Search for: All records

Specific to the topic of oxidation–reduction (redox), teachers are obligated by the discipline to prioritise symbolic traditions such as writing equations, documenting oxidation states, and describing changes ( e.g. , what undergoes oxidation/reduction). Although the chemistry education research community endorses connecting the vertices of Johnstone's triangle, how symbolic traditions undermine chemistry concept development, especially during lesson planning and teaching, is underexplored. To clarify this gap, we use the Mangle of Practice framework to unpack the clash between symbolic vs. particulate-focused instruction. We investigate teachers’ ( n = 3) co-planning and micro-teaching of a redox learning design at the VisChem Institute-2 using a narrative approach and video research methods. Our results show that the traditions of redox instruction are problematically entrenched in chemistry symbols. Mnemonics, the single replacement reaction scheme, and the written net ionic equation all constrain instruction focused on chemical mechanism and causality in various ways. We assert that the nature of redox knowledge in terms of what is worth teaching and learning must first be re-evaluated for reform-based efforts to succeed. Implications and suggestions for chemistry teaching and research at both secondary and tertiary levels are discussed. 

Despite years of research and practice inspired by chemistry education research, a recent report shows that US secondary instruction is not aligned with current national reform-based efforts. One means to mitigate this discrepancy is focusing on pedagogical conceptual change, its precursors (higher self-efficacy and pedagogical discontentment), and the subtleties of its mechanisms (assimilation and accommodation). In this study, we investigate the final reflections of participants ( N = 35) who completed our professional development program known as the VisChem Institute (VCI). Our results show that Johnstone's triangle as well as evidence, explanations, and models can be conducive for stimulating pedagogical discontentment among VCI teachers who exhibit higher self-efficacy. In addition, how VCI teachers assimilate and/or accommodate reform-based chemistry teaching ideas problematizes conventional assumptions, broadens application of novel theories, and is germane to introductory chemistry learning environments across the world. Implications and recommendations for chemistry instruction and research at both secondary and tertiary levels are discussed. 

Social distancing can reduce the infection rates in respiratory pandemics such as COVID-19. Traffic intersections are particularly suitable for monitoring and evaluation of social distancing behavior in metropolises. Hence, in this paper, we propose and evaluate a real-time privacy-preserving social distancing analysis system (B-SDA), which uses bird’s-eye view video recordings of pedestrians who cross traffic intersections. We devise algorithms for video pre-processing, object detection, and tracking which are rooted in the known computer-vision and deep learning techniques, but modified to address the problem of detecting very small objects/pedestrians captured by a highly elevated camera. We propose a method for incorporating pedestrian grouping for detection of social distancing violations, which achieves 0.92 F1 score. B-SDA is used to compare pedestrian behavior in pre-pandemic and during-pandemic videos in uptown Manhattan, showing that the social distancing violation rate of 15.6% during the pandemic is notably lower than 31.4% prepandemic baseline. 

Researchers have typically identified and characterized teachers’ knowledge bases ( e.g. , pedagogical content knowledge and subject matter knowledge) in an effort to improve enacted instructional strategies. As shown by the Refined Consensus Model (RCM), understanding teacher learning, beliefs, and practices is predicated on the interconnections of such knowledge bases. However, lesson planning (defined as the transformation of subject matter knowledge to enacted pedagogical content knowledge) remains underexplored despite its central position in the RCM. We aim to address this gap by developing a conceptual framework known as Pedagogical Chemistry Sensemaking (PedChemSense). PedChemSense theoretically expands upon the RCM that generates actionable guidelines to support chemsistry teachers’ lesson planning. We incorporate the constructs of sensemaking, Johnstone's triangle, and the models for perspective to provide a lesson-planning mechanism that is specific, accessible, and practical, respectively. Lesson examples from our own professional development contexts, the VisChem Institute, demonstrate the efficacy of PedChemSense. By leveraging teachers’ sensemaking of the limitations and utility of models, PedChemSense facilitates teachers’ designing for opportunities to advance their students’ chemistry conceptual understanding. Implications and recommendations for chemistry instruction and research at secondary and undergraduate levels are discussed. 

Social media platforms are playing increasingly critical roles in disaster response and rescue operations. During emergencies, users can post rescue requests along with their addresses on social media, while volunteers can search for those messages and send help. However, efficiently leveraging social media in rescue operations remains challenging because of the lack of tools to identify rescue request messages on social media automatically and rapidly. Analyzing social media data, such as Twitter data, relies heavily on Natural Language Processing (NLP) algorithms to extract information from texts. The introduction of bidirectional transformers models, such as the Bidirectional Encoder Representations from Transformers (BERT) model, has significantly outperformed previous NLP models in numerous text analysis tasks, providing new opportunities to precisely understand and classify social media data for diverse applications. This study developed and compared ten VictimFinder models for identifying rescue request tweets, three based on milestone NLP algorithms and seven BERT-based. A total of 3191 manually labeled disaster-related tweets posted during 2017 Hurricane Harvey were used as the training and testing datasets. We evaluated the performance of each model by classification accuracy, computation cost, and model stability. Experiment results show that all BERT-based models have significantly increased the accuracy of categorizing rescue-related tweets. The best model for identifying rescue request tweets is a customized BERT-based model with a Convolutional Neural Network (CNN) classifier. Its F1-score is 0.919, which outperforms the baseline model by 10.6%. The developed models can promote social media use for rescue operations in future disaster events. 

Transformers provide a class of expressive architectures that are extremely effective for sequence modeling. However, the key limitation of transformers is their quadratic memory and time complexity O(L2) with respect to the sequence length in attention layers, which restricts application in extremely long sequences. Most existing approaches leverage sparsity or low-rank assumptions in the attention matrix to reduce cost, but sacrifice expressiveness. Instead, we propose Combiner, which provides full attention capability in each attention head while maintaining low computation and memory complexity. The key idea is to treat the self-attention mechanism as a conditional expectation over embeddings at each location, and approximate the conditional distribution with a structured factorization. Each location can attend to all other locations, either via direct attention, or through indirect attention to abstractions, which are again conditional expectations of embeddings from corresponding local regions. We show that most sparse attention patterns used in existing sparse transformers are able to inspire the design of such factorization for full attention, resulting in the same sub-quadratic cost (O(L log(L)) or O(L√L)). Combiner is a drop-in replacement for attention layers in existing transformers and can be easily implemented in common frameworks. An experimental evaluation on both autoregressive and bidirectional sequence tasks demonstrates the effectiveness of this approach, yielding state-of-the-art results on several image and text modeling tasks.