To estimate the model, we used a bivariate probit analytical technique. A bivariate probit model is a method often used in analyzing two binary choices and allows one to assess whether these choices are correlated with each other. The bivariate probit model provides consistent estimates of parameters of the two binary choices, as do the estimates derived from a simple univariate probit model, which examines each choice model separately rather than jointly. The bivariate method gains efficiency because one might expect the error terms across the two equations to be contemporaneously correlated (Baum, 2006).