UCL DIVISION OF PSYCHOLOGY AND LANGUAGE SCIENCES

Speaker-Controlled Variability in Children's Speech in Interaction

• Home
• Overview
• Aims & Methods
  • Example of the 'spot the difference' game
  • Examples of speech in the noisy conditions
• Progress
• Publications
• Public engagement
• Taking part
• Contact Us
• Getting to Us

Aims & Methods

We aim to recruit a total of 120 children and adolescents aged 9-14 years. We will examine development in three age groups, 9-10yrs, 11-12 and 13-14, with 40 participants in each group.

Our research questions are as follows:

Study 1

Do children/young people adapt their speech when talking to someone who is listening to them through noise, even if they do not experience the noise themselves? Do they do this in a way which is most effective for the specific kind of interference they have to overcome? Finally, are they able to do this in a similar manner to adults?

We will use a technique which is designed to collect conversational speech from participants and which has been used with adults in our previous research. Two participants play 'spot the difference' games (you can see pictures and listen to an example of children doing this task here). As they are in two different rooms and communicate through headsets and microphones, we are able to vary the quality of the signal for each player. There are three different conditions: no interference, i.e. a clear signal for both players, a babble noise condition, where one of the players hears the other in a background of multiple talkers (as if speaking to someone in a crowded room), and a vocoder condition, where one player's speech is distorted to sound as if heard through a cochlear implant.  You can hear what the noisy conditions sound like here.  The roles of listening and speaking through noise are reversed within each condition so that participants experience both.

We will record and analyse the global acoustic properties of participants’ speech to see whether they differ between noisy and clear conditions. We will also look at whether these changes are specific to each type of noise, or whether they are the same. Finally, we will compare the kind and degree of changes children made in response to interference to those made by adults in our previous study.

Study 2

What does the capacity to adapt their speech to noisy conditions rely on in children and adolescents? Do those who vary less in the way they produce speech find this an easier task? Or is it more to do with how well children/young people themselves are able to understand speech in noise?

The stability of speech production will be examined by recording participants naming pictures, and by carrying out detailed acoustic analyses on the repetitions. Detailed phonetic analyses will also be applied to certain key words in the 'spot-the-difference' game. A computer game tests the ability of participants to recognise speech in noisy conditions.

Statistical analyses will be used to evaluate the relationship between speech production, speech recognition in noise and the degree to which participants change their speech when communicating through noise.

Study 3

What is the effect on speech clarity of the acoustic changes made in response to noisy conditions?

Because we want to know how much of an effect the acoustic changes have on speech clarity, we will ask adult listeners who are not familiar with this study to rate anonymous extracts of participants' speech in terms of clarity. We will compare these ratings with the degree to which children/adolescents changed the acoustic properties of their speech when having to overcome noise.