Interpretation of the errors on Minuit parameters

It often happens that the solution of a minimization problem using Minuit is itself straightforward, but the calculation or interpretation of the resulting parameter uncertainties is considerably more complicated. The purpose of this chapter is to clarify the most commonly encountered difficulties in parameter error determination. These difficulties may arise in connection with any fitting program, are discussed here with Minuit terminology. The most common causes of misinterpretation may be grouped into three categories:

1.

Proper normalization of the user-supplied chi-square or likelihood function, and appropriate ERROR DEF.

2.

Non-linearities in the problem formulation, leading to different errors being calculated by different techniques, such as MIGRAD, HESSE and MINOS.

3.

Multiparameter error definition and interpretation.

All these topics are discussed in some detail in Eadie et al.[5], which may be consulted for further details.