…continued from Which Voice Is Best?

What makes an automated voice featured in an interactive voice response system, or IVR, appealing to users?  Are there certain tones, pitches, inflections, emotions, or even genders that users perceive in a more positive manner than others?  If users do perceive certain positive and negative characteristics, what sort of causal behaviors are associated with these perceptions?

Researchers from Rice University attempted to answer these questions.  Twenty-three individuals representing a variety of different demographics (male, female, professional voice talent, amateur individuals reading from a script) were asked to record excerpts of a preselected medical survey meant to assess post-chemotherapy symptoms.  The survey included a variety of scale questions for the respondents to answer, and test subjects were instructed to listen to a variety of voice recordings and rate each voice on a personality scale.

So what did the researchers find? Based on a series of experiments, findings showed that “disclosure is not influenced by a particular voice personality or gender…This stability in rating indicates that IVR designers do not have to concern themselves with locating a particular voice talent when seeking a voice for an IVR that may involve disclosure.”  They go on to state that it is not necessary to supply multiple IVR voices based on the gender of the caller.

The authors aptly point out that while these findings appear to suggest that companies developing IVR systems should be less concerned with elements like the voice inflection or gender of the voice talent, the participants in this study are population outliers, and there are some drawbacks to the method used.

Stay tuned for part three of this three-part post, Voicing Your Opinion…