Skip to content
Snippets Groups Projects

Draft: add trials factor clarification

Open Andrew Toivonen requested to merge andrew.toivonen/userguide:trials_factor into main
1 file
+ 6
1
Compare changes
  • Side-by-side
  • Inline
+ 6
1
@@ -217,7 +217,12 @@ targets **significant** candidates from all online pipelines (**RAVEN**).
We thus conservatively estimate a trial factor of **6** for :term:`stellar mass`
:term:`CBC` searches and **3** for Burst searches. Similarly we have two
independent :term:`SSM` :term:`CBC` searches (**GstLAL SSM** and **MBTA SSM**).
Hence we use a trials factor of 2 for :term:`SSM` search alerts.
Hence we use a trials factor of 2 for :term:`SSM` search alerts. The significance
threshold in the section above can be found by taking the estimated rate of false alarms
found just below and diving by the trials factors reported in this paragraph. As
previously stated, the reported false alarm rates for events do not include this trial
factor, but the significance threshold employed does. This keeps the estimated rates
of false alarms the same regardless of the number of searches.
The estimated rate of false alarms marked as **significant** is :math:`3.9
\times 10^{-7}` Hz (one per month) for :term:`CBC` searches
Loading