Mandarin Speech Test Materials (MSTMs)

after several international conferences, some friends asked me about my research. i put one of my researches?chinese speech tests here as your references. i also introduced myself.

if you wanted to know more, please send mail to me. thank you.

 

zhang hua  ([e-mail])北京同仁医院耳鼻咽喉头颈外科张华

at beijing, china

 

 mandarin speech test materials
(mstms)

 

zhang hua, m.d.

 

capital medical university

beijing tongren hospital

beijing institute of otolaryngology

 

preface

 

speech audiometry  is generally regarded as more acceptable for identifying patients with poor auditory capability  and evaluating the communication ability in their daily lives

 

different kinds of english materials have been used in clinics for a long time, such as the pb-50 word lists, ab word lists, w-22, nu-6, cid everyday sentences, spin tests

 

speech test history in china

why so slowly in china？

edited materials too old, some words are seldom used now

 

the old materials were edited by a few different institutes and not standardized yet. almost no  choices

 

hard to compare the results among institutes and clinics

 

no commercials materials could be purchased in the market

why so slowly in china？

audiology is not a popular profession yet

 

the homogeneity, equivalence, validity, reliability and maneuverability need strict evaluation

 

all these needs a long time of research and cooperation with different professions

 

 

the mandarin speech test materials (mstms)  in china

study procedures

designing of mstms

recording and calibration of mstms

evaluating of mstms

further studies

step 1  designing of mstms

monosyllabic materials

 

     10 lists , 50 words per list

     all the words are in common use and have real meaning. chosen from authorized linguistic books

    the test words in each list are 3-d  phonologically balanced (vowel, consonant and tone)

step 1  designing of mstms

  disyllabic (spondaic) materials

    10 lists , 50 words per list

 

    all words are common in daily life and phonologically balanced.

    avoid using two 3rd-tone monosyllabic words to compose the test word.

    the disyllabic word with “一” (yī) or “不” (bù) as the first syllable can not be selected

    all words were stressed equally, and were excluded with neural tone in the end

step 1  designing of mstms

sentence materials

 

     30 lists with 10 sentences in each list , 50 keywords per list

    the test sentences are representative of conventional spoken mandarin (not written chinese language) in most areas in china

    the intelligibility of each list is homogeneous

step2  recording and calibration of mstms

the materials were digitally recorded in the china national radio station by an experienced male mandarin broadcaster

 

the pronunciation  was clear and natural, the intensity of speech sounds was at a similar level

 

the calibration tone (a 1000 hz pure tone , 30 seconds ) was inserted at the beginning of the recording

 

the speech material in each track was within ±0.5 db of the standard reference 1000 hz tone

 

the presentation level of speech materials via the headphone from the audiometer was calibrated at 20 db spl, which equates to 0 db hl

step3  evaluation of mstms

equivalence analysis of mandarin materials on normal-hearing subjects

 

performance-intensity (pi) function on normal-hearing subjects

 

disyllabic words performance-intensity function on hearing-impaired subjects

equivalence analysis with normal-hearing subjects

 

subjects ： normal hearing subjects including 25 males and 35 females ，5 subjects  for the pilot test and 60 subjects  for the  formal test

               

  method： the subject listened to the materials and required to repeat loudly, guessing was encouraged . there was a practice list in the beginning of each  session

 

presentation level        decided by the pilot  test

monosyllabic words       pta+ 10 db

 disyllabic words         pta+ 2 db

 sentences               fixed at 5 db hl

                                                                                

 equivalence analysis with normal-hearing subjects

7 monosyllabic, 9 disyllabic and 15 sentence lists are equivalent

   monosyllabic words ：list 1,4,5,6,7,8,9

   disyllabic words list：1,2,3,4,6,7,8,9,10

   sentences list：1,2,4,6,7,8,12,13,14,15,16,18,21,22,25

performance-intensity (pi) function on normal-hearing subjects

 p-i function on normal-hearing subjects

 

     for the monosyllabic, disyllabic and sentence materials, all   the regression analysis showed a high correlation between correct score and presentation level. the regression function follows           

           monosyllables:   y=34.56x³-7.53x2+0.43x-0.0052

          disyllable:       y=-0.14x²+13.4x-200.9

           sentences:       y=124.29x³+7.5931x2-0.0001

 

p-i function on normal-hearing subjects

  monosyllabic p-i function on hearing-impaired

 

subjects:

  45 patients with bilateral sensorineural hearing loss (pilot study 5, formal study 40)

 average age: 58

 pta (0.5, 1k, 2k and 4k) for better ear ? 60 db hl

  monosyllabic p-i function on hearing-impaired

  presentation level

  two groups for shorting test time: 20 each group and 5 presentation levels

  monosyllabic  p-i function on hearing-impaired

disyllabic p-i function on hearing-impaired subjects

 subjects

    40 subjects (23 males and 17 females) with mild to moderate sensorineural hearing loss (5 for pilot and 35 for formal), the mean age is  58 years old. the average pta threshold at 500, 1000 and 2000 hz is 44.6 db hl

 

     presentation level ： srt+15 db，srt+10 db，srt+5 db，srt，srt-5 db，srt-10 db were selected as the presentation levels

 

 disyllabic p-i function on hearing-impaired subjects

 for disyllabic materials, the quadratic regression analysis

     (y= -270.3+12.2x-0.11x2) showed a very high correlation (0.997) between correct score and presentation level.

    the srt of these subjects was 44.6 db. the average pta at 0.5, 1k, 2k and 4k hz was 43.7 db hl. there was thus a very good agreement between the srt and pta threshold for hearing-impaired subjects using these mandarin disyllabic materials

 disyllabic p-i function on hearing-impaired subjects

purpose for short lists

 

 it is perfect to use 50-word list but takes longer time and not good for clinical usage

 

editing of the short  lists

 set up the subgroup of consonants and vowels according to their location and method of pronunciation

 only chose the representatives of the phonemes in the same group

 combine selected phonemes: reduce word numbers and make sure to test necessary phonetic characteristics

edit of short monosyllable lists

equivalent evaluation of short monosyllable lists

 24 short monosyllable lists, 20 words in  each list

 subjects: 48 normal hearing young students, half male/female

 method: repeat after presentation, encourage guessing. practice before test

 presentation level: one level for one list. every subjects listened all 24 lists successively. from 20 to 5 db hl, reducing the level with 5 db each step. each circle has 6 lists with 6 presentation levels

equivalent evaluation of short monosyllable lists

 

after analysis, 16 lists are equivalent

 

got p-i curves of 24 lists by nonlinear          as the logistic regression

 

the slope of linear part for 20-80% discrimination scores is 3.2%/db

the slope of linear part for 40-50% discrimination scores is 22.5%/db

 

further studies

mstms tests in noise

nonsense lists      

computerization for all lists

 

     hope to  cooperate with you

 

funds

national natural science foundation of china

beijing natural science foundation

    national 10th 5-year scientific research program

capital medical developing foundation

 

acknowledgement

§ prof. zhang jialu, acoustician

§ prof. wang hongjun, linguistician

§ associate prof. cao wen, linguistician

§ prof. liu zhaoxiong and song xinqiao, linguistician

§ prof. ma binrong, statistician

§ prof. li lin, audiologist

§ prof. wu zongji, linguistician

§  prof. robert h. mannell (special visit for this topic)

§  prof. philip newall (regular visiting professor of our school)

§  ms. fiona sutherland  (stayed with us  for 10 months)

§  prof. jack katz

§  prof. gus mueller

§  dr. harvey dillon

§  dr.  gus mueller 

§  auditec of st. louis 的 dr. william f. carver

§ 

§ 

curriculum vitae

 

 name: hua zhang, m.d.,

     male, born  in shandong province

        nationality: chinese

 work:

     beijing institute of otolaryngology

     clinical audiology center

     beijing tongren hospital

 

title：associated professor

position: director of applied science department and clinical audiology center         

 

1977.11: medical student in nanjing, 5 years.

1982.12: assistant professor, resident of ent

1985：graduate student. peaking union medical college. supervisor:  prof. zhizhong wang. director of china national committee of otolaryngology. finished to edit “the minimal auditory capabilities in chinese (macc)”, the current main speech evaluation materials of cochlear implants in china.

1988: attending ent doctor in peaking union medical college hospital.

curriculum vitae

1994：portland, or. passed tests: license of dispensing hearing aids in oregon, bc-his, his membership. visited many professional places and attended congresses of his and aaorl-hns.

1995: minneapolis, mn. starkey labs. inc. hearing aid fitting supervisor and regional manager.

curriculum vitae

1996：first one to make and use cic in china. organize many professional classes to train hearing aid dispensers.

2000：beijing institute of otolaryngology and           clinical audiology center of   beijing tongren hospital. insist to enhance audiology education in china. introduce binaural digital fitting of hearing aids. attended many international and national seminars and congress.

 

professional membership

 

acoustic society china                                          1989

biomedical engineering society of china                         1993

board certified hearing instrument specialist of international hearing society                                                         1994

member of international hearing society                         1994

licensed hearing consultant of oregon                           1994

member of american auditory society                             1994

member of american academy of audiology                         1996

china association of rehabilitation of disabled persons, member of hearing and speech rehabilitation committee               2003, 1997

unhs expert group of beijing                                    2003

expert of audiology development foundation of china            2004

expert of public bidding of hearing aids for china disabled persons’ federation                                                     2004

《chinese archives of otolaryngology-head and neck surgery》editorial member                                                         2004

《chinese scientific journal of hearing and speech rehabilitation》standing editor                                                2003

《journal of audiology and speechpathology》 editorial member  2003

《otolaryngology foreign medical sciences》audiology editor    2004

 

research

clinical audiology

§speech audiometry，esp. standardization of chinese test list for adults.

§hearing aid fitting, speech recognition in noise

§hearing screening, national hearing evaluation standards.

otology

§common ear diseases

§inner ear diseases

§ 

contact

dr. hua zhang

beijing institute of otolaryngology

 

17 hou gou hu tong

beijing, 100005

p. r. china

 

e-mail: [e-mail]