords: Eigenvalues, Eigenvectors, Transformation, Unrotated and Rotated Factors, Principal Components Analysis (PCA) and factor analysis (FA)