Perception and Acoustics of Empathy – Results & Analysis
To reiterate, the study asked listeners to judge, based on varying vocal qualities, whether a recorded greeting (“hello”) was directed toward a face with a happy, sad, or neutral expression. The data-collecting phase of our research is now complete. Three sets were analyzed. Listeners were presented with a forced choice of either happy or sad (H/S) for the first set, happy or neutral (H/N) for the second set, sad or neutral (S/N) for the third. About twenty female participants judged each set, for a total of 60 participants. The data revealed a trend in line with our predictions: perceptual ability aligns with subtle acoustical differences. This was true for listeners asked to discriminate between a “happy-directed hello” and a “sad-directed hello,” and for those who discriminated between a “happy-directed hello” and a “neutral-directed hello.” In both of these sets, participants performed above chance, choosing the correct option more frequently than not. This tells us that the nuanced vocal—specifically pitch—variations we use as speakers when responding empathetically are noticeable even to a third party. However, participants performed at-or-below chance when discriminating between “sad-directed-” and “neutral-directed hellos,” as there was no pitch difference between them.
We are currently working preparing a poster that will show the salience of pitch in emotional expressions.
The premise of our study states that speakers will vary the tone of their voice to adjust for their perception of a listener’s emotional state. Specifically, when greeting a listener who possesses an outwardly cheerful appearance, a speaker will adjust specific characteristics of her voice, and the greeting will sound somewhat different than if she were greeting a listener with a dejected, or otherwise neutral/indifferent, look. We know by our own experiences as effective communicators that we respond to listeners—our perception of them—with variations in voice. These changes are very subtle. Our study aims to determine whether a third party listener, after hearing a pre-recorded greeting, can conclude whom that speaker was addressing – a listener with a “happy”, “sad” or “neutral” appearance? Are there elements in her voice that our participants, the third party listeners, can pick up on to distinguish greetings made to a happy-looking person from greetings made to a sad-looking person from those made to a “neutral”-looking person? Furthermore, is there anything about our participants themselves that makes them more successful at picking out these subtle differences, something about their personality, perhaps?
A lot more goes into conducting research than I had imagined! To start, I had to get my IRB certification in ethical conduct toward participants. Dr. Karthikeyan and I had to precisely prepare our materials and formulate our instructions, so that each participant receives uniform treatment. This was, of course, to avoid as much as possible any extraneous variables that have the potential to skew results. Finding participants was a challenge in itself. Initially, the response to our flyers and online posts were strong; however, as with all people-oriented tasks, scheduling conflicts arose. Although we were able to get some participants from outside Pace during the summer break, the majority of our participants will now be from the student pool at the university.
Currently, we are still in the data-gathering stage of our study. As we conclude this phase in the coming month or two, I look forward to moving on to analyzing the data—looking for patterns, determining significant variables, and gaining insight into the validity of our hypothesis. In the final stages, I hope to assist in writing a paper about our findings and presenting it at a conference in the Spring.