 |
|
|
|
|
Download full report (130kb)
Summary of the report
Using Computer-based Technologies to Support Farmers' Decision Making
A report for the Cooperative Venture For Capacity Building
July 2007
RIRDC Publication No 07/104 RIRDC Project No CSW-38A
Executive Summary
What this report is about
Many Australian farmers
own and use computers, yet decision support software is not integral to
the management of more than a fraction of Australian family farms, despite
the availability and affordability of such packages. This study aimed to
generate insights to help developers of decision support software avoid
marketing failures associated with low adoption of these technologies.
Who is the report targeted
at?
The report is targeted at
developers for computer based decision support systems, and those who fund
and use such systems.
Background
A portfolio of projects
known as FARMSCAPE employed an approach that involved scientists engaged
directly with groups of farmers to explore matters of tactical risk management
and strategic planning in dryland cropping systems. Integral to this was
the use of a computer-based cropsimulation model known as APSIM. Two key
strategies have been tried to deliver the FARMSCAPE approach more widely,
and in a cost-effective and commercially-sustainable manner: a) FARMSCAPE
Training and Accreditation and b) Yield-Prophet® on-line. There are
no guarantees that these strategies will result in widespread use of simulation,
and a recent contribution to the diffusion literature (Moore, 1999) provided
the stimulus for an investigation into the adoption potential of cropping
systems simulation.
Moore (1999) proposed two distinct markets for complex technology products, separated by a ‘chasm’ between the early adopter (also termed visionary) and early majority (pragmatist) adoption categories.
The chasm refers to a breakdown in social referencing between the two markets. Social referencing is a process in which one person utilizes another person’s interpretation of the situation (or technology) to formulate his or her own interpretation of it (Feinman, 1992). This behaviour arises when there is uncertainty concerning the innovation and when one’s own intrinsic appraisal processes cannot be used. In referencing, one person serves as a base of information for another, and it is a key process in technology diffusion – i.e. the process in which an innovation is communicated through certain channels, over time, among the members of a social system (Rogers 2003). The ‘chasm’ refers only to discontinuous technologies – i.e. technologies that require the adopter to change their behaviour or modify infrastructure in order to use them.
If the ‘chasm’ theory applies to the use of cropping-systems simulation, the implication is that successes with innovative farmers will not automatically result in widespread use of this technology, and there is a risk that the pragmatist market will be left behind. This exposes the possibility that cropping systems simulation, which has been highly valued by some collaborating farmers, will remain as niche products, only attractive and accessible to a very small fraction of farmers, and possibly not providing the critical market volume to allow agribusiness to viably retain this as a commercial service.
Aims
In the context of the two
commercial delivery approaches introduced above, this study aimed to provide
new knowledge about market segments and evaluate the significance of Moore’s
‘chasm’ in the diffusion of cropping systems simulation. Likewise, the
‘chasm’ was also explored in a third case study, which examined the use
satellite technology designed to support pasture management decisions.
The technology is branded ‘Pastures from Space’.
The study also aimed to identify the critical issues for effectively implementing computer-mediated decision support among a sufficient segment of the farming community to enable a viable commercial agribusiness service.
Methods used
The project adopted a case-study
methodology and interviewing was the key method used for data collection.
The research was conducted in two stages, and the report is structured
accordingly.
Stage 1 of the research explored the adoption of cropping systems simulation and ‘Pastures from Space’ satellite technology through three case studies. In addition to these focus technologies, a number of other complex, or computer-based, technologies that were believed to fit the description of a ‘discontinuous technology’ were explored, including personal computers, precision agriculture technologies, and other decision support packages. The research entailed classifying farmers according to their visionary / pragmatist adoption category and investigating influences on their adoption decisions.
Stage 2 of research, planned after the first was completed, focused solely on the use of Yield Prophet® in four states of Australia. This stage of research investigated the cognitive, in additional to the social, processes influencing adoption.
Key findings
Was a ‘chasm’ in social
referencing evident?
The test of Moore’s ‘chasm’
thesis is – are there some technology users that potential users of the
technology would not reference? The findings are inconclusive. Not all
pragmatists agreed that it was a prerequisite to reference other pragmatist
farmers before making a decision to adopt the technologies investigated
and there were cases where some self-assessed pragmatists referenced visionaries.
There were, however, individual cases that provided a degree of support
to the notion of chasm in social referencing between visionaries and pragmatists.
Is farmer-to-farmer social
referencing likely to be an effective diffusion mechanism for the technologies
investigated?
‘Adoption’ by interviewees
of the focus technologies depended on the sense that the ‘virtual world’
created by the technology is relevant to the physical world it represents
and an experience that outcomes are significant to farm management practice
– i.e. benefits sufficiently outweigh the costs (including non-monetary
costs) of adopting. The findings indicated that the process of social referencing
was not likely to be sufficient as a means to facilitate farmers’ appreciation
of the relevance and significance of the key technologies investigated.
In relation to the focus technologies, they key conditions for relevance were that the technology adequately represented the structure and behaviour of the biophysical production system and that the technologies’ output adequately relates to that experienced on the adopter’s own farm. In other words, relevance was established in the context of one’s own farm and the technology had to be credible and flexible. A personal learning experience (albeit socially facilitated in group events) also appeared to be a factor in achieving relevance.
Although a social referencing process can communicate potential relevance, resulting in an attitude of openness towards the innovation, it cannot be relied on to provide the subjective appreciation that goes the next step to establish the conditions for relevance described above. This launched a second stage of enquiry for this study, which sought to identify what interventions, if necessary, could facilitate the cognitive experiences for farmers that expedite adoption of technologies.
Value of cropping systems simulation in commercial use
These correspond with
3 of the 4 paths for model-based information systems identified by McCown
(2002). In relation to Path 1, Yield Prophet’s value is the provision of
a flexible tool for managing climatic uncertainty by forecasting production
outcomes in relation to contemplated tactical management alternatives.
To date, this has largely involved Nitrogen (N) management, and there has
been little use of Yield Prophet for other management decisions (e.g. pre-season
planning) – a situation largely explained by the limited promotion of the
latter. While demand for tactical management support would be expected
on a continuous basis, it is questionable whether Yield Prophet will generate
enough value to warrant ongoing use by subscribers if it is used solely
for N management. Although N is an important management decision – it is
a costly but necessary input, and the only in-season management option
identified by interviewed farmers for dealing with climate variability
– farmers varied in the degree to which their experience and judgment was
augmented by Yield Prophet’s explicit presentation of probabilities of
certain crop outcomes occurring. The interviews revealed that Yield Prophet
was not the only method for informing the N management decision. Not all
farmers received the same value from Yield Prophet – for some, the value
was a low marginal benefit, whereas other reported high value.
Among the clearest and most convincing reports of the value of Yield Prophet’s probabilistic yield forecasts were those from subscribers who previously had no way of making structured comparisons between yield outcomes under different seasonal climate forecasts. While this points to a ‘unique and compelling value’ for Yield Prophet, there was poor understanding and/or trust in confidence in seasonal climate forecasts to guide tactical management decisions. This presents a challenge for Yield Prophet, but not a barrier given that interviewees clearly viewed seasonal climate forecasts as a component that could be separated from the rest of Yield Prophet.
Is cropping systems simulation
cognitively compatible with the way farmers see their farming environment?
The results demonstrated
that while interviewees had no conceptual representations of the farm production
system that were fundamentally incompatible with Yield Prophet’s representation
of the production system, there was variation in the way interviewees mentally
represented and measured the performance of their farming systems. Some
interviewees clearly articulated a change in conceptualisation of their
production systems that was a result of and facilitated their appreciation
of Yield Prophet – this illustrates the nature of Path 2, above. Experienced
Yield Prophet users and those who provided a clear description of a valuable
Yield Prophet experience typically discussed the production system using
objective, scientific measures. The shifts and differences in production
system conceptualisation can be represented along a continuum between two
forms of cognition – analysis and intuition.
Analysis or analytical thought involves working with an environment represented in abstract terms through a data interface, and this is required for Yield Prophet use. This means that for Yield Prophet to make sense to farmers they must, to some degree, appreciate the logic (i.e. conceptual constructs, though not the mathematical representation) used by APSIM. In addition, to participate in setting up a simulation of a crop and to discuss simulation outputs, familiarity with quantitative measures is often necessary, and it may involve the farmer dealing with terms and measurements not normally used in the discourse of practical farming. The findings suggested that developing analytical understanding – i.e. scientific representations of the production system – among farmers in the Yield Prophet domain has facilitated an appreciation of its relevance.
Role of consultants and
farmer groups in technology diffusion
Generally interviewees did
not delegate crop management decisions in the Yield Prophet domain to consultants.
However, to varying degrees, interviews revealed consultants’ abilities
to compensate for farmer deficiencies in analytical understanding and to
facilitate farmer learning in the key steps of soil characterisation and
monitoring, conducting APSIM runs, interpreting crops yields in relation
to water and nutrient supply, and interpreting model output, presented
as probabilities, for farmer action.
The research supports the view that farm advisers are obvious candidates for delivering simulation as decision support in relation to Path 1. In relation to the value of Yield Prophet to facilitate farmer learning (Path 2), previous FARMSCAPE activities have identified that decision support and learning is most effective in a participatory process that combines the strengths of practical knowledge and scientific knowledge, and this was also evident in this study. Given the high cost of providing facilitation on an individual-subscriber basis, group processes appear to be an effective way of introducing these technologies and providing user support.
Recommendations
Farm advisers are key
candidates for delivering simulation as decision support, and could be
targeted as delivery agents by developers of DSS. These results confirm
the key learnings from the FARMSCAPE experience.
 
|