(function(doc, html, url) {
var widget = doc.createElement("div");
widget.innerHTML = html;
var script = doc.currentScript; // e = a.currentScript;
if (!script) {
var scripts = doc.scripts;
for (var i = 0; i < scripts.length; ++i) {
script = scripts[i];
if (script.src && script.src.indexOf(url) != -1) break;
}
}
script.parentElement.replaceChild(widget, script);
}(document, '
What is it about?
This article compares two tasks that can be used for the evaluation of speech quality: a speech intelligibility test and a speech comprehension test. The speech intelligibility test consists in asking some listeners to write down the words they hear, whereas the speech comprehension test implies a communication context and requires the listeners to interpret the spoken messages and to respond to them. The main result of this study is that there is only a weak correlation between the scores issued from the two tasks.
Why is it important?
For many application domains (e.g. disordered speech, room acoustics...) only speech intelligibility tests are conducted in order to evaluate speech quality. Our study shows that the relationship between speech intelligibility scores and the ability of people to understand speech in a particular context (e.g. when ordering food in a noisy restaurant) is not straightforward. In other words, speech comprehensibility in a communication context can not be directly predicted from speech intelligibility scores alone.