Comparison of the Statistical Efficiency of Single Bounded and Double Bounded Dichotomous Choice Contingent Valuation

Contingent valuation is one of the standard methods is utilized to estimate the value of non-market goods and services. Recently, this method is used in Iran frequently. Much more studies were designed using DBDC but analyzing their data was done using logit model that is suitable for analyzing SBDC data. Also, there are some studies designed by SBDC and analyzed using logit model. The main purpose of using DBDC is to increase statistical efficiency of the estimates, but efficiency does not increase when logit model is used for its data analysis. The main goal of this paper is to analyze SBDC data using logit model and DBDC data using seemingly unrelated bivariate probit regression (SURBP) and compare statistical efficiency of these estimates. The results showed that the variance of the coefficients of all variables in the SURBP model is less than the logit model. Also, the amount of willingness to pay in SBDC and DBDC methods was estimated 6524 and 6437 Rials, respectively; and the estimated confidence interval of the willingness to pay in the DBDC method is less than the SBDC. It is suggested that if the DBDC elicitation method is used to estimate the willingness to pay, the data be analyzed using the SURBP model.