Emotion Facial Coding Timeseries
The primary output from facial coding is a set of time-series for the 6 Eckman emotion facial codes (Happiness, Surprise, Puzzlement, Disgust, Fear, Sadness).
A timeseries for one participant contains measurements of each of these codes every 200 milliseconds during Stim exposure.
We aggregate these timeseries over all participants to identify group trends and effects such as sudden shifts.
We translate this data into single statistics for the video by looking at peaks in the time series (e.g., the peak level of happiness). We look at peaks because of Daniel Kahneman’s peak-end rule: people tend to remember (and be influenced by) the emotional peak of an experience. For example, when you go on holiday you will remember the jet-skis but less so the airport.