Sampling in tourism: The length of stay bias

Abstract

Tourist research heavily relies on quantitative approaches, particularly on-site surveys, to gather data on behaviour, preferences, motivations, and satisfaction. While on-site surveys offer immediate and context-specific responses, their location significantly impacts data accuracy. This study highlights the bias introduced when surveys are conducted at tourist attractions rather than border points, as selection probabilities vary based on tourists' lengths of stay. The research demonstrates how this bias distorts estimations of tourist behaviour. The findings challenge the validity of past studies using non-border surveys and propose correction methods. Ultimately, the study calls for a reconsideration of empirical tourism research methodologies to ensure unbiased and reliable data collection or to weigh properly the data when the sample is biased.