Normative Performance Functions for the Modified Connected Speech Test.

Area of Science:

• Audiology
• Speech-Language Pathology
• Psychoacoustics

Background:

• The Connected Speech Test (CST) evaluates understanding of running speech in noise.
• An updated CST version by Saleh et al. (2020) uses a General American accent and modern recording for reduced accent effects and noise.
• This study aims to establish normative data for the updated CST.

Purpose of the Study:

• To collect normative data for the updated Connected Speech Test (CST).
• To characterize performance on the updated CST across different listening conditions.
• To investigate subjective ratings of speech intelligibility and listening effort.

Main Methods:

• 40 normal-hearing, native English-speaking adults participated.
• Speech stimuli were presented amidst multitalker babble at various signal-to-babble ratios (SBRs).
• Participants were tested in co-located and separated loudspeaker conditions, with performance scored on keyword identification and subjective ratings collected.

Main Results:

• Participants performed significantly better in the separated loudspeaker condition, showing spatial release from masking.
• Increased SBR led to improved performance and subjective speech intelligibility across both conditions.
• Higher SBR was also associated with reduced listening effort.

Conclusions:

• The study provides essential normative data for the updated CST.
• Findings characterize expected performance on the updated test, aiding clinical and research applications.
• The updated CST offers a more refined measure of speech understanding in noise.