The understanding of how households make decisions may improve the success of an economic development program and enhance targeted training efforts. Technology adoption and farm management decisions depend on household decision making. If a relevant decision maker can be clearly identified and specifically trained to meet his or her needs, the development program may be enhanced. Many approaches have been developed to help understand household decision making processes and the responses to household surveys provide the basis for many such approaches. Survey questions are often asked of a single person, and proxy responses are commonly used. Though potential bias from proxy responses is well documented, there is less information regarding the relationship between the proxy and his or her characteristics and the veracity of responses to subjective questions like who makes decisions within the household or who is in charge of major responsibilities. This paper employs the methods of mining contrast-set (Bay and Pazzani, 1999, 2001) and association rule (Agrawal et al., 1993) to answer the general question of whether and under what conditions proxy responses to survey questions are acceptable in a rural area of Ecuador. It also analyzes how factors such as gender of the respondent matters and how other factors affect the suitability of using proxy responses.

The findings show that gender matters for farm-household decision making in rural Ecuador. For instance, more male than female respondents are likely to claim that they are responsible for household decision-making. Respondents answer differently not only to some subjective questions such as who sells crops, but also to objective survey question such as the number of female workers in a family. Factors such as the age of the respondent are found to influence responses about certain activities such as preparing and applying pesticides. The pattern of responses to both objective and subjective questions as well as the effect on responses by characteristics differ by areas where the survey is conducted, etc..