Behavioral Interest Identification for Farm Mechanization Development using Path Analysis and Neuro-fuzzy Models

This paper studies the behavioral side of people’s interest regarding farm mechanization development. The objectives were to identify and explain the predictor and the most important variables of perceptional and behavioral characteristics of young people to the interest in farming jobs and farm machines in a region. Path analysis and neuro-fuzzy models were developed to take advantage of both techniques to explain the causal reasoning, nonlinear representation, and the human-likeness reasoning of the imprecise behavioral and perceptional data. The data used for this research were students observed from three upper secondary schools in North Sulawesi Province, Indonesia, using questionnaires we designed. The path analysis model identifies that the gender variable is the direct positive predictor variable of the interest in farming jobs. The interest in farming jobs, the willingness to take jobs related to farming and the gender variables are the predictor variables of the interest in hand tractors. The neuro-fuzzy approach identifies that the perception of risk and the ease perception of the load of overall farming activities are the important variables for the interest in farming jobs, whereas the interest in farming jobs and the ease perception of the load of overall farming activities are the most important prediction variables for the interest in hand tractors. The models and information gathered support a behavioral consideration for incorporating it with the technical and economical farm machine selection system in such a region.