From the Mara to Michigan: How our field observations turn into computerized data

A year ago from the day I sat down to write this post, I was in Kenya sitting in front of a den full of rambunctious little hyena cubs. Now, my love of the hyena project and the lifelong friends I made in the Mara has brought me to Michigan to help out on the data side of our research.

As a research assistant out in the field, I honed my skills at taking field observations and translating them into a consistent code of behaviors so that I could write them up in my notes and send them back to MSU. However, I had little to no idea what actually happened to the notes after that.

My job in the lab right now is to “session” the field notes from Fig Tree. Sessioning involves taking parts of the written notes and entering them into a computer database that graduate students and other researchers can use to do larger calculations. Sessioning is the backbone for data analysis in the hyena lab; it gives a reference number for every observed event that researchers record. Using this foundation, we can calculate (for example) demographics of the various clans, how the location of the clans changes over time, observation rates in the field, and when we add in other behaviors to the sessioned notes, we can also calculate things like rate of aggressions or other social behaviors. In this way, sessioned notes become a powerful research tool.

In many ways, this transition from the page to the computer feels similar to the transition from real life hyenas to observational notes in the field. It all involves translating data from one form to another with the ultimate goal of taking the complexity and unpredictability of a natural system and turning it into patterns that we can analyze and use to understand more about behavior, ecology, and evolution in the world around us.

To get a sense of how this process works all the way from the field to the database, I thought I’d illustrate the different steps of the process.

Here’s a video from one day when Hadley, Dave, and I went on observations and saw natal female Crimson and immigrant male Juba acting strangely. This was one of Hadley’s first transcriptions, and Dave walked both of us through what was happening. Crimson was a subadult female in the clan who was approaching the age at which many hyenas start to have cubs, and Juba was clearly interested in courting her, hyena-style:

However, part of our job as RAs in the field was to take these observations and turn them into notes that someone who wasn’t there would still understand. Hadley’s transcription of the event looked like this:

She listed the time and location that the events occurred at, and then any behaviors that we observed, noting extra details or when we might have missed anything. So when Crimson (recorded as CRMS in the notes) was attacking Juba, these behaviors get translated into abbreviations like lk (look) or snap, indicating an angry glare or an attempted bite, with the t-level indicating how serious the aggression was.

Finally, at MSU, we can take these notes and session them so that the database includes which hyenas were present, where they were, and a general category of what they were doing. Other graduate students and their undergraduate student assistants will pull out the specific behaviors that they need from the notes and enter them into different tables. All of the notes get printed out and put into binders in the lab like this:

Here is how we mark up the written records to be entered into the computer:

Each hyena gets circled, and the entire interaction, or session, is given a number and a code to indicate whether it was at a den, a carcass, etc. In this case, the situation is just “o” for “other” but we could still make a note that they might have been mating.

And here is what it finally looks like in the database:

The locations, hyenas, time, date, and session number are all included.

Now this session is part of the giant hyena lab database of approximately 83861 sessions in Talek, and that doesn’t even include the other clans! As it becomes part of such a long history of data it enables us to learn more about these amazing animals.