Peer assessment actively engages peers in the formative assessment and evaluation of work produced by a peer. This thesis explores how social processes, such as classroom talk, influence the quality of children’s learning in more interactive contexts of PA. This focus is needed since children often find PA challenging as they may not have the interpersonal skills to collaborate effectively leading them to use talk ineffectively as a tool for learning. This research was interventionist and children in the year three and four classes I taught received Thinking Together lessons as a strategy to enhance the quality of their talk in contexts of peer assessment. Methods used to examine the impact of the talk intervention, and to gain greater insights into the role that the social context plays in peer assessment, included transcribed digital audio recordings, open ended observations, semi-structured interviews, mind maps and children’s work. Qualitative data were analysed using thematic coding analysis whilst data in transcripts were quantitatively analysed to calculate the frequency of words and phrases associated with exploratory talk before and after the intervention. Findings suggest that children’s characteristics influence the way they communicate in contexts of PA and some of the most challenging learners seemed to benefit most from the talk intervention in terms of its influence on their ability to collaborate, hypothesise and reason throughout the peer assessment tasks. The findings also draw attention to previously under-researched PA social processes such as discussion, negotiation and peer questioning that lead to outcomes for learners such as self assessment. The main conclusions drawn are that more interactive kinds of peer assessment might be viewed as a differentiated and discursive practice where teachers consider the various needs of learners, based on their individual characteristics, and provide appropriate support so they are able to collaborate and use language for mediating effective PA practice.

Automated short-answer marking cannot 'guarantee' 100% agreement between the marks generated by a software system and the marks produced separately by a human. This problem has prevented automated marking systems from being used in high-stake short-answer marking. Given this limitation, can an automated short-answer marking system have any practical application? This thesis describes how an automated short-answer marking system, called IndusMarker, can be effectively used to improve learning and teaching.The design and evaluation of IndusMarker are also presented in the thesis. IndusMarker is designed for factual answers where there is a clear criterion for answers being right or wrong. The system is based on structure matching, i.e. matching a pre-specified structure, developed via a purpose-built structure editor, with the content of the student's answer text. An examiner specifies the required structure of an answer in a simple purpose-designed language called Question Answer Markup Language (QAML). The structure editor ensures that users construct correct required structures (with respect to QAML's syntax and informal semantics) in a form that is suitable for accurate automated marking.

No description available.

This thesis documents the use of Natural Language Processing (NLP) in Computer Assisted Language Learning (CALL) and its contribution to the learning experience of students studying Arabic as a foreign language. The goal of this project is to build an Intelligent Computer Assisted Language Learning (ICALL) system that provides computational assistance to learners of Arabic by teaching grammar, producing homework and issuing students with immediate feedback. To produce this system we use the Parasite system, which produces morphological, syntactic and semantic analysis of textual input, and extend it to provide error detection and diagnosis. The methodology we adopt involves relaxing constraints on unification so that correct information contained in a badly formed sentence may still be used to obtain a coherent overall analysis. We look at a range of errors, drawn from experience with learners at various levels, covering word internal problems (addition of inappropriate affixes, failure to apply morphotactic rules properly) and problems with relations between words (local constraints on features, and word order problems). As feedback is an important factor in learning, we look into different types of feedback that can be used to evaluate which is the most appropriate for the aim of our system.

Submitted by Livia Mello (liviacmello@yahoo.com.br) on 2016-09-28T14:16:07Z No. of bitstreams: 1 TeseRVR.pdf: 17301603 bytes, checksum: ca1c2799fb4bf31f2af4bdd75cd2a815 (MD5) / Approved for entry into archive by Ronildo Prado (ronisp@ufscar.br) on 2016-09-30T19:52:52Z (GMT) No. of bitstreams: 1 TeseRVR.pdf: 17301603 bytes, checksum: ca1c2799fb4bf31f2af4bdd75cd2a815 (MD5) / Approved for entry into archive by Ronildo Prado (ronisp@ufscar.br) on 2016-09-30T19:53:02Z (GMT) No. of bitstreams: 1 TeseRVR.pdf: 17301603 bytes, checksum: ca1c2799fb4bf31f2af4bdd75cd2a815 (MD5) / Made available in DSpace on 2016-09-30T19:59:08Z (GMT). No. of bitstreams: 1 TeseRVR.pdf: 17301603 bytes, checksum: ca1c2799fb4bf31f2af4bdd75cd2a815 (MD5) Previous issue date: 2014-11-25 / Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / Training aims to improve the professional competences (knowledge, skills and attitudes) and it is essential particularly in risk activities to the life, property and environment. Due to the risk and cost of traditional practical training, serious games, which reproduce systems, processes and real-world objects in a safe virtual environment has been used as an effective tool for learning and training or motivation to learning, and support for assessment during and post-action. On the other hand, the development of serious games is a complex process, has development high cost (time, money, resources), need highly qualified professionals, who hold requirements of learning, assessment, simulation and game itself. Different methodologies have emerged for developing serious games, however, in general, they do not include the different dimensions and need to be combined and balanced with other methodologies to achieve the training and assessment objectives. This thesis presents a novel iterative methodology (named DevJSTA) for the development of serious games for training and human performance assessment (named JSTA), which integrates multidisciplinary teams by providing guidelines, artifacts, game elements and simulation models (formalisms of discrete events, automata and Fuzzy logic) to integrate and include concepts such as effective and experiential learning in serious games, besides providing feedback during the simulation and at the end. This methodology DevJSTA describes the professional roles, the development life cycle, the generated products and the technologies and tools that can be used to create these products, as well as a support architecture and generic components. The main differential of this methodology is its holistic approach that integrates pedagogical principles of learning and assessment in the development of simulation. The methodology was evaluated through interviews with software engineering experts, serious games developers and domain experts. The results were positive and confirmed the need of these new methodology that is integrative. To validate the methodology, one JSTA was created to train firefighters to control the cooking gas leak, and to evaluate the trainee’s perception and performance. Serious game fidelity and effectiveness were also validated. Moreover, the serious game development, with the use of the methodology DevJSTA, allowed to evaluate an opportunity for improvement in the current protocol that is used in the Fire Department. / Um treinamento visa aprimorar as competências dos profissionais no âmbito do conhecimento, habilidades e reforço de boas atitudes e é essencial particularmente em atividades que envolvem risco à vida, ao patrimônio e ao meio ambiente. Devido ao risco e custo de treinamentos práticos tradicionais, jogos sérios de treinamento que reproduzem sistemas, processos e objetos do mundo real em um ambiente virtual seguro vêm sendo usados como instrumento efetivo de aprendizagem e treinamento ou motivação ao aprendizado, além de apoio para avaliações durante e pós-ação. Por outro lado, o desenvolvimento de jogos sérios é um processo complexo, com alto custo de desenvolvimento (recursos humanos, materiais, financeiros, espaço e tempo), que requer profissionais qualificados, que detêm conhecimentos de aprendizagem, avaliação, simulação e de jogos em si. Diferentes metodologias emergiram para o desenvolvimento de jogos sérios que, de maneira geral, não contemplam as suas diferentes dimensões, necessitando serem combinadas e equilibradas com outras metodologias para atingirem os objetivos de treinamento e avaliação. Este trabalho apresenta uma metodologia iterativa e integradora (DevJSTA) para o desenvolvimento de Jogos Sérios para Treinamento e Avaliação do desempenho humano (JSTA), que integra equipes multidisciplinares, provendo desde orientações e artefatos até elementos de jogos e modelos de simulação (com modelos formais de eventos discretos, autômatos e lógica Fuzzy), para integrar e inserir conceitos de aprendizagem efetiva e experiencial aos jogos sérios, além de avaliação do desempenho humano durante e pós-ação. Na metodologia DevJSTA são descritos os papeis dos profissionais envolvidos, cada processo do ciclo de vida de desenvolvimento, os produtos gerados e as tecnologias e ferramentas que podem ser usadas para criar estes produtos, além de uma arquitetura de suporte e componentes genéricos. O principal diferencial desta metodologia é sua abordagem holística que integra princípios pedagógicos de aprendizagem e avaliação no desenvolvimento da simulação. A metodologia foi submetida à avaliação, por meio de entrevistas com especialistas das áreas de engenharia de software, desenvolvimento de jogos e domínio de aplicação. Os resultados positivos corroboram com a necessidade desta nova metodologia que é integradora. Para validar a metodologia, foi criado um JSTA para treinar bombeiros no controle de vazamento de gás de cozinha, e avaliar a percepção e o desempenho de cada aprendiz. O JSTA foi também validado quanto a sua fidelidade e efetividade. Além dos resultados positivos, a criação do JSTA, com a aplicação da metodologia DevJSTA, permitiu avaliar uma oportunidade de melhoria no protocolo atual que é utilizado no Corpo de Bombeiros do Estado de São Paulo.