We stack WMAP 7-year temperature data around extragalactic point sources, showing that the profiles are consistent with WMAPs beam models, in disagreement with the findings of Sawangwit & Shanks (2010). These results require that the source samples s
election is not biased by CMB fluctuations. We compare profiles from sources in the standard WMAP catalog, the WMAP catalog selected from a CMB-free combination of data, and the NVSS catalog, and quantify the agreement with fits to simple parametric beam models. We estimate the biases in source profiles due to alignments with positive CMB fluctuations, finding them roughly consistent with those biases found with the WMAP standard catalog. Addressing those biases, we find source spectral indices significantly steeper than those used by WMAP, with strong evidence for spectral steepening above 61 GHz. Such changes modify the power spectrum correction required for unresolved point sources, and tend to weaken somewhat the evidence for deviation from a Harrison-Zeldovich primordial spectrum, but more analysis is required. Finally, we discuss implications for current CMB experiments.
