Fig. 4. Scatter plot illustrating the relationship between model predictions and livelihood impact scores, pooling the case studies together. Model predictions represent the probability values (represented as percentages) for the five states of nodes representing the changes in availability of assets required for livelihoods, treating different asset types separately. For these predictions, the network was instantiated with factor values for all NTFP case studies pooled together. Livelihood impact scores represent independent measures of the impact of NTFP commercialization on livelihoods assessed using the CIFOR indicators, pooling household- and community-level responses together, but treating each capital asset type separately.