ABSTRACT
Many statistical analyses are based on a specic model for a set of data, where this consists of one or more equations that describe the observations in terms of parameters of distributions and random variables. For example, a simple model for the measurement X made by an instrument might be
X = θ + ε
where θ is the true value of what is being measured, and ε is a measurement error that is equally likely to be anywhere in the range from −0.05 to +0.05.
