Abstract
We employ two different statistical tests to examine whether, in the framework of the Constrained MSSM, the experimentally determined values of $BR(b \to s \gamma)$ and the anomalous magnetic moment of the muon, $(g-2)_\mu$, are consistent with each other. Our tests are designed to compare the theoretical predictions of the CMSSM in data space with the actual measurements, once all of the CMSSM free parameters have been integrated out and constrained using all other available data. We investigate the value of $(g-2)_\mu$ as obtained by using $e^+e^-$ data alone (which shows a $\sim 3\sigma$ discrepancy with the SM prediction) and as obtained based on $\tau$ decay data (which shows a much milder, $1\sigma$ discrepancy). We find that one of our tests returns either a statistically inconclusive result or shows weak evidence of tension between $BR(b \to s \gamma)$ and the $e^+e^-$ data based value of $(g-2)_\mu$. On the other hand, our second test, which is more stringent in this application, reveals that the joint observations of $BR(b \to s \gamma)$ and $(g-2)_\mu$ from $e^+e^-$ data alone are incompatible within the CMSSM at the $\sim 2\sigma$ level. On the other hand, for both tests we find no significant tension between $BR(b \to s \gamma)$ and the value of $(g-2)_\mu$ evaluated using $\tau$ decay data. These results are only weakly dependent on the three different priors that we employ in the analysis. We conclude that, if the discrepancy between the Standard Model and the experimental determinations of $(g-2)_\mu$ is confirmed at the $3\sigma$ level, this could be interpreted as strong evidence against the CMSSM.
Introduction
The search for physics beyond the Standard Model (SM) often centers on the Constrained Minimal Supersymmetric Standard Model (CMSSM), a framework that addresses theoretical gaps such as the hierarchy problem and the nature of dark matter. However, the viability of the CMSSM depends on its ability to simultaneously accommodate a wide range of experimental observations. A notable tension has emerged between two specific observables: the branching ratio of the B-meson decay, $BR(\bar{B} \to X_s \gamma)$, and the anomalous magnetic moment of the muon, $(g-2)_\mu$.
While $BR(\bar{B} \to X_s \gamma)$ measurements typically favor the "focus point" region of the CMSSM characterized by heavy supersymmetric particles, $(g-2)_\mu$ discrepancies with SM predictions generally require relatively light sparticles. This study quantifies this tension using sophisticated Bayesian statistical tests to determine if the CMSSM remains a consistent theoretical framework.
Methodology
Moving beyond traditional $\chi^2$ grid scans, which may overlook theoretical uncertainties or fine-tuning issues, this analysis employs a Bayesian framework. Two primary statistical tests are utilized:
- The $\mathcal{L}$-test (Predictive Likelihood Ratio): This test evaluates the probability of observing specific values for the target observables given all other available data. It compares observed data against the model's optimal predictions in data space.
- The $\mathcal{R}$-test (Model Comparison): This test uses the Bayes factor to weigh two hypotheses: $H_0$ (data compatibility within the model) versus $H_1$ (data incompatibility).
The interpretation of the $\mathcal{R}$-test results is based on the magnitude of the log-Bayes factor, as summarized in the following table:
Calculations were performed using the SuperBayeS package and the MultiNest algorithm to efficiently compute Bayesian evidence.
Experimental Data and Constraints
The study integrates several key datasets:
- Target Observables: $BR(\bar{B} \to X_s \gamma)$ and $(g-2)_\mu$. Two interpretations of $(g-2)_\mu$ are tested: one based on $e^+e^-$ annihilation data ($\sim 3\sigma$ SM discrepancy) and another based on $\tau$ decay data ($\sim 1\sigma$ SM discrepancy).
- Global Constraints: WMAP 5-year dark matter density ($\Omega_\chi h^2$), W-boson mass ($M_W$), and LEP limits on Higgs and sparticle masses.
- Nuisance Parameters: Uncertainties in SM inputs, such as the top quark mass ($M_t$) and the strong coupling constant ($\alpha_s$), are accounted for.
Results and Discussion
The analysis reveals that the choice of $(g-2)_\mu$ data is critical. When utilizing $e^+e^-$ based values, the $\mathcal{L}$-test indicates an incompatibility with $BR(\bar{B} \to X_s \gamma)$ at approximately the $95\%$ significance level ($\sim 2\sigma$). Conversely, the $\tau$ decay data shows no significant tension ($< 1\sigma$).
A key finding is the robustness of these results across different "priors" (initial parameter distribution assumptions), including flat, log, and fine-tuning-penalized priors. The $\mathcal{L}$-test proved more sensitive than the $\mathcal{R}$-test, particularly in identifying tensions within degenerate regions of the data space. While the $\mathcal{R}$-test often yielded "inconclusive" results (as defined in Table 1), the more direct $\mathcal{L}$-test provided clearer evidence of internal model stress.
Conclusion
The study concludes that the CMSSM faces significant pressure from the combined requirements of B-physics and the muon's magnetic moment, specifically when $e^+e^-$ data is used. If the $3\sigma$ discrepancy in $(g-2)_\mu$ is confirmed, it would constitute strong evidence against the CMSSM. While more general versions of the MSSM might reconcile these data points, the constrained version appears increasingly strained. Future clarity depends on resolving the experimental discrepancies between $e^+e^-$ and $\tau$ datasets.
Summary updated