Robust estimators and bootstrap confidence intervals applied to tourism spending

Statistical research carried out over the past few years evaluating important aspects of tourism has shown that supply and demand for tourism products have risen. This study focuses on various methods of evaluating a fundamental variable of tourist expenditure: average daily expenses per tourist. When analysing this variable, extreme values that invalidate the average location parameter are not uncommon. The presence of skewed values and the asymmetry of distribution justify using alternative methods for parameter estimation. Using data collected from the tourist expenditure survey taken in the Balearic Islands in 2001, this study presents results obtained from different robust location estimators, placing special emphasis on Huber and one-step's M-estimators, accompanied by calculating confidence intervals. Additionally, results were obtained by using a resampling method called the bootstrap estimation.