In additions to the methodology concerns you raised, I have to question the use of a "linear" trend line, particularly when one axis is arbitrary buckets of une

In additions to the methodology concerns you raised, I have to question the use of a "linear" trend line, particularly when one axis is arbitrary buckets of unequal size.