Perivolaropoulos et al [1] have argued that the residual torque data in the Eöt-Wash experiment shows evidence for an oscillating potential, which could be a signature of non-local modified gravity theories. We independently assess the viability of this claim by analyzing the same data. We fit this data to three different parametrizations (an offset Newtonian, Yukawa model, oscillating model) and assess the significance of the oscillating model using four distinct model comparison techniques: frequentist, Bayesian, and information theoretic criterion such as AIC and BIC. We find that the frequentist test favors the Newtonian model over the oscillating one. The other techniques on the other hand favor the oscillating potential. However, only the BIC test decisively favors the oscillating parametrization as compared to a constant offset model. Our analysis codes have been made publicly available.

We evaluate the statistical significance of the DAMA/LIBRA claims for annual modulation using three independent model comparison techniques, viz frequentist, information theory, and Bayesian analysis. We fit the data from the DAMA/LIBRA experiment to both cosine and a constant model, and carry out model comparison by choosing the constant model as the null hypothesis. For the frequentist test, we invoke Wilk's theorem and calculate the significance using ∆χ 2 between the two models. For information theoretical tests , we calculate the difference in Akaike Information Criterion (AIC) and Bayesian Information criterion (BIC) between the two models. Finally, we compare the two models in a Bayesian context by calculating the Bayes factor. This is the first proof of principles application of AIC, BIC as well as Bayes factor to the DAMA data and can be easily extended to the results from other direct detection experiments looking for annual modulation.

We perform an independent search for annual modulation in the recently released COSINE-100 data, which could be induced by dark matter scatterings. We test the hypothesis that the data contains a sinusoidal modulation against the null hypothesis, that the data consists of only background. We compare the significance using frequentist method, information theoretic techniques (such as AIC and BIC), and finally a Bayesian model comparison technique. Both the frequentist and Bayesian techniques reveal no significant differences between the two hypotheses, whereas the null hypothesis is slightly favored according to AIC and BIC-based tests. This is the first proof of principles demonstration of application of Bayesian and information theory based techniques to COSINE-100 data to assess the significance of annual modulation.

We reconstruct the history of reionization using Gaussian process regression. Using the UV luminosity data compilation from Hubble Frontiers Fields we reconstruct the redshift evolution of UV luminosity density and thereby the evolution of the source term in the ionization equation. This model-independent reconstruction rules out single power-law evolution of the luminosity density but supports the logarithmic double power-law parameterization. We obtain reionization history by integrating ionization equations with the reconstructed source term. Using the optical depth constraint from Planck cosmic microwave background observation, measurement of UV luminosity function integrated until truncation magnitude of −17 and −15, and derived ionization fraction from high redshift quasar, galaxies, and gamma-ray burst observations, we constrain the history of reionization. In the conservative case we find the constraint on the optical depth as τ = 0.052 ± 0.001 ± 0.002 at 68% and 95% confidence intervals. We find the redshift duration between 10% and 90% ionization to be 2.05 − 0.21 − 0.30 + 0.11 + 0.37 . Longer duration of reionization is supported if UV luminosity density data with truncation magnitude of −15 is used in the joint analysis. Our results point out that even in a conservative reconstruction, a combination of cosmological and astrophysical observations can provide stringent constraints on the epoch of reionization.

We perform an independent search for sinusoidal-based modulation in the recently released ANAIS-112 data, which could be induced by dark matter scatterings. We then evaluate this hypothesis against the null hypothesis that the data contain only background, using four different model comparison techniques. These include frequentist, Bayesian, and two information theory-based criteria (Akaike and Bayesian information criteria). This analysis was done on both the residual data (by subtracting the exponential fit obtained from the ANAIS-112 Collaboration) as well as the total (non-background subtracted) data. We find that according to the Bayesian model comparison test, the null hypothesis of no modulation is decisively favored over a cosine-based annual modulation for the non-background subtracted dataset in the 2–6 keV energy range. None of the other model comparison tests decisively favor any one hypothesis over another. This is the first application of Bayesian and information theory techniques to test the annual modulation hypothesis in ANAIS-112 data, extending our previous work on the DAMA/LIBRA and COSINE-100 data. Our analysis codes have also been made publicly available.