I've been reading the comments to date with interest.
Although I admire Dr. Lavie's meticuousness in cataloguing
discrepancies and ommissions (too bad he wasn't a reviewer
on this paper), I don't feel these negate being
able to discuss the findings. For example, in figure 4,
although he's right that there is a blatant mistake
in either the numbers on the horizontal axes or
their labeling as simple level rather that interaction
coefficeints, I tend to assume that the
relationships between symptoms and events are being
accurately represented. And although, yes, it would be
interesting to know more about cause of death of
relatives, it is still interesting to see how this variable
interacts with dream recall--"violence and accidents"
rules out the natural causes that would account for most of
the variations not due to the occupation.
In response to Kathy's comments about the depression
findings, I wanted to say that since that although paper &
pencil dperessions measures have reasonable correlations
with clinical depression in a population without major
significant external losses, the items include many things
about sadness and crying that pick up normal grief reactions
also and in this population, unless you looked at a few
particular items about low self-esteem, you may have
something that is more sadness that "depression" in the
maladaptive sense. And although "anxiety" is the natural
reaction to some of these traumatic events, sadness may be
the more natural one to death of a relative and other of the
loss-related events.
I know the focus was intentionally on dream recall, but it
seems the author has only scratched the surface of a very
rich data base. I'd love to see future analyses of themes
such as Cartwright's distinction between repetitive
unpleasant dreams vs. ones that attempt mastery or Terr's of
literal re-enactments vs. incorporating traumas into ongoing
concerns in terms of whether these distinguish sypmtomatic
vs. nonsymptomatic children
-Deirdre Barrett
