One of the main applications of a BBN is the development of tools to support decision-making. In this case, an additional set of nodes was added to the network to enable impacts on livelihoods at both household and community levels to be predicted. Two nodes, respectively representing household- and community-level impacts, were linked to each of the nodes representing the change in availability of assets as a result of commercialization. Each was assigned five possible states signifying the impact on livelihoods, respectively labelled ‘Very negative’, ‘Negative’, ‘Neutral’, ‘Positive’ and ‘Very positive’. The CPTs for these nodes were adjusted so that the outputs of these nodes precisely mirrored the probability distributions of the CIFOR impact scores, when the factor nodes were instantiated with the probability distributions for all NTFP cases combined.
An interactive decision-support tool was developed using this version of the BBN, incorporating default factor scores for the 19 NTFP case studies combined, by construction of an interface in Java. This tool (the ‘CEPFOR Decision Support Tool’, CDST) has been made freely available (as a downloadable file; http://quin.unep-wcmc.org/forest/ntfp/outputs.cfm). The CDST enables predictions to be made regarding the impact of commercialization on livelihoods, by entering factor scores for the NTFP in question. The CDST also exploits one of the advantages of BBNs, by allowing impacts to be inferred even where information is lacking on particular factors. Furthermore, if information about a particular factor is uncertain, then this can be entered into the network by instantiating the relevant factor node with a probability value of less than 1.