[Babase] ADLIB data -- consortships and more

Karl O. Pinc babase@www.eco.princeton.edu
Fri, 21 Oct 2005 22:27:47 +0000


I looked at the psion point sample data loader
(uppoints) and it only put grooming,
agonisim, request, and approach adlibs into
INTERACT.  It explictily does not put consortships
into INTERACT.  I belive there was a reason
for this, although whether it was technical,
perhaps related to double counting or other
sorts of how many of what are supposed to go
where rules, or whether it was procedural
related to what sort of data was expected to
be collected when, I couldn't say.

Ah, I believe I've found the problem.  The
email below explains things.  Basically,
psion consortship data does not fit the
Babase consortship model.  No doubt
we'll figure out how to pound the pieces
until they fit together, but we didn't
do so at the time.

Karl <kop@meme.com>
Free Software:  "You don't pay back, you pay forward."
                  -- Robert A. Heinlein

-------------snip-----------------
Date: Tue, 4 Sep 2001 09:45:41 -0400
To: "Karl O. Pinc" <kop@meme.com>
From: Susan Alberts <alberts@duke.edu>
Subject: Re: Consortship point samples
Cc: altj@princeton.edu
Content-Type: text/plain; charset="us-ascii" ; format="flowed"
Content-Length: 1928
Lines: 38
Status: RO
X-Status:
Status: RO

Well, the consortship record in the point sample is entered NOT as an
additional piece of babase data, rather it's because the behavior of
the focal animal is presumed to be affected by the fact that she is
in a consortship.  A more sensible way to record this piece of
information would be to have a header field that indicated whether
the sample was or was not a sample of a female in a consortship, but
this seemed like too big a change from the old, handwritten method of
data collection, and so I didn't do this when I implemented the Psion
data collection.  That is, consortships have always been recorded as
"ad lib" data attached to a behavioral sample, even though this isn't
quite what they are.

Related to this is the old versus new format of the ad lib records.
When jeanne and I did a quick look at the ad lib records that had
been collected as part of Psion points between 1999 and 2001, we
noted that they varied so greatly that it would be rather hard to
convert them easily into agonism and grooming and consort records
(which we really care about) versus all the different ways that the
observers had of saying "missed sample."  One possibility is that we
do the "extraction" of ag, groom and consort records manually from
the first two years of Psion data.  Extraction from aug 2001 on
should be trivial.

hope this helps, let me know if I've muddied the waters,
Susan

> Robert's design spec indiciates that consortship is recorded in the
> point samples.  Presently, consortship is recorded in INTEARACT as a
> single event occuring over a time interval.  How is this supposed to
> integrate with the instantainious meansurements of point samples?
> 
> Karl <kop@meme.com>