Article
The Jena Screening Test For Speech Comprehension In Sentences (JESSCom): Development and preliminary validation
Search Medline for
Authors
Published: | March 5, 2024 |
---|
Outline
Text
Dependable screening for speech comprehension is vital in addressing hearing loss, offering a quantifiable metric for researchers and clinicians. Unfortunately, conventional speech comprehension tests often demand significant time and financial resources, posing a challenge for widespread implementation. Our aim was to create a screening test for sentence-based speech comprehension tailored for cochlear implant (CI) users, with minimal resource requirements and full open access. We developed JESSCom using stimuli from the Jena Speaker Set (JESS, 120 speakers) [1], selecting 10 sentences from one young female and 10 sentences from one young male speaker, with speakers chosen based on minimal dialect ratings M=1.96 and 1.98 on a 6-point scale from 1 to 6, respectively). To achieve a balance between structural standardization of stimuli, test sensitivity, and everyday relevance of materials, we included everyday sentences both with higher (e.g.,“The dog buries the bone”) and lower (e.g., “The teacher receives the message”) content redundancy. Two JESSCom versions were created, both following the same adaptive testing procedure (starting at 65 dB, adjusting the dB level per trial based on participant performance, aiming at approx. 75% correct). One version comprises original JESS recordings at varying dB levels, the other features the same recordings at varying dB levels with added artificial breaks (~0.125 s) between words for improved parsing and comparability to the OLSA [2]. In both versions, participants are asked to listen to a sentence and repeat each word. To assess construct validity of JESSCom, we evaluated speech comprehension in 26 CI users (21 F, 5 M; MAge=57.73, SDAge=16.45) using both versions of JESSCom and the OLSA. We calculated the speech recognition threshold (SRT) in quiet for each participant, separately for both JESSCom versions, by adding the dB values of the second half of sentences (#11 to #20) and dividing the resulting sum by 10. We then correlated the SRTs we measured with both versions of JESSCom with the SRTs in quiet, as measured by the OLSA. A moderately strong positive correlation was found between the OLSA SRTs and the SRTs of JESSCom with original sentences r=.731, p< .001), indicating a consistent relationship. A strong positive correlation emerged between the OLSA SRTs and the SRTs of JESSCom with artificial breaks (r=.827, p<.001), underscoring a robust association. In addition, both versions of JESSCom showed strong intercorrelation (r=.826, p<.001). JESSCom provides a valuable and accessible tool for evaluating speech comprehension in CI users.
References
- 1.
- Zäske R, Skuk VG, Golle J, Schweinberger SR. The Jena Speaker Set (JESS)-A database of voice stimuli from unfamiliar young and old adult speakers. Behav Res Methods. 2020 Jun;52(3):990-1007. DOI: 10.3758/s13428-019-01296-0
- 2.
- Wagener K, Kühnel V, Kollmeier B. Entwicklung und Evaluation eines Satztests für die deutsche Sprache I: Design des Oldenburger Satztests. Z. Audiol. 1999;38(1):4-15.