"Product" Versus "Process" Measures in Assessing Speech Recognition Outcomes in Adults With Cochlear Implants

Abstract

Hypotheses: 1) When controlling for age in postlingual adult cochlear implant (CI) users, information-processing functions, as assessed using "process" measures of working memory capacity, inhibitory control, information-processing speed, and fluid reasoning, will predict traditional "product" outcome measures of speech recognition. 2) Demographic/audiologic factors, particularly duration of deafness, duration of CI use, degree of residual hearing, and socioeconomic status, will impact performance on underlying information-processing functions, as assessed using process measures.

Background: Clinicians and researchers rely heavily on endpoint product measures of accuracy in speech recognition to gauge patient outcomes postoperatively. However, these measures are primarily descriptive and were not designed to assess the underlying core information-processing operations that are used during speech recognition. In contrast, process measures reflect the integrity of elementary core subprocesses that are operative during behavioral tests using complex speech signals.

Methods: Forty-two experienced adult CI users were tested using three product measures of speech recognition, along with four process measures of working memory capacity, inhibitory control, speed of lexical/phonological access, and nonverbal fluid reasoning. Demographic and audiologic factors were also assessed.

Results: Scores on product measures were associated with core process measures of speed of lexical/phonological access and nonverbal fluid reasoning. After controlling for participant age, demographic and audiologic factors did not correlate with process measure scores.

Conclusion: Findings provide support for the important foundational roles of information processing operations in speech recognition outcomes of postlingually deaf patients who have received CIs.