Qualitative Facial Expression Recognition Technology: How to use it

20|20 is very proud to be selected as the preferred partner for Affectiva’s new technology for qualitative research, Affdex Discovery.  Why is this important to qualitative researchers?  Put simply, it provides the first vivid, easy to use behavioral analysis tool that can be used in qualitative research in real time.  The technology actually informs the interviewer in during an interview allowing the interviewer to probe thoughts and emotions behind the reactions.

Facial recognition technology has been available for quantitative research for some time.  However, real-time qualitative is a very different method and a very different technology application.  Affectiva came to 20|20 to help them modify their quantitative facial recognition and coding platform to apply to face-to-face qualitative interviews.  20|20 tested the technology with live interviews in several settings.  Affectiva used our learning and expertise in qualitative research to develop the Affdex Discovery technology for qualitative.  Also, during the testing, 20|20 developed the logical capabilities (project set-up, support, training and rollout procedures) to ensure researchers could use the technology anywhere they wanted to apply it.

Along the way, we learned about best practices.  As a researcher, I believe it is a terrific tool for in-depth interviews, particularly for advertising testing.  The technology compliments rather than replaces current ad testing capabilities.  Here is an example of a 3-part methodology that I would use with Affdex Discovery.

1.  Ad Viewing:  Prior to the interview, the respondent should use the Affdex Discovery system to view the ad video.  Results will be stored for moderator use in the actual interview.

2.   Interview:  The interview begins with a discussion of the ad using typical questioning protocols (e.g., memorability, message recall, brand impact, etc.).  The moderator can use the Affdex Discovery technology to probe deeper.  It worked well for me to show the respondent their results from the facial coding.  I played the video while the respondent and I watched the their results graph.  When the graph showed positive or negative emotions, I stopped the video to probe specific points with the respondent as well as pointed questions around their emotional responses.

3.   Reporting:  Affdex Discovery provides outstanding opportunity for analysis and reporting.  Not only can the researcher review each individual response to the video, the responses can be aggregated for trending.  In addition, Affdex provides feedback on 7 different emotions that may have been expressed while watching the video.  Also, a researcher cannot discount the impact such a visual representation of the qualitative findings will have on the report and its acceptance by management.

Affectiva developed the facial recognition and coding technology; 20|20 developed the technology, processes and capabilities to make it available virtually anywhere. To get more information, go to our Affdex Discovery web page at: https://staging.2020research.com/tools/affdex-discovery/

One Comment
  1. You are such a gifted young man and your new adventure thrills us. You and your family and company are in our prayers. BJ/MP

Leave a reply