This study, covering the period between 1968 and 1991, employed the ordinary least square (OLS) multiple regression technique to estimate the elasticities of international tourism demand from four generating countries (Japan, USA, Hong Kong and South Korea) to Taiwan for a set of potential important determinants: per capita income and population in the tourist-generating countries, relative prices, exchange rate, and trade volume within the tourist-generating countries and Taiwan. The log-linear model was used in this study to explain both tourist arrivals and tourist expenditures. The C (p), R2, Model significance level (a), Durbin-watson d statistic, regression coefficients of variables and variable significance levels are reported for each regression analysis. The criteria used to determine the "best" tourism demand models for each country are: conformity to regression and theoretical concepts and best empirical explanatory ability. For Japan, the tourist arrival model is best explained by trade volume, and the tourist expenditure model is best explained by relative prices and exchange rates. For the USA and South Korea, both the tourist arrival models and tourist expenditure models are best explained by relative prices and trade volume. For Hong Kong, the tourist arrival model is best explained by population and exchange rate; tourist expenditure model is best explained by income. / Master of Science

Thesis (MPhil (Curriculum Studies))--Stellenbosch University, 2008.