Effect of set size and method on speech reception thresholds in noise

Area of Science:

• Audiology
• Speech-Language Pathology
• Psychoacoustics

Background:

• Speech reception thresholds (SRTs) are influenced by signal and response uncertainty.
• The standard definition of SRT as a fixed percent-correct score becomes problematic when set size (M) varies.
• This variability affects the interpretation of threshold measures across different set sizes (M=2 to 16).

Purpose of the Study:

• To investigate the impact of set size (M) on speech reception thresholds (SRTs).
• To compare two different testing strategies for determining SRTs under varying set sizes.
• To evaluate the influence of set size on the signal-to-noise ratio (S/N) at threshold.

Main Methods:

• SRTs were measured in 80 dB SPL white noise for word sets of 2, 4, 8, and 16 words.
• Two procedures were employed: one using a fixed percent-correct criterion and another using d' = 1.00.
• Participants were 12 young women with normal hearing and limited prior audiological testing experience.

Main Results:

• When using a fixed percent-correct criterion, the S/N at threshold was significantly dependent on set size (F=3.333, p=0.031).
• Conversely, when threshold was defined by d' = 1.00, the S/N at threshold was independent of set size.
• This indicates that d'-based measures offer a more stable assessment across different set sizes.

Conclusions:

• For clinical SRT testing with set sizes smaller than recommended guidelines, criterion-free measures are preferable.
• Using a d'-based criterion ensures SRTs are independent of the tested word set size.
• This approach enhances the reliability and comparability of SRT measurements in audiology.