Lin-Log Model for Forecasting International Tourism Income in Sri Lanka: Post-War Period

Abstract

Income forecasting is an essential discipline in planning and other decision-making processes within any business. Fitting a suitable Lin-log model for forecasting international tourism income in Sri Lanka is the objective of the study. Monthly income data utilized from January 2009 to December 2013. Data obtained from annual reports of Sri Lanka Tourism Development Authority. Autoregressive Distributed Lag Model with log transformation (Lin-Log Model) was tested on forecasting income at different lags. One way Analysis of Variance (ANOVA) technique was used for overall model testing and t-test was used for individual parameter testing. Residual plots and Anderson-Darling test for residuals and the Durbin-Watson test was used as a model validation criterion. Forecasting ability of the models was assessed by considering adjusted R2 and three measurements of errors. Box and whisker plot showed no outliers in the data set. Results revealed that Lin-log model was significant at lag 1, 2 and 3. MAPE’s of the model in model fitting and verification were 14.86% and 16.63% respectively. Adjusted R2 of the model was 83.6 %. Residual plots and Anderson-Darling tests confirmed the normality of residuals. Also, residuals Vs fits confirmed the independence of residuals. Durbin-Watson statistic confirms the same. It was concluded that the Lin- log model is suitable for forecasting international tourism income in Sri Lanka. It is recommended to test Auto Regressive Integrated Moving Average (ARIMA) models also on forecasting international tourism income in Sri Lanka.