Correlation between Identification Accuracy and Response Confidence for Common Environmental Sounds

Report No. ARL-TR-8383
Authors: Kelly Dickerson; Ashley Foots; Alecia Moser; Jeremy Gaston
Date/Pages: June 2018; 21 pages
Abstract: One of the difficulties in studying environmental sound perception is determining what factors lead to successful identification. Environmental sounds as a class cover a broad range of acoustic and semantic attributes, thus challenging researchers who seek to balance the need for a representative set of environmental sounds with stimulus control and precision. The present study is one in a series of efforts to provide a baseline evaluation of a set of environmental sounds that is representative of the everyday environment. Fifteen listeners were presented with 41 different environmental sounds from six broad categories: household items, alarms, animals, human generated, mechanical, and vehicle sounds. Each sound was presented five times and participants had to generate a label for each of the samples. After typing their response, participants were then asked to rate their confidence in the accuracy of the label on a 7-point Likert scale. Participants were most accurate labeling alarms and human-generated sounds, consistent with previous studies using similar categories. Further, participants' confidence was highest when an accurate label was provided, suggesting that feelings of uncertainty were related to the ability to generate an accurate label.
Distribution: Approved for public release
  Download Report ( 0.325 MBytes )
If you are visually impaired or need a physical copy of this report, please visit and contact DTIC.

Last Update / Reviewed: June 1, 2018