Assessment of the suitability of grafted model in forecasting rice consumption trend in nigeria.

Abstract

This paper was designed to examine the predictive power of grafted polynomial functional form in forecasting rice consumption in Nigeria with a linear model used as bench mark. Data on rice consumption trend in Nigeria from 1961 to 2011 elicited from the United States department of Agriculture foreign Agricultural services were utilized in this study. It  was observed that rice consumption did not correlate linearly with trend over the entire sample period and this necessitated the grafting of a Linear – Quadratic – Linear model. The ex-post forecasts of the grafted model had lesser forecast error and were closer to the observed values than that of the linear model which had greater forecast error. Therefore, the predictive performance of the grafted model is more reliable in forecasting future trend of rice consumption in Nigeria for planning purposes.