[Babase] Demography Notes in BaBase Members

Catherine Markham babase@www.eco.princeton.edu
Wed, 20 Apr 2005 17:45:58 -0400


Demography Notes in BABASE Members

Demography notes were established to capture information in BABASE from 
one of 3 sources that were not otherwise being incorporated formally 
into BABASE (although they were often used, e.g. in assigning death or 
emigration or immigration dates).  The primary motivation for 
establishing Demography Notes in BABASE was for tracking male movements. 
  The sources are:
1. Demography note in a study group notebook
2. Other group note in a study group notebook
3. Comments or censusing of known individuals (males) in a non-study 
group (e.g. in Ositeti Group census)

Emigrations, immigrations, and deaths may or may not have associated 
notes in these sources.  Demography notes will not appear in BABASE if 
there is no note in one of these sources.  If there is information in 
these sources, a note will sometimes, but not always, be entered.  The 
question is when a note from these sources should be entered in BABASE.

We’ll come back to deaths in a subsequent email.  Here we follow up on 
the situation of emigrations/disappearances.

Disappearances/Emigrations
I was thinking presenting a hypothetical example might help work out 
some of the details for when we want to enter disappearances/emigrations 
as demography notes.  If you all are with me, consider the following:

In a particular month, a male is observed in two groups.  To start with, 
let’s make things easy and suppose both groups are censused on the same 
days at regular intervals (the 1st, 4th, 7th, 10th, etc.)  Say the male 
is present in Group A on the first and absent the rest of the month 
(according to the census records for Group A).  In Group B, this male is 
first marked present on the census sheets on the 10th (absent on census 
days up until that date).

Because Members is not going to interpolate through an absence, I think 
would get the following in the Members table if we did not use any 
demography notes:

1st – Group A (census)
2nd – Group A (interpolate)
9th – Group B (interpolate)
10th – Group B (census)

(Possibly the 3rd and 8th would also be interpolated as Group A and 
Group B, respectively – I’m not sure.)

But if we put a demography note in for the 4th and the 7th saying the 
individual was absent from Group A and placing him in “unknown” group, I 
think we get the following in Members:

1st – Group A (census)
2nd – Group A (interpolate)
3rd – Unknown Group (interpolate)
4rd – Unknown Group (demography note)
5th – Unknown Group (interpolate)
6th – Unknown Group (interpolate)
7th – Unknown Group (demography note)
8th – Unknown Group (interpolate)
9th – Group B (interpolate)
10th – Group B (census)

What’s nice about this is that we no longer have that gap in members 
between the 2nd and the 9th.  So in cases like this, I think having a 
demography note marking the disappearance of a baboon is really handy. 
But I definitely don’t think that means that you will necessarily be 
documenting all disappearances from a study group with a demography 
note.  One reason why is that an absence in one group might coincide 
nicely with appearance in another, so the baboon’s presence is already 
captured between the two group censuses.  Explaining this scenario is a 
bit more complicated, but bear with me . . .

Suppose the two groups that the male is observed in are not necessarily 
censused on the same days.  Group A is censused on the 1st and the 4th 
and Group B is censused on the 1st and 5th.  The male is present in 
Group A on the 1st but absent on the 4th.  The same male is first 
observed in Group B on the 5th (absent on census days up until then).

Without any demography notes for disappearances from Group A, I think 
Members would interpolate as follows:

1st – Group A (census)
2nd – Group A (interpolate)
3rd – Group A or B – a toss-up with interpolation (interpolate)
4th – Group B (interpolate)
5th – Group B (census)

I like the looks of that.  But if you had put in the demography note 
saying the guy was absent from Group A on the 4th and placing him as 
“unknown” for that day, I think you’d get:

1st – Group A (census)
2nd – Group A (interpolate)
3rd – Unknown Group (interpolate)
4th – Unknown Group (demography)
5th – Group B (census)

My guess is that what happens in Members without the demography note 
here is a much cleaner view of the male’s movement between groups, 
though I could be wrong.

So, where does this leave me.  For one, I should not have written in my 
original email launching this discussion that disappearances/emigrations 
do not warrant demography notes.  That is simply not the case and was a 
mistake on my part.  But, I do not think the converse is true – not all 
disappearances get a demography note.

Pending revisions to the Members table (namely not skipping any calendar 
days for an individual between birth and death) will change how the 
above scenarios play out.  I think the adjustments will be for the 
better, but I’m not sure at what point they are going to be incorporated 
into BaBase.  Is this something we’re shooting for with the conversion?