Phonemically Balanced Research Articles

A Hierarchical Approach to Efficient Assessment of Functional Intelligibility: The Modular FITI (Functional Importance to Intelligibility) Phrase List

Purpose: Gurevich and Kim (2023a) developed speech stimuli with full phonetic coverage and phonemic balance to comprehensively examine production and perception of consonants. They identify a hierarchy of positional variants with functional importance to intelligibility (FITI). The FITI stimuli are well suited for assessing intelligibility. Multiple studies report that clinicians prefer nonstandard materials (e.g., reading passages, word lists) over formal tools (e.g., Frenchay Dysarthria Assessment–Second Edition [FDA-2]; Enderby & Palmer, 2008), citing time constraints among reasons. Having analyzed the phonetic coverage of the three standard passages (Gurevich & Kim, 2022), the purpose herein is to provide a similar analysis for the FDA-2 and to construct an alternative list of stimuli that achieves strategic coverage of priority intelligibility targets and is adaptable to time constraints. Method: The FDA-2 word list was analyzed in terms of its coverage of the 466 allowable contexts for the 24 English consonants and its coverage of priority FITI targets. A subset of the stimuli from Gurevich and Kim (2023a) was used to construct phrases organized in modules from highest to lowest FITI priority. Results: The full FDA-2 word list covers 35.41% of the allowable 466 contexts, compared to 83.05% by the 308 FITI word list from Gurevich and Kim (2023a). It also underperforms in covering priority FITI targets but does capture some of the hierarchy if all 115 words are used. Fifteen modules were constructed from the 308 FITI words with one to five phrases per module, presented in order of highest to lowest FITI priority. Conclusions: The newly constructed list of phrases has comprehensive phonetic coverage, follows the FITI hierarchy, and is built in autonomous modules. Using the modular FITI phrases to investigate speech is adaptable to time constraints in a manner that prioritizes intelligibility targets.

Development of Novel Speech Stimuli With Phonetic Coverage and Phonemic Balance

Purpose: Gurevich and Kim (2022) call for new speech materials that provide not only phonemic but also allophonic coverage so that production and perception of consonants can be examined as a function of their positional contexts. The purpose of this work was to construct such materials. Method: A total of 466 allowable contexts for the 24 English consonants were determined by considering positional contexts in terms of word and syllable positions, lexical stress, and proximity to other phonemes. These contexts were filled using 308 unique tokens composed of frequent content words ensuring both phonemic and allophonic coverages. In addition, the words were combined into 37 declarative and interrogative phrases to incorporate prosodic elements of speech. Twenty healthy adults were recorded reading these phrases to assess time and effort required for the task. Results: The newly constructed speech materials achieved high-overall phonetic coverage at 83.05% of all allowable English contexts, including coverage of consonant categories and positional contexts of particular importance for clinical and research purposes. Moreover, these materials attained the balance of consonant frequencies in typical spoken English. This was achieved with 37 phrases that take an average of 3 min, 6 s to read at an average of 5.02 s per phrase. Conclusions: Compared with the commonly used standard reading passages examined in Gurevich and Kim (2022), the newly constructed materials provide significantly superior phonetic coverage, including important linguistic features known to affect production and perception of consonants. The new materials are therefore well suited for eliciting speech representative of General American English for investigating consonant production and perception.

INDO-SPRITT: DEVELOPMENT AND NORMS OF AN INDONESIAN SPEECH RECOGNITION THRESHOLD TEST FOR CHILDREN

IntroductionWell-designed audiometric speech tests for Indonesian children are not currently available. This paper describes the development of the Indonesian Speech Recognition Threshold Test (INDO-SPRITT).Material and methodsA list of Indonesian words with response foils and pictures was developed. Presentation level was varied and the 50% recognition threshold was calculated as the average of the midpoints of each reversal. A normative reference was established using a sample of 118 normal hearing participants,16 children with severe to profound hearing loss, and 25 adults. The effects of age on speech reception thresholds and test reliability were also assessed.ResultsINDO-SPRITT material was found to be appropriate for children older than 4 years and 6 months. The speech reception threshold (SRT) improved on average from 18 dB HL for 4 to 5 year old children to 13 dB for children aged 10 to 13 years, providing a normative reference against which the SRT of children with unknown hearing status can be compared. Five reversals are enough to estimate the SRT.ConclusionsSuitable words, phonemic balance, and pictures have been created for Indonesian children. The reliability of different lengths of the test was similar, with 5 reversals being enough to estimate the SRT. The mean SRT decreased with age, but did not vary with the number of reversals.

OPRACOWANIE, STANDARYZACJA I WALIDACJA DWUSYLABOWEGO, FONEMATYCZNIE ZRÓWNOWAŻONEGO TAMILSKIEGO TESTU SŁOWNEGO W CISZY I W SZUMIE

BackgroundThe present study aimed to develop and standardize a phonemically balanced bisyllabic word test in Tamil for adult listeners.Material and methodsIn total, 1015 bisyllabic Tamil words were collected from different sources; 20 Tamil speakers rated the words for familiarity and 5 experts validated the content. Based on the familiarity rating and content validation, 760 words were shortlisted for phonemic balancing. Then 25 phonemically-balanced lists were prepared with 25 words in each. The prepared lists were presented to 100 normal-hearing listeners at 40 dB SL in quiet, and 30 listeners in noise at −5 dB SNR for the standardization process. The lists were also presented at different sensation levels (SLs) in quiet to 30 listeners to obtain a psychometric function.ResultsThe mean speech identification scores (SISs) in adults was 99.8% in quiet. The results revealed no significant difference in SIS across the 25 word lists, indicative of list equivalency. The scores increased as the level increased from 10 to 40 dB SL for all the lists, suggesting homogeneity in difficulty and audibility. However, in noise, only 23 lists were equivalent to each other.ConclusionsAll the test lists can be utilized for testing during audiological evaluation in quiet, and 23 word lists are useful in noise.

Development of a Thai phonetically balanced monosyllabic word recognition test: Derivation of phoneme distribution, word list construction, and response evaluations

This paper proposes a test tool for Thai word recognition, the Thammasat University Phonetically Balanced Word List 2014 (TU PB’14), standardized on several major criteria: phonemic balance, familiarity, reliability, list equivalency, and homogeneity. Phoneme distributions from the largest written Thai corpus (InterBEST) were obtained and used to construct five phonetically balanced word lists, each containing 25 frequently occurring monosyllabic words. Listeners’ percent correct discrimination scores from test and re-test sessions were not significantly different, confirming test reliability. Detailed analysis of listeners’ errors revealed that perceptual errors occurred predominantly for initial sound only, final only, and initial together with final. In terms of list equivalency and homogeneity, derived psychometric function slopes of TU PB’14 ranged from 0.0941 to 0.1155, while intensities required for 50% intelligibility ranged from 41.0279 to 41.3697. Two-way Chi-Square analysis performed on both parameters indicated that there was no significant difference among the word lists.

Test–retest reliability of Kannada time-compressed speech test and time-compressed monosyllables test

Objective: The study attempts to determine the test–retest reliability for time-compressed phonemically balanced (PB) words in Kannada and monosyllables across different compression ratios.Study design: The study was carried out on 30 normal hearing individuals, 15 males and 15 females. The PB words were time-compressed at 50%, 60% and 70% compression ratios and monosyllables were time-compressed at 50%, 60%, 70% and 80% compression ratios. The compressed material was administered three times (baseline, after 15 min break and after 1 week) at all the compression ratios for PB words and monosyllables.Results: The results of the study show that scores were higher for monosyllables at high compression ratios. The scores were more reliable at 60% compression ratio for both monosyllables and PB words. The study also shows that the reliability of the test does not vary across gender.Conclusion: The study recommends using 60% compression ratio for PB words and monosyllables while testing a clinical population.

WPŁYW WIEKU NA PERCEPCJĘ MOWY SKOMPRESOWANEJ W CZASIE I PERCEPCJĘ MOWY W SZUMIE U OSÓB Z NORMALNYM SŁUCHEM

BackgroundThis work studied, in individuals with normal hearing, the perception of time-compressed speech, compression ratio, and speech perception in noise using phonemically balanced (PB) word lists in Kannada. The effect of age was also studied.Material and MethodsParticipants were 29 young and middle-aged adults (native Kannada speakers) aged 15 to 50 years. The Speech Perception in Noise (SPIN) test was administered at signal to noise ratios (SNRs) of +5 dB, 0 dB, –5 dB, and in quiet. The Time-Compressed Speech Test (TCST) was administered on all participants at compression ratios of 0%, 40%, 50%, and 60%.ResultsYoung adults performed better than middle-aged adults in both the TCST and SPIN tests. As difficulty levels rose, performance decreased in both tests.ConclusionsEven in individuals with normal hearing, ageing has a deleterious effect on the performance of the central auditory processing system, which is necessary for speech perception under adverse listening conditions.

Effect of Compression Ratio on Perception of Time Compressed Phonemically Balanced Words in Kannada and Monosyllables.

The present study attempted to study perception of time-compressed speech and the effect of compression ratio for phonemically balanced (PB) word lists in Kannada and monosyllables. The test was administered on 30 normal hearing individuals at compression ratios of 40%, 50%, 60%, 70% and 80% for PB words in Kannada and monosyllables.The results of the study showed that the speech identification scores for time-compressed speech reduced with increase in compression ratio. The scores were better for monosyllables compared to PB words especially at higher compression ratios. The study provides speech identification scores at different compression ratio for PB words and monosyllables in individuals with normal hearing. The results of the study also showed that the scores did not vary across gender for all the compression ratios for both the stimuli. The same test material needs to be compared the clinical population with central auditory processing disorder for clinical validation of the present results.

Development of a Turkish Monosyllabic Word Recognition Test for Adults

MATERIALS and METHODS: The most frequently used Turkish monosyllabic words were selected from the corpus. Thirty-six young adults with normal hearing were divided into two groups and asked to listen to words from the word pool; the words were given twice at six different intensity levels. The least and most frequently known words were identified and eliminated in order to provide homogeneity. Three word lists, each composed of 50 phonemically balanced words, were developed to be used in the tests. These lists were divided into two according to the phonemic balance criteria. RESULTS: No statistically significant difference was found among the word lists. Furthermore, the internal reliability of each list was analyzed using KR-20 and was found to be above 98% for all lists. CONCLUSION: The lists derived from the Turkish language were ascertained to be appropriate for use. As a result of using the developed lists to test individuals with various auditory pathologies, it will be possible to assess the lists’ capability to distinguish pathological cases according to the location of the pathology.

Development of a script of phonemically balanced monosyllable lists of Mandarin–Chinese

ObjectivesTo develop a set of monosyllable audiometry test material that can be used in speech recognition testing in Mandarin Chinese. Materials and MethodsIn an effort to improve reliability and efficiency in clinical practice, a set of 30 test lists of 25 monosyllable test items each was designed with consideration of the following: length of the list, phonemical balance (PB), word utility frequency, coverage of common words, and equivalency among lists. The 30 lists contained 750 monosyllabic words and were constructed on the basis of Chinese phoneme distribution probability summarized in the Manual of Acoustic. The occurrence incidence of the 22 consonants, 36 vowels, and 4 tones of the Chinese Mandarin characters in the 30 test lists were calculated. To achieve PB, 489 monosyllables were selected from the 2500 most commonly used Chinese characters and the 4000 most commonly used Chinese phrases to compile the 30 monosyllable test lists using a computer algorithm with manual adjustment. ResultsThirty phonemical balanced word lists of common Mandarin Chinese monosyllabic words were compiled. Each list consisted of 25 monosyllables. A total of 489 phonemes (consonants, vowels and tones) were included in the set. ConclusionsThis set of lists can potentially be used as the basic lists for future development of Mandarin PB monosyllable speech test materials.

Development of a Mandarin Monosyllable Recognition Test

This study aimed to develop a set of 50- and 25-item word lists that exhibited familiarity, homogeneity, and phonemic balance for a Mandarin monosyllable recognition test (MMRT). To achieve the design goal of high subject familiarity with the test material, we selected the 700 most frequently occurring monosyllables to be the test material. The homogeneity of the test material was achieved by evaluating five psychometric characteristics of these 700 monosyllables to obtain 348 homogeneous monosyllables with similar psychometric functions for constructing the word lists. The phonemic balance of the 50-item word lists was achieved by deriving the desired numbers of initials, finals, and tones in these lists according to their occurrence frequencies in 4733 monosyllabic words. The phonemic balance of the 25-item word lists was achieved by equally dividing the desired numbers of initials, finals, and tones in the 50-item word lists into two groups, called half-A and half-B lists. Three half-A lists and three half-B lists were constructed from the 348 homogeneous monosyllables, and they could be paired to form nine 50-item word lists. Accordingly, all of the MMRT word lists are familiar, homogeneous, and phonemically balanced. The homogeneity of the MMRT word lists was examined by evaluating the interlist equivalence, interitem variability, and intersubject variability. A chi test was used to evaluate the interlist equivalence among the MMRT word lists; that is, to determine whether the values of percent correct recognition were distributed identically at 12 different presentation levels among the six 25-item word lists and nine 50-item word lists. No statistically significant differences were found among the word lists: chi(55) = 12.004 and p = 1.000 for the six 25-item word lists, chi(88) = 7.695 and p = 1.000 for the nine 50-item word lists, and chi(154) = 19.699 and p = 1.000 for the six 25-item word lists plus nine 50-item word lists. The chi tests indicated that the six 25-item word lists and nine 50-item word lists exhibited interlist equivalence. The mean interitem variability (2.0 dB) and mean intersubject variability (2.3 dB) are lower for the MMRT word lists than for previously reported word recognition tests. This study applied an innovative approach to design MMRT word lists from familiar and homogeneous monosyllables, in which the familiarity, homogeneity, and phonemic balance of the six 25-item and nine 50-item word lists were strictly controlled. These word lists exhibit interlist equivalence with respect to their psychometric functions and five psychometric characteristics; moreover, their interitem and intersubject variability are lower than those of previously reported lists. Future clinical experiments should examine whether such a design approach can improve the reliability and diagnostic sensitivity of word recognition tests for hearing-impaired listeners.

Predicted and measured speech recognition performance in noise with linear amplification.

The purpose of this study was to investigate the applicability of the Speech Intelligibility Index (SII) in hearing aid fitting. It was hypothesized that estimated speech intelligibility, based on the SII, could be a more reliable measure than real speech recognition results for comparing hearing aid characteristics. The test subjects were 29 elderly persons (66 to 80 yr) with mild-to-moderate hearing loss, who were using monaurally fitted linear hearing aids. They were selected from the files at Sahlgrenska hearing clinic. Speech recognition scores were obtained at fixed speech-to-noise ratios with Phonemically Balanced (PB) words in speech-weighted noise and in low-frequency noise. A Just-Follow-Conversation (JFC) test was performed with connected speech presented in the same background noises. The subjects were tested without hearing aid and with their hearing aids set at three different frequency responses. Predicted speech recognition scores were calculated for each condition based on the SII, complemented with a correction for sensorineural hearing impairment. The calculations involved speech and noise spectra, pure tone thresholds and insertion gain responses. For each condition, the measured speech recognition scores were, on average, well predicted by the calculated scores. The intra-individual standard deviation of the predicted scores was estimated to be about one percent unit. The group results of the JFC test were in agreement with the word recognition results for the aided conditions, but a floor effect was observed for the unaided conditions. Speech intelligibility prediction based on the modified SII is a valid estimate of speech recognition performance of hearing-impaired persons with mild-to-moderate hearing loss. Estimated intelligibility based on the SII is more reliable than actually measured speech recognition performance, for comparing amplification conditions within subjects.

KIDS: A database of children’s speech

A database of children reading age- and reading-level-appropriate text aloud has been collected. This labeled data, to be distributed in the near future, was primarily intended to be used in CMU’s LISTEN tutor, which employs speech recognition to monitor children’s reading and then helps correct errors. The speaker population was therefore chosen to represent good and poor readers and to incorporate dialects of the speakers for whom the reading coach is intended. Phonemic balance could not be achieved (although it has been calculated) since the primary concern in recording children reading is to present sentences that can effectively be read by first through third graders. The text is a series of sentences that was adapted from text in the Weekly Reader series—most of the adaptation concerned the lack of the accompanying images. The text was chosen for its intrinsic interest and widespread use. Several trial recording sessions were used to develop a protocol that kept extraneous noises produced by the children at a minimum. This and other problems inherent in recording children reading will be discussed. Novel techniques developed for labeling this kind of speech will also be presented. [This work was funded by NSF Grant No. IRI-9528984.]

Low-frequency hearing loss in presbycusis. A central interpretation.

Elderly subjects with a central phonemically balanced (PB) and synthetic sentence identification (SSI) pattern show greater low-frequency (LF) sensitivity loss than elderly subjects with a peripheral PB-SSI pattern. This LF hearing loss is not the simple consequence of age, nor does the LF hearing loss account for depressed SSI performance. We conclude that LF sensitivity loss and a central PB-SSI pattern are both components of central aging. We discuss the implications for auditory rehabilitation of elderly subjects.

Resident's Page

<h3>AUDIOLOGIC QUIZ CASE</h3><h3>Deborah Hayes, MA, Houston</h3> A 35-year-old woman was evaluated by the audiology service for tinnitus in the left ear. She also noticed a decrease in sensitivity in that ear of one month's duration, but still used the left ear on the telephone. Results of pure-tone audiometry are shown in Fig 1. The right ear shows relatively normal sensitivity with an apparent air-bone gap in the high frequencies. The left ear shows a mild sensorineural loss. Results of impedance audiometry are shown in Fig 2. Both right and left tympanograms are normal (type A). Acoustic reflexes are present, although slightly elevated (95 to 105 dB HL) with sound in right ear, probe in left ear. However, acoustic reflexes are absent at 110 dB HL, with sound in left ear, probe in right ear. Speech audiometry at 30 dB SL yielded phonemically balanced (PB) word and synthetic sentence identifi

Whole-Word Scoring Versus Phoneme Scoring in Speech Audiometry

This paper reports on the relationship between the scores (%) resulting from two scoring methods, namely whole-word scoring and phoneme scoring, employed in speech audiometric procedures with phonemically balanced (PB) word lists. Seventy-two adult subjects, 30 with normal hearing, 17 with conductive and 25 with sensorineural hearing impairment participated in this investigation. Their speech discrimination abilities whilst listening to Fry's PB word lists presented both through earphones and in a free-field situation were quantified first with the one and then with the other method and the relationship between the two groups of resulting scores was ascertained. The total number of word lists administered was just over 13,000 and the scores derived from the two scoring methods showed consistent differences (calculated at the linearly rising section of the speech discrimination curve) varying from 20-25%, phoneme scoring giving the higher percentage values. When these differences in scores were translated ...

Diagnostic speech audiometry.

The scope of speech audiometry can be expanded to provide useful diagnostic information. Comparison of performance vs intensity (Pl) functions for both phonemically balanced (PB) words and synthetic sentence identification (SSl) yields pattersn useful in differentiating among peripheral and central sites of auditory disorder.

A test for speech discrimination composed of CNC monosyllabic words.

In an attenlpt to develop two highly equivalent forms of a test of phonemic discrimination, two lists, each composed of 50 CNC monosyllabic words and conforming to the phonemic balance favored by Lehiste and Peterson (1959), were recorded on magnetic tape in six different word orders. These lists, designated as N. U. Auditory Test No. 4, were administered monaurally on a test-retest basis to 16 Ss with normal hearing, 16 with conductive losses, and 16 with sensori-neural impairments. An interval of 1 to 2 wk. separated the two test sessions for a given S. During each test session six presentation levels were employed. These were arranged in a progression of ascending intensities in order to keep the effects of familiarity with the test vocabulary at a minimum. The total range was from -4 db to 40 db sensation level. All three types of Ss yielded articulation functions which were characterized by a linear segment at low presentation levels and a curvilinear segment at high levels. In the linear portion of the function for the normal and conductive groups the articulation score increased ac the race of 6% per decibel increase in signal intensity. In the sensori-neural group the slope of the linear segment of the function was only 3% per decibel. For all three groups, List I yielded slightly higher discrimination scores than List 11. In addition, a small improvement in discrimination scores was found from test to retest. This was true for all groups and both lists. However, in all groups, scores between lists as well as those from test to retest showed relatively high positive Pearson product-momenc correlations. The general conclusions are that: (1) Lists I and I1 are sufficiently interdependent to be considered as alternate forms of a single exploratory tool; (2) the tool itself appears sufficiently reliable to be viewed as a particularly promising method for determining the slope of the linear portion of an individual's articulation function.