how large is the area that undergoes anomalous processes before a large earthquake, and how long an interval does it do it on?

one answer is that such a nucleation zone may be only meters across and start to fracture only at the beginning of the earthquake. There is little evidence to contradict this idea.