Tesis sobre el tema "Reading ability"
Crea una cita precisa en los estilos APA, MLA, Chicago, Harvard y otros
Consulte los 50 mejores tesis para su investigación sobre el tema "Reading ability".
Junto a cada fuente en la lista de referencias hay un botón "Agregar a la bibliografía". Pulsa este botón, y generaremos automáticamente la referencia bibliográfica para la obra elegida en el estilo de cita que necesites: APA, MLA, Harvard, Vancouver, Chicago, etc.
También puede descargar el texto completo de la publicación académica en formato pdf y leer en línea su resumen siempre que esté disponible en los metadatos.
Explore tesis sobre una amplia variedad de disciplinas y organice su bibliografía correctamente.
Fuller, Frank D. (Frank Davidson). "The association between reading ability and test performance among adults of limited reading ability". Thesis, University of North Texas, 1990. https://digital.library.unt.edu/ark:/67531/metadc332849/.
Texto completoSimpson, Lauree Smith. "Multiple intelligences and reading ability". CSUSB ScholarWorks, 1996. https://scholarworks.lib.csusb.edu/etd-project/1159.
Texto completoManning, Margaret. "Laterality, reading and ability in children". Thesis, University of Leicester, 1990. http://hdl.handle.net/2381/34667.
Texto completoMcCartney, Rebecca Eisenberg. "Rapid Automatized Naming and Reading Ability". Digital Archive @ GSU, 2008. http://digitalarchive.gsu.edu/psych_diss/48.
Texto completoYoshida, Tomoko. "Phonological awareness and reading ability in Japanese children". Thesis, Click to view the E-thesis via HKUTO, 2005. http://sunzi.lib.hku.hk/hkuto/record/B31937287.
Texto completoJia, Fanlu y 贾凡路. "Gender differences of reading ability in Chinese children". Thesis, The University of Hong Kong (Pokfulam, Hong Kong), 2013. http://hub.hku.hk/bib/B50639420.
Texto completopublished_or_final_version
Linguistics
Master
Master of Arts
Alsawar, Radah. "Using The Qualitative Reading Inventory to Assess A Saudi Reader’s Reading Ability". University of Toledo / OhioLINK, 2017. http://rave.ohiolink.edu/etdc/view?acc_num=toledo1493394444213539.
Texto completoFrey-Toompere, Linet. "Suppression in L1 and L2 reading comprehension". Thesis, University of Cambridge, 2010. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.608937.
Texto completoNicholson, Ann L. "The Effects of Repeated Readings on the Fluency Scores of Low Ability Third Grade Readers". Bowling Green State University / OhioLINK, 2006. http://rave.ohiolink.edu/etdc/view?acc_num=bgsu1163077388.
Texto completoBarnie, Charlene. "Parent and teacher estimates of children's reading ability". Thesis, National Library of Canada = Bibliothèque nationale du Canada, 2001. http://www.collectionscanada.ca/obj/s4/f2/dsk3/ftp04/MQ58320.pdf.
Texto completo勞皓珍 y Ho-chun Rebecca Lo. "The relationship between phonological awareness and reading ability". Thesis, The University of Hong Kong (Pokfulam, Hong Kong), 2003. http://hub.hku.hk/bib/B2682274X.
Texto completoWass, Malin. "Children with Cochlear Implants : Cognition and Reading Ability". Doctoral thesis, Linköpings universitet, Institutet för handikappvetenskap (IHV), 2009. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-51735.
Texto completoAvhandlingens övergripande syfte var att studera kognitiva förmågor hos barn med grav hörselskada eller dövhet som fått cochleaimplantat (CI). Auditiv stimulering från CI i tidig ålder påverkar de flesta kognitiva funktioner som en följd av hjärnans plasticitet hos små barn. Det är viktigt att förstå de kognitiva konsekvenserna av auditiv stimulering från CI för att kunna ge dessa barn bästa möjliga stöd. Avhandlingen undersökte tre specifika aspekter av kognitiv förmåga (arbetsminne, fonologiska förmågor och lexikal aktivering), samt läsförmåga hos barn med CI, i jämförelse med barn med normal hörsel i samma åldrar. Relationerna mellan kognitiva förmågor och läsförmåga studerades också, liksom sambanden mellan demografiska faktorer (t ex implantationsålder och kommunikationssätt) och kognitiva förmågor samt läsfärdigheter. Barnen med CI hade generellt lägre prestationsnivå än barnen med normal hörsel i uppgifter som mäter fonologiskt och generellt arbetsminne, fonologiska förmågor och lexikal aktivering. De hade specifika problem i uppgifter som i hög grad belastar fonologiskt arbetsminne, medan deras visuospatiala arbetsminneskapacitet var jämförbar med den hos barnen med normal hörsel. Majoriteten av barnen med CI hade läsfärdigheter i nivå med normalhörande barn, för både avkodning och läsförståelse. Sambanden mellan demografiska faktorer och kognitiva förmågor och läsförmåga varierade mellan studierna. Resultatmönstren diskuteras utifrån teorier om arbetsminne, fonologiska färdigheter och lexikal aktivering.
Ali, Sumaya Nader. "Reading ability and diglossia in Kuwaiti primary schools". Thesis, University of Leeds, 2014. http://etheses.whiterose.ac.uk/7457/.
Texto completoVon, Spiegel Jacqueline Janice. "Reading ability and summarization effects on metacomprehension accuracy". The Ohio State University, 2006. http://rave.ohiolink.edu/etdc/view?acc_num=osu1400152139.
Texto completoLorenson, Susan Beth. "Phonemic Awareness and Reading Ability in Literate Adults". Diss., The University of Arizona, 2014. http://hdl.handle.net/10150/316779.
Texto completoThompson, Meri Dawn. "Authentic reading assessment: The reading portfolio". CSUSB ScholarWorks, 1995. https://scholarworks.lib.csusb.edu/etd-project/1134.
Texto completoChan, Lik-hang y 陳力行. "Effectiveness of using reading assessment to enhance the grade 8 students' reading ability". Thesis, The University of Hong Kong (Pokfulam, Hong Kong), 2010. http://hub.hku.hk/bib/B50175312.
Texto completopublished_or_final_version
Education
Master
Master of Education
黃珮詩 y Pui-sze Catherine Wong. "Hong Kong secondary three students' reading comprehensionprocess". Thesis, The University of Hong Kong (Pokfulam, Hong Kong), 2002. http://hub.hku.hk/bib/B31963274.
Texto completoKeidel, Lora L. "Third grade teachers' instructional groupings for reading and improvement of Idaho reading indicator scores /". ProQuest subscription required:, 2003.
Buscar texto completoLi, Wing-yan Michelle y 李穎昕. "Understanding the relationship between rapid automatized naming and reading in Chinese". Thesis, The University of Hong Kong (Pokfulam, Hong Kong), 2014. http://hdl.handle.net/10722/209658.
Texto completopublished_or_final_version
Educational Psychology
Master
Master of Social Sciences
Henson, Janice M. "Procedures for recognition and transformation of nonefficacious personal constructs about reading and learning /". free to MU campus, to others for purchase, 2002. http://wwwlib.umi.com/cr/mo/fullcit?p3060104.
Texto completoYAMASHITA, JUNKO. "Transfer of L1 Reading Ability to L2 Reading : An Elaboration of the Linguistic Threshold". 名古屋大学大学院国際言語文化研究科, 2001. http://hdl.handle.net/2237/7937.
Texto completoWang, Wei. "Reading tutors using the iPod to enhance and motivate struggling literacy learners' reading ability". [Ames, Iowa : Iowa State University], 2008.
Buscar texto completoMorles, Armando. "The improvement of leaming ability by means of reading". Pontificia Universidad Católica del Perú, 2013. http://repositorio.pucp.edu.pe/index/handle/123456789/99772.
Texto completoEl presente estudio discute la importancia de la instrucción en la capacidad para aprender el contenido de textos escritos. Al inicio se define el aprendizaje a través de la lectura para luego continuar con la identificación de los factores que condicionan este aprendizaje. Luego se suministra lineamientos para mejorar esa capacidad. Finalmente se aportan ideas acerca de cómo conducir el entrenamiento de los estudiantes, incluyendo las modalidades y su ubicación dentro del contexto educativo.
McLain, Katherine Victoria Mayer. "Effects of two comprehension monitoring strategies on metacognitive awareness and reading achievement in third and fifth grade students". Virtual Press, 1990. http://liblink.bsu.edu/uhtbin/catkey/720162.
Texto completoDepartment of Elementary Education
Bourne, R. J. (Roger John). "Two studies on the effectiveness of contiguous, graphemic and phonological interventions on measures of reading and spelling". Master's thesis, Faculty of Education, 2002. http://hdl.handle.net/2123/6270.
Texto completoLau, Sin-Yee. "Cognitive and metacognitive strategies of English as second language readers in Hong Kong primary schools". Thesis, University of Bristol, 2002. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.251558.
Texto completoRoloff, Vera Lucia Posnik. "Foreign language reading comprehension: Text representation and the effects of text explicitness and reading ability". Thesis, University of Ottawa (Canada), 1999. http://hdl.handle.net/10393/8791.
Texto completoRoloff, Vera Lúcia Pósnik. "Foreign language reading comprehension, text representation and the effects of text explicitness and reading ability". Thesis, National Library of Canada = Bibliothèque nationale du Canada, 1999. http://www.collectionscanada.ca/obj/s4/f2/dsk2/ftp03/NQ36793.pdf.
Texto completoGuo, Lin. "Explore the relationship between metacognition, L1 reading ability, L2 language proficiency and L2 reading comprehension". The Ohio State University, 2014. http://rave.ohiolink.edu/etdc/view?acc_num=osu1405090119.
Texto completoGibeault, Kimberly. "The use of ability grouping and flexible grouping within guided reading". Online version, 2008. http://www.uwstout.edu/lib/thesis/2008/2008gibeaultk.pdf.
Texto completoParker, Cynthia T. "An evaluation of student reading attitudes : does ability affect attitude? /". Electronic version (PDF), 2004. http://dl.uncw.edu/etd/2004/parkerc/cynthiaparker.pdf.
Texto completoWolfe, Christopher B. "Motor control and reading fluency contributions beyond phonological awareness and rapid automatized naming in children with reading disabilities /". unrestricted, 2007. http://etd.gsu.edu/theses/available/etd-11282007-113346/.
Texto completoTitle from file title page. Rose A. Sevcik, committee chair; MaryAnn Romski, Rihana Williams-Smith, Robin D. Morris, committee members. Electronic text (99 p. : ill.) : digital, PDF file. Description based on contents viewed Jan. 28, 2008. Includes bibliographical references (p. 90-99).
Brez, Sharon. "Adult learners' perspectives on screening reading ability for patient teaching". Thesis, University of Ottawa (Canada), 1995. http://hdl.handle.net/10393/9879.
Texto completoSperring, Rachael. "Magnocellular processing and reading ability : the effect of test sensitivity". Thesis, University of Reading, 2013. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.603524.
Texto completoBoon, Joyce Linda. "Stroke sequence, working memory and Chinese reading ability in Chinesechildren". Thesis, The University of Hong Kong (Pokfulam, Hong Kong), 2010. http://hub.hku.hk/bib/B45150011.
Texto completoMurry, David D. "Reading ability vs. readability in a sex offender treatment program". Online version, 1999. http://www.uwstout.edu/lib/thesis/1999/1999murryd.pdf.
Texto completoTsangari, Victoria. "An interactive software program to develop pianists' sight-reading ability". Diss., University of Iowa, 2010. https://ir.uiowa.edu/etd/3393.
Texto completoWesterveld, Marleen Frederike. "Examining the relationship between oral narrative ability and reading comprehension in children with mixed reading disability". Thesis, University of Canterbury. Communication Disorders, 2007. http://hdl.handle.net/10092/1389.
Texto completoTaylor, Jaimée. "Working memory and reading comprehension: The relationship between central executive functioning and discourse-level reading ability". Thesis, Taylor, Jaimée (2017) Working memory and reading comprehension: The relationship between central executive functioning and discourse-level reading ability. Honours thesis, Murdoch University, 2017. https://researchrepository.murdoch.edu.au/id/eprint/40825/.
Texto completoCastle, M. "A study of methods endeavouring to improve reading readiness with regard to sex differences and social deprevation". Thesis, University of Bath, 1988. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.234638.
Texto completoChang, Shu-chieh y 張淑婕. "The influences of Dialogic Reading and Flashcards Readingon Children’s Reading Ability and Interest". Thesis, 2013. http://ndltd.ncl.edu.tw/handle/39813771491485471002.
Texto completo國立臺南大學
幼兒教育學系碩士班
101
This study discusses the influence of "Dialogic Reading" and "Flashcards Reading" on the reading ability and interest of children. The research method the study applies is "pretest-posttest control group design". The object of study is 28 children of the top class in a public kindergarten in Tainan City. They were divided randomly into the experimental group and the control group. A week before the experiment, the children of these two groups received the pretest of PPVT-R (Peabody Picture Vocabulary Test-Revised) Form L, Preliminary Graded Chinese Character Recognition Test, target words in sketchbook test, and reading attitude test. The experiment lasted for ten weeks. A week later after the experiment, the children of two groups received the posttest of the four tests given above and story comprehension test in order to know the difference between the pretest and posttest of vocabulary comprehension, story comprehension, characters recognition ability and reading interest. One and a half months later, a tracking test of the five tests above was applied for the difference of abilities listed above that the children in these two groups had and for whether the test results possessed continued influences. The data was collected and analyzed by means of ANCOVA, t-test. The study results are as follows: 1. The flashcard group improved recognizing characters in sketchbooks, which was better than the dialogic group and the influence continued. 2. The flashcard group improved recognizing Chinese characters better than the dialogic group, but it had no continuous influence. 3. The dialogic group improved vocabulary comprehension better than the flashcard group and the influence continued. 4. The dialogic group improved story comprehension ability better than the flashcard group and the influence continued. 5. The dialogic group improved reading interest better than the flashcard group and the influence continued.
Leou, Yea-mei y 柳雅梅. "Influence of Balanced Reading Instruction on Students’ Reading Ability and Reading Motivation". Thesis, 2005. http://ndltd.ncl.edu.tw/handle/74582787726923052248.
Texto completo國立臺南大學
教育經營與管理研究所博士班
93
Abstract This study was designed to find out the principles for designing a balanced reading program for the elementary school students in Taiwan, and to test how this balanced reading program influence on the students’ reading ability and reading motivation. The participants were seventy-two fifth graders. Thirty-six students were in the experimental group, including 18 boys and 18 girls, and 36 fifth- grade students in the controlled group, including 18 boys and 18 girls. The experimental teaching lasted for nine weeks. The reading ability in this study meant the ability of word recognition and reading comprehension. The reading subtest and the spelling subtest of The Wide Range Achievement Test, the sound subtest of the English Word Recognition and the meaning subtest of the English Word Recognition were used to test the students’ word recognition ability. The sentence comprehension subtest of the English Achievement Test was used to evaluate the students’ comprehension ability. The Motivation for Reading Questionnaire was used for reading motivation. Pretests and posttests were used before and after the intervention. Ancova was used to analyze the data. Besides, questionnaires of learning, questionnaires of reading, interviews and the teaching journals were used to gather further information about the program. From the review of literature, the consideration of the national education policy and the practical teaching in the classrooms, the researcher suggested that five factors should be taken into consideration when creating a balanced reading program. First, as for environment, the classroom should be abundant with learning materials and posters. There should be a reading corner for the kids. Second, as for the choosing of the reading materials, connecting with the children’s life experience, predicative, fitting the kids’ language ability, having useful and attractive pictures, being interesting, having suitable length and clear layout, being able to act as a play and being able to extend the school learning were the principles. Third, as for the principles of phonics, the combinations between the most common words announced by the Ministry of Education and Fry (2001) common phonics rules were good choices. Fourth, as for the strategies of reading comprehension, prediction, asking, looking for related clues, summary and review, and ministering were strategies worthy of teaching. The results showed that the experimental group scored significantly higher than the controlled group in reading subtest, sound subtest, meaning subtest, comprehension subtest, and reading motivation test after the intervention. Besides, the main findings from the data of the questionnaires, interviews and the teaching journals were consistent to the above statistic results. This study proved that a balanced reading instruction had positive influences on students’ reading ability and reading motivation. Suggestions were provided to the administrators and English teachers according to the research results.
吳烈文. "To improve reading ability and teaching reading:a study between a teacher and three reading underachievers". Thesis, 2006. http://ndltd.ncl.edu.tw/handle/71116985882453236074.
Texto completo國立新竹教育大學
人資處輔導教學碩士班
95
This study aims to enhance underachievers’ reading ability through an action research conducted by a sixth-grade teacher in an elementary school. The subjects included the author and three reading low-achievers in class A, six grade in Hsin Hsin Elementary School. These three students went quite well with their peers; however, they had underachievement in terms of reading performance. Since the uniform teaching method in a large-sized classroom worked ineffectively to improve the students’ reading ability, the author decided to apply a three-in-one strategy, that is knowledge, action and thought, to improve their reading ability. Initially, the teacher referred to related literature regarding reading theory, action research and underachievement, then the teacher refined and narrowed down the research focus by presenting the drafted project and counseling in person with experts and dialogue partners. As a result, with collected literature and counseling with experts, the author discovered two important factors for reading ability enhancement: one is teachers’ guidance for students to apply reading strategies; the other, worthy of note, is teachers’ demand and tolerance on students’ under-achieved reading performances. These three underachievers, through the reading action research, performed even more actively in reading activities, were competent to discuss the questions regarding the reading texts, became more involved in reading assignment, and were able to share ideas with the teacher and the peers. In addition, the students started trying to read with reading strategies and methods. Moreover, the teacher found out the appropriate teaching approaches for reading low-achievers.
Sung, She-Ping y 宋詩蘋. "The Effects of Sustained Silent Reading on The Reading Motivation And Reading Ability". Thesis, 2015. http://ndltd.ncl.edu.tw/handle/63546462939241265038.
Texto completo華梵大學
工業工程與經營資訊學系碩士班
103
The purpose of this research is to investigate the effect of sustained silent reading (SSR) on reading comprehension, reading attitude, and ability of word recognition of elementary school students in the 5th grade. Students from two classes participated in the experiments. One class, as the experimental group, was implemented SSR every day, while the other class, as the control group, received conventional teaching courses. 1200 minutes of reading experiments were conducted by 15 minutes daily, 5 days a week for 16 weeks. The results are summarized as follows: 1. SSR can effectively improve the students’ “reading motivation”, “ability of reading comprehension”, but not “ability of word recognition”. 2. The students’ gender and the intensity of internet usage do not have significant difference in “reading motivation”, “ability of reading comprehension”, and “ability of word recognition” after implementing SSR. 3. Students give positive comments on SSR.
Su, Huei-Sian y 蘇慧嫻. "Chinese Reading Ability and Mathematical Ability among Seventh Graders in Taiwan". Thesis, 2012. http://ndltd.ncl.edu.tw/handle/33024623373802315260.
Texto completo國立臺中教育大學
特殊教育學系碩士在職專班
100
The purposes of the study are to analyze the correlation between Chinese reading ability (character recognition and reading comprehension) and mathematical ability (procedural knowledge test and problem solving test) for the seventh graders in Taiwan, and to investigate the difference in mathematical ability between the students with different Chinese reading abilities. The participants of study were 129 seventh graders with 64 males and 65 females. The main findings of this research are: (1) There is a significant difference between male and female students in character recognition, but there were no significant differences in reading comprehension, procedural knowledge test and problem solving test. (2) The performance of students with different Chinese reading abilities showed significant differences in procedural knowledge and problem solving. By comparing group performance, the high level group performed better than the middle and low level groups, and the middle level group performed better than the low level group. (3) There was a significantly positive correlation between Chinese reading ability (character recognition and reading comprehension) and mathematical ability (procedural knowledge and problem solving ) between all tests. (4) For the part of procedural knowledge, a significant prediction was only found for character recognition after controlling the factors of age and intelligence of the seventh graders; for the part of problem solving, both character recognition and reading comprehension showed significant prediction with a higher effect of character recognition than reading comprehension, which showed character recognition has its own influence on problem solving.
Huang, Lin-Hui y 黃琳惠. "The Effects of Peer Instruction on Improving Reading Comprehension Ability for Elementary School Students with Low Reading Ability". Thesis, 2009. http://ndltd.ncl.edu.tw/handle/84938917759271742198.
Texto completo國立臺北教育大學
特殊教育學系碩士班
98
The fundamental of this research was to evaluate the enhance effectiveness of peer instruction for reading comprehension ability toward the elementary school students with low reading ability. The independent variable of this research was peer instruction which focused on seven important story grammar components including time, character, place, initial event, process, consequence, and thought. The dependent variables of this research were the abilities of oral reading fluency and story grammar. The research adopted single subject research design with A-B design method and performed continuous ten week instruction on five low reading ability students. Through chart display, and questionnaire analysis, the conclusions could be described as follows: 1. After the experimental instruction on five students with low reading ability, one student achieved the expected target standard, two students demonstrated improvement, and the other two students only demonstrated momentous improvement on the first two weeks, but the performance did not stable on the rest of eight weeks. 2. Students with low reading ability increased the frequency of using coherent elements in the story grammar expression as experiment processed. Some students with low reading ability decreased the frequency of using phonetic symbols or writing errata. 3. Students with low reading ability and parents expressed satisfaction for the peer instruction. The results of this research offered some suggestion for future relevant research and teaching strategies for elementary students with low reading ability: 1. Students with medium and high reading abilities should be monitored at the same time. 2. Experimental instruction time should be extended. 3. Experimental materials should be added phonetic notations and modified properly 4. Instruction time should be chosen on early studies by oneself and noon break.
Geier, B. Kelly. "MORPHOLOGICAL AWARENESS, READING ABILITY, AND THE READING OF MULTI-MORPHEMIC WORDS". Thesis, 2010. http://hdl.handle.net/1974/5567.
Texto completoThesis (Master, Education) -- Queen's University, 2010-04-19 18:50:46.661
Lee, Chung-li y 李中莉. "PROMOTING READING COMPREHENSION ABILITY AND VOCABULARY LEARNING THROUGH COLLABORATIVE STRATEGIC READING". Thesis, 2003. http://ndltd.ncl.edu.tw/handle/74867727757856222370.
Texto completo國立台北師範學院
兒童英語教育研究所
91
English education in primary schools has been highly valued in recent years. However, the reading instruction has not been put on as much emphasis as the speaking and listening instructions. Since reading is the central means for independent learning especially for EFL learners, the study is intended to conduct a reading strategy instruction, which is called the Collaborative Strategic Reading, to promote primary school students’ reading proficiency and to understand how they apply reading strategies. The Collaborative Strategic Reading (CSR), which is proposed by J. K. Klingner and S. Vaughn in 1999, is found effective on enhancing reading comprehension ability, academic performance, and content learning on ESL students in the United States. CSR combines the principles of cooperative learning and reciprocal teaching. It is aimed to help students apply reading strategies through cooperative group discussion. In the CSR instructional context, students first learn reading strategies from their teacher and then apply the strategies through small, heterogeneous group discussion without their teacher’s directing. They cooperate and help each other dealing with the difficult words or concept in text. Besides, they record their discussion in their reading logs. There were sixty-three fifth-grade students participating in this study. They were pretested and postested on the proficiency of word-naming, vocabulary, and reading comprehension. Students practiced CSR in the experimental group, and their conversation in cooperative group discussion was transcribed and analyzed three times. In the control group, students received the traditional teacher-directed reading instruction. All the subjects read seven storybooks and two worksheets in fourteen weeks in this study. After posttesting, a questionnaire was used to examine students’ attitudes towards CSR. The findings of the study are summarized as the following: (1) CSR promotes students’ reading comprehension ability effectively, however, it does not enhance students’ word-naming and vocabulary abilities effectively in this study. (2) Students in the CSR context have progress in using reading strategies over time. (3) CSR facilitates students’ positive reading attitudes.
Proença, Jorge Daniel Leonardo. "Automatic assessment of reading ability of children". Doctoral thesis, 2018. http://hdl.handle.net/10316/83815.
Texto completoThe work detailed in this thesis proposes solutions to automatically evaluate the reading ability of children, targeted towards European Portuguese. Contributions were made towards the state of the art of reading assessment by using sentences and pseudowords, proposing sentence utterance segmentation strategies that consider disfluencies and by proposing multiple features both for mispronunciation classification and overall reading performance estimation. Knowing how to read is one of the most important markers of a child's cognitive development. Teachers usually have to expend a large effort to properly evaluate a child's reading aloud performance on a 1-on-1 basis, manually taking notes for accuracy and time. A tool that records and automatically analyzes reading tasks could be an important complement for reading evaluation. The objectives of this work were to develop methods that support an automatic reading evaluation of children 6 to 10 years old. Providing an overall reading aloud level score can be useful to quickly get an appreciation of a child's ability and follow their evolution along time and to combine the information of several metrics that teachers take into account when evaluating a child. A large European Portuguese database of children reading aloud was collected to have sufficient data to train acoustic models and to have a large number of examples of reading disfluencies. The reading tasks presented to children were carefully designed by selecting appropriate sentences and generating pseudowords, distributed throughout tasks according to a difficulty metric. Several types of disfluencies were identified, with the most common being mispronunciations, false-starts, repetitions and intra-word pauses. Consequently, these were the ones targeted for automatic detection. Several strategies were developed to automatically detect reading disfluencies and get automatic annotation of utterances. All followed the same two-step basis: segmentation that detects extra content, and mispronunciation classification. First, segmentation is achieved by constrained decoding lattices based on the ideal pronunciation of the prompt text but allowing freedom of repetition and syllable-based false starts. The best approach uses syllables as units and allows optional silence between each syllable to address the problem of intra-word pauses. Decoding an utterance results in word candidate segments that will be classified as correctly pronounced or not. The best performing feature to classify mispronunciations was a log-likelihood ratio between the ideal pronunciation and a free-phone-loop filler model, done in a word-spotting manner. Additional features of likelihoods of individual phonemes and Levenshtein distances between correct pronunciation and recognized phonemes are combined in multi-feature models. Elementary school teachers were asked to rate the overall reading level score of children as 0-5, resulting in a ground truth of reading score. Regression models to estimate these scores were trained based on performance features extracted from the automatic annotation, with separate features for the reading of sentences or pseudowords. Gaussian process regression models achieved the best results from automatic annotation, with results closely approaching the use of features extracted from manual annotation. Two applications of the developed work were built: a demo and a prototype website. The demo application showcases the methodology applied to the children of the collected dataset. The prototype website is a platform for teachers where they can assign reading tasks to several students, tell children to read tasks using a microphone, and analyze the automatically given performance score and utterance annotations.
O trabalho detalhado nesta tese propõe soluções para automaticamente avaliar a capacidade de leitura de crianças, tendo como alvo o Português europeu. Foram feitas contribuições para o estado da arte de avaliação de leitura ao usar frases e pseudopalavras, ao propor estratégias de segmentação de locuções de frases que consideram disfluências e ao propor vários parâmetros quer para classificação de pronunciações incorretas como para estimação de capacidade de leitura geral. Saber ler é um dos mais importantes marcadores do desenvolvimento cognitivo de uma criança. Os professores têm habitualmente de despender um grande esforço a avaliar decentemente a capacidade de leitura em voz alta de uma criança, um a um, manualmente tirando notas de exatidão e tempo. Uma ferramenta que grave e automaticamente analise tarefas de leitura poderá ser um importante complemento à avaliação de leitura. Os objetivos deste trabalho foram desenvolver métodos que suportem uma avaliação automática da leitura de crianças de 6 a 10 anos de idade. Fornecer um valor de nível geral de leitura em voz alta pode ser útil para rapidamente obter um apreciação da capacidade de uma criança e seguir a sua evolução ao longo do tempo e para combinar a informação de várias métricas que os professores têm em conta quando avaliam uma criança. Uma grande base de dados de Português europeu de crianças a ler em voz alta foi adquirida para ter dados suficientes para treino de modelos acústicos e para ter um largo número de exemplos de disfluências da leitura. As tarefas de leitura apresentadas às crianças foram cuidadosamente construídas seleccionando frases apropriadas e gerando pseudopalavras, distribuídas ao longo das tarefas de acordo com uma métricas de dificuldade. Vários tipos de disfluências foram identificados, com os mais comuns sendo pronunciações incorretas, pré-correções, repetições e pausas intra-palavra. Consequentemente, estes foram o alvo de detecção automática. Foram desenvolvidas várias estratégias para automaticamente detetar disfluências da leitura e obter anotação automática de locuções. Todas seguiram a mesma base de duas fases: segmentação que deteta conteúdo extra, e classificação de pronunciações incorretas. Primeiro, segmentação é conseguida por gramáticas de descodificação restritas baseadas na pronunciação ideal do texto mas permitindo liberdade de repetição e pré-correções baseadas em sílabas. O melhor método usa sílabas como unidades e permite silêncio opcional entre cada sílaba para responder ao problema de pausas intra-palavra. Descodificar uma locução resulta em segmentos candidatos de palavra que serão classificados como correta ou incorretamente pronunciados. O melhor parâmetro para classificar pronunciações incorretas foi uma razão de verosimilhança logarítmica entre a pronunciação ideal e um modelo de enchimento com todos os fones em paralelo, feita de uma forma semelhante a deteção de palavras. Parâmetros adicionais de verosimilhanças individuais de fonemas e de distâncias de Levenshtein entre pronunciação correta e fonemas reconhecidos foram combinados em modelos de múltiplos parâmetros. Professores do 1º ciclo do ensino básico foram convidados a avaliar o nível geral de leitura de crianças em 0-5, resultando em valores de referência. Modelos de regressão que estimem estes valores foram treinados baseados em parâmetros extraídos da anotação automática, com parâmetros diferentes para a leitura de frases e de pseudopalavras. Modelos de regressão de processos Gaussianos obtiveram os melhores resultados com anotação automática, resultados que se aproximaram do uso de parâmetros extraídos da anotação manual. Duas aplicações do trabalho desenvolvido foram construídas: uma demo e um website protótipo. A demo apresenta a metodologia desenvolvida aplicada às crianças da base de dados adquirida. O website protótipo é uma plataforma para professores onde estes podem atribuir tarefas de leitura a vários alunos, dizer às crianças para lerem tarefas usando um microfone, e analisar o nível de leitura automático e a anotação automática.