Analysis of Human Observations

Careful Analysis of Aviary Data

A lot of behavioral data exists that can be used to test hypotheses or identify new ones for future experiments. Often this comes down to wrangling old data and looking at it in new ways.

In order to prepare for the data we would get from our Smart Aviary I worked with long-time collaborator David White to identify prior years of behavioral observations that could provide a template for the sort of data we would receive. Dave has decades of scan-sampling data, documenting pairwise interractions in large, social flocks. This was a dream data set to work on.

One of the questions that interested me was the function of female ‘chatter’ (repetitive vocalizations that females produce in a vareity of contexts.). It was often assumed that chatter served to strengthen pairbonds, however, I noticed early in the data that females chattered quite often to non-pairbonds.

Then lab member H. Luke Anderson wondered whether chatter might actually represent a form of signal jamming, and we’ve worked together to combine focal observations, analysis of old data, and new experiments to test this idea. Our paper just got accepted to Animal Behaviour and is going through final formatting revisions, so expect it early 2021.

This has also led to some fun ways to visualize (or sonify) behavioral data. (Check out NestPlotter and DataSound repos over on code.) Having spent so much time organizing and parsing this old behavioral data, I’m now following up on some remaining questions, such as what predicts pairbond formation, and how males and females respond to aviary disruptions.