Large-scale power in the CMB and new physics: An analysis using Bayesian model comparison

One of the most tantalizing results from the WMAP experiment is thesuggestion that the power at large scales is anomalously low when compared tothe prediction of the ``standard'' Lambda-CDM model. The same anomaly, althoughwith somewhat larger uncertainty, was also previously noted in the COBE data.In this work we discuss possible alternate models that give better fits onlarge scales and apply a model-comparison technique to select amongst them. Wefind that models with a cut off in the power spectrum at large scales areindeed preferred by data, but only by a factor of 3.6, at most, in thelikelihood ratio, corresponding to about ``1.5 sigma'' if interpreted in thetraditional manner. Using the same technique, we have also examined thepossibility of a systematic error in the measurement (or prediction) of thelarge-scale power. Ignoring other evidence that the large-scale modes areproperly measured and predicted, we find this possibility somewhat more likely,with roughly a 2.5 sigma evidence.Comment: A curve added in fig.1 depicting the correlation function from the pixelized map; a few comments added. Results unchanged. This version to be published in Physical Review