standardized to per hectare and converted to income.
Labor costs were not included as all farmers used family
labor in the farm operations. For each province the
means of the control plots,experimental plots and the
mean differences were computed for participants from
each province in each season. The total variable costs of
seeds,fertilizers,insecticides and fungicides,incomes
from rice yields were computed using prevailing price
rates. Herbicide costs of all farmers and all other costs
were constant in both plots. The gross margins were
thus computed by deducting the variable input costs
from the incomes in all cases.
All statistical analyses were performed using SPSS
version 11.5 (SPSS,2002). Means between the control
and experimental plots were compared using the paired
t-test. The differentials between control and experimen-
tal plots were used to explore relationships. Pearson’s
correlation and the linear multiple regression analyses
were used for hypotheses testing with the gross margins
differential as the dependent variable. The stepwise
regression method was adopted to explore model
relationships between gross margins and six independent
variables.