Environmental sound similarity and identification performance
Loading...
Links to Files
Permanent Link
Author/Creator
Author/Creator ORCID
Date
2017-10-25
Type of Work
Department
Towson University. Department of Audiology, Speech-Language Pathology and Deaf Studies
Program
Citation of Original Publication
Rights
Subjects
Abstract
Prior to completing the experiments, listeners received a quick hearing test. Two experiments investigated listeners’ performance related to classification and similarity ratings of environmental sounds. In Experiment 1, listeners were presented with a single stimulus to identify out of a closed set of 18 stimuli. Classification scores were above chance for all stimuli. In Experiment 2, listeners were presented with all possible pairs of stimuli from Experiment 1, after which they were prompted to choose a similarity rating using a 7-point Likert rating scale where 1 represented “as similar as possible” and 7 represented “as different as possible”. The ratings from Experiment 2 were useful in developing a two-dimensional multidimensional scaling (MDS) solution. The Euclidean map showed separation between the three broad categories of sounds: power tools, household appliances, and vehicles. Sounds from similar sources were plotted closer together than sounds from different sources. A short survey was given upon completion of the experiments for listeners to report their familiarity with and frequency of use related to each stimulus. The survey used a 7-point Likert rating scale where 1 represented “not familiar” and “never use” and 7 represented “very familiar” and “use daily”. The survey results indicated push mower, upright vacuum, and bathroom fan were the most familiar sounds; Mosquito, Mustang Piper, and Bell 212 were the least familiar. Bathroom fan, upright vacuum, and box fan were the most used while Bell 212, Mosquito, and Jet Ranger were the least used. Correlational analyses of familiarly and frequency of use indicated there was a significant moderately-strong positive correlation for all stimuli. Frequency of use and classification data from Experiment 1 had a significant and moderately positive correlation for bathroom fan, hand vacuum, and sawzall. Familiarity and classification data from Experiment 1 had a significant and moderately positive correlation for bathroom fan, Bell 212, box fan, circular saw, and weed eater.