Diagnostic Precision of Open-Set Versus Closed-Set Word Recognition Testing

Area of Science:

• Audiology
• Speech-Language Pathology
• Hearing Science

Background:

• Accurate assessment of hearing ability is crucial for timely intervention.
• Word recognition (WR) tasks are standard tools for evaluating speech understanding.
• Comparing the precision of different WR task formats is essential for optimizing hearing assessments.

Purpose of the Study:

• To evaluate and compare the precision of closed-set (forced-choice) and open-set (open-ended) word recognition tasks.
• To determine which task format is more effective in identifying changes in hearing ability.
• To assess the efficiency of different WR tasks under various speech degradation conditions.

Main Methods:

• Seventy listeners with normal hearing participated in the study.
• Word recognition performance was measured using closed-set (4 and 6 choices) and open-set tasks.
• Speech stimuli were degraded using either noise or a sine wave vocoder across different signal-to-noise ratios and channel numbers.

Main Results:

• Both noise and vocoder conditions produced similar psychometric functions, with closed-set tasks showing shallower slopes and higher scores than open-set tasks.
• Computer simulations and analysis of individual data using confidence intervals revealed the following efficiency order (most to least): open-set (phoneme), open-set (word), closed-set (6 choices), and closed-set (4 choices).

Conclusions:

• While closed-set word recognition testing offers practical advantages for implementation, its precision in detecting hearing changes is lower than that of open-set tasks.
• Open-set tasks, particularly at the phoneme level, demonstrate superior precision for identifying subtle changes in hearing.
• The choice between open-set and closed-set tasks should consider the trade-off between implementation ease and diagnostic precision.