A study of factors affecting the adoption and use of computer software in Canterbury (New Zealand) and Florida (Uruguay) dairy farming

This study is about understanding farmers' information management behaviour, particularly in relation to the software adoption and use process, and consequently how farm information systems might be improved. Two dairy farming communities, one in Canterbury (New Zealand) and the other in Florida (Uruguay), were used to study and test a theoretical model developed to explain farmer behaviour. The research method was based on studying both farmers using and not using software for information management, asking them the reasons for using software, and relating farmers' behaviour to the farmers' and farm characteristics. On average, software-adopting Canterbury farmers are more educated, younger, milking bigger herds, involve more (non-family) people in their management, use different information sources, and present more developed information skills than Florida farmers. Results from the path analysis showed that farm software adoption behaviour can be explained similarly for both Canterbury and Florida regions, using a transactional model. In this explanation, farm and farmer characteristics contribute to explaining software use both directly and indirectly through mediating variables (farmer goals and information management practices). In a situation where software adoption involved a low number of farmers (17%), factors related to age, experience and education seem to have a strong influence on software use. On the other hand, in a situation where software use involved a large number of farmers (64%), it appears that factors related to farmer personality, such as the learning style, have more influence on software uptake. In both cases, farm size had a strong influence on the software use decision. It seems that location differences (New Zealand and Uruguay) mainly affected software use through the farmers' education levels and the better economic environment for promoting investment and farm growth in the dairy industry in New Zealand. Additionally, the high standard of New Zealand fiscal requirements also creates incentives for financial software use. A model including three factors related to software development was able to explain 42% of the variation in the "successfulness" ranking assigned by Canterbury farmers to their software packages. Software "successfulness" is positively related to the new systems similarity to the past work environment, and its user-friendliness, though only the last relationship shows statistical significance. On the other hand, the relationship between software success and "decision system matching" is negative and statistically significant. For the Florida case, the proposed factors explained very little of the variation in the "successfulness" ranking assigned by farmers to the software packages. This may be due to the small number of observations, or because of the lack of standards in software use compared with Canterbury users. In both cases, an univariate analysis between software usefulness and its theoretically related factors with the explanatory variables used to explain software use proved to be useful in understanding software usefulness. For those wondering whether to improve their information systems, the research results suggest that the key element is their current information systems. Research results also confirm the need for developers to focus more on software users. This involves not only user information requirements, but also farmers' characteristics, such as the education levels, predominant learning approaches, and current information management practices. Overall, this research has proved that the farmer her/himself is a very important part of the successful computer use equation, and that each farmer will be different, though training is likely to help overcome some of these differences