Sample Abstract

Can the web-based survey be employed on empirical studies?


A major concern regarding the use of web-based survey in empirical studies lies on the reliability of information. Economic values estimated using such technique, under certain circumstances, can provide the estimates that are associated with some level of strategic and self-selection bias. We employ a stochastic frontier model to calibrate the online survey information.


This study examines whether the web-based surveys can be calibrated to obtain reliable estimates in empirical studies.

Data and Method

The data for this study is collected using online and onsite survey methods. Slightly over 8% observations are gathered using onsite survey method. A total of 2,691 responses are collected using both types of survey methods. The set of questionnaire contains the information on individuals characteristics as well as other information related to the recreational site, including average cost incurred in recreation and related activities.


The likelihood ratio test for the presence of one-sided error term supports the presence of inefficiency on web-based data but not in conventionally used face to face interview. The result shows the variance associated with the one-sided error term is highly significant. This supports the idea of calibrating online values. We estimate the frontier expenditure using a log-linear form of the function. Most of the explanatory variables are contributing significantly to the predicted value of recreational expenditure for all categories of the cost. The impact analysis result shows that the most benefiting sector of the local economy was art and entertainment sectors which are directly related to the recreation.


The preliminary analysis shows the presence of inefficiency on online survey data. Our result suggested environmental valuation studies to consider direct and indirect effects of visitor's spending on local economic sectors while providing a numerical value to the natural resource.