TL;DR
This paper investigates how our assumptions about unknown parameters (priors) and the choice of experimental data influence our conclusions about the fundamental parameters of the Constrained Minimal Supersymmetric Standard Model (CMSSM). They find that current data is not yet strong enough to definitively pin down these parameters, highlighting a significant tension between different experimental measurements.
What Problem Does This Paper Address?
The Standard Model (SM) of particle physics has been incredibly successful, but it doesn't explain everything, like dark matter or the hierarchy problem. Supersymmetry (SUSY) is a leading theoretical extension, and the Constrained Minimal Supersymmetric Standard Model (CMSSM) is a popular, simplified version. To test these theories, physicists compare experimental data with theoretical predictions. However, when analyzing complex models like the CMSSM, inferring fundamental parameters from data is challenging. This paper tackles two key issues:
- Prior Dependence: In Bayesian statistics, we start with "priors" – our beliefs about parameter values before seeing data. For complex models, the final conclusions can be heavily influenced by these initial assumptions, making it hard to say what the data itself is telling us.
- Observable Influence and Tension: Different experimental observables (like measurements of particle masses or decay rates) constrain the model in different ways. This paper explores how combinations of these observables affect our conclusions and, crucially, investigates tensions where different observables seem to point to contradictory parameter values.
Methods
The authors employ a Bayesian statistical framework, using a sophisticated sampling algorithm called MultiNest. This algorithm is an advancement over previous methods, significantly increasing computational efficiency.
- Bayesian Inference: This approach combines prior knowledge with likelihood (how well the data fits the model for given parameters) to produce a posterior probability distribution for the parameters.
- MultiNest Algorithm: This is a novel algorithm for exploring complex parameter spaces and calculating the Bayesian evidence. It's more efficient than traditional Markov Chain Monte Carlo (MCMC) methods, especially for high-dimensional problems.
- Profile Likelihood: As an alternative to the full Bayesian posterior, they also analyze the profile likelihood. This method maximizes the likelihood over "nuisance" (less fundamental) parameters, offering a potentially less prior-dependent view.
- Information Content (KL Divergence): They use the Kullback-Leibler divergence to quantify how much information each observable provides about the CMSSM parameters.
Datasets
The paper uses a comprehensive set of experimental data and theoretical constraints relevant to the CMSSM. These include:
- Collider Data: Limits on Higgs boson and superpartner masses from experiments like LEP and Tevatron.
- Cosmological Data: The relic abundance of dark matter from the WMAP satellite.
- Flavor Physics Data: Measurements of rare B-meson decays (e.g., BR(B → Xsy)) and the anomalous magnetic moment of the muon (g-2)μ.
- Electroweak Precision Data: Measurements of W and Z boson properties.
Key Insights and Ideas
- Prior Dependence is Significant: The choice of prior (e.g., flat vs. logarithmic distribution for parameters) has a substantial impact on the inferred CMSSM parameter values, indicating that current data is not yet strong enough to fully override these assumptions.
- Dark Matter Dominates Constraints: The cosmological dark matter abundance measurement provides the largest portion (at least 80%) of the total constraining power of all observables.
- Tension Between Observables: A key finding is the tension between the anomalous magnetic moment of the muon (g-2)μ and the decay rate of B → Xsy. The former favors lower superpartner masses, while the latter favors higher masses (the "focus point" region). This tension significantly complicates parameter inference.
- Profile Likelihood as a Robust Measure: The profile likelihood appears to be less sensitive to prior choices than the full Bayesian posterior, offering a more robust assessment of favored regions.
Key Results
- CMSSM Parameters Not Well-Constrained: Despite using a wide array of data, the authors conclude that current experimental constraints are not yet sufficient to draw statistically robust conclusions about the CMSSM parameters.
- Preference for Lower m1/2: In general, values of m1/2 (a key parameter related to SUSY particle masses) below 2 TeV are preferred.
- (g-2)μ vs. BR(B → Xsy) Tension: The (g-2)μ constraint strongly favors lower masses, while BR(B → Xsy) favors higher masses, leading to a compromise region that is discrepant with at least one of them.
- Promising Prospects for Detection: Despite the uncertainties, the paper maintains that prospects for detecting dark matter and discovering superpartners at the LHC remain encouraging.
Limitations and Future Work
- Data Insufficiency: The primary limitation is that current data is not sufficiently constraining to overcome prior dependence and resolve tensions between observables.
- Future Work: The authors emphasize the need for better data to definitively test the CMSSM and other SUSY models. They also suggest that their formalism and MultiNest algorithm can be readily applied to a wider class of models. The paper also notes a recent study with differing conclusions, highlighting the need for further detailed comparisons.
Summary updated