ABSTRACT
Identifying the locations and settings where technologies are most likely to have important effects can make the most of development or extension efforts. In the context of development and applied ecology, decisions must often be made by policy makers and donors about where to implement projects designed to improve management. Implementation in some regions may provide substantially higher payoffs to investment, and higher quality information may help to target the high-payoff locations. The value of information (VOI) in this context is formalized by comparing the benefits from decision making guided by a set of information and the results of acting without taking the information into account. We present a framework for management performance mapping and for evaluating the value of information for decision making about geographic priorities in regional intervention strategies. In our case studies of Andean and Kenyan potato seed systems, we evaluate seed health and yield information from farms, plots, and individual plant observations. We use Bayesian networks and recursive partitioning to efficiently characterize the relationship between these performance measures and the environmental and management predictors used in studies aimed at understanding seed degeneration. These analyses return the expected performance of an intervention for predictor variables mapped across the landscape. We link the scientific process and the learning cycle to the value of information assessments to support a culture of continuous improvement that informs strategic agricultural development. Assessment of the value of information demonstrates the value of science as an integral part of targeted development programs.
Footnotes
karengarrett{at}ufl.edu