OPINION27 March 2012

Putting emotions in context

Biometric data gives researchers the power to measure emotional reactions. But ABC TV’s Justin Fromm reminded delegates at the ARF’s Re:think conference in New York of the importance of using these tools alongside more conventional methods.

Biometric data gives researchers the power to measure emotional reactions. But ABC TV’s Justin Fromm reminded delegates at the ARF’s Re:think conference in New York of the importance of using these tools alongside more conventional methods.

Fromm, who leads sales and strategy research at ABC, shared results of research on TV shows and ads conducted at Disney’s advertising research lab in Austin, Texas, using electrodermal activity (EDA) measurement to gauge emotion, eye-tracking to see where people were looking, and follow-up surveys to get their explicit responses.

In the first study, which looked at the effect of product placement and advertising in an edition of Extreme Makeover: Home Edition, the survey results chimed with the biometric data. Participants were shown three edits of the show with different elements included or removed in each. EDA showed a stronger emotional response to the advertised product when the placement was included as well as ads, and, sure enough, surveys showed higher reported levels of awareness and purchase intent.

But a second study on a trailer for Modern Family showed how results could diverge. Eye-tracking results suggested that a text banner at the top of the screen attracted more attention than a banner at the bottom, or no banner at all. But survey results suggested that fewer people, rather than more, had actually noticed the banner when it was at the top. “People seemed to be looking at the banner but not recalling what was there,” said. More ‘looking’, as it turns out, doesn’t necessarily mean more attention.

A third study that compared people’s responses to three different hosts of another show, revealed how biometrics could lead you astray if taken out of context. One host got much higher levels of emotional engagement than the other two, according to EDA measurements, which seemed like good news. But when asked to rate the hosts in surveys, participants preferred the other two. The problem is that EDA doesn’t tell you whether an emotion is positive or negative – and it seems that what it registered in this case was a feeling of dislike for the host.

It all goes to show that, while biometric data can be hugely valuable, it needs to be triangulated with other measures. As Fromm said, never underestimate the importance of “smart analysis on the back end”.

@RESEARCH LIVE

0 Comments