Development, Standardization and Evaluation of A Close-set Speech Recognotion Test for Persian Speaking Children of 4 to 6 Years old

• Main Authors: Mahin Sedaei, Dr. Yahya Modaresi, Dr. Gholamreza Babei, Mansoure Adelghahraman
• Format: Article
• Language: fas
• Published: Tehran University of Medical Sciences 2000-05-01
• Series: Audiology
• Subjects: Close-set Speech Recognotion Test; Persian Speaking Children; Pure Tone Audiometry; Immittance Audiometry; Speech Recognition Test
• Online Access: http://journals.tums.ac.ir/PdfMed.aspx?pdf_med=/upload_files/pdf/13726.pdf&manuscript_id=13726

Materials and Methods: This study was performed in two parts: included development and standardization of the test and the second part was the evaluation stage. In the first section, Watsons crieria fpr valid speech recognition tests for children was considered. In order to balance the lists phonemically, it was necessary to determine Persian phonemes frequency -of- occurences. Thus, the frequency -of- occurences for 29 phonemes of Persian language in children's and adults' speech were separately determined. There were significant differences between some phonemes frequencies in the two samples. So their frequencies in childrens speech was used to design the test. It should be mentioned that each list could contains only 25 vowels, while almost 32 ones were meeded according to the phonemically balancing procedure. Therefore, vowels and consonants were separately balanced. The required monosyllabic words were extracted from the books written for 4 to 6 years old children. Finally, four closed set (3- picture matrix) 25 word lists were prepared. The lists were recorded on magnetic cassette in one of the IRIB studios. To assess the test validity and reliabity, 33 normal hearing children were randomly selected from kindergartens over the city of Tehranand were tested at the Audiology Clinic of the Rehabilitation Faculty, Tehran University of Medical Sciences in 1998. The tests included: otoscopy, pure tone & immittance audiometry, establishing speech reception threshold, and speech recognition test - retest by these lists. Findings: 1- Normal hearing children obtained 92-100 percent scores for each list at their most comfortable loudness levels through test - retest. 2- No significant difference was observed in test - retest scores in each list (p>0.05). 3-No significant difference obtained between the lists test or retest scores (p>0.O5). 4-No significant difference was observed between children's scores in terms of their age and gender (p>O.OS) Conclusion: this study, therefore, is reliable and valid. The lists are phonemically balanced and equal in difficulty. Due to the conditions under which the test was designed (including, phonemic balance, necessity of using the words that could be pictured and of the presence of tow other word in children's vocabulary which is similar in auditory aspects), only the content validity for word familiarity decreased a little that could be overcome by introducing them through pictures prior to the start of the test.