Search

Access to body-based resources has been shown to augment cognitive processes, but not all movements equally aid reasoning. Interactive technologies, like dynamic geometry systems (DGS), potentially amplify the link between movement and geometric representation, thereby deepening students' understanding of geometric properties. This study investigated the influence of access to a DGS on students' geometric reasoning and gesture production compared to the impact of only spontaneous gestures. Access to simulations did not significantly increase the likelihood of students producing correct mathematical insights compared to spontaneous gestures. However, when controlling gesture production, DGS access became a significant predictor for transformational proof production. We also found that access to DGS influenced the type of gestures students produced during geometric reasoning tasks. This study underlines the vital role of gestures in geometric reasoning and suggests that while DGS offers benefits, spontaneous gestures remain a crucial embodied resource for exploring geometric concepts. [This paper was published in: "Proceedings of the 18th International Conference of the Learning Sciences - ICLS 2024," edited by R. Lindgren et al., International Society of the Learning Sciences, 2024, pp. 131-38.]

In this paper duality between real world phenomenon and mathematics will be discussed. This duality exists for many years which considerably remarked in the history of mathematics curriculum. One of the good potential for filling the gap between real world phenomenon and mathematical ideas would be modelling tasks which require performing the modelling cycle. In this paper after discussion about modelling in more details, two modelling cycles will be elaborated in the context of mathematical tasks that are related to everyday life. Main ideas behind different modelling cycles is starting point that would be from real world situation. Then gradually mathematical elements of real world phenomenon identified and mathematics problem will have shaped in a form that could be solved through mathematical problem solving techniques. Furthermore, it would be important phase that call interpretation of mathematics answer and check it in front of real world situation. This paper will be followed by discussion about learning theories that support the idea of modelling. Finally, two educational challenges (Design good modelling tasks and assessing of students performance) in mathematical modelling activity will be discussed. [For the complete proceedings, see ED655360.]

Although some research has been conducted on the importance of mediation in language learning and teaching (Dendrinos, 2006; González-Davies, 2020; Piccardo, 2012, 2020; Scarino, 2016), there is still scarce research on the integration of plurilingualism, mediation, and technology. Through qualitative and quantitative methodology, this paper investigates teachers' abilities in Italy and Spain to integrate plurilingualism, mediation, and technology. A survey has been distributed in order to explore teachers' attitudes towards the use of Information and Communication Technologies (ICTs) and their awareness of the definition and importance of mediation. Furthermore, some mediation tasks performed by teachers in which they had to integrate mediation, plurilingualism, and technology were analysed. [For the complete volume, "Intelligent CALL, Granular Systems and Learner Data: Short Papers from EUROCALL 2022 (30th, Reykjavik, Iceland, August 17-19, 2022)," see ED624779.]

Computer-mediated communication tools facilitate international collaboration projects between foreign language learners and peers abroad (O'Dowd, 2018). Social Virtual Reality (VR) applications allow for synchronous interactions and task-based communication in which learners can experience telepresence and immersion and conversate in a foreign language. Based on previous pilot experiences (Jauregi-Ondarra, Gruber, & Canto, 2020, 2021), this Virtual Exchange (VE) project aims to investigate how the specific affordances of Social High-immersion VR (SHiVR) in conjunction with designed tasks influence interaction patterns, and learning episodes. The VE took place between two groups of university students in the Netherlands (N=15) and Cyprus (N=14) through SHiVR in March 2022. The main aims of the tasks were to raise student intercultural awareness, stimulate task-based communication processes using English as a lingua franca and digital pedagogical competences of language education students. Different sources of data were gathered and analysed. In this paper, we describe and present the pedagogical experience and the initial results. [For the complete volume, "Intelligent CALL, Granular Systems and Learner Data: Short Papers from EUROCALL 2022 (30th, Reykjavik, Iceland, August 17-19, 2022)," see ED624779.]

This paper explores use of a set of diagrammatic tools for representation and analysis of the moves a teacher makes implementing a mathematical modelling task. The focus here is on identifying what the teacher did so we can subsequently interrogate this, as to the why. Data include pre and post lesson teacher interviews and transcripts of a video and audio-recorded task implementation. The analytical tools developed, with one teacher and one task early in a three-year project were particularly useful in ascertaining what the teacher moves were as we subsequently sought to determine reasons for these.

This paper presents an initial analysis of 10 Preschool children's responses to a look-draw-trace-draw task. The findings suggest that figure-tracing helped half the children to produce a more accurate representation of the geometric figure presented to them, in their second drawing.

There is an educational expectation that children's natural drawing will develop into proficient mathematical representations and formal diagrams, yet there is little research available to guide the assessment and development of children's mathematical drawing skills. The aim of this paper is to explore how Year 2 children (approx. 7 years) chose to represent their interpretations of a simple story that is suggestive of the take-away subtraction process, and what changes occurred when the drawing task was repeated 6 months later. Analysis of 13 pairs of drawings revealed changes in what the children drew (categories of number representations) and how they drew it (style). The findings suggest that substantial change in children's representational ability can occur in within 6 months.

Climate change is a complex issue that requires collective actions from various disciplines, including educating future generations about sustainability. Some countries such as Indonesia and Australia have included sustainability into their school curricula, which require teachers to design and implement tasks related to a sustainability issue. However, teachers may not have the capability and understanding to do so. The challenges are difficult for mathematics teachers because designing a mathematics task pertaining to sustainability is rarely explored. Therefore, this paper draws from mathematics education literature to provide and discuss a new framework for designing sustainability-related, mathematics tasks (green mathematic tasks).

The present exploratory study investigated the efficacy of secondary face-to-face classroom-based English as a Foreign Language (EFL) instruction digitally supported by the FeedBook, an interactive computer assisted language learning web-based suite of exercises providing item-level scaffolded feedback. Seventy-seven native (L1) German seventh-grade students used the FeedBook during four two-week training periods (cycles, here we analyse Cycle 2 and 3). Classroom and FeedBook practice occurred in parallel, except for grammar constructions for which only FeedBook practice was provided (controls). At the end of Cycle 3, students engaged in a classroom-based communicative task for which the practised constructions were relevant. Custom-designed pre- and post-tests, administered via the FeedBook, assessed language accuracy in each cycle. Mixed-effect models revealed significant pre-/post-test accuracy gains independent of the learners' proficiency in English, but not in controls. Gains from digitally supported instruction were also positively related to accurate use of EFL grammar constructions in the communicative task. Overall, the results indicate learning benefits for face-to-face classroom-based instruction supported by the FeedBook, evidenced both in grammar tests and in communicative activities. [For the complete volume, "Intelligent CALL, Granular Systems and Learner Data: Short Papers from EUROCALL 2022 (30th, Reykjavik, Iceland, August 17-19, 2022)," see ED624779.]

The aim of the study was to investigate whether telecollaboration is a suitable environment for the acquisition of the speech act of apologies. Participants were aerospace engineering students from the Universitat Politècnica de València (Spain) who performed six open role-plays to elicit apologies. The Control Group (CG, n=17) carried out the task in pairs with their Spanish classmates in a Face-To-Face (FTF) setting, while the Experimental Group (EG, n=7) conducted the task with first language (L1) or highly proficient speakers of English from the University of Bath (UK) through synchronous Zoom sessions. The results of the descriptive analysis revealed a higher tendency of improvement in the EG, which also used a higher number of strategies compared to the CG. Findings from the quantitative analysis carried out through an Eta coefficient revealed a significant correlation (r=0.71) between the number of strategies used and the modality where they were performed. [For the complete volume, "Intelligent CALL, Granular Systems and Learner Data: Short Papers from EUROCALL 2022 (30th, Reykjavik, Iceland, August 17-19, 2022)," see ED624779.]

The current study investigated the role of visual short-term memory, working memory, and declarative memory as individual differences in the earliest stages of vocabulary and syntactic learning in "Duolingo"-based language instruction. Thirty-eight L1-Greek adults completed memory tasks and engaged in learning Navajo on "Duolingo." Subsequently, vocabulary and syntax were assessed respectively via a word recognition, a word translation, and a grammaticality judgement task. Multiple regression analyses revealed an advantage for distributed practice both in vocabulary and syntax, after controlling for amount of practice. Further, declarative memory played a significant role in learning syntax and vocabulary, when measured in a word translation task. Extending the analysis for the first time to app-based environments, the results of the present study confirm the importance of declarative memory and distributed practice in adult acquisition of L2 vocabulary and syntax. [For the complete volume, "Intelligent CALL, Granular Systems and Learner Data: Short Papers from EUROCALL 2022 (30th, Reykjavik, Iceland, August 17-19, 2022)," see ED624779.]

The aim of this study is to compare how two different Virtual Exchanges (VE) contribute to language learning in a Japanese university context. Task-Based Language Teaching (TBLT) using VE has been a focus in this study. There are huge advantages of using technology in language teaching such as increasing language use opportunities, feedback, and others (Lee, 2016). However, many recent TBLT and VE studies have focused on students' interaction in the classroom (Hagley, 2020). This study investigated whether two different types of VEs have the potential to promote 'real-world communication' in a university context. The findings show that there was no significant difference regarding lexical complexity between both modes. The result suggests that both modes are unlikely to enhance students' vocabularies. [For the complete volume, "Intelligent CALL, Granular Systems and Learner Data: Short Papers from EUROCALL 2022 (30th, Reykjavik, Iceland, August 17-19, 2022)," see ED624779.]

Cloze items are commonly used for both assessing learning and as a learning activity. This paper investigates the selection of sentences for cloze item creation by comparing methods ranging from simple heuristics to deep learning summarization models. An evaluation using human-generated cloze items from three different science texts indicates that simple heuristics substantially outperform summarization models, including state-of-the-art deep learning models. These results suggest that sentence selection for cloze item generation should be considered a distinct task from summarization and that continued advances on this task will require large datasets of human-generated cloze items. [This paper was published in: "Joint Proceedings of the Workshops at the 14th International Conference on Educational Data Mining," Vol. 3051, LDI-6, edited by T. W. Price and S. San Pedro, CEUR-WS.org, 2021.]

Automated scoring of student language is a complex task that requires systems to emulate complex and multi-faceted human evaluation criteria. Summary scoring brings an additional layer of complexity to automated scoring because it involves two texts of differing lengths that must be compared. In this study, we present our approach to automate summary scoring by evaluating a corpus of approximately 5,000 summaries based on 103 source texts, each summary being scored on a 4-point Likert scale for seven different evaluation criteria. We train and evaluate a series of Machine Learning models that use a combination of independent textual complexity indices from the ReaderBench framework and Deep Learning models based on the Transformer architecture in a multitask setup to predict concurrently all criteria. Our models achieve significantly lower errors than previous work using a similar dataset, with MAE ranging from 0.10-0.16 and corresponding R[superscript 2] values of up to 0.64. Our findings indicate that Longformer-based models are adequate for contextualizing longer text sequences and effectively scoring summaries according to a variety of human-defined evaluation criteria using a single Neural Network. [This paper was published in: "AIED 2022, LNCS 13355," edited by M. M. Rodrigo et al., Springer Nature Switzerland, 2022, pp. 756-761.]

Collaborative learning is a complex process during which two or more learners exchange opinions, construct shared knowledge, and solve problems together. While engaging in this interactive process, learners' satisfaction toward their partners plays a crucial role in defining the success of the collaboration. If intelligent systems could predict peer satisfaction early during collaboration, they could intervene with adaptive support. However, while extensive studies have associated peer satisfaction with factors such as social presence, communication, and trustworthiness, there is no research on automatically predicting learners' satisfaction toward their partners. To fill this gap, this paper investigates the automatic prediction of peer satisfaction by analyzing 44 middle school learners' interactions during collaborative coding tasks. We extracted three types of features from dialogues: (1) linguistic features indicating semantics; (2) acoustic-prosodic features including energy and pitch; and (3) visual features including eye gaze, head pose, facial behaviors, and body pose. We then trained several regression models to predict the peer satisfaction scores that learners received from their partners. The results revealed that head position and body location were significant indicators of peer satisfaction: lower head and body distances between partners were associated with more positive peer satisfaction. This work is the first to investigate the multimodal prediction of peer satisfaction during collaborative problem solving, and represents a step toward the development of real-time intelligent systems that support collaborative learning. [For the full proceedings, see ED623995.]

In this paper, we explore student solutions to a free response mathematical assessment task which had opportunities for students to notice structural properties in the context of number systems. In total, 308 students aged between 10 years to 13 years participated in the study. Their responses were analysed to determine whether they noticed algebraic structures in a task using equivalent equations. Findings indicate that students were able to recognise equation pairs that drew on both the associative and distributive properties. A limited number of students were able to notice the general structure and draw on number properties to support their claims, moving beyond using algorithmic thinking.

This paper will elaborate five levels of algebraic generalisation based on an analysis of students' responses to Reframing Mathematical Futures II (RMFII) tasks designed to assess algebraic reasoning. The five levels of algebraic generalisation will be elaborated and illustrated using selected tasks from the RMFII study. The five levels will be matched against the eight zones identified in the RMFII study supported by its Rasch analysis. We identify two shifts where students' capacity to generalise appear difficult to navigate. The first being where students move from noticing and describing regularities to formalising these regularities into verbal or symbolic expressions. The second is where students use their understanding of equivalence based on relational thinking to write and recognise equivalent algebraic expressions.

Building from the evidence-based learning progression in geometric reasoning from the RMFII project, this paper presents data from students' solutions to three problems in geometry and measurement situations to identify key components needed to nurture reasoning. To show emerging analytical reasoning students must coordinate multiple pieces of information and demonstrate cognitive flexibility in their use of visualisation, diagrams, language, and symbols.

Covariational reasoning and the creation and interpretation of graphs of covariational situations are important skills in math and science. Unfortunately, research shows that students often struggle to make meaningful connections between graphs and the covariational situations they represent. Educational activities designed to help students overcome this struggle tend to use either student-generated or automatically-generated graphs, and have students either act out covariational situations or more passively observe them. In this paper, we present the design of a tool and task that enabled two students to simultaneously embody both the creation of a graph and the covariational actions that the graph represents. Through a process of collaborative instrumentation, the students made meaningful connections between their motions and the embodied traces they created as they reasoned about the covarying quantities of height and time. [For the complete proceedings, see ED630210.]

This paper describes how fifth-grade English Learner students (ELs) in an urban school district develop the mathematics register during a problem-solving lesson. It provides examples of students' work to illustrate how they use the mathematics register to communicate their mathematical ideas orally and in writing. The teacher implemented teaching practices such as mathematics discourse to facilitate their students' development of the mathematics register during the problem-solving lesson. Students were engaged in a problem-solving task that involved fractions. Findings provide insights into EL students' challenges when learning the mathematics register and inform instruction about the importance of incorporating teaching practices such as paraphrasing and assessing others' reasoning to support students in learning the mathematics register through problem-solving. [For the complete proceedings, see ED630210.]

How can we test whether state-of-the-art generative models, such as Blender and GPT-3, are good AI teachers, capable of replying to a student in an educational dialogue? Designing an AI teacher test is challenging: although evaluation methods are much-needed, there is no off-the-shelf solution to measuring pedagogical ability. This paper reports on a first attempt at an AI teacher test. We built a solution around the insight that you can run conversational agents in parallel to human teachers in real-world dialogues, simulate how different agents would respond to a student, and compare these counterpart responses in terms of three abilities: speak like a teacher, understand a student, help a student. Our method builds on the reliability of comparative judgments in education and uses a probabilistic model and Bayesian sampling to infer estimates of pedagogical ability. We find that, even though conversational agents (Blender in particular) perform well on conversational uptake, they are quantifiably worse than real teachers on several pedagogical dimensions, especially with regard to helpfulness (Blender: [delta] ability = -0.75; GPT-3: [delta] ability = -0.93). [For the full proceedings, see ED623995.]

Social Virtual Reality (VR) applications enable real-time interpersonal conversation and allow users to perform activities together. They have the potential of changing the ways learners practise speaking a foreign language. Following a previous study (Jauregi Ondarra, Gruber, & Canto, 2020), we designed the present study to explore how presence, immersion, and interactivity affect overall social experience. Students from Germany and the Netherlands engaged in High-immersion VR (HiVR) virtual exchange sessions, using Spanish as a lingua franca at A2 level. International dyads carried out four interaction tasks in AltspaceVR, using head-mounted devices. To examine students' HiVR virtual exchange experiences, different sources of data were gathered: questionnaires, reflection diaries, recordings, and focus group interviews. The preliminary results, based on the surveys and reflection journals, show that students liked to use a social VR app to communicate in the target language with peers from other countries, as they felt completely immersed and co-present in the social interactive VR space. This might enhance engagement and lower anxiety levels. [For the complete volume, "CALL and Professionalisation: Short Papers from EUROCALL 2021 (29th, Online, August 26-27, 2021)," see ED616972.]

Despite the unquestionable academic interest on corpus-based approaches to language education, the use of corpora by teachers in their everyday practice is still not very widespread. One way to promote usage of corpora in language teaching is by making pedagogically appropriate corpora, labelled with different types of problems (for instance, sensitive content, offensive language, structural problems), so that teachers can select authentic examples according to their needs. Because manually labelling corpora is extremely time-consuming, we propose to use crowdsourcing for this task. After a first exploratory phase, we are currently developing a multimode, multilanguage game in which players first identify problematic sentences and then classify them. [For the complete volume, "CALL and Professionalisation: Short Papers from EUROCALL 2021 (29th, Online, August 26-27, 2021)," see ED616972.]

Spatial analytics receive increased attention in educational data mining. A critical issue in stop detection (i.e., the automatic extraction of timestamped and located stops in the movement of individuals) is a lack of validation of stop accuracy to represent phenomena of interest. Next to a radius that an actor does not exceed for a certain duration to establish a stop, this study presents a reproducible procedure to optimize a range parameter for K-12 classrooms where students sitting within a certain vicinity of an inferred stop are tagged as being visited. This extension is motivated by adapting parameters to infer teacher visits (i.e., on-task and off-task conversations between the teacher and one or more students) in an intelligent tutoring system classroom with a dense layout. We evaluate the accuracy of our algorithm and highlight a tradeoff between precision and recall in teacher visit detection, which favors recall. We recommend that future research adjust their parameter search based on stop detection precision thresholds. This adjustment led to better cross-validation accuracy than maximizing parameters for an average of precision and recall (F1 = 0.18 compared to 0.09). As stop sample size shrinks with higher precision cutoffs, thresholds can be informed by ensuring sufficient statistical power in offline analyses. We share avenues for future research to refine our procedure further. Detecting teacher visits may benefit from additional spatial features (e.g., teacher movement trajectory) and can facilitate studying the interplay of teacher behavior and student learning. [For the complete proceedings, see ED630829.]

Computer-mediated social learning contexts have become increasingly popular over the last few years; yet existing models of students' cognitive-affective states have been slower to adopt dyadic interaction data for predictions. Here, we explore the possibility of capitalizing on the inherently social component of collaborative learning by using keystroke log data to make predictions across conversational partners (i.e., using person A's data to make prediction about if person B is mind wandering). Log files from 33 dyads (total N = 66) were used to examine: (a) how mind wandering (defined here as task-unrelated thought) during computer-mediated conversations is related to critical outcomes of the conversation (trust, likability, agreement); (b) if task-unrelated thought can be predicted by the keystrokes of one's partner; and (c) how much data is needed to make predictions by testing various window-sizes of data preceding task-unrelated thought reports. Results indicated a negative relationship between task-unrelated thought and perceptions of the conversation, suggesting that attention is an important factor during computer mediated chat conversations. Finally, in line with our hypothesis, results from mixed effects models showed that one's level of task-unrelated thought was predicted by the keystroke patterns of their conversational partner, but only using small window sizes (5s worth of data). [For the complete proceedings, see ED630829.]

Vocabulary proficiency diagnosis plays an important role in the field of language learning, which aims to identify the level of vocabulary knowledge of a learner through his or her learning process periodically, and can be used to provide personalized materials and feedback in language-learning applications. Traditional approaches are widely applied for modeling knowledge in science or mathematics, where skills or knowledge concepts are well-defined and easy to associate with each item. However, only a handful of works focus on defining knowledge concepts and skills using linguistic characteristics for language knowledge proficiency diagnosis. In addressing this, we propose a framework for vocabulary proficiency diagnosis based on neural networks. Specifically, we propose a series of methods based on our framework that uses different linguistic features to define skills and knowledge concepts in the context of the language learning task. Experimental results on a real-world second-language learning dataset demonstrate the effectiveness and interpretability of our framework. We also provide empirical evidence with ablation testing to prove that our knowledge concept and skill definitions are reasonable and critical to the performance of our model. [For the complete proceedings, see ED630829.]

In this study, the authors investigate the ways in which young students demonstrate their critical mathematical thinking (CMT). Students aged 5-6 who are beginning their first formal year of education participated in the study. Data is presented from individual clinical interviews undertaken with 16 students. These interviews were analysed using the Critical Mathematical Thinking for Young Students (CMTFYS) framework to identify common patterns in the responses. The findings suggest that these young students beginning school, most often rely on providing explanations and more specifically, justifying, to demonstrate their CMT.

The primary school mathematics syllabuses in Singapore as of the year 2020 reinforces that Big ideas are central to the learning of mathematics. In support of the push to teach for big ideas, a research study is presently underway. A part of it is on the professional development (PD) of primary school mathematics teachers. As part of the PD teachers attempted a mathematical task as measure of the big idea, Equivalence, in an online environment at the start of their PD. Data from the task show that teachers, were generally not cognisant of the big idea of equivalence when solving the task. They were also unable to distinguish between a heuristic (diagrams) and a mathematical idea about relationships, specifically equivalence as in the mathematical task.

Two case studies of Australian primary school students tracked changes in their data interpretation and representation over three years. Students were engaged in predictive reasoning tasks based on their interpretation of a data table showing temperature change over time. Students' explanations and graphical representations were collected at the beginning of Years 3 and 4 and the end of Years 4 and 5. The first case study was a student mathematically weaker than her peers while case study two was within the average range for her year. Despite differences in starting points, both case studies followed a similar developmental sequence of predicting, interpreting and representing, with the first case generally lagging one stage behind the second case. Similarities and contrasts between the two students are discussed.

In this investigation we propose new machine learning methods for automated scoring models that predict the vocabulary acquisition in science and social studies of second grade English language learners, based upon free-form spoken responses. We evaluate performance on an existing dataset and use transfer learning from a large pre-trained language model, reporting the influence of various objective function designs and the input-convex network design. In particular, we find that combining objective functions with varying properties, such as distance among scores, greatly improves the model reliability compared to human raters. Our models extend the current state of the art performance for assessing word definition tasks and sentence usage tasks in science and social studies, achieving excellent quadratic weighted kappa scores compared with human raters. However, human-human agreement still surpasses model-human agreement, leaving room for future improvement. Even so, our work highlights the scalability of automated vocabulary assessment of free-form spoken language tasks in early grades. [This paper was published in: "Proceedings of the Tenth ACM Conference on Learning @ Scale (L@S '23), July 20-22, 2023, Copenhagen, Denmark," ACM, 2023.]

The "Framework" and NGSS bring to the forefront the role of language in doing science and in learning from doing science. Yet, most existing science assessments for elementary learners do not integrate or attend to aspects of scientific language and literacy that are essential components of science proficiency. Accordingly, there is a need for high-quality classroom resources capable of providing teachers with information on where students are in the process of developing proficiency in science and scientific-specific literacy. In this paper, we present our design process for developing assessment resources -- tasks, rubrics, and teacher support materials -- and describe initial insights from using the process to develop resources for elementary grade classrooms. We highlight challenges and opportunities that have emerged from our work attending to the role of literacy assessing multi-dimensional science learning, incorporating equity and inclusion considerations in assessment design, and embracing the affordances of technology. [This paper was published in: "ICLS2022 Proceedings," International Society of the Learning Sciences, 2022, pp. 1237-1240.]