| |
 |
| Paper: |
Setting the Statistical Significance of Detections of Non-Gaussianity in the WMAP Data |
| Volume: |
371, Statistical Challenges in Modern Astronomy IV |
| Page: |
18 |
| Authors: |
Cayon, L. |
| Abstract: |
Several works have claimed the detection of deviations from Gaussianity in the WMAP data. Different test statistics have been used in the multiple analyses carried out in real, spherical harmonic and wavelet spaces. The statistical significance of these detections has, in most cases, been set either by considering a single configuration or by defining a χ2 function that takes into account the multiple tests performed. I. Szapudi raises a very important question in his contribution. He introduces the application of the False Discovery Rate as a method for massive hypothesis testing in CMB data analyses. Could this be a viable method to set the statistical significance of the claimed non-Gaussian detections?, is it possible to define a method to set the statistical significance taking into account all tests performed?. We will comment on these issues and will present some examples of the application of False Discovery Rate in this context. |
|
|
|
|
|
