List of sentences in Portuguese: speech perception evaluation in children

Murari, Tatiana Cristina; Frederigue-Lopes, Natália Barreto; Santos, Flávia Rodrigues dos; Delgado-Pinheiro, Eliane Maria Carrit

doi:10.1590/1982-0216/20222459122

ABSTRACT

Purpose: to develop lists of sentences in Brazilian Portuguese based on daily vocabulary to evaluate auditory speech perception in children.

Methods: a Brazilian Portuguese corpus with adaptations to attend the criteria established for this study and analyzed by judges was used to obtain the sentences. Thirty-seven lists, each composed of thirteen sentences with fifty phonological keywords, were constructed. The material was presented to ten children, aged between 6 and 10 years, with no hearing or communication complaints. The Friedman test was applied to verify possible differences among the lists. The level of significance adopted was 5% (0.05).

Results: the mean speech recognition index presented by the participants was 96.8% for all lists. The thirty-seven lists were considered statistically similar (p=0.140).

Conclusion: the developed material proved to contain sentences that represent situations of children’s day-to-day communication and its application is easy and quick.

Keywords: Speech Perception; Auditory Perception; Child; Hearing Tests; Hearing Loss

RESUMO

Objetivo: elaborar listas de sentenças no Português Brasileiro, com vocabulário baseado em situações de comunicação do dia a dia, para a avaliação da percepção auditiva dos sons da fala em crianças.

Métodos: o material utilizado para a obtenção das sentenças foi um corpus do Português Brasileiro, com adequações aos critérios estabelecidos nesta pesquisa e analisado por juízes. Foram elaboradas 37 listas, compostas por 13 sentenças e 50 vocábulos fonológicos, que foram denominados palavras-chave. O material foi aplicado em 10 crianças sem queixas auditivas ou de comunicação, na faixa etária entre seis e 10 anos. Foi aplicado o teste de Friedman para verificar possíveis diferenças entre as listas. O nível de significância adotado foi de 5% (0,05).

Resultados: a média do índice de reconhecimento de fala apresentada pelos participantes nas 37 listas de sentenças foi de 96,8%. As 37 listas foram estatisticamente semelhantes (p= 0,140).

Conclusão: o material elaborado possui sentenças que representam situações de comunicação do dia a dia de crianças, é de fácil e rápida aplicação.

Descritores: Percepção da Fala; Percepção Auditiva; Criança; Testes Auditivos; Perda Auditiva

Introduction

The perception of speech sounds involves auditory skills, everything from detection to understanding, and hearing loss may compromise this process, thus impacting the individual’s language¹.

For children who have hearing loss and use spoken language in their communication, it is essential to know the ability these children have in relation to speech recognition².

The recognition of sentences used daily in oral communication is a skill that can bring valuable information about the perception of children's speech sounds, especially for the group of children with hearing loss who, with electronic devices, have access to speech sounds.

Technological advances have provided the possibility of using electronic devices, such as the hearing aid (HA) and the cochlear implant (CI), which are able to minimize the impact of hearing loss, providing audibility of speech sounds³.

However, although audibility is related to speech perception skills, it is not determinant to predict the development of these skills⁴. Aspects such as the time of diagnosis, the type and degree of hearing loss, the time of sensory deprivation and the appropriate intervention, focusing on the development of auditory abilities and oral language should be considered⁵.

The evaluation of speech perception has its importance in the selection, indication, adjustment and monitoring of the benefit of the electronic devices, in addition to providing information regarding the level of development of auditory skills and guiding hearing rehabilitation programs⁵^-⁹.

To this extent, it is necessary that, linked to the technological resource, the construction process of the different auditory abilities is constantly evaluated and monitored⁷. However, due to the complexity that involves speech perception, as well as the difficulty of obtaining a single evaluation procedure that covers all auditory skills, the availability of several procedures in the same language, aiming to expand the possibilities of evaluation is of fundamental importance⁶.

The procedures described, both in the national and international literature, use several types of stimuli: phonetically balanced monosyllabic and dissyllabic words, meaningless words, sentences with and without the presence of noise, among others.

Speech tests that use sentences as a stimulus came about in order to evaluate speech perception through a situation similar to that experienced daily by the individual, since they contain contextual and spectral properties closer to conversation⁹^-²⁴. However, this type of material for children in Brazilian Portuguese is scarce²⁰^-²⁴.

In this context, we highlight the need for instruments that present diversified sentences as a stimulus for the evaluation and monitoring of auditory skills in children who present communication difficulties, including those with hearing loss and have the possibility of access to speech sounds through electronic devices.

In view of the above, the aim of the study was to elaborate lists of sentences in the Brazilian Portuguese with vocabulary based on day-to-day communication situations for the evaluation of speech perception in children.

Methods

The study was approved by the Ethics Committee for The Analysis of Research Projects - Clinical Board of the Hospital das Clínicas and the Medicine School of the University of São Paulo, Brazil (protocol no. 900/02) and carried out according to the criteria established by the current resolution. Those responsible for the children participating in this study signed a Free and Informed Consent Form.

The criteria established for the preparation of sentences were: appropriate vocabulary for children; content of sentences taken from day-to-day situations, excluding proper names, slang, or proverbs and avoiding stereotyped structures; words related to phonological words; affirmative and negative sentences, with simple and compound periods; sentences with an extension of two to six words (phonological words); and lists made up of 13 sentences with each list using 50 keywords⁹^-²⁴.

The material used to obtain sentences with day-to-day vocabulary was extracted from the Child Language Data Exchange - (CHILDES) data base²⁵. The corpus consisted of speech samples from 180 children aged between 5 years and 9 years and 6 months⁶.

The Word List tool from the WordSmith Tools program was used for the quantitative description of the initial corpus where only the data from the statistical list was considered. The number of tokens (total items, words, or occurrences) and types was specified, which are the different words found in this initial corpus, data obtained from the statistical list.

For the delimitation of the corpus, two analyses were performed at this stage in order to exclude shifts of the evaluator and emissions of children inadequate to the proposal in question. Substitutions and/or additions of words were also necessary in the sentences produced by the children. In addition, the grammatical correction was performed so that it did not compromise orality. The reduction of the corpus was carried out individually for each child in the attempt to maintain the same pattern of modification for each one of them (Chart 1).

The characteristic words of the regional dialect, words that were little known or characterized by slang, proper names or sentences that did not present subjects were replaced by words that exercised the same function in the sentence. These words were extracted from another corpus, from the Brazilian Portuguese word bank composed of a list elaborated by the WordSmith Tools program of 50,000 words ordered by frequency, from newspapers, magazines, conversation, classes, meetings, business documents, theses, and academic articles²⁶. Speech therapists with experience in the area of auditory rehabilitation analyzed the list and selected the words with a greater possibility of being part of the vocabulary of children with hearing loss.

Three speech therapists judged a first analysis of the sentences. Next, the sentences were organized through the Word List tool in the WordSmith Tools program, and from the lists, the Concord tool was used, which produces agreements or listing of the occurrences of a specific item for the elimination of repeated or similar sentences. In the case of the study, all verbs on the list that presented frequency higher than one were search words. Two other speech therapists conducted a new analysis. Chart 1 describes the changes made by the judges.

For the last analysis, the Word List and Concord tools were again used in the final list of sentences elaborated in the present study. Chart 1 also presents the modifications made in this stage.

Thumbnail

Chart 1:
Description of the stages performed to develop the sentences

Before the final elaboration of the lists, the sentences were applied to three children aged between six and ten years, with normal hearing confirmed by audiological evaluation. The purpose of that application was to obtain a prior result on the most appropriate sentences. The sentences were presented in free field, on speakerphone, at 0º azimuth, at the fixed intensity of 50 dBA, with the participant one meter from the speaker, with competitive noise presented in the same speaker, in order to obtain a fixed signal/noise ratio of -5 dB. Masking noise presented the same frequency spectrum of sentences, as proposed in the Hearing in Noise Test procedure²³. Sentences that presented answers with any type of error by two of the three children were then eliminated.

A linguistics professional analyzed the selected sentences so that the words were characterized as phonological words²⁷.

After this analysis, the sentences were distributed in lists where each list consisted of thirteen sentences and fifty phonological words, which were also called keywords. Each sentence presented a balanced distribution, so that the lists were composed of different numbers of words.

For the application of the lists of sentences, ten participants were recruited, aged between six and ten years, with no hearing alterations, confirmed by audiological evaluation. The children received instructions that there would be a competitive noise and should repeat the sentences the way they heard them, and that they could request an interruption interval between the application of one list and another.

The sentences were applied by the same evaluator, in an acoustic booth, using a two-channel audiometer and supra-aural headphones, on speakerphone, at an intensity of 50 dBA bilaterally, in the presence of competitive noise²³ concomitant to speech stimulus (ipsilaterally), which was presented at the intensity of 53 dBA, which resulted in a fixed signal/noise ratio of -3 dB. The use of the speakerphone stimulus and competitive noise aimed to simulate a situation closer to daily communication.

To verify the applicability of the lists, the speech recognition index (SRI) of each child for each list was obtained from the sum of the total number of correct answers of the keywords, multiplied by two, considering that each list contained fifty keywords.

The results were analyzed, considering the SRI score presented by each participant, in order to verify the performance in each sentence list.

The equivalence between the lists was analyzed by the application of the Friedman test, aiming to identify possible differences between the lists considered. The level of significance adopted was 5% (0.05). The lists that did not present equivalence were excluded.

In order to obtain the mean value of the participants' score, the overall average of the children in the various sentences was calculated.

Results

The initial corpus consisted basically of 128,737 tokens (total number of items, words or occurrences) and 4,780 types (words). Adult productions were excluded, totaling 34,965 items. The number of sentences was not specified, as the conventions used to define such units were not available in the corpus before the modifications. The judges modified 245 of the 1,756 selected sentences for further analysis.

As described in the methodology, words related to the regional dialect, slang, proper names and sentences that did not present subjects were substituted based on the analysis of the judges, thus 20,000 words with a greater possibility of being part of the vocabulary of children with hearing loss were selected.

The words replaced, because they are part of the dialect or because they are slang, were, for example: cara, botou, botei, mano, camarada, cafajeste, refri, super. A total of thirty-six modifications were made in this aspect. In relation to the proper names, forty-one words were added for substitution. Regarding the use of words as subjects, there were 126 substitutions.

Considering the number of sentences (n= 1,756) there was a tendency to make as few modifications as possible (n= 245), aiming to maintain the general characteristics of the corpus, so as not to compromise the naturalness of the sentences.

Three judges conducting the analysis selected 898 sentences and excluded 855.

It is worth noting that the total number of sentences excluded and selected is not equivalent to the sum, because in both situations, only the sentences that had agreement among the judges were accepted, that is, of the 1,756 sentences presented to them, the three judges selected the 898 final sentences. The same occurred with the 855 sentences excluded.

Based on the 898 sentences, a frequency list was made, and from this list, the Concord tool was used with verbs that presented frequency higher than one. Next, 213 repeated or very similar sentences were deleted. The search words were 138 verbs (assuming that most of the time the verb is the element with the highest semantic load in sentences), whose frequencies ranged from 2 (0.05%) to 52 (1.32%).

In the final analysis, seventy-five sentences or phrases inadequate according to the criteria of this study were excluded, resulting in 610.

Seventeen sentences that contained expressions such as “o pai - a mãe - o vô - a vó” were changed to “papai, mamãe, vovô e vovó”. (“father - mother - grandfather - grandmother" were changed to "Daddy, Mommy, Grandpa and Grandma".) After the selection made by the judges, the pronoun "I" presented an occurrence of 398 (8.5%) and, through the changes, this occurrence was modified to 223 (7.04%).

After conducting all the stages of analysis, 610 sentences remained: thirty-one with eight words; seventy-nine with seven words; 138 with six; 153 with five; 147 with four and sixty-two with three words.

In the prior application of the sentences, only one interval during the application was necessary for two of the children, and for one of them, the application had to occur on different days due to the attention span, the unavailability and/or tiredness presented by the child. At this stage, sixty-eight sentences were eliminated because two or three of the children presented some kind of alteration.

Developing the Lists of Sentences

Phonological analysis of the remaining 542 sentences was performed. The number of sentences according to the number of phonological words was as follows: eleven sentences with two words, 185 with three words, 201 with four words, 115 with five words, twenty-nine with six and one with seven phonological words.

Considering the words as keywords, thirty-seven lists were drawn up, each containing thirteen sentences and fifty keywords (Figures 1 to 5). The remaining sentences would be used for any modifications that were necessary and the sentence of seven words was excluded because it was unique.

Figure 1:
Lists of sentences 1 to 8

Figure 2:
Lists of sentences 9 to 16

Figure 3:
Lists of sentences 17 to 24

Figure 4:
Lists of sentences 25 to 32

Figure 5:
Lists of sentences 33 to 37

In the application of the lists of sentences, the children received the instructions once and none of them showed any difficulties in understanding. Of the ten participants evaluated, seven took the interval and three chose to do the procedure without interruption. The duration of the application of all sentences was approximately 40 minutes, disregarding the interval time, which resulted in approximately one minute per list.

The verification of the applicability of the lists was performed from the calculation of the average scores of the speech recognition index presented by the participants in the thirty-seven lists of sentences, the value being equivalent to the global average of correct answers of 96.8%.

The application of the Friedman test to identify possible differences between the 37 lists revealed that all lists were statistically similar (p= 0.140). Thus, none of them were eliminated, and therefore the final number of lists was thirty-seven.

Discussion

This study aimed to develop lists of sentences in the Brazilian Portuguese, with vocabulary based on day-to-day communication situations, for the clinical speech-language pathology evaluation of auditory speech perception in children.

Regarding the initial corpus, it was possible to observe its representativeness, considering the number of words from the material characterized as tokens (n=128,737). The importance of using a corpus with the characteristics presented in the present study is directly linked to the fundamental aspects pointed out in the literature for the elaboration of sentences to be used in the evaluation of speech perception. These aspects are related to the familiarity of words, naturalness and extension of sentences, grammatical structure, phonetic content, among others⁹^-²⁴.

In fact, it is possible to observe that there has been progress in the search for the development of evaluation procedures that use sentences as a stimulus in an attempt to meet some of the criteria mentioned⁹^-²⁴.

With regard to the reduction of the corpus, it is important to highlight that the least possible changes were made to maintain the general characteristics of the corpus, in order to preserve the naturalness of the sample in the modification of the oral to written pattern, in the aspects such as vocabulary, syntax, semantics, phonetics, phonology and pragmatics. Thus, the path leading to the elaboration of the final sentences became complex, characterizing it as the most extensive and most difficult stage of the study.

After the analysis of the sentences selected by the judges, it was verified that there was a tendency to eliminate the most extensive ones. The criterion of extension of the sentences is an aspect pointed out by the literature, since, in the previous studies, a maximum number of twelve words was observed in the sentences, a number that varied according to each study cited¹⁶^-¹⁹.

The number of sentences is also of great importance when considering that the tests for speech sound perception are used in several evaluation situations and, therefore, a small number of sentences have an effect on the memory of the individual submitted to the evaluation and, consequently, on their performance²¹.

Regarding the preparation of the sentences, considering the words as keywords, thirty-seven lists were drawn up, each with thirteen sentences and fifty keywords. A previous study elaborated lists of sentences, using fifty keywords per list which was considered a sufficient number of items¹⁰.

In addition, the number of thirty-seven lists of sentences can be considered satisfactory when compared to the similar procedures available in Brazilian Portuguese¹^,²^,²⁸^,²⁹. When compared to international tests, it is possible to observe a greater proximity between the number of lists of such procedures and that of the present study⁹^-¹⁴^,¹⁷^,¹⁸.

With regards to the application of the sentences, disregarding the interval time of approximately one minute per list, 40 minutes was the average time to apply all of the sentences. Based on the literature, it is possible to observe that the procedure of this application was rather quick¹⁵^,¹⁹^,²².

The overall average of the speech recognition index scores for the participants in the 37 lists of sentences was 96.8%, with 37 lists being statistically similar (p= 0.140) and, thus, none of them were eliminated.

The methodological procedure adopted in this study made it possible to elaborate lists with diversified sentences, composed of vocabulary based on situations of day-to-day communication and suitable for children. These lists allow the evaluation of hearing capacity in the clinical context of speech therapy, as it proved to be an easy and fast application procedure.

There was a concern to consider harmoniously the various aspects involved in the perception of speech sounds, emphasizing in this context the fundamental importance of the material used. In fact, this represented the spontaneous conversation of hearing children between 5 and 9 years old. Since the target age of this study was children aged 6 to 10 years, it is possible to observe a sample that included an age group younger than this, with the purpose of covering the possible difficulties of the child with hearing loss when compared to the hearing child⁶.

Given the heterogeneity among the infant population with hearing loss and the impossibility of covering all aspects of speech perception in a single test, it is observed the importance of using other procedures to assess speech perception so that individual differences and diverse levels of auditory skills are evaluated¹^,²^,²⁸^-³¹.

One may apply these lists in the clinical context in several ways: with supra-aural headphones or in a free-field system; on speakerphone or in recorded material; with or without the presence of noise.

There is a need for further research that performs the application of the lists of sentences proposed in this study with recorded speech stimulus. It is suggested that other studies evaluate the perception of speech sounds with material elaborated in children with hearing loss.

Conclusion

This study allowed the elaboration of diversified lists representing day-to-day situations experienced by children, and whose application is easy and quick. As future perspectives, the group is developing the recording of these lists.

Acknowledgments

We are grateful to Dr. Maria Cecilia Bevilacqua (in memoriam), for her guidance and supervision during the development of the research.

REFERENCES

¹ Bevilacqua MC, Tech EA. Elaboração de um procedimento de avaliação de percepção de fala em crianças deficientes auditivas profundas a partir de cinco anos de idade. In: Marchesan IQ, Zorzi JL, Gomes IC, editors. Tópicos em fonoaudiologia. São Paulo: Lovise; 1996. p.411-33.
² Ciscare GKSS, Zabeu JS, Santos DR, Morettin-Zupelari M, Delgado-Pinheiro EMC, Frederigue-Lopes NB. List of words to evaluate speech perception: recording and verification of applicability. Rev. CEFAC. 2020;22(5):e2820.
³ Ângelo TCS, Bevilacqua MC, Moret ALM. Speech perception in pre-lingual deaf users of cochlear implant. Pró-Fono R. Atual. Cient. 2010;22(3):275-80.
⁴ Barzaghi L, Madureira S. Percepção de fala e deficiência de audição: elaboração de um procedimento de avaliação da percepção auditiva das plosivas do português brasileiro. Distúrb. Comun. 2005;17(1):87-99.
⁵ Ciscare GKS, Mantello EB, Fortunato-Queiroz CAU, Hyppolito MA, Reis ACMB. Auditory speech perception development in relation to patient's age with cochlear implant. Int. Arch. Otorhinolaryngol. 2017;21(3):206-12.
⁶ Murari TC. Elaboração de sentenças em português para avaliação da percepção da fala em crianças [dissertation]. São Paulo (SP): Universidade de São Paulo; 2004.
⁷ Ferreira K, Moret ALM, Bevilacqua MC, Jacob RST. Translation and adaptation of functional auditory performance indicators (FAPI). J Appl Oral Sci. 2011;19(6):586-98.
⁸ Camargo N, Mendes BCA, Novaes BCAC. Relationship between hearing capacity and performance on tasks of speech perception in children with hearing loss. CoDAS. 2020;32(1):e20180139.
⁹ Schafer EC, Pogue J, Milrany T. List Equivalency of the AzBio Sentence Test in Noise for Listeners with normal-hearing sensitivity or cochlear implants. J Am Acad Audiol. 2012;23(7):501-9.
¹⁰ Bench J, Kowal A, Bamford J. The BKB (BamfordKowal-Bench) sentence lists for partially-hearing children. Br J Audiol. 1979;13(3):108-12.
¹¹ Bench RJ, Doyle JM. The BKB/A (Bamford-Kowal-Bench/Australian Version) Sentence Lists for Hearing-impaired Children. (La Trobe University, Victoria). 1979.
¹² Etymotic Research. Bamford-Kowal-Bench Speech-in Noise Test (Version 1.03)-User manual; 2005. Available at: https://www.etymotic.com/downloads/dl/file/id/260/product/160/bkb_sintm_user_manual.pdf Accessed 2020 jun 19.
» https://www.etymotic.com/downloads/dl/file/id/260/product/160/bkb_sintm_user_manual.pdf
¹³ Killion MC, Niquette PA, Gudmundsen GI, Revit LJ, Banerjee S. Development of a quick speech-in-noise test for measuring signal-to-noise ratio loss in normal hearing and hearing-impaired listeners. J Acoust Soc Am. 2004;116(4):2395-405.
¹⁴ Blamey PJ, Zargarbashi M, Blamey JK, Saunders E. The Australian SpINTM speech in noise test. 2016. Available at: http://www.assta.org/sst/2016/papers/Blamey_SST2016.pdf Accessed 2020 jun 19.
» http://www.assta.org/sst/2016/papers/Blamey_SST2016.pdf
¹⁵ Cameron S, Dillon H. Development of the Listening in Spatialized Noise-Sentences Test (LISN-S). Ear Hear. 2007;28(2):196-211.
¹⁶ Dawson PW, Hersbach AA, Swanson BA. An adaptive Australian Sentence Test in Noise (AuSTIN). Ear Hear. 2013;34(5):592-600.
¹⁷ Spahr AJ, Dorman MF, Litvak LM, Wie SV, Gifford RH, Loizou PC et al. Development and validation of the AzBio Sentence Lists. Ear Hear. 2012;33(1):112-7.
¹⁸ Spahr AJ, Dorman MF, Litvak LM, Cook S, Loiselle LM, DeJong MD et al. Development and validation of the Pediatric AzBio Sentence Lists. Ear Hear. 2014;35(4):418-22.
¹⁹ Boyle PJ, Nunn TB, O'Connor AF, Moore CJ. STARR: A speech test for evaluation of the effectiveness of auditory prostheses under realistic conditions. Ear Hear. 2013;34(2):203-12.
²⁰ Oliveira ST. Avaliação da percepção de fala utilizando sentenças do dia a dia [dissertation]. São Paulo (SP): Pontifícia Universidade Católica; 1992.
²¹ Costa MJ, Iorio MCM, Mangabeira-Albernaz PL. Speech recognition: development of a list of sentences in Portuguese. Acta Awho. 1997;16(4):164-73.
²² Valente SLO. Elaboração de listas de sentenças construídas na língua portuguesa [dissertation]. São Paulo (SP): Pontifícia Universidade Católica; 1998.
²³ Bevilacqua MC, Banhara MR, Da Costa EA, Vignoly AB, Alvarenga KF. The Brazilian Portuguese hearing in noise test. Int J Audiol. 2008;47(6):364-65.
²⁴ Pinheiro MMC, Vieira MG, Vieira LM, Koerich I, Rosseto I, Lazzarotto-Volcão C et al. Updating sentences lists for assessment speech perception. CoDAS. 2022;34(1):e20200301.
²⁵ CHILDES [database online]. Mahwah, NJ: Lawrence Erlbaum Associates; 1999.
²⁶ Sardinha TB. A linguística de corpus. São Paulo, BR: Manole; 2004.
²⁷ Câmara Júnior JM. Estrutura da língua portuguesa. Petrópolis, BR: Vozes; 2015.
²⁸ Delgado EMC, Bevilacqua MC. Lista de palavras como procedimento de avaliação da percepção dos sons da fala para crianças deficientes auditivas. Pró-Fono R. Atual. Cient. 1999;11(1):59-64.
²⁹ Castiquini EAT, Bevilacqua MC. Escala de Integração Auditiva Significativa: procedimento adaptado para a avaliação da percepção da fala. Rev Soc Bras Fonoaudiol. 2000;4(6):51-60.
³⁰ Leandro FSM, Costa EC, Mendes BCA, Novaes BCAC. LittlEars(tm) - Hearing questionnaire: semantic and cultural adaptation of the version of the Littlears(tm) questionnaire in Portuguese in families of children with hearing loss. Audiol., Commun. Res. 2016;21:e1640
³¹ Levy CCAC, Rodrigues-Sato LCCB. Questionnaire validation - PEACH on Brazilian Portuguese. CoDAS. 2016;28(3):205-11.

Study carried out in Faculdade de Medicina da Universidade de São Paulo - FMUSP, São Paulo, São Paulo, Brasil.

Publication Dates

Publication in this collection
16 Dec 2022
Date of issue
2022

History

Received
04 Oct 2022
Accepted
20 Nov 2022

This is an open-access article distributed under the terms of the Creative Commons Attribution License

[1] ¹ Bevilacqua MC, Tech EA. Elaboração de um procedimento de avaliação de percepção de fala em crianças deficientes auditivas profundas a partir de cinco anos de idade. In: Marchesan IQ, Zorzi JL, Gomes IC, editors. Tópicos em fonoaudiologia. São Paulo: Lovise; 1996. p.411-33.

[2] ² Ciscare GKSS, Zabeu JS, Santos DR, Morettin-Zupelari M, Delgado-Pinheiro EMC, Frederigue-Lopes NB. List of words to evaluate speech perception: recording and verification of applicability. Rev. CEFAC. 2020;22(5):e2820.

[3] ³ Ângelo TCS, Bevilacqua MC, Moret ALM. Speech perception in pre-lingual deaf users of cochlear implant. Pró-Fono R. Atual. Cient. 2010;22(3):275-80.

[4] ⁴ Barzaghi L, Madureira S. Percepção de fala e deficiência de audição: elaboração de um procedimento de avaliação da percepção auditiva das plosivas do português brasileiro. Distúrb. Comun. 2005;17(1):87-99.

[5] ⁵ Ciscare GKS, Mantello EB, Fortunato-Queiroz CAU, Hyppolito MA, Reis ACMB. Auditory speech perception development in relation to patient's age with cochlear implant. Int. Arch. Otorhinolaryngol. 2017;21(3):206-12.

[6] ⁶ Murari TC. Elaboração de sentenças em português para avaliação da percepção da fala em crianças [dissertation]. São Paulo (SP): Universidade de São Paulo; 2004.

[7] ⁷ Ferreira K, Moret ALM, Bevilacqua MC, Jacob RST. Translation and adaptation of functional auditory performance indicators (FAPI). J Appl Oral Sci. 2011;19(6):586-98.

[8] ⁸ Camargo N, Mendes BCA, Novaes BCAC. Relationship between hearing capacity and performance on tasks of speech perception in children with hearing loss. CoDAS. 2020;32(1):e20180139.

[9] ⁹ Schafer EC, Pogue J, Milrany T. List Equivalency of the AzBio Sentence Test in Noise for Listeners with normal-hearing sensitivity or cochlear implants. J Am Acad Audiol. 2012;23(7):501-9.

[10] ¹⁰ Bench J, Kowal A, Bamford J. The BKB (BamfordKowal-Bench) sentence lists for partially-hearing children. Br J Audiol. 1979;13(3):108-12.

[11] ¹¹ Bench RJ, Doyle JM. The BKB/A (Bamford-Kowal-Bench/Australian Version) Sentence Lists for Hearing-impaired Children. (La Trobe University, Victoria). 1979.

[12] ¹² Etymotic Research. Bamford-Kowal-Bench Speech-in Noise Test (Version 1.03)-User manual; 2005. Available at: https://www.etymotic.com/downloads/dl/file/id/260/product/160/bkb_sintm_user_manual.pdf Accessed 2020 jun 19.
» https://www.etymotic.com/downloads/dl/file/id/260/product/160/bkb_sintm_user_manual.pdf

[13] ¹³ Killion MC, Niquette PA, Gudmundsen GI, Revit LJ, Banerjee S. Development of a quick speech-in-noise test for measuring signal-to-noise ratio loss in normal hearing and hearing-impaired listeners. J Acoust Soc Am. 2004;116(4):2395-405.

[14] ¹⁴ Blamey PJ, Zargarbashi M, Blamey JK, Saunders E. The Australian SpINTM speech in noise test. 2016. Available at: http://www.assta.org/sst/2016/papers/Blamey_SST2016.pdf Accessed 2020 jun 19.
» http://www.assta.org/sst/2016/papers/Blamey_SST2016.pdf

[15] ¹⁵ Cameron S, Dillon H. Development of the Listening in Spatialized Noise-Sentences Test (LISN-S). Ear Hear. 2007;28(2):196-211.

[16] ¹⁶ Dawson PW, Hersbach AA, Swanson BA. An adaptive Australian Sentence Test in Noise (AuSTIN). Ear Hear. 2013;34(5):592-600.

[17] ¹⁷ Spahr AJ, Dorman MF, Litvak LM, Wie SV, Gifford RH, Loizou PC et al. Development and validation of the AzBio Sentence Lists. Ear Hear. 2012;33(1):112-7.

[18] ¹⁸ Spahr AJ, Dorman MF, Litvak LM, Cook S, Loiselle LM, DeJong MD et al. Development and validation of the Pediatric AzBio Sentence Lists. Ear Hear. 2014;35(4):418-22.

[19] ¹⁹ Boyle PJ, Nunn TB, O'Connor AF, Moore CJ. STARR: A speech test for evaluation of the effectiveness of auditory prostheses under realistic conditions. Ear Hear. 2013;34(2):203-12.

[20] ²⁰ Oliveira ST. Avaliação da percepção de fala utilizando sentenças do dia a dia [dissertation]. São Paulo (SP): Pontifícia Universidade Católica; 1992.

[21] ²¹ Costa MJ, Iorio MCM, Mangabeira-Albernaz PL. Speech recognition: development of a list of sentences in Portuguese. Acta Awho. 1997;16(4):164-73.

[22] ²² Valente SLO. Elaboração de listas de sentenças construídas na língua portuguesa [dissertation]. São Paulo (SP): Pontifícia Universidade Católica; 1998.

[23] ²³ Bevilacqua MC, Banhara MR, Da Costa EA, Vignoly AB, Alvarenga KF. The Brazilian Portuguese hearing in noise test. Int J Audiol. 2008;47(6):364-65.

[24] ²⁴ Pinheiro MMC, Vieira MG, Vieira LM, Koerich I, Rosseto I, Lazzarotto-Volcão C et al. Updating sentences lists for assessment speech perception. CoDAS. 2022;34(1):e20200301.

[25] ²⁵ CHILDES [database online]. Mahwah, NJ: Lawrence Erlbaum Associates; 1999.

[26] ²⁶ Sardinha TB. A linguística de corpus. São Paulo, BR: Manole; 2004.

[27] ²⁷ Câmara Júnior JM. Estrutura da língua portuguesa. Petrópolis, BR: Vozes; 2015.

[28] ²⁸ Delgado EMC, Bevilacqua MC. Lista de palavras como procedimento de avaliação da percepção dos sons da fala para crianças deficientes auditivas. Pró-Fono R. Atual. Cient. 1999;11(1):59-64.

[29] ²⁹ Castiquini EAT, Bevilacqua MC. Escala de Integração Auditiva Significativa: procedimento adaptado para a avaliação da percepção da fala. Rev Soc Bras Fonoaudiol. 2000;4(6):51-60.

[30] ³⁰ Leandro FSM, Costa EC, Mendes BCA, Novaes BCAC. LittlEars(tm) - Hearing questionnaire: semantic and cultural adaptation of the version of the Littlears(tm) questionnaire in Portuguese in families of children with hearing loss. Audiol., Commun. Res. 2016;21:e1640

[31] ³¹ Levy CCAC, Rodrigues-Sato LCCB. Questionnaire validation - PEACH on Brazilian Portuguese. CoDAS. 2016;28(3):205-11.

Stage: Reduction of the corpus
First analysis:		Second analysis:
Excluded items:	Other alterations:	Excluded items:	Other alterations:
• Shifts of the evaluator; • Codes used in the transcription; • Meaningless phrases; • Incomplete sentences (children’s responses to the evaluator’s speech); • Phrases repeated by the same child; • Phrases that represented the demonstration, a drawing made by the child; • Phrases that narrated the figures in sequence (resulted in many repeated sentences, because the same sequence of figures was used); • Phrases produced as a function of reading (since they are not spontaneous productions of the children); • Words repeated in the same sentence; • Expressions characteristic of spoken language (aí terminou, acontece que, só que, assim, ó) (then it ended, it turns out that, only that, like that).	• Separation of paragraphs into sentences, considering the beginning marked by an introductory word (daí, então, um dia, depois) (hence, then, one day, after) and the final points, characterizing the beginning of a new sentence or fact; • Separation of long compound periods.	• Phrases with transitive verb, but meaningless and without complement; • Paragraphs in which a child was encouraged to repeat to another child what they had told the evaluator; • Interrogative phrases; • Very extensive sentences, with no possibility of reduction, because, when possible, the sentence was divided; • Confusing phrases or those that only made sense from the context.	• Substitution of words characteristic of the regional dialect; • Substitution of little-known words or those that characterized slang; • Addition of a word that performed the same function in the sentence without altering the agreement and verbal tense when the sentence did not present a subject (e.g.: the child’s answers to the evaluator’s question); • Replacing proper names with words that exercised the same function in the sentence.
Stage: Analysis of the sentences by the judges
First analysis:		Second analysis:
Judges:	Alterations:	Judges:	Alterations:
• Three judges (speech therapists)	• Only sentences not excluded by any of the judges were selected	• Two judges (speech therapists)	• Exclusion of phrases considered inappropriate for the purpose of the study
Stage: Final analysis
Excluded items:		Other alterations:
• Phrases unrelated to the study criteria; • The pronoun “I” in some sentences, so that it would not harm the phrasal structure, that is, the verbal conjugation was not altered, and the phrase should contain some element that indicated the subject “I” (example: my, mine).		•Correction of Portuguese so that it does not compromise the naturalness of the sentences; • Modification of uncommon expressions (“o pai”, “a mãe”, “o vô”, “a vó” were changed to “papai”, “mamãe”, “vovô” e “vovó”) (“father - mother - grandfather - grandmother” were changed to “Daddy, Mommy, Grandpa and Grandma”).