[Babase] Monitoring Points dataset
Catherine Markham
amarkham at princeton.edu
Mon Aug 27 15:29:44 EDT 2007
Hi Susan, Tabby, and Lacey,
I'm writing to let you all know that we have just completed the entry
and proofing of another historic dataset called Monitoring Points
(hooray!). The dataset spans 1982-1995 and records data for all social
groups monitored during that time period. At least at Princeton, the
datasheets are filed in their own section of the group monitoring
notebooks. The data were entered in June and July by a recent EEB grad,
Melissa Bowlin, and then proofed this month (August) by me.
Monitoring Points captures both behavioral and locational data, and is
quite similar to elements from point samples and SWERB half-hourly
readings. For each record in Monitoring Points, researchers recorded
information on: group, date, time, name of focal animal, focal's nearest
neighbor, focal's activity, food code (if feeding), and location
(typically quad given the time period of data collection, but some
coordinate locations provided in later years). Data were typically
collected each half-hour, and focal animal (usually) varied from one
record to the next. See collection instructions in Monitoring Guide
version about 1983 for details.
Jeanne and I feel that certain components of the Monitoring Points
dataset are well suited for integration into the Samples-related tables
in Babase whereas other elements mesh very well with SWERB. With this
in mind, we suggest breaking apart this dataset and weaving the
individual fields into each already-existing dataset (Samples vs.
SWERB). The dividing lines is very logical (see below). By utilizing a
unique numerical ID field, we could easily reintegrate the data if
needed. This seems the most efficient solution to maximize the database
infrastructure already established with Babase.
Specifically, the Monitoring Points elements applicable to the Samples
tables are:
Group
Date
Time
Focal
Neighbor
Activity
Food Code
The Monitoring Points elements applicable to the SWERB tables are:
Group
Date
Time
Location
Obviously the overlap of group, date, and time information creates an
issue with our database design principles (not wanting information
repeated in multiple tables). Jeanne and I feel we will eventually be
able to avoid this by keeping these fields editable in Samples, and
presenting them in SWERB through the use of a view (again, the datasets
will be easily linked through the ID field).
I'm currently working with the data fields relevant to SWERB (this was
actually the reason for pushing the entry and proofing this summer).
However, I'll maintain management of the "working copy" of the entire
dataset until one of the database managers is ready to tackle the
clean-up and integration of the behavioral data fields into Babase. This
way, if I need to make corrections to "shared" columns for group, date,
and time, I can do so easily and with reference to a single file.
Regarding the clean-up of the Samples-related fields that will be needed
prior to inclusion in Babase, the data have been entered as recorded on
the datasheets but have NOT been proofed beyond visually matching the
electronic copy to our photocopied records (in other words, no logic
testing for things like be certain that SNAME is valid, individual was
in group on specified date, individual not his/her own nearest neighbor,
etc.). Obviously, there will eventually be some need to check for
mistakes as well as consistency with codes, protocols, etc. by the managers.
Finally, I have a readme file of the entry protocols that is ready at
any time to be passed along.
Whew. I think that's it. Please let me know if you have any questions.
Otherwise, just know that the Samples data is ready for to be handed
over at any time.
Tabby, I am placing a back-up disk of the Excel file and the Word readme
file in with the other back-up CDs in the spindle by your desk.
Thanks all,
Catherine
More information about the Babase
mailing list