Effect of reward in a new localization test method for children under five.

D.A. McCartney, C.J. Church & B.U. Seeber
MRC Institute of Hearing Research, Nottingham, UK

Absolute localization ability in children under 5 has rarely been measured. 
One of the reasons for this is the difficulty of keeping the child engaged 
long enough in the task to obtain sufficient data. This was addressed by 
developing an uninstructed game‐like localization task in which children must 
find the location of sounds. Correct localization responses, defined as a head 
or eye movement toward the sound’s location, are reinforced with visual rewards. 
These are presented on video walls surrounding the child. This study 
investigated how varying the location of the visual reward affected the number 
of correct responses obtained. The study used three visual reward strategies: 
(I) the reward always presented at 0 deg (front), (II) the reward presented at 
the sound’s location, and (III) the reward varied randomly ±20 deg about the 
sound’s location. Preliminary analysis of data from 24 children (15–56 months) 
shows a significant effect of reward strategy and age. More correct responses 
were generally obtained using reward strategies (II) and (III). Reward strategy 
(III) is preferred for use in a refined localization test because the reward is 
not presented at a fixed set of locations, helping to avoid visual learning of 
the test directions.