Pim van Meurs writes:
How can you claim it to be a better estimate when the same
data show inflated statistics (often 10 fold) in several
other areas as well ? How can you claim that at survey which
restricts definition of gun used in self defense ends up
finding far more than ever found before ? Not that much was
changed in the methodology to account for such a jump and
certainly the study should be compared to other surveys like
the NCS.
Kleck deals with the jump by "adjusting" the earlier estimates.
Apparently the Hart poll now implies exactly 1,797,461 defensive. gun uses.
(page 182 of Kleck's paper). Kleck seems to be impressed by how close
the new, improved, Hart estimate is to the results of his latest
survey. I am puzzled as to why he continues to attack the NCVS
estimate. Surely it would be simpler to adjust it so that it agrees
with his latest survey?