Successful Learning Research Articles

Retrieval-style In-context Learning for Few-shot Hierarchical Text Classification

Abstract Hierarchical text classification (HTC) is an important task with broad applications, and few-shot HTC has gained increasing interest recently. While in-context learning (ICL) with large language models (LLMs) has achieved significant success in few-shot learning, it is not as effective for HTC because of the expansive hierarchical label sets and extremely ambiguous labels. In this work, we introduce the first ICL-based framework with LLM for few-shot HTC. We exploit a retrieval database to identify relevant demonstrations, and an iterative policy to manage multi-layer hierarchical labels. Particularly, we equip the retrieval database with HTC label-aware representations for the input texts, which is achieved by continual training on a pretrained language model with masked language modeling (MLM), layer-wise classification (CLS, specifically for HTC), and a novel divergent contrastive learning (DCL, mainly for adjacent semantically similar labels) objective. Experimental results on three benchmark datasets demonstrate superior performance of our method, and we can achieve state-of-the-art results in few-shot HTC.

Use of interactive technologies in an innovative educational environment

The article analyzes the content and system of interactive technologies in an innovative educational environment, substantiates the components of interactive technology, and reveals the main pedagogical idea and main requirements in the mode of interactive technology for successful learning. Types of active methods in the mode of interactive technology are grouped and analyzed for successful learning and stimulation of the natural activity of students. Ways of creating an innovative educational environment in a higher education institution in the mode of interactive technology are shown. The most effective technologies that we recommend in the interactivity mode have been analyzed. The stages of development of interactive learning technology in the innovative educational environment of a higher school are shown. The impact of the active implementation of information and communication technologies, modern digital tools, and new technological means of learning on the possibilities of using interactive technologies in an innovative educational environment is proven. This is due to the need for innovative training of a generation of new specialists capable of competently and quickly applying the results of scientific and technical progress in professional and educational activities, in particular, the Internet, web technologies, smartphones, digital educational platforms, cloud services, electronic educational resources, artificial intelligence, and other modern devices.

THE ENHANCEMENT OF CRITICAL THINKING SKILLS THROUGH THE IMPLEMENTATION OF ACADEMIC WRITING IN ENGLISH LESSONS

The article delves into the significance of higher-order thinking skills in the context of academic writing, exploring various methodologies aimed at enhancing students' competencies in this area. It emphasizes the role of critical thinking, which is a key component of interdisciplinary knowledge as per Bloom's taxonomy strategy, encompassing the stages of remembering, understanding, applying, analyzing, evaluating, and creating. The study's main purpose is to analyze the effectiveness of the implementation of academic writing in the development of student's critical thinking skills. Accordingly, the research seeks to investigate and nurture students' critical thinking skills through academic writing. It demonstrates that students can effectively acquire critical thinking skills through specially structured educational activities in a phased learning approach. The study utilized a mixed-method design, collecting data through observation, written documents, and online surveys. The data obtained from these sources were then analyzed both quantitatively and qualitatively. The findings of the research reveal various stages of teaching critical thinking, facilitated by the instructor's efforts to actively engage students in the learning process. Based on the results of theoretical and expert analysis, the quantitative and qualitative data were compared and illustrated positive growth in learners’ academic writing skills. The processed-based approach was determined as a means of successful learning methodology in the development of higher-order thinking skills. The study also highlights the importance of well-designed writing assignments and group activities in encouraging students to revisit concepts, critically evaluate assumptions, and make substantial revisions to their written work.

Elementary learning styles and application of the Dunn and Dunn learning style inventory model

This study explored the diverse learning styles of elementary pupils in an elementary school in the Philippines. It examined the application of the Dunn and Dunn Learning Style Inventory Model (DDL-SIM) to enhance educational outcomes. Recognizing the unique learning preferences of learners, the research aimed to bridge the gap in learning styles by implementing tailored teaching methods and strategies. Utilizing an explanatory mixed-methods research design, the study engaged all pupils using the complete enumeration sampling technique to ascertain their learning styles across five domains: environmental, emotional, sociological, physiological, and psychological. The survey questionnaire for the quantitative phase and the semi-structured interview guide questionnaire for the qualitative phase were employed, which were subjected to validity and reliability tests by the expert panel of evaluators. The study's data analyses utilized descriptive statistics, Spearman Rho correlation, and thematic analysis. As per the DDL-SIM, quantitative findings revealed significant emotional and physiological preferences, with significant relationships between demographic profiles and their learning styles. Qualitative insights underscored a holistic approach to learning as the generated major theme, emphasizing the interconnectedness of minor themes such as emotional intelligence, social dynamics, physical health, and environmental factors in the learning experiences. The study suggested that educators create tailored learning plans and gender-sensitive approaches and include emotional and physiological aspects in teaching. In support of varied learning styles, it promoted the DDL-SIM of educator’s professional development and parent and community participation. Hence, the DDL-SIM recognizes and addresses each learner's unique preferences, promoting a more inclusive and successful learning environment.

The Impact of Learning Behavior and Student Engagement on Learning Outcomes in Geography Subjects at Junior School Level

Geography is a science that deals with all activities and symptoms found on earth. At the junior high school level, geography is integrated into integrated social studies subjects. Geography is also an important social science to be implemented in lower secondary schools, but in its application there are also factors for its success, including learning behavior and student involvement. The objectives of this research are1) Analyzing the effect of learning behavior on student geography learning results, 2) Analyzing the effect of student engagement on geography learning results, and 3) Analyzing the effect of learning behavior and student engagement on geography learning outcomes. This research used quantitative method. The population in this study were students from grade 7 to grade 9 at SMP Negeri 2 Kemalang totaling 584 students. The sample in this study consisted of 237 students. SThe sampling was conducted using the stratified random sampling technique. Data were obtained using observation, documentation, and questiinnaires. The data analysis techniques used in this study include descriptive statistic, simple correlation tests, and multiple correlation analysis. The result shiwed that 1) students have a level of learning behavior in the high category, 2) student learning engagement in geography subjects categorized into high category, and 3) There is no significant influence of learning behavior and student engagement on geography learning success. It can be observed that the r value is smaller than the r tabke value 0.059 0.134) and the significance value is above 0,05 (0.416 0.05).

Students Confidence in Increasing Learning Motivation

The study shows that students' self-confidence in increasing learning motivation in physical education class VIII in participating in Physical Education learning at SMP Negeri 2 Harian in the 2023/2024 Academic Year is in the moderate category. This study was conducted at SMP Negeri 2 Harian with a total of 30 students consisting of 17 boys and 14 girls. The questionnaire items in this study were 30 statements prepared in Google Form https://forms.gle/w86uMf1yTbZntFnF6 . The data analysis technique used in this study is a quantitative descriptive analysis technique. The results of the study show that students' self-confidence in increasing learning motivation in physical education class VIII in participating in Physical Education learning at SMP Negeri 2 Harian in the 2023/2024 Academic Year is in the moderate category. The success of Physical Education learning must be supported by students who have self-confidence in participating in learning, and the results of this study state that the self-confidence of class VIII students of SMP Negeri 2 Harian in participating in Physical Education learning is in the moderate category, so teachers need to pay attention and try to improve the learning process that has been carried out, for example teaching methods that are in accordance with student interests, so that the level of student confidence increases in participating in Physical Education learning can be maintained, and even will increase.

Representation in best-selling preschool storybook apps in the United States

ABSTRACT In this study, we examine preschool storybook apps to identify the types of characters represented in these stories and how they relate to design quality and opportunities for children to build their vocabularies – a significant predictor of later success in learning to read. Specifically, we sought to understand (a) how storybook apps portray gender, race, and disability status; (b) how these characteristics compare in apps that contain the most and least opportunities to learn vocabulary; and (c) how these characteristics compare in apps with design features that might provide additional support for children’s word learning. To do so, we conducted a content analysis of 70 best-selling preschool storybook apps in the United States. Results indicate that most protagonists in these apps are animals, while human characters are overwhelmingly white, able-bodied, and male, particularly in apps with the highest percentage of advanced vocabulary and design qualities that are most likely to support children’s word learning. Thus, apps with the greatest learning potential lack human diversity, raising significant concerns for inclusive education.

Exploring Language Proficiency: Dynamics of Acquiring Arabic Speaking Skills from Stephen Krashen's Perspective At Islamic Universities

Understanding the dynamics of acquiring Arabic speaking skills is crucial in the context of Islamic universities, where language proficiency is integral to academic success and personal development. This study aims to explore the process of acquiring Arabic speaking skills from Stephen Krashen's perspective at two undisclosed Islamic universities in Indonesia. Using a qualitative case study design, the research employs interviews with lecturers and students as the primary data collection technique. The analysis is conducted descriptively, involving stages of data collection, reduction, presentation, and conclusion drawing. The findings reveal that Arabic speaking skills are acquired through two main approaches. Within the classroom, intensive learning is facilitated through a direct method emphasizing language practice, supported by a curriculum focused on daily themes and language reinforcement programs. Outside the classroom, students engage in active communication in Arabic in various environments, such as dormitories, mosques, and cafeterias, while participating in activities that support language acquisition. These factors, combined with student motivation and a conducive learning environment, significantly contribute to the acquisition of Arabic speaking skills. The study concludes that the different learning approaches at the two universities effectively enhance students' Arabic speaking proficiency. Key factors contributing to successful language learning include the presence of Arabic language environments, appropriate curricula, active learning methods, individual motivation, and practical experiences outside the classroom, such as internships or student exchange programs. The research suggests that educators and students should focus on creating immersive language environments and incorporating practical experiences to maximize language acquisition.

An Experimental and Clinical Physiological Signal Dataset for Automated Pain Recognition

Access to large amounts of data is essential for successful machine learning research. However, there is insufficient data for many applications, as data collection is often challenging and time-consuming. The same applies to automated pain recognition, where algorithms aim to learn associations between a level of pain and behavioural or physiological responses. Although machine learning models have shown promise in improving the current gold standard of pain monitoring (self-reports) only a handful of datasets are freely accessible to researchers. This paper presents the PainMonit Dataset for automated pain detection using physiological data. The dataset consists of two parts, as pain can be perceived differently depending on its underlying cause. (1) Pain was triggered by heat stimuli in an experimental study during which nine physiological sensor modalities (BVP, 2×EDA, skin temperature, ECG, EMG, IBI, HR, respiration) were recorded from 55 healthy subjects. (2) Eight modalities (2×BVP, 2×EDA, EMG, skin temperature, respiration, grip) were recorded from 49 participants to assess their pain during a physiotherapy session.

Resisting technological solutionism in open universities in the time of global digital convergence

PurposeThis purpose of this paper is to explore the future of open universities (OUs), particularly in developing Asia, focusing on the potential risks associated with technological solutionism, or the popular belief that educational problems can be readily solved through technological means. This exploration is pertinent as the initial advantages that OUs once held have significantly diminished and competing higher education institutions worldwide are converging on a digital future.Design/methodology/approachThe qualitative method of empathy-based stories (MEBS) was employed as the primary research method for drawing data from sampled OU teachers. The obtained data were parsed via textual analysis and the lens of critical discourses on technology and higher education.FindingsThe study shows that, in the future envisioned by the sampled OU teachers, students’ lack of privileged access to frontier technologies is no barrier to learning success. It also emphasises that learning outcomes have causality beyond technological determinism, and that educational problems, which OUs may deem as requiring technological solutions, are often non-technological in nature and require no technological fixes.Originality/valueCritiques of techno-solutionism, such as the present study, are virtually absent in the context of OUs in developing Asia. For this very reason, this study is vital and serves as a guardrail while these OUs seek to reformulate their respective value propositions in the time of global digital convergence.

White matter connectivity linked to novel word learning in children.

Children and adults are excellent word learners. Increasing evidence suggests that the neural mechanisms that allow us to learn words change with age. In a recent fMRI study from our group, several brain regions exhibited age-related differences when accessing newly learned words in a second language (L2; Takashima et al. Dev Cogn Neurosci 37, 2019). Namely, while the Teen group (aged 14-16years) activated more left frontal and parietal regions, the Young group (aged 8-10years) activated right frontal and parietal regions. In the current study we analyzed the structural connectivity data from the aforementioned study, examining the white matter connectivity of the regions that showed age-related functional activation differences. Age group differences in streamline density as well as correlations with L2 word learning success and their interaction were examined. The Teen group showed stronger connectivity than the Young group in the right arcuate fasciculus (AF). Furthermore, white matter connectivity and memory for L2 words across the two age groups correlated in the left AF and the right anterior thalamic radiation (ATR) such that higher connectivity in the left AF and lower connectivity in the right ATR was related to better memory for L2 words. Additionally, connectivity in the area of the right AF that exhibited age-related differences predicted word learning success. The finding that across the two age groups, stronger connectivity is related to better memory for words lends further support to the hypothesis that the prolonged maturation of the prefrontal cortex, here in the form of structural connectivity, plays an important role in the development of memory.

Leap, learn, earn: exploring academic risk taking and learning success across gender and socioeconomic groups

Abstract Background The positive learning effects of academic risk taking (ART) in higher education has been discussed since the 1980s. However, this may not apply equally for all social groups. Men and women may differ in the way they use ART to construct their gender identity. Students with different socioeconomic status (SES) may differ in their ability to navigate academic risks due to differences in available cultural capital. Aims This study examines gender and SES disparities in ART and their impact on learning success. It explores if ART mediates and is moderated by gender and SES effects. Additionally, it assesses if ART directly predicts learning success. Sample A sample of N = 381 German university students was used. Methods Data was analyzed following a structural equation modeling approach. Results Men show more ART on the seminar group dimension, whereas women show more ART on the peer dimension. Being male indirectly predicts higher learning success via the seminar group dimension of ART. Furthermore, SES and gender moderate the effect between ART and learning success. Both ART dimensions directly predict students’ learning success. Conclusions Our research contributes to understanding the mechanisms of social disparities within higher education and offers implications for the development of inclusive teaching strategies and research on aspects of intersectionality.

Perceiving depth beyond sight: Evaluating intrinsic and learned cues via a proof of concept sensory substitution method in the visually impaired and sighted.

This study explores spatial perception of depth by employing a novel proof of concept sensory substitution algorithm. The algorithm taps into existing cognitive scaffolds such as language and cross modal correspondences by naming objects in the scene while representing their elevation and depth by manipulation of the auditory properties for each axis. While the representation of verticality utilized a previously tested correspondence with pitch, the representation of depth employed an ecologically inspired manipulation, based on the loss of gain and filtration of higher frequency sounds over distance. The study, involving 40 participants, seven of which were blind (5) or visually impaired (2), investigates the intrinsicness of an ecologically inspired mapping of auditory cues for depth by comparing it to an interchanged condition where the mappings of the two axes are swapped. All participants successfully learned to use the algorithm following a very brief period of training, with the blind and visually impaired participants showing similar levels of success for learning to use the algorithm as did their sighted counterparts. A significant difference was found at baseline between the two conditions, indicating the intuitiveness of the original ecologically inspired mapping. Despite this, participants were able to achieve similar success rates following the training in both conditions. The findings indicate that both intrinsic and learned cues come into play with respect to depth perception. Moreover, they suggest that by employing perceptual learning, novel sensory mappings can be trained in adulthood. Regarding the blind and visually impaired, the results also support the convergence view, which claims that with training, their spatial abilities can converge with those of the sighted. Finally, we discuss how the algorithm can open new avenues for accessibility technologies, virtual reality, and other practical applications.

Testing the Construct Validity and Reliability of the Student Learning Interest Scale Using Confirmatory Factor Analysis (CFA)

Learning interest is one of the determining factors of student learning success. The low interest of students in learning process is caused by several problems that occur. Thus, learning interest is very important in learning progress, namely to bring out students' interest and activeness in learning. To support learning and increase students' interest in learning, teachers should not only master the material but also create interesting learning to help arouse curiosity, enjoyment and interest in learning in them. This study aims to test the validity and reliability of the construct of the developed learning interest scale. A researcher needs 250 to 500 respondents. This study involved 300 random samples of fifth grade elementary school students in Berbah District, Sleman, Yogyakarta, Indonesia. Data were analyzed with second-order confirmatory factors assisted by IBM SPSS version 26 software. This study measures the validity and reliability of elementary school students' interest in learning science. So the data analysis in this study measures validity and reliability. So this study uses second order confirmatory factor analysis (nd order CFA). The results showed that the developed learning interest scale had met the criteria of suitability, convergent validity, discriminant validity, construct reliability. Therefore, the developed scale is feasible to use.

Improved GAN-based deep learning approach for strain field prediction and failure analysis of precast bridge slab joints

Nonlinear finite element analysis (FEA) is crucial for understanding complex stress-strain behaviors and assessing potential structural failure risks. Building on the recent success of deep learning (DL) methods in physical field predictions, there has been growing interest in forecasting internal force fields of structures beyond the existing data scope. In this study, a data-driven strain field analysis model based on an improved Generative Adversarial Network (GAN) is proposed. To enrich the input data, this novel approach incorporates the Signed Distance Field (SDF) derived from the characteristics of the original Finite Element (FE) models. The generator, based on U-Net, is enhanced with channel attention mechanisms and loss penalty terms, forming the core of the improved GAN. The effectiveness of the proposed method is demonstrated using two parallel datasets comprising 3-Headed bar tension models across two planes. Subsequently, the generated strain field is subjected to feature extraction using the 2D-Principal Component Analysis (2D-PCA) method, followed by structural failure determination through the K-Nearest Neighbor (KNN) algorithm. The results indicate that the proposed improved approach achieves a significant reduction in Root Mean Square Error (RMSE) of 20.71 % and 13.21 % across the two datasets, with the structural failure determination method reaching an F1-score of 0.971. Thus, the intelligent strain field analysis method introduced here effectively predicts strain distributions and promptly identifies failure states, offering valuable insights for further numerical analysis and structural design.

Fair and optimal prediction via post‐processing

AbstractWith the development of machine learning algorithms and the increasing computational resources available, artificial intelligence has achieved great success in many application domains. However, the success of machine learning has also raised concerns about the fairness of the learned models. For instance, the learned models can perpetuate and even exacerbate the potential bias and discrimination in the training data. This issue has become a major obstacle to the deployment of machine learning systems in high‐stakes domains, for example, criminal judgment, medical testing, online advertising, hiring process, and so forth. To mitigate the potential bias exhibited by machine learning models, fairness criteria can be integrated into the training process to ensure fair treatment across all demographics, but it often comes at the expense of model performance. Understanding such tradeoffs, therefore, is crucial to the design of optimal and fair algorithms. My research focuses on characterizing the inherent tradeoff between fairness and accuracy in machine learning, and developing algorithms that can achieve both fairness and optimality. In this article, I will discuss our recent work on designing post‐processing algorithms for fair classification, which can be applied to a wide range of fairness criteria, including statistical parity, equal opportunity, and equalized odds, under both attribute‐aware and attribute‐blind settings, and is particularly suited to large‐scale foundation models where retraining is expensive or even infeasible. I will also discuss the connections between our work and other related research on trustworthy machine learning, including the connections between algorithmic fairness and differential privacy as well as adversarial robustness.

Success of machine learning and statistical methods in predicting landslide hazard: the case of Elazig (Maden)

Success of machine learning and statistical methods in predicting landslide hazard: the case of Elazig (Maden)

Meta-Analysis of Collaborative Learning Approaches in Educational Management and Their Impact on Student Performance

This meta-analysis investigates the effects of collaborative learning approaches within the framework of educational management and their impact on student performance. Collaborative learning, as a student-centered instructional method, has been widely adopted across various educational levels and settings. However, its effectiveness in enhancing academic achievement remains a topic of debate. This study synthesizes findings from 12 empirical studies conducted between 2022-2024, covering primary, secondary, and higher education contexts globally. Using a random-effects model, the meta-analysis quantifies the overall effect size of collaborative learning approaches on student performance, while accounting for variations in geographical region, subject matter, and educational level. The results reveal a high to strong positive effect of collaborative learning on student academic performance (mean effect size = 0.819). Additionally, subgroup analyses show that the impact is more pronounced in subjects requiring higher-order thinking skills, such as science and mathematics, and in educational settings where collaborative learning is systematically supported by institutional management. The analysis also highlights key factors that influence the success of collaborative learning, including group dynamics, teacher facilitation, and the integration of technology.

Upaya Meningkatkan Minat Belajar Peserta Didik Melalui Permainan Berburu Binatang Dalam Pembelajaran PJOK di SMPN 28 Surabaya

Comfort becomes very important when used as a driver to carry out learning that requires full concentration in receiving all materials. This factor also plays an important role in the success of learning. It can be seen from the comfort of each student. This classroom action research focuses on collecting information on how animal hunting games can increase students' interest in learning. This action was carried out in two cycles. The sample in this study was 30 students of class 9F SMPN 28 Surabaya. The results of the actions of cycle I and cycle II through observation and questionnaires were that there was an increase in students' interest in learning when using animal hunting games.

The influence of Kahoot-based TBLT in improving the critical reading skills in higher education

This study aimed to assess the effectiveness of Task-based language Teaching (TBLT) approach through Kahoot in improving EFL students’ critical reading skills. The study used a quasi-experimental method with a pre-experiment design involving a pre and post-test for one group. The research was conducted with 70 students (15 males and 65 females) from the Department of English Education at a Private University in Indonesia. The Kahoot-based TBLT approach was used to assess students' critical reading skills. The data were analyzed utilizing paired t-tests and N-gain to determine the effectiveness. The result showed a significant improvement in critical reading skills from the pre-test to the post-test after using the Kahoot-based TBLT model. the findings indicate that the Kahoot-based TBLT model is a successful approach in improving students' critical reading skills. Therefore, it can be concluded that Kahoot-based TBLT is effective in improving EFL students’ critical skill. Future research recommends the usage of Kahoot-based TBLT by educators to increase academic success in English language learning for higher education students.