The National Travel Survey (NTS) published annually figures about journeys and patterns of mobility of the Dutch population. For a few important variables the corresponding 95% confidence intervals based on estimated variances are published, too. This assumes that these point estimates are approximately normally distributed. In this report the plausibility of this assumption is investigated for a few number of variables by means of the bootstrap method. For most of these variables this assumption appears to be correct if the target population concerns the resident population of The Netherlands. However, prudence is called if these variables are itemized to smaller areas and/or more specific categories. The 95% confidence intervals based on the normal distribution for estimates concerning such detailed information are shifted to the left and to small.
Abstract