[Babase] Conversion errors for SOD and FRA
Catherine Markham
amarkham at princeton.edu
Wed Feb 15 13:20:21 EST 2006
Hi all,
I just had a chance to get back into the remaining 2 male conversion
errors (see my email to the listserv on 9 September 2005). Leah's text
researching issues and documenting what changes should be made to the
database is inserted below. Note that my comments after Leah and
discussed the errors on the phone are added as well.
Catherine
1. SOD
ERROR: SOD: 1993-06-11: Statdate of live individuals must be =
1993-10-20, the last census date
Details: In members SOD is found in group 1.1 from birth (03/02/1990)
until 2nd last day that he shows up in members (19/10/93). On last day
in members (20/10/93) he is in group 1.3. Statdate is listed as
11/06/93. SOD was censused in the field in 1.3 on 11/06/93. However, he
was also field censused in 1.3 before this time and after (until
26/01/95). The field censuses for February 1995 do not look complete.
Not sure if SOD was there or not. SOD is not found in Demography notes,
OGN, or OG censuses through 2003.
Solution: Statdate: Change to 26/01/95 * see note
Members and Census to delete: NONE
Members and Census to change: Change to grp 1.3 between (and including)
01/06/92-19/10/93, change origin to “R” on 01/06/92 * see note
Members and census to add: add records between (and including)
21/10/1993 -26/01/95 placing SOD in grp 1.3. Status of the new rows in
census will be “R”, and origin of the new rows in members will be “R”
Biograph.status: NO CHANGE
NOTE: The original statdate of 11/06/93 is definitely incorrect. He was
in 1.3 until 26/01/95. This is the last complete census I could find for
Proton’s.
SOD was flip-flopping between 1.1 and 1.3 a lot around this time but we
have very scattered data. [Text for "We need a detailed note in demog
explaining that this was during the fission" deleted per conversation
with Leah on 15/02/06]. He was born during the fission and his mother
died when he was 1.5 years old. A solution is to place him in grp 1.3
(Proton’s) as of 01/06/92. This is a somewhat arbitrary date (picked by
Susan) and this is definitely a stop gap solution. There is very little
data during this time period and there are incomplete censuses. This
situation will be remedied if the fission is eventually accounted for in
more detail in Babase.
2. FRA
ERROR: FRA: BIOGRAPH.Statdate cannot be less than
DISPERSEDATES.Dispersed unless individual is alive
Details: Disperse date (29/12/93) is after statdate (20/12/93). This
disperse date most likely occurred due to a typo. The statdate is
calculated as the midpoint between 20/10/93 (group censused that day)
and 19/02/94 (the next time the group was censused). FRA was censused
as absent on 19/02/94. So statdate is 20/12/93 but disperse date was
listed as 29/12/93.
Solution: Statdate: Change to 20/10/93 * see note
Members and Census to delete: delete rows after 20/10/93 for members.
Not
sure how you deal with the census rows that have “A”? (Delete rows
between 20/10/93 and the first “A?)”
Biograph.status: Change to alive (0) instead of dead (1).
Biograph.dcause: Delete dcause of 8
Biograph.disperse: delete * see note
NOTE: Statdate is 20/10/93 (the last day censused as present) and
disperse date is 20/12/93 (the midpoint between last day censused
present and the first day censused absent). Susan decided to calculate
the disperse date this way. It is more consistent with how she has
calculated them in the past. This is during a time period where there is
very little data (no census data in between October 1993 and February
1994). Because Babase will not allow a disperse date after a statdate we
will not have a disperse date for FRA.
Catherine's comments: The current file for disperse dates is NOT in
Babase at the momment - it exists as a seperate file maintained at Duke.
Leah is going to delete FRA's disperse date from that file. I did not
need to make any changes to the disperse date in Babase, but did make
the necessary changes to MEMBERS, CENSUS, and BIOGRAPH.
More information about the Babase
mailing list