Group: Sylvain Fichet, Grégory Moreau,… [please add up your name if interested !]

In the fits of the Higgs signal strengths based on LHC data, the theoretical uncertainties (and possibly other systematics) are usually added in quadrature to the experimental errors.

A more appropriate treatment of theoretical uncertainties - due mainly to QCD at the Higgs production level - is via a bias which finds a more natural framework, as a prior, in the Bayesian approach than in the frequentist way.

The goal here is to point out possible differences in the final determination of the Higgs couplings when comparing the Bayesian and frequentist approaches.

Analytical computation of the Likelihoods in that case :

---

Related works:

- Measuring the Higgs Sector http://fr.arxiv.org/abs/0904.3866