|
Authored by: Anonymous on Sunday, June 24 2012 @ 01:00 PM EDT |
Dunno. Admittedly not a perfectly relevant study - but
perhaps the best available.
If your car goes missing - a bluebook value would be a
better measure.
But, if your never-sold or appraised Picasso goes missing -
it might be kind of reasonable to look at a distribution of
sales values for other Picasso's and then argue about where
your Picasso would have fit in. (House appraisers appear to
do something similar.)
Similarly, it is sort of reasonable to look at a
distribution of patent values, notice that their values are
highly skewed, and then argue that the patents-in-suit are
likely to be the most highly valued in the portfolio. OTOH,
it'd be better to do a larger study - and test for
confounding variables. Obvious ones would be patent type
and number of successful lawsuits with a given patent.
However, it is an ancient (albeit perhaps not ideal)
practice to cite results from studies that are similar to
the question of interest even if they don't exactly answer
it.*
It would also be perfectly reasonable to point out that
trying to value based on outliers is a highly fiddly
technique and likely to produce extreme results. I would
guess that judges tend to buy this argument because anything
that reduces the number of patents being considered is
beneficial.
--Erwin
*For an off-topic result, consider the China study - which,
overall, answered the question: are vegetarians healthier
than non-vegetarians? (Yes.) Please note - it did not
answer the question: Is a _vegetarian diet_ healthier than a
non-vegetarian diet? (Probably not.) It is important to
test for confounding variables.
[ Reply to This | Parent | # ]
|
|
|
|
|