A parameter that is of no interest but must be estimated to draw inference about the parameter of interest $\theta$.

A parameter that is equal to zero under the null hypothesis.

A parameter that must be fixed to a constant value for the model to be identifiable.

A parameter that represents the error variance.