學齡前聽損幼兒在聽覺作業的聆聽費力度主觀感受及客觀生理指標

研究目的：本研究欲了解華語聽損幼兒在不同聽覺作業下的聆聽費力度，藉由測量幼兒在不同訊噪比的聽覺作業過程中的聆聽費力度主觀感受及客觀生理指標，探討聽損幼兒在不同訊噪比的聽覺作業中，其正確率、反應時間、費力度主觀感受及客觀生理指標的變化是否受到訊噪比的影響，並檢視聽損幼兒各項反應是否與聽常同儕相當，以更全面評估聽損幼兒的聽能表現。研究背景：聽力檢查能測得聽損幼兒的聽察覺反應及語音聽辨分數，評估聽損幼兒輔具效益的第一步為評估其聲場聽力圖，然而即使是學齡兒童，由於專注力較短暫，在完成聽力圖評估後，再進行語音測驗來交互比對實為困難；但若單純以正確率來評估，亦容易錯估學生的聽能表現，且不同作業難度會使幼兒的聆聽費力度改變，因而影響學生聆聽時的表現。然而目前臨床及聽語復健並未針對幼兒之聆聽費力度進行量測，無法得知聽損幼兒在進行評估或復健時的表現是否受到聆聽費力度的影響。故本研究探討學齡前聽損及聽常幼兒在進行聽覺作業時的聆聽費力度，並分析聆聽費力度的主觀感受及客觀生理指標之間的相關性及預測力，希望及早偵測學齡前聽損幼兒在聽覺作業下的主觀及客觀聆聽費力度，以利臨床人員將其與聽力圖及其他表現整合，更全面地評估聽損幼兒之聽能表現。研究方法：十名4~6歲之學齡前聽常幼兒參與預試，另外25名聽常及25名聽損幼兒參加正式測驗，以配對方式控制兩組之語言、年齡等人口學變項。測試內容包含純音聽力檢查、三種訊噪比之下（安靜、SNR 0、SNR-5）兩項聽覺作業的表現（區辨正確率、辨識正確率、區辨反應速度、辨識反應速度）、聆聽費力度主觀感受尸（問卷）、聆聽費力度客觀生理指標（膚電、膚溫、心率），以了解兩組幼兒在進行聽覺作業時之主客觀聆聽費力度為何，並探討費力度主觀感受、客觀生理指標及聽覺作業表現之間的相關性及預測性。研究結果：聽損幼兒在區辨及辨識作業中的正確率均顯著較聽常幼兒低，且反應時間較慢，顯示聽損幼兒雖然與聽常同儕有相當的語言理解能力且配戴合適的輔具，其聽覺表現仍不如聽常同儕。雖然兩組幼兒在聽覺作業的表現有顯著差異，但兩組幼兒的費力度主觀感受卻無顯著差異；兩組幼兒在區辨作業中的三個訊噪比之主觀費力度兩兩相比，主觀費力度感受無顯著差異，表示兩組幼兒在進行區辨作業時，對於不同訊噪比的主觀感受雷同。在辨識作業中，聽損組在三個訊噪比之主觀費力度兩兩相比，均有顯著差異，聽常幼兒僅在安靜環境與SNR-5相比的情況下，主觀費力度感受有顯著差異，其他情境則無顯著差異，結果顯示出兩組幼兒對於不同訊噪比的主觀感受不同。而本研究中的客觀生理指標則顯示出聽損幼兒在區辨及辨識作業中的三種訊噪比的「膚電反應」顯著較聽常幼兒高，說明聽損幼兒雖然主觀感受與聽常幼兒相似，但其聽覺作業表現較差，且客觀生理回饋指標也呈現出較高的聆聽費力度。另外，兩組幼兒的心率與膚溫均無顯著組間差異，顯示心率與膚溫在評估幼兒在區辨及辨識作業之下的聆聽費力度較不敏感。斯皮爾曼相關係數結果顯示，「膚電與區辨正確率」之間的相關性在區辨作業的安靜環境及SNR0都達顯著水準；而「膚電與辨識正確率」之間的相關性則在辨識作業的所有訊噪比中達顯著水準。迴歸結果顯示，區辨及辨識作業中，均無可以預測主觀費力度感受之指標。此結果再次強調欲評估幼兒的聆聽費力度，應同時評估主觀費力度及費力度客觀生理指標。最後，多元迴歸結果顯示，「反應時間」可預測區辨正確率，而「膚電反應」可以預測辨識正確率，然而由於本研究結果顯示之預測力較低，故仍建議在進行聽覺作業時，仍需記錄幼兒之正確率。結論：本研究結果呈現聽損幼兒無論在何種訊噪比之下的聽覺反應正確率均顯著較聽常幼兒低、反應時間較長，且膚電反應顯著高於聽常幼兒，說明聽損幼兒雖具備與聽常幼兒相當的語言理解能力，且配戴合適的輔具，其聽覺表現仍較聽常同儕差，且客觀的聆聽費力度生理反應亦較聽常幼兒高。區辨及辨識作業中，均無可以預測主觀費力度感受之指標。多元迴歸結果顯示，「反應時間」可預測區辨正確率，而「膚電反應」可以預測辨識正確率。在區辨及辦識作業中，主觀費力度量表結果顯示兩組無顯著差異。本研究結果顯示，欲評估學齡前幼兒之聆聽費力度，宜測量聽覺作業正確率、反應時間及膚電反應，再加上主觀聆聽費力度測量，以更全面評估學齡前聽損幼兒的聆聽費力度及聽能表現。

關鍵字

學齡前聽損幼兒；聆聽費力度；生理回饋；膚電反應；語音區辨；語音辨識

並列摘要

Purpose：The present study aimed to examine the listening effort in different listening tasks for preschool children with hearing loss. By measuring the subjective listening effort and objective biofeedback of listening tasks, the present study investigated whether the signal-to-noise ratio (SNR) affects the percent correct, response time, subjective listening effort and objective biofeedback for children with hearing loss and typical hearing. Background：Audiometry examination could evaluate the detection threshold and speech recognition score of children with hearing loss. However, the short span of concentration of children makes it difficult to complete the pure tone audiometry and the speech test in a one-time assessment. Using the percent corrects of listening tasks to assume the auditory ability of children with hearing loss may under-or over-estimate the performance of children. Recent studies indicated that the task difficulty alters the listening effort and influences the listening performance. However, routine auditory examination and auditory rehabilitation clinics did not include the measurement of listening effort. Therefore, the present study aimed to investigate the listening effort of preschool children with hearing loss and typical hearing, and further analyzing the correlation and predictive power between subjective listening effort and objective biofeedback, in order to detect the listening effort elicited by different listening tasks for preschool children. The results of the present study would help the clinicians to integrate the audiogram, speech perception score and listening effort to evaluate the auditory performance of children with hearing loss more comprehensively. Methods：Ten children with typical hearing and ages between 4 to 6 participated in the piloting test. Another 25 children with typical hearing and 25 children with hearing loss participated in the formal test. Children in two groups were matched by the chorongical age and language comprehension ability. Formal test included the puretone audiometry test, speech discrimination and identification test, subjective listening effort questionnaire, and objective biofeedback measurement of listening effort in the listening condition with three levels of SNRs (quiet, SNR 0, SNR-5). The subjective listening effort and objective biofeedback between two groups were analyzed by two-way mixed designed ANOVA and Chi-Square Test. Spearman’s correlation, multiple logistic regression and multiple linear regression were conducted for analyzing the correlation and predictive power between subjective listening effort and objective measurements. Results：The results indicated that after controlling the language comprehensive score, children with hearing loss performed lower scores and slower response time in listening tasks than that of children with typical hearing, but the subjective listening effort between two groups are similar. Comparing the subjective effort of the three SNRs between the two groups of children in the discrimination task, there was no significant difference in the perception of subjective effort. In the identification task, children with hearing loss had a significant difference in the subjective effort of the three SNRs. Children with typical children only experienced the significant difference of subjective effort between the SNR-5 and in quiet environment, no significant difference in other situations. The results emphasized that the two groups of children have different subjective perceptions of listening efforts in the different SNRs. The objective biofeedback indicators measured in this study showed that the skin conductance response of the three SNRs in the speech discrimination and identification tasks of children with hearing loss is significantly higher than that of children with typical hearing. In addition, heart rate and skin temperature had no significant difference between the two groups of children, indicating that heart rate and skin temperature were less sensitive in assessing children's listening effort under discrimination and identification tasks. The results of Spearman's correlation coefficients showed that the skin conductance response and percent correct were significantly correlated in the conditions of quiet environment and SNR0 of the discrimination task; while the correlation between skin conductance response and percent correct were correlated in all the SNR conditions in the identification task. The multiple logistic regression results showed that there were no indicators that could predict individuals’ subjective effort in the discrimination and identification tasks. The multiple linear regression results indicated that the reaction time can predict the percent correct of speech discrimination ability, and the skin conductance response can predict the percent correct of speech identification. Conclusions：The lower accuracy score, longer response time and higher skin conductance response were observed in children with hearing loss when comparing to the age-matched typical children. In discrimination and identification tasks, there was no significant difference between the subjective listening effort for the two groups. And there is no variable that can predict the subjective listening effort. On the other hand, reaction time can predict the percent correct of discrimination task, and skin conductance response can predict the percent correct of identification task. This study suggests that to assess the listening effort of preschool children, measuring the percent correct, reaction time, and skin conductance response, as well as the subjective listening effort are recommended to conduct a more comprehensive evaluation of the listening effort in preschool children with hearing loss.