One additional somewhat related question.
My observed values are actually aggregated data coming from a sample. For each observed value I know the sample size used to derive the observed value.
I want to use that sample size information to consider the uncertainty of the observed value. My initial thoughts for this were to assume each observed value comes from a normal distribution with standard deviation dependent upon the sample size and have the mean be the observed value.
I would then like to feed this information into the Gamma distribution, but it’s not clear how to do this since the values from that Normal distribution are no longer observed.
So my question is how can I include the uncertainty related to the sample size?