Student Project - Creating a Custom Expression for Engagement

How do you keep customers engaged? How can a movie trailer or a commercial trigger engagement? That is what every marketeer wants to know. But how do you know if someone is engaged? It would be very useful if you can read this from someone’s face. That is exactly what Dayenne Sarkol-Teulings, a graduate student in Applied Cognitive Psychology tried to find out.

Watching trailers while measuring engagement

As part of her studies at the University of Utrecht in The Netherlands she did an internship at Noldus, the exclusive distributor of FaceReader for Windows. Her project involved researching engagement extensively and running an experiment and validation study. In the experiment, participants watched several different trailers while their emotional response was analyzed. Interestingly, those that were most engaged showed more neutral expressions. In addition, some action units were activated less.

Creating a custom expression

She then used the new custom expression module of FaceReader to develop a custom expression for engagement. This module allows you to create your own expressions based on Action Units and other inputs (see a simple example in the image below). Creating a temporal algorithm for engagement was difficult, but it was a very relevant learning experience with many interesting aspects for follow-up analyses remaining. This project also shows how FaceReader Online and FaceReader can complement each other.

For a more elaborate discussion of the results, see her guest blog on the Noldus website. Are you interested in using FaceReader Online in your student project? Let us know .