BISoN: A Bayesian framework for inference of social networks

Animal social networks are often constructed from point estimates of edge weights. In many contexts, weights inferred observational data, and the uncertainty around can be affected by various factors. Though this has been acknowledged in previous work, methods that explicitly quantify have not yet widely adopted remain undeveloped for common types data. Furthermore, existing unable to cope with some complexities found do propagate subsequent statistical analyses. We introduce a unified Bayesian framework modelling based on This framework, which we call BISoN, accommodate capture confounds model effects at level observations is fully compatible popular used network analysis. show how applied data downstream analyses performed, including non-random association tests regressions properties. Our opens up opportunity test new hypotheses, make full use datasets, increase reliability scientific inferences. made both an R package example scripts available enable adoption framework.