[Babase] mstatus E

Daphne Onderdonk babase@www.eco.princeton.edu
Fri, 09 Apr 2004 13:39:10 -0400


Hi everyone,

I've been pulling togather some data for Jeanne on male ages at
maturation, and we came across an issue with the use of "E" for
Mstatus.  In the protocol, an "E" in Mstatus is defined as an estimated
matured date that is 4.5 years from birth for females and 5.5 years from
birth for males (earlier for Lodge).  But I looked at several males with
Mstatus of "E", and their matured dates were not 5.5 years from birth,
so I did some investigating.  Most of the matured dates were in fact
estimated, but because they were interpolated when data was missing.
Some were estimated because only a few months were missing.  Some were
estimated because a whole year was missing, or there was extremely
little information.

So this means that "E" currently includes two different types of
estimation.  Obviously, to get calculate ages at maturation, we want to
use individuals with known, not estimated, birth and matured dates, and
we want to be able to distinguish between known and estimated in
Biograph.  So Jeanne's thinking was that we should keep Es for both
types of estimation, and change to definition of E to include both
types.

Currently, we have been interpolating when a few months of maturation
data are missing without using an E.  So the Es when just a few months
were missing are inconsistent with what we're doing now.  Jeanne and I
thought that these cases should NOT have Es.  But we need to decide a
cut-off point between what doesn't get an E and what does.  Jeanne
thought something like 4 or more months of missing data.

This issue with "E" came up when Karen and I were looking at the matured
dates of females for a question of Karl's.  We came across a couple of
cases where the matured dates were estimated by interpolation when there
was a gap in the data, and we decided NOT to enter Es in these cases so
that we didn't have Es in Biograph meaning two different things.  (See
emails re: female matured dates on 5 & 6 Feb.)  But now I see the value
of having E mean both types of estimation, so that we can distinguish
between known and estimated.  So if we decide to keep Es for
interpolation estimating, I'll have to put the ones for those females in
too.

Thoughts?
Daphne