where i represents the farmers sampled. Themodel parameters
are estimated by maximising this log-likelihood function with
respect to the yet unknown parameters. A likelihood ratio test
was used to test the hypothesis (H0: ρ=0). If the hypothesis is
rejected (correlated errors), the bivariate probit model with
selection gives unbiased parameters for the two equations; if
the hypothesis cannot be rejected, independent probit models
are sufficient to obtain the parameters for each equation.