Teacher Student Learning Research Articles

Doing Things Differently: Responding to the 'policy problem' of Teacher Education in Wales

Responding to a clear justification for the reform of initial teacher education (ITE) in Wales (Furlong, 2015) the Welsh government has committed itself to strengthening provision through 'a truly collaborative system, where universities and schools work in robust partnership, supported by the consortia, recognising the importance of research' (Williams, 2017: 1). The publication of the national criteria for the accreditation of ITE programmes in Wales (Welsh Government, 2017) set out the requirements for all ITE programmes, emphasising the need for an integrated approach to student teacher learning within collaborative models which are research-informed at all levels. Using Vidovich's (2007) dynamic model of policy analysis, which takes into account influences on policy text production at the macro, intermediate and micro levels, the paper examines the process by which these particular reforms (covering a six-year period from 2013 to 2019) have been proposed and set in motion and the ways in which recently-accredited providers in Wales have begun to respond to the challenging agenda that they represent. The paper concludes by offering a view as to what the opportunities and challenges for ITE providers in Wales might be in terms of further developing models of research-informed clinical practice.

Direct Speech-to-Image Translation

Direct speech-to-image translation without text is an interesting and useful topic due to the potential applications in human-computer interaction, art creation, computer-aided design. etc. Not to mention that many languages have no writing form. However, as far as we know, it has not been well-studied how to translate the speech signals into images directly and how well they can be translated. In this paper, we attempt to translate the speech signals into the image signals without the transcription stage. Specifically, a speech encoder is designed to represent the input speech signals as an embedding feature, and it is trained with a pretrained image encoder using teacher-student learning to obtain better generalization ability on new classes. Subsequently, a stacked generative adversarial network is used to synthesize high-quality images conditioned on the embedding feature. Experimental results on both synthesized and real data show that our proposed method is effective to translate the raw speech signals into images without the middle text representation. Ablation study gives more insights about our method.

The development of student teachers’ sense of professional agency in the classroom during teacher education

ABSTRACT Teacher education needs to facilitate the development of student teachers’ capacity for active and skilful learning in classroom interaction, referred to as sense of professional agency in the classroom. It consists of motivation to learn, self-efficacy beliefs about learning, and strategies promoting one’s own and pupils’ learning in the classroom. However, not much is known about how student teachers’ sense of professional agency in the classroom develops during teacher education. This longitudinal study investigated the interrelations between the components of Finnish student teachers’ (N = 268) sense of professional agency in the classroom during teacher education. Structural equation modelling was used to analyse the data that was collected annually during the bachelor’s degree phase of primary school teacher education. The results showed that the construction of student teachers’ sense of professional agency in the classroom was based on being able to interpret and analyse classroom interaction and became more strongly related to their abilities to use peers, more experienced teachers, and pupils as resources in constructing supportive collaborative learning environments and developing functional engaging teaching methods. The findings suggest that professional agency in the classroom is a functional integrative concept for investigating student teacher learning and developing teacher education.

3D LiDAR Map Compression for Efficient Localization on Resource Constrained Vehicles

Large scale 3D maps constructed via LiDAR sensor are widely used on intelligent vehicles for localization in outdoor scenes. However, loading, communication and processing of the original dense maps are time consuming for onboard computing platform, which calls for a more concise representation of maps to reduce the complexity but keep the performance of localization. In this paper, we propose a teacher-student learning paradigm to compress the 3D point cloud map. Specifically, we first find a subset of LiDAR points with high number of observations to preserve the localization performance, which is regarded as the teacher of map compression. An efficient optimization strategy is proposed to deal with the massive data in original map. With the supervision of compressed map, a student model is built by training a random forest model fed with geometric feature descriptors of each point. As a result, the student model is able to compress the map without referring to the expensive numerical optimization. Additionally, by incorporating the features, the innovative student model can be generalized to other new maps while no re-training is required. We conduct thorough experiments on multi-session dataset and KITTI dataset to demonstrate the effectiveness and efficiency of the proposed learning paradigm, and the comparison with other map compression methods. The final results show that the learned student model can achieve efficient map compression with comparable LiDAR based localization performance to the original map at the same time.

The Balanced Loss Curriculum Learning

Curriculum Learning (CL) selects the training samples from the easy to difficult to boost the classification results. Most existing variates of CL measure the difficulty level of an example in an intuitive way, i.e., the loss value between the prediction and the ground truth. This way ignores the different distances to every class boundary, which is implied in the prediction vectors. In this paper, we propose a novel CL framework, named Balance Loss Curriculum Learning(BLCL), to reveal the comprehensive difficulty level of an example and improve the curriculum process based on the deep architectures. We follow the teacher-student learning module and train the teacher & student networks simultaneously. The predictions learned by the teacher network indicates the boundary information and is set as the difficulty measurement to guide the student network. The student network focuses on the easy samples according to the boundary information by decreasing the loss from difficult examples as well as increasing the loss from easy examples. Through extensive and rigor experiments, we demonstrate the effectiveness and generality of our framework.

Como fazer um telejornal universitário em redena pandemia:O caso doESPM no Ar

This article maps elements of laboratory production of the ESPM no Ar newscast conducted by ESPM / SP during the pandemic period of the new coronavirus. The context of remote education promoted the integrated work of subjects from the Journalism courses in São Paulo, Porto Alegre andRio de Janeiro in a program shown live. The case study is adopted as methodology in order to observe aspects of the construction of the informative audiovisual product in the light of the changes stressed by technology and media convergence. Changes in the production routine and changesin the teacher-student learning relationship are identified in the face of the social distance and entirely online classes.

What does the research evidence base tell us about the use and impact of sustainability pedagogies in initial teacher education?

Many sustainability education scholars argue that active, participatory and experimental pedagogies, known as sustainability or ESD pedagogies, provide the most effective strategies for developing learners’ understanding, thinking and ability to act for sustainability. However, factors that impact learning are complex and the extent to which some pedagogies are more effective than others for promoting student learning is contested. The work on sustainability pedagogies has tended not to be problematized. This paper interrogates the literature on sustainability pedagogies in initial teacher education to determine the extent to which the current level of evidence supports pedagogical assumptions held by sustainability education scholars. The findings suggest that factors that impact student teacher learning for sustainability require additional exploration in order to draw reliable conclusions and inform pedagogical decisions that can best support the development of prospective teachers’ understanding, thinking and ability to employ sustainability pedagogies.

Speech Enhancement Based on Teacher–Student Deep Learning Using Improved Speech Presence Probability for Noise-Robust Speech Recognition

In this paper, we propose a novel teacher-student learning framework for the preprocessing of a speech recognizer, leveraging the online noise tracking capabilities of improved minima controlled recursive averaging (IMCRA) and deep learning of nonlinear interactions between speech and noise. First, a teacher model with deep architectures is built to learn the target of ideal ratio masks (IRMs) using simulated training pairs of clean and noisy speech data. Next, a student model is trained to learn an improved speech presence probability by incorporating the estimated IRMs from the teacher model into the IMCRA approach. The student model can be compactly designed in a causal processing mode having no latency with the guidance of a complex and noncausal teacher model. Moreover, the clean speech requirement, which is difficult to meet in real-world adverse environments, can be relaxed for training the student model, implying that noisy speech data can be directly used to adapt the regression-based enhancement model to further improve speech recognition accuracies for noisy speech collected in such conditions. Experiments on the CHiME-4 challenge task show that our best student model with bidirectional gated recurrent units (BGRUs) can achieve a relative word error rate (WER) reduction of 18.85% for the real test set when compared to unprocessed system without acoustic model retraining. However, the traditional teacher model degrades the performance of the unprocessed system in this case. In addition, the student model with a deep neural network (DNN) in causal mode having no latency yields a relative WER reduction of 7.94% over the unprocessed system with 670 times less computing cycles when compared to the BGRU-equipped student model. Finally, the conventional speech enhancement and IRM-based deep learning method destroyed the ASR performance when the recognition system became more powerful. While our proposed approach could still improve the ASR performance even in the more powerful recognition system.

Experimenting with data and analysis in researching the writing practices of student teachers

ABSTRACTPrimarily methodological in its orientation, this paper offers a presentation of ‘research outcomes’ in ways that challenge and disrupt commonplace notions of data and analysis. In an attempt to write against the grain of conventional qualitative research practice and to experiment with alternative encounters with data and analysis, I present ‘data/analysis’ in the form of a play (or imagined performance) that writes into being two Deleuzo-Guattarian principles – the critique of the self-conscious ‘I’ and desire. The play draws on a wider study that examined the potential of writing as a tool for learning for undergraduate student teachers in England. As such, the paper also contributes to debates on the practice and purpose of writing as a method of (professional) inquiry.

Super Vision: The Role of Remote Observation in The Professional Learning of Student Teachers and Novice Placement Tutors

Developments in the application of remote live technology are reshaping conventional methods of classroom observation. Lessons can now be observed through digital networks without the need to physically visit school sites. This study sought to investigate if remote observation could support the professional learning of student teachers and novice placement tutors on an Initial Teacher Education programme, and to examine the perceived challenges. A case study approach was employed, based on qualitative semi-structured interviews with eight students teachers and two novice placement tutors. Findings indicate that professional learning for both student teachers and placement tutors was facilitated by a number of factors. Remote observation reduced reactivity to create more authentic teaching and learning contexts. It also facilitated social learning through dialogue, the provision of detailed feedback from multiple sources and the prompt implementation of recommendations in subsequent remotely observed classes. Limitations and potential further applications of remote live observation are also outlined.

Crossing the boundaries of imagination: The role of a public library exhibition in global learning for student teachers and teacher educators

In this paper, we utilise the theoretical concept of ‘boundary crossing’ to explore how, as Australian university teacher educators, we worked with library curators and a class of student teachers to mount a public exhibition of their group work. We consider how the students crossed ‘boundaries of imagination’ in symbolic representation and critical analysis by creating artistic installations to express global education themes. We reflect on what we, the university educators, learned about crossing institutional and pedagogical boundaries to mutually facilitate new learning. We argue for the importance of shared public pedagogy as offering new avenues for teacher education and of connecting with local communities.

Compressing CNN-DBLSTM models for OCR with teacher-student learning and Tucker decomposition

Integrated convolutional neural network (CNN) and deep bidirectional long short-term memory (DBLSTM) based character models have achieved excellent recognition accuracies on optical character recognition (OCR) tasks, along with large amount of model parameters and massive computation cost. To deploy CNN-DBLSTM model in products with CPU server, there is an urgent need to compress and accelerate it as much as possible, especially the CNN part, which dominates both parameters and computation. In this paper, we study teacher-student learning and Tucker decomposition methods to reduce model size and runtime latency for CNN-DBLSTM based character model for OCR. We use teacher-student learning to transfer the knowledge of a large-size teacher model to a small-size compact student model, followed by Tucker decomposition to further compress the student model. For teacher-student learning, we design a novel learning criterion to bring in the guidance of succeeding LSTM layer when matching the CNN-extracted feature sequences of the large teacher and small student models. Experimental results on large scale handwritten and printed OCR tasks show that, using teacher-student learning alone achieves 9.90 × footprint reduction and 15.23 × inference speedup yet without degrading recognition accuracy. Combined with Tucker decomposition method, we can compress and accelerate the model further. The decomposed model achieves 11.89 × footprint reduction and 22.16 × inference speedup while suffering no or only a small recognition accuracy degradation against the large-size baseline model.

Incorporating lesson study in ITE: organisational structures to support student teacher learning

ABSTRACTLesson study is a collaborative model of professional development that provides teachers with the opportunity to learn from the planning, conduction and reflection of a specifically designed research lesson. Lesson study is increasingly incorporated in initial teacher education (ITE) and calls have been made to further detail the lesson study structures that support student teacher learning. In this paper the core organisational elements of incorporating lesson study in ITE are investigated as a comparative case study situated in two distinct educational and cultural settings: one in Ireland (post-primary) and the other in Norway (primary). Qualitative data, including student teachers’ planning notes, research lesson designs, and observation notes from live research lessons, were analysed and findings reveal three structural factors for the incorporation of lesson study in ITE. First, defined roles of mentor and university teachers as knowledgeable others; second, the engagement of student teachers at each phase of lesson study, with the inclusion of case pupils; and third, the conduction of the lesson study cycle on both university campus and in the classroom. These findings contribute to the growing literature of lesson study in ITE and may support the inclusion of lesson study in ITE programmes.

The Universities and initial teacher education; challenging the discourse of derision. The case of Wales

ABSTRACTFor nearly 40 years the quality and value of the contribution of universities to initial teacher education has been brought into question. This is particularly so in England where the ongoing ‘discourse of derision’ has resulted in universities no longer being seen as necessary partners in the process. More recently, similar challenges have taken place in other countries such as USA and Australia. However in 2013, when the Welsh Government turned its attention to the apparent low quality of its current provision, rather than challenging the role of universities, it chose to strengthen their contribution. There were however to be important changes that insisted that universities put the student teacher learning at the heart of course planning, that universities clarify their own distinctive contribution and that they work in close collaboration with schools. While this approach to initial teacher education is not new, this is the first time that such a model has been implemented on a national scale. This paper outlines the nature, rationale and underlying research for the reforms in Wales. It concludes by speculating on their likely impact in raising the quality of provision and securing the future contribution of universities to teachers’ learning.

Opportunities for professional learning about mathematics instruction: the role of joint work in student-teaching triads

This paper reports about professional learning related to mathematics instruction in student-teaching triads. The study focuses on how the joint work of triads offers opportunities for the professional learning of student teachers (interns), mentor teachers, and supervisors. Two elementary student-teaching triads reported different areas of and opportunities for professional learning related to mathematics instruction, with more areas and richer opportunities reported by the triad with more extensive collaboration among its members. Findings suggest the need for professional development activities in which mentors and supervisors can examine their stances toward joint work and professional learning in triads.

Constructing learning spaces – knowledge development in work-based learning

Purpose The purpose of this paper is to discuss how kindergarten, as a learning arena equal to a university college, creates learning spaces that engage or intervene in the professional learning of student teachers in early childhood education. Design/methodology/approach This paper is based on narratives from students in work-based education. Findings The paper addresses the complexity of education by outlining how the concept of learning is applied in earlier research on work-based learning (WBL). Research limitations/implications This earlier understanding is complemented this with two theoretical lenses (sociocultural and sociomaterial thinking) to analyse a constructed narrative from the students. Originality/value The two theoretical positions open up to examine knowledge development and potentially enrich the picture of learning spaces in experiential WBL, going beyond the student as an individual learner.

Robust Student Network Learning.

Deep neural networks bring in impressive accuracy in various applications, but the success often relies on heavy network architectures. Taking well-trained heavy networks as teachers, classical teacher-student learning paradigm aims to learn a student network that is lightweight yet accurate. In this way, a portable student network with significantly fewer parameters can achieve considerable accuracy, which is comparable to that of a teacher network. However, beyond accuracy, the robustness of the learned student network against perturbation is also essential for practical uses. Existing teacher-student learning frameworks mainly focus on accuracy and compression ratios, but ignore the robustness. In this paper, we make the student network produce more confident predictions with the help of the teacher network, and analyze the lower bound of the perturbation that will destroy the confidence of the student network. Two important objectives regarding prediction scores and gradients of examples are developed to maximize this lower bound, to enhance the robustness of the student network without sacrificing the performance. Experiments on benchmark data sets demonstrate the efficiency of the proposed approach to learning robust student networks that have satisfying accuracy and compact sizes.

基于“互联网+”背景下师生群组学习模式的构建与案例研究

基于“互联网+”背景下师生群组学习模式的构建与案例研究

Experiential strategies and learning in environmental education: lessons from a teacher training college in Tanzania

ABSTRACTThe teaching of environmental topics is still the main challenge. In this paper we discuss how student teachers learning of experiential teaching strategies influenced their teaching of EE in primary schools in Tanzania. Student teachers learning to use tree-planting as example of outdoor learning for the teaching of environmental topics were the main sources of data. The first author collected data by interview and participant observation and then transcribed and organized the data as part of the analysis process. The results revealed that the student teachers teaching of environmental topics were active and particptory. We discuss the results in line with the quality of facilitation process. The study demonstrates the potential of experiential learning strategies in promoting actions related to environmental sustainability among student teachers and hopefully, their future students and the wider community.

Development of problem-solving skill by using active learning for student teachers in Introductory Physics

This research was to study student teachers’ problem-solving skill after they took active learning Introductory Physics class in academic year 2017. There were 101 student teachers taking the course as a compulsory for 17 weeks. The one group pre-test and post-test design was used as quantitative research and also qualitative research. Twelve active learning activities were used throughout the whole semester. The IPST problem-solving assessment and rubrics were used to develop the research instrument to evaluate students’ problem-solving skills. The focus group interview was conducted to collect qualitative data on student teachers’ opinions of active learning activities. The paired-sample t-test was applied to compare problem-solving scores before and after they took in the class. The interview data, content analysis was used to summarise students’ opinions. The results indicated that their problem-solving skills were significantly higher after learning through active learning activities at .05 confident levels. Students’ opinions mostly reflected the positive feeling toward active learning activities and they recommended for other classes to use the same pedagogy.