Bias on the fit parameters

Roofit and RooStats
1

Dear Experts,

I am performing a simple signal +Bkg fit using RooFit. My signal is modelled with asymmetric double-sided CB function (defined in AsymBifurcatedCB.cpp ) and the background is with polynomial function. I am check the bias on the fit parameters from the few thousands of toy. I observe the mean and sigma of the signal is mainly biased by more than 20%. So my question is, are the observed bias from the signal model definition or from the Root fit method?

Many thanks in advance for your help.

I use ROOT 6.30/04.
The fitting code: MassFit_study_more_onlyTwoComp.C

Bests
Chandi
AsymBifurcatedCB.cpp (2.5 KB)
AsymBifurcatedCB.h (1.4 KB)
MassFit_study_more_onlyTwoComp.C (6.7 KB)

2

Dear @cpkar ,

Thanks for reaching out to the forum! Let me invite our statistical experts @Jonas @moneta to aid you with your question.

Cheers,
Vincenzo