<p>Using dye intensity measurements at synaptic terminals to examine neurotransmitter release leads to the problem of estimating an expectation when, instead of observing independent random variables with the same expectation, one observes, with error, independent random sums of independent variables with the same expectation--but with the number of summands in the random sums unobserved. Here, a relatively convenient nonparametric approach to estimation is presented. Data from an experiment in which cationic styrylpyridinium dye FM4-64 was used in cultured mouse hippocampal neurons are used to illustrate the approach.</p>
