Data Collection

We collect three main types of data: group scans, focal follows, and ad lib data. All of these are written down using a special “monkey shorthand” composed of 2-letter codes for different types of behaviors, 3-letter codes for names of monkeys, and additional codes for names of objects, all arranged in a complicated syntax. When the project first started 30 years ago, we narrated all of these actions onto a recording device and then wrote them out later. But now, with years of experience, we type the monkey shorthand into a special application on our phones, which puts a timestamp on each line of code.  The first letter of the line tells us what kind of data is going to follow, and this tells our “parser” (a computer program) how to divide the information into the columns of tables needed to analyze our data.
         In group scans, we write the name of the individual monkey being scanned, his activity, and then list the monkeys who are nearby.  For example the code PMBB YFFBP DCAMB 3MHH means that the monkey named Mombassa (MBB) is engaged in the activity FFBP (forage fruit, species Bromelia pinguin), and she is carrying on her back her baby AMB (Amboseli). Her sister Mooch (MHH) is in proximity category 3, i.e. 5-10 monkey body lengths away from her.  We use groups scans to get quick snapshots of what monkeys in the group are doing at a particular time.
         Our primary form of data collection is the focal follow, in which we select a single monkey to follow for a pre-determined amount of time.  Everything that this focal monkey does, and everything that other monkeys do to it, is recorded continuously during the focal follow.  In addition, every 2.5 minutes, we say what the focal monkey is doing, and which other monkeys are nearby (and how close they are).  Below is the data from a focal follow of Avocado (code name AVV), with English translation to the right:
HFMO I – H lines are about handling objects: handle fruit mora, ingest
HFMO II – handle fruit mora, ingest, ingest
HFMO D – handle fruit mora, drop
SLEMO –  S lines are general behaviors, not directed at a particular monkey. In this case it means “focal leaves the mora tree”
FABBMM – F lines are about behaviors directed by the focal animal to someone else. In this case, the focal approaches to 5 body lengths of Bedlam
FAABMM– focal approaches to 1 body length of Bedlam
FPUBMM – focal play pounces on Bedlam
MPWBMM – M lines are about mutually initiated behaviors: mutual play wrestle with Bedlam
RPBBMM – R lines are about actions directed towards the focal animal by someone else: receive play bite Bedlam
FPWBMM – focal play wrestles Bedlam
RPHBMM – receive play hit from Bedlam
RABYBB – receive approach to 5 lengths from Ybobo
RAAYBB – receive approach to 1 body length from Ybobo
RPUYBB – receive play pounce from Ybobo
FPWBMM – focal play wrestle Bedlam
FPFYBB – focal play flee Ybobo
FLAYBB – focal leaves one body length of Ybobo
FLABMM – focal leaves one body length of Bedlam
FLBYBB – focal leaves 5 body lengths of Ybobo
FLBBMM – focal leaves 5 body lengths of Bedlam
YSP – Y lines are about activities at the point sample when the watch beeps every 2.5 minutes. During this point sample, the focal is engaged in social play (SP)
XCCYBBBMM – X lines tell us the physical distance of other monkeys to the focal monkey at the time the watch beeps: the focal is in 10 body lengths of Ybobo (YBB) and Bedlam (BMM)
XBBMUU – and in 5 body lengths of Mikumi (MUU)

Finally, we use ad lib data to collect information about particularly interesting events that happen outside the context of group scans or focal follows.  Ad lib is short for the Latin expression ad libitum, which means “at one’s pleasure.” 
Ad lib data looks like this:
A (to denote that it is ad lib), a 2-letter behavior code, a 3-letter actor code, and a 3-letter recipient code. If it is a 3-way social interaction, there is an additional column for the partner.
E.g. AGTCAAMBB means that Cassie (CAA) made a threatening face (GT) at the monkey named Mombassa (MBB). 
AHOCAAMBBTQQ means that Casssie (CAA) stacked himself on top of (HO) the monkey named Totoro (TQQ) in a coalition against Mombassa (MBB).
In addition to group scans, focal follows, and ad lib, which are all types of behavioral data, we also census the monkeys each day we are with them.  Censusing just means checking off all the monkeys that are present in the group that day, and noting any changes in their health, like new wounds or evidence of pregnancies.  Keeping these records can be really helpful for a lot of reasons, as you’ll see if you read the Story of Bedlam.

Comments

Popular Posts