ABSTRACT
This chapter discusses the available statistical models and their degree of applicability to CNS synapses as typically studied, and the physiological framework in which models may be critically assessed. Heterogeneity requires more complex statistical models to interpret parameter changes at individual synapses than either a simple or compound binomial model allows, yet in many instances data has been analyzed blindly or without regards to such complicating features. Guidelines for suitability of physiological data sets have also been established, based on simulation approaches. Once these data sets were generated the Bayesian analysis was performed in a manner identical to that used for the physiological data sets. However, further simulations indicate that unequal site parameters also are resolved nearly as well as these equal site parameters, suggesting that Bayesian site analysis may be helpful in assessing whether or not a binomial model may be appropriate for an individual data set.
