Application of Transformation of Variables in Remedying Heteroscedasticity in Nigeria GDP, Conditioning and Some Fiscal Variables

Abstract

Heteroscedasticity is a significant problem in regression analysis and occur in any situation where the error variance is not constant. The presence of heteroscedasticity is problematic, as changes in the dependent variable cannot be accurately attributed to individual explanatory variables. It can cause estimated coefficients to be unstable and have high variances and thus be potentially inaccurate to guide management policy. The main objective of this research work is to detect the violation of the constancy of variance in the data set and provide remedial measures to remove it in the classical linear regression model. The two estimators were compared using standard error of regression and coefficient of determination. A number of key findings are identified such as B-P and Koenker test, white test with the following statistics: p-value of 0.00, 50.031 before transformation and 0.718,0.405 and 22.338 after transformation respectively. Based on these figures, we conclude that transformed least squares (TLS) outperformed ordinary least squares (OLS) in the presence of heteroscedasticity. It should be used as method of analysis when the error variance is not constant. However, government should spend much especially on expenditure on economic service and expenditure on administration so as to increase the GDP.