Statistical inference

Once the probability model Φ has been defined, it is necessary, in order to deter­mine numerical values for its parameters, to introduce a statistical model. This can be defined as follows (cf. Spanos 1986: 216-8).

12.8.1 Definition

It is particularly interesting to notice that answering the second question pre­supposes that the statistical model postulated (let alone the axioms of probability theory) is valid, “and considers various forms of inference relating to unknown parameters θ". This is typical of T-theoretical terms, indeed.

12.9