Good listeners: cocktail party cognition

07 October 2016

Interview with

Daniel Oberfeld, University of Mainz

Cocktail party experimentImagine for a moment that you've just sat down with someone for a quiet chat. It's just the two of you, so the only sound you're going to hear is the voice of the other person. Meaning that following what they're saying isn't at all tricky. But now, let's transport ourselves to a busy restaurant or even a cocktail party. Now, your ears are being assailed by the conversations of many others. Yet, remarkably we manage to zone out that interference and continue to focus exclusively on the speech of the person we're conversing with. But some people find this easier to do than others including people with perfectly normal hearing. So could it be because they are less good at paying attention. Chris Smith listened intently to experimental psychologist Daniel Oberfeld...

Daniel -   My research focus is auditory perception. Studies during the last 10 years or so showed that persons really differ a lot in the capability of understanding in an acoustically difficult situation. In particular, these difference even apply to listeners with normal hearing. That was the motivation for our study to gain a closer look on this and to identify potential factors which cause these differences between persons.

Chris - So what did you do and who did you do it to?

Daniel - We conducted these experiments in 50 relatively young listeners with normal hearing. Our basic idea was that we need to look not only at basic hearing capabilities but also to more cognitive processes because singling out a single person's voice from an acoustic soundscape requires selective attention. So we use a so-called virtual acoustic setup where the subject has headphones then we recreate a spatial sound scene where one listener is presented from directly in front of the listener and there's a target speaker. This target speaker produces very simple sentence like, "Peter buys 8 wet knives." And then there are two other speakers, so-called interfering or distracting speakers which are located slightly to the left and to the right. They produce similar sentences and the challenge for the listener is to ignore these interfering speakers and then to repeat the words this target speaker produced.

Chris - What happened? Do people perform differently?

Daniel - Absolutely. So, we saw a pretty large variability in percentage correct in the speech identification task. The percentage of correctly identified words ranged between less than 50 per cent to almost 100 per cent.

Chris - How do you relate their accuracy to their ability to pay attention to the task?

Daniel - Right. We used a very simple acoustic task namely intensity discrimination where you listen to two pure tones and one of the pure tones is a slightly higher sound pressure level. And you always have to identify the louder sound essentially. After each of these two tones, an additional tone is played.

Chris - You've effectively got to hold both of those sounds in working memory and then disregard the effect of that subsequent masking sound in order to make a valid comparison between the two stimuli, haven't you?

Daniel - Absolutely. So that is something which is a situation where you need selective attention.

Chris - If you now take the results of the people who are very good or very bad at doing those discrimination tasks and you ask, are they the same people who do or don't do well on the cocktail party effect? Do you see a relationship?

Daniel - Yeah. That was the interesting and also predicted finding. The important conclusion of our study is that cocktail party listening, the ability to understand speech and noises indeed related to hearing capabilities but also, we found a significant influence of a more cognitive capability to focus attention on a target stimulus while there are other informations around which you need to ignore. There seems to be an important influence of higher processing stages as we call it the cognitive processing in the sense of selective attention.

Add a comment

This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.