Computers and Education Artificial Intelligence最新文献_第8页

Diverging perceptions of artificial intelligence in higher education: A comparison of student and public assessments on risks and damages of academic performance prediction in Germany 对高等教育中人工智能的不同看法：德国学生和公众对学习成绩预测的风险和损害的评估比较

Q1 Social Sciences

Computers and Education Artificial Intelligence

Pub Date : 2024-09-30 DOI: 10.1016/j.caeai.2024.100305

Marco Lünich, Birte Keller, Frank Marcinkowski

The integration of Artificial Intelligence (AI) into higher education, particularly through Academic Performance Prediction (APP), promises enhanced educational outcomes. However, it simultaneously raises concerns regarding data privacy, potential biases, and broader socio-technical implications. Our study, focusing on Germany–a pivotal player in shaping the European Union's AI policies–seeks to understand prevailing perceptions of APP among students and the general public. Initial findings of a large standardized online survey suggest a divergence in perceptions: While students, in comparison to the general population, do not attribute a higher risk to APP in a general risk assessment, they do perceive higher societal and, in particular, individual damages from APP. Factors influencing these damage perceptions include trust in AI and personal experiences with discrimination. Students further emphasize the importance of preserving their autonomy by placing high value on self-determined data sharing and explaining their individual APP. Recognizing these varied perceptions is crucial for educators, policy-makers, and higher education institutions as they navigate the intricate ethical landscape of AI in education. This understanding can inform strategies that accommodate both the potential benefits and concerns associated with AI-driven educational tools.

将人工智能（AI）融入高等教育，特别是通过学业成绩预测（APP），有望提高教育成果。然而，它同时也引发了人们对数据隐私、潜在偏见以及更广泛的社会技术影响的担忧。我们的研究以德国--欧盟人工智能政策制定的重要参与者--为重点，旨在了解学生和公众对 APP 的普遍看法。一项大型标准化在线调查的初步结果表明，人们对APP的看法存在分歧：虽然与普通大众相比，学生在一般风险评估中并不认为APP具有更高的风险，但他们确实认为APP会造成更高的社会损害，尤其是个人损害。影响这些损害认知的因素包括对人工智能的信任和个人遭受歧视的经历。学生们进一步强调了通过高度重视自主决定的数据共享和解释其个人APP来维护其自主性的重要性。认识到这些不同的看法对于教育工作者、政策制定者和高等教育机构来说至关重要，因为他们要驾驭人工智能在教育领域错综复杂的伦理环境。这种认识可以为制定战略提供依据，从而兼顾与人工智能驱动的教育工具相关的潜在益处和担忧。

{"title":"Diverging perceptions of artificial intelligence in higher education: A comparison of student and public assessments on risks and damages of academic performance prediction in Germany","authors":"Marco Lünich, Birte Keller, Frank Marcinkowski","doi":"10.1016/j.caeai.2024.100305","DOIUrl":"10.1016/j.caeai.2024.100305","url":null,"abstract":"<div><div>The integration of Artificial Intelligence (AI) into higher education, particularly through Academic Performance Prediction (APP), promises enhanced educational outcomes. However, it simultaneously raises concerns regarding data privacy, potential biases, and broader socio-technical implications. Our study, focusing on Germany–a pivotal player in shaping the European Union's AI policies–seeks to understand prevailing perceptions of APP among students and the general public. Initial findings of a large standardized online survey suggest a divergence in perceptions: While students, in comparison to the general population, do not attribute a higher risk to APP in a general risk assessment, they do perceive higher societal and, in particular, individual damages from APP. Factors influencing these damage perceptions include trust in AI and personal experiences with discrimination. Students further emphasize the importance of preserving their autonomy by placing high value on self-determined data sharing and explaining their individual APP. Recognizing these varied perceptions is crucial for educators, policy-makers, and higher education institutions as they navigate the intricate ethical landscape of AI in education. This understanding can inform strategies that accommodate both the potential benefits and concerns associated with AI-driven educational tools.</div></div>","PeriodicalId":34469,"journal":{"name":"Computers and Education Artificial Intelligence","volume":"7 ","pages":"Article 100305"},"PeriodicalIF":0.0,"publicationDate":"2024-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418636","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Overview and confirmatory and exploratory factor analysis of AI literacy scale 人工智能素养量表概述及确认性和探索性因素分析

Q1 Social Sciences

Computers and Education Artificial Intelligence

Pub Date : 2024-09-27 DOI: 10.1016/j.caeai.2024.100310

Martin J. Koch , Carolin Wienrich , Samantha Straka , Marc Erich Latoschik , Astrid Carolus

Comprehensive concepts of AI literacy (AIL) and valid measures are essential for research (e.g., intervention studies) and practice (e.g., personnel selection/development) alike. To date, several scales have been published, sharing standard features but differing in some aspects. We first aim to briefly overview instruments identified from unsystematic literature research in February 2023. We identified four scales and one collection of items. We describe the instruments and compare them. We identified common themes and overlaps in the instruments and developmental procedure. We also found differences regarding scale development procedures and latent dimensions. Following this literature research, we came to the conclusion that the literature on AI literacy measurement was fragmented, and little effort was undertaken to integrate different AI literacy conceptualizations. The second focus of this study is to test the factorial structures of existing AIL measurement instruments and identify latent dimensions of AIL across all instruments. We used robust maximum-likelihood confirmatory factor analysis to test factorial structures in a joint survey of all AIL items in an English-speaking online sample (N=219). We found general support for all instruments' factorial structures with minor deviations from the original factorial structures for some of the instruments. In a second analysis step, to address the issue of fragmented research on AI literacy conceptualization and measurement, we used principal axis exploratory factor analysis with oblique rotation to identify latent dimensions across all items. We found four correlated latent dimensions of AIL, which were mostly interpretable as the abilities to use and interact with AI, to design/program AI (incl. in-depth technical knowledge), to perform complex cognitive operations regarding AI (e.g., ethical considerations), and a common factor for the abilities to detect AI/differentiate between AI and humans and manage persuasive influences of AI (i.e., persuasion literacy). Our findings sort the multitude of AIL instruments and reveal four latent core dimensions of AIL. Thus, they contribute importantly to the conceptual understanding of AIL that has been fragmented so far.

全面的人工智能素养（AIL）概念和有效的测量方法对于研究（如干预研究）和实践（如人员选拔/发展）都至关重要。迄今为止，已经发布了几种量表，它们具有共同的标准特征，但在某些方面存在差异。我们首先简要介绍 2023 年 2 月从非系统文献研究中发现的工具。我们确定了四个量表和一个项目集。我们对这些工具进行了描述和比较。我们发现了工具和编制程序中的共同主题和重叠之处。我们还发现了量表编制程序和潜在维度方面的差异。经过文献研究，我们得出结论，人工智能素养测量方面的文献是支离破碎的，几乎没有人努力整合不同的人工智能素养概念。本研究的第二个重点是检验现有人工智能素养测量工具的因子结构，并确定所有工具中人工智能素养的潜在维度。我们采用稳健的最大似然确证因子分析方法，在英语在线样本（N=219）中对所有人工智能素养项目进行联合调查，以检验因子结构。我们发现，所有工具的因子结构都得到了普遍支持，部分工具的因子结构与原始因子结构略有偏差。在第二步分析中，为了解决人工智能素养概念化和测量研究零散的问题，我们使用了主轴探索性因子分析和斜向旋转来识别所有项目的潜在维度。我们发现了人工智能素养的四个相关潜维度，主要可解释为使用人工智能并与之互动的能力、设计/编程人工智能的能力（包括深入的技术知识）、执行有关人工智能的复杂认知操作的能力（如伦理考虑），以及检测人工智能/区分人工智能与人类和管理人工智能说服影响的能力（即说服素养）的共同因子。我们的研究结果对众多人工智能素养工具进行了分类，并揭示了人工智能素养的四个潜在核心维度。因此，它们对迄今为止零散的人工智能语言的概念理解做出了重要贡献。

{"title":"Overview and confirmatory and exploratory factor analysis of AI literacy scale","authors":"Martin J. Koch , Carolin Wienrich , Samantha Straka , Marc Erich Latoschik , Astrid Carolus","doi":"10.1016/j.caeai.2024.100310","DOIUrl":"10.1016/j.caeai.2024.100310","url":null,"abstract":"<div><div>Comprehensive concepts of AI literacy (AIL) and valid measures are essential for research (e.g., intervention studies) and practice (e.g., personnel selection/development) alike. To date, several scales have been published, sharing standard features but differing in some aspects. We first aim to briefly overview instruments identified from unsystematic literature research in February 2023. We identified four scales and one collection of items. We describe the instruments and compare them. We identified common themes and overlaps in the instruments and developmental procedure. We also found differences regarding scale development procedures and latent dimensions. Following this literature research, we came to the conclusion that the literature on AI literacy measurement was fragmented, and little effort was undertaken to integrate different AI literacy conceptualizations. The second focus of this study is to test the factorial structures of existing AIL measurement instruments and identify latent dimensions of AIL across all instruments. We used robust maximum-likelihood confirmatory factor analysis to test factorial structures in a joint survey of all AIL items in an English-speaking online sample (<em>N</em>=219). We found general support for all instruments' factorial structures with minor deviations from the original factorial structures for some of the instruments. In a second analysis step, to address the issue of fragmented research on AI literacy conceptualization and measurement, we used principal axis exploratory factor analysis with oblique rotation to identify latent dimensions across all items. We found four correlated latent dimensions of AIL, which were mostly interpretable as the abilities to use and interact with AI, to design/program AI (incl. in-depth technical knowledge), to perform complex cognitive operations regarding AI (e.g., ethical considerations), and a common factor for the abilities to detect AI/differentiate between AI and humans and manage persuasive influences of AI (i.e., persuasion literacy). Our findings sort the multitude of AIL instruments and reveal four latent core dimensions of AIL. Thus, they contribute importantly to the conceptual understanding of AIL that has been fragmented so far.</div></div>","PeriodicalId":34469,"journal":{"name":"Computers and Education Artificial Intelligence","volume":"7 ","pages":"Article 100310"},"PeriodicalIF":0.0,"publicationDate":"2024-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418683","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Latent space bias mitigation for predicting at-risk students 减少潜空间偏差，预测问题学生

Q1 Social Sciences

Computers and Education Artificial Intelligence

Pub Date : 2024-09-27 DOI: 10.1016/j.caeai.2024.100300

Ali Al-Zawqari, Dries Peumans, Gerd Vandersteen

Researchers have observed the relationship between educational achievements and students' demographic characteristics in physical classroom-based learning. In the context of online education, recent studies were conducted to explore the leading factors of successful online courses. These studies also investigated how demographic features impact student achievement in the online learning environment. This motivates the use of demographic information alongside other features to predict students' academic performance. Since demographic features include protected attributes, such as gender and age, evaluating predictive models must go beyond minimizing the overall error. In this work, we analyze and investigate the use of neural networks to predict underperforming students in online courses. However, our goal is not only to enhance the accuracy but also to evaluate the fairness of the predictive models, a problem concerning the application of machine learning in education. This paper starts by analyzing the available solutions to fairness in predictive models: bias mitigation with pre-processing and in-processing methods. We show that the current evaluation is missing the case of partial awareness of protected features, which is the case when the model is aware of bias on some protected attributes but not all. The in-processing method, specifically the adversarial bias mitigation, shows that debiasing in some protected features exacerbates the bias on other protected features. This observation motivates our proposal of an alternative approach to enhance bias mitigation even in the partial awareness scenario by working with latent space. We implement the proposed solution using denoising autoencoders. The quantitative analysis used three distributions from The Open University Learning Analytics dataset (OULAD). The obtained results show that the latent space-based method offers the best solution as it maintains accuracy while mitigating the bias of the prediction models. These results indicate that in the case of partial awareness, the latent space method is considered superior to the adversarial bias mitigation approach.

研究人员观察了基于物理课堂学习的教育成就与学生人口特征之间的关系。在在线教育方面，最近开展了一些研究，以探索成功在线课程的主导因素。这些研究还调查了人口特征如何影响学生在在线学习环境中的成绩。这就促使我们使用人口统计学信息和其他特征来预测学生的学业成绩。由于人口统计学特征包括受保护的属性，如性别和年龄，因此评估预测模型必须超越最小化整体误差的范畴。在这项工作中，我们分析并研究了如何利用神经网络来预测在线课程中表现不佳的学生。然而，我们的目标不仅是提高准确性，还要评估预测模型的公平性，这是机器学习在教育领域应用的一个难题。本文首先分析了预测模型公平性的现有解决方案：通过预处理和内处理方法减轻偏差。我们发现，目前的评估缺少对部分受保护特征的认识，即模型能认识到某些受保护属性的偏差，但不能认识到所有属性的偏差。内处理方法，特别是对抗性偏差缓解方法表明，某些受保护特征的去偏差会加剧其他受保护特征的偏差。这一观察结果促使我们提出了一种替代方法，通过使用潜空间来加强偏差缓解，即使在部分感知场景下也是如此。我们使用去噪自动编码器来实现所提出的解决方案。定量分析使用了开放大学学习分析数据集（OULAD）中的三种分布。结果表明，基于潜在空间的方法提供了最佳解决方案，因为它既能保持准确性，又能减轻预测模型的偏差。这些结果表明，在部分认知的情况下，潜空间方法被认为优于对抗性偏差缓解方法。

{"title":"Latent space bias mitigation for predicting at-risk students","authors":"Ali Al-Zawqari, Dries Peumans, Gerd Vandersteen","doi":"10.1016/j.caeai.2024.100300","DOIUrl":"10.1016/j.caeai.2024.100300","url":null,"abstract":"<div><div>Researchers have observed the relationship between educational achievements and students' demographic characteristics in physical classroom-based learning. In the context of online education, recent studies were conducted to explore the leading factors of successful online courses. These studies also investigated how demographic features impact student achievement in the online learning environment. This motivates the use of demographic information alongside other features to predict students' academic performance. Since demographic features include protected attributes, such as gender and age, evaluating predictive models must go beyond minimizing the overall error. In this work, we analyze and investigate the use of neural networks to predict underperforming students in online courses. However, our goal is not only to enhance the accuracy but also to evaluate the fairness of the predictive models, a problem concerning the application of machine learning in education. This paper starts by analyzing the available solutions to fairness in predictive models: bias mitigation with pre-processing and in-processing methods. We show that the current evaluation is missing the case of partial awareness of protected features, which is the case when the model is aware of bias on some protected attributes but not all. The in-processing method, specifically the adversarial bias mitigation, shows that debiasing in some protected features exacerbates the bias on other protected features. This observation motivates our proposal of an alternative approach to enhance bias mitigation even in the partial awareness scenario by working with latent space. We implement the proposed solution using denoising autoencoders. The quantitative analysis used three distributions from The Open University Learning Analytics dataset (OULAD). The obtained results show that the latent space-based method offers the best solution as it maintains accuracy while mitigating the bias of the prediction models. These results indicate that in the case of partial awareness, the latent space method is considered superior to the adversarial bias mitigation approach.</div></div>","PeriodicalId":34469,"journal":{"name":"Computers and Education Artificial Intelligence","volume":"7 ","pages":"Article 100300"},"PeriodicalIF":0.0,"publicationDate":"2024-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418623","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Enhancing python learning with PyTutor: Efficacy of a ChatGPT-Based intelligent tutoring system in programming education 利用 PyTutor 加强 Python 学习：基于 ChatGPT 的智能辅导系统在编程教育中的功效

Q1 Social Sciences

Computers and Education Artificial Intelligence

Pub Date : 2024-09-25 DOI: 10.1016/j.caeai.2024.100309

Albert C.M. Yang , Ji-Yang Lin , Cheng-Yan Lin , Hiroaki Ogata

Programming is regarded as a focal point in the current rapidly evolving educational landscape. To aid learning in this domain, we developed PyTutor, an innovative intelligent tutoring system (ITS) that is designed to assist beginners in Python programming. PyTutor utilizes the ChatGPT model to offer continuous guidance, problem-solving hints, and detailed code explanations. It features a structured hint system for each question, covering pseudocode, cloze, basic, and advanced coding solutions. In our 11-week experiment, we compared 35 students who used PyTutor with 36 students who did not. The results indicated the effectiveness of PyTutor, particularly for students with weak foundations in programming. Those with lower initial knowledge exhibited higher engagement, completion rates, and success rates in in-class and after-class programming exercises. Nevertheless, we observed a potential risk of overreliance on PyTutor among students, which may impede the development of independent problem-solving skills. Thus, we recommend the balanced usage of PyTutor. In conclusion, PyTutor is a valuable ITS in programming education that considerably improves the learning outcomes of beginners. Its tailored approach renders it a promising tool for bridging knowledge gaps and enhancing overall educational experiences in the field of programming.

在当前快速发展的教育领域，编程被视为一个焦点。为了帮助这一领域的学习，我们开发了 PyTutor，这是一个创新的智能辅导系统（ITS），旨在帮助 Python 编程初学者。PyTutor 利用 ChatGPT 模型提供持续指导、问题解决提示和详细的代码解释。它的特点是每个问题都有一个结构化的提示系统，涵盖了伪代码、掐头去尾、基础和高级编码解决方案。在为期 11 周的实验中，我们对使用 PyTutor 的 35 名学生和未使用 PyTutor 的 36 名学生进行了比较。结果表明，PyTutor 非常有效，尤其是对于编程基础薄弱的学生。初始知识水平较低的学生在课上和课后编程练习中表现出更高的参与度、完成率和成功率。不过，我们也注意到学生中存在过度依赖 PyTutor 的潜在风险，这可能会阻碍独立解决问题能力的发展。因此，我们建议均衡使用 PyTutor。总之，PyTutor 是编程教育中一种有价值的智能学习工具，能显著提高初学者的学习效果。其量身定制的方法使其成为一种很有前途的工具，可以弥补编程领域的知识差距，提升整体教育体验。

{"title":"Enhancing python learning with PyTutor: Efficacy of a ChatGPT-Based intelligent tutoring system in programming education","authors":"Albert C.M. Yang , Ji-Yang Lin , Cheng-Yan Lin , Hiroaki Ogata","doi":"10.1016/j.caeai.2024.100309","DOIUrl":"10.1016/j.caeai.2024.100309","url":null,"abstract":"<div><div>Programming is regarded as a focal point in the current rapidly evolving educational landscape. To aid learning in this domain, we developed PyTutor, an innovative intelligent tutoring system (ITS) that is designed to assist beginners in Python programming. PyTutor utilizes the ChatGPT model to offer continuous guidance, problem-solving hints, and detailed code explanations. It features a structured hint system for each question, covering pseudocode, cloze, basic, and advanced coding solutions. In our 11-week experiment, we compared 35 students who used PyTutor with 36 students who did not. The results indicated the effectiveness of PyTutor, particularly for students with weak foundations in programming. Those with lower initial knowledge exhibited higher engagement, completion rates, and success rates in in-class and after-class programming exercises. Nevertheless, we observed a potential risk of overreliance on PyTutor among students, which may impede the development of independent problem-solving skills. Thus, we recommend the balanced usage of PyTutor. In conclusion, PyTutor is a valuable ITS in programming education that considerably improves the learning outcomes of beginners. Its tailored approach renders it a promising tool for bridging knowledge gaps and enhancing overall educational experiences in the field of programming.</div></div>","PeriodicalId":34469,"journal":{"name":"Computers and Education Artificial Intelligence","volume":"7 ","pages":"Article 100309"},"PeriodicalIF":0.0,"publicationDate":"2024-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142356972","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Preservice teachers’ behavioural intention to use artificial intelligence in lesson planning: A dual-staged PLS-SEM-ANN approach 职前教师在备课中使用人工智能的行为意向：双阶段 PLS-SEM-ANN 方法

Q1 Social Sciences

Computers and Education Artificial Intelligence

Pub Date : 2024-09-25 DOI: 10.1016/j.caeai.2024.100307

Bernard Yaw Sekyi Acquah , Francis Arthur , Iddrisu Salifu , Emmanuel Quayson , Sharon Abam Nortey

In the ever-changing landscape of education, the integration of technology has become an inevitable force that reshapes the foundations of teaching and learning. Amidst this transformative wave, the concept of Artificial Intelligence (AI) has taken center stage, promising innovative approaches, and increased efficiency. Within this context, the exploration of preservice teachers' behavioural intention to employ AI in lesson planning has emerged as a critical issue for examination. This study used a descriptive cross-sectional survey design and employed a purposive sampling technique to recruit 783 preservice teachers. By employing a cutting-edge dual-staged partial least squares structural equation modelling-artificial neural network (PLS-SEM-ANN) approach, this study investigated the influence of the following essential variables on preservice teachers' intentions to incorporate AI into their lesson planning endeavours: performance expectancy, effort expectancy, habit, hedonic motivation, social influence, and facilitating conditions. Social influence emerged as the most significant positive predictor of preservice teachers' behavioural intention to use AI in lesson planning. Additionally, habit, performance expectancy, effort expectancy, and facilitating conditions substantially positively influenced preservice teachers' behavioural intention to use AI in lesson planning. Conversely, hedonic motivation did not significantly affect preservice teachers’ behavioural intention to use AI in lesson planning. This study not only enhances our understanding of technology integration in pedagogy from a theoretical standpoint but also provides practical recommendations for refining educational curricula and instructional strategies that promote effective AI integration.

在不断变化的教育环境中，技术的整合已成为重塑教学基础的必然力量。在这股变革浪潮中，人工智能（AI）的概念占据了中心位置，有望带来创新的方法和更高的效率。在此背景下，探究职前教师在备课中使用人工智能的行为意向成为一个需要研究的关键问题。本研究采用描述性横断面调查设计，采用目的性抽样技术，招募了 783 名职前教师。通过采用最先进的双阶段偏最小二乘结构方程建模-人工神经网络（PLS-SEM-ANN）方法，本研究探讨了以下基本变量对职前教师将人工智能纳入备课工作的意愿的影响：绩效预期、努力预期、习惯、享乐动机、社会影响和便利条件。社会影响是对职前教师在备课中使用人工智能的行为意向最重要的积极预测因素。此外，习惯、成绩预期、努力预期和便利条件也对职前教师在备课中使用人工智能的行为意向产生了重大的积极影响。相反，享乐动机对职前教师在备课中使用人工智能的行为意向没有显著影响。这项研究不仅从理论上加深了我们对教学法中技术整合的理解，还为完善促进有效人工智能整合的教育课程和教学策略提供了切实可行的建议。

{"title":"Preservice teachers’ behavioural intention to use artificial intelligence in lesson planning: A dual-staged PLS-SEM-ANN approach","authors":"Bernard Yaw Sekyi Acquah , Francis Arthur , Iddrisu Salifu , Emmanuel Quayson , Sharon Abam Nortey","doi":"10.1016/j.caeai.2024.100307","DOIUrl":"10.1016/j.caeai.2024.100307","url":null,"abstract":"<div><div>In the ever-changing landscape of education, the integration of technology has become an inevitable force that reshapes the foundations of teaching and learning. Amidst this transformative wave, the concept of Artificial Intelligence (AI) has taken center stage, promising innovative approaches, and increased efficiency. Within this context, the exploration of preservice teachers' behavioural intention to employ AI in lesson planning has emerged as a critical issue for examination. This study used a descriptive cross-sectional survey design and employed a purposive sampling technique to recruit 783 preservice teachers. By employing a cutting-edge dual-staged partial least squares structural equation modelling-artificial neural network (PLS-SEM-ANN) approach, this study investigated the influence of the following essential variables on preservice teachers' intentions to incorporate AI into their lesson planning endeavours: performance expectancy, effort expectancy, habit, hedonic motivation, social influence, and facilitating conditions. Social influence emerged as the most significant positive predictor of preservice teachers' behavioural intention to use AI in lesson planning. Additionally, habit, performance expectancy, effort expectancy, and facilitating conditions substantially positively influenced preservice teachers' behavioural intention to use AI in lesson planning. Conversely, hedonic motivation did not significantly affect preservice teachers’ behavioural intention to use AI in lesson planning. This study not only enhances our understanding of technology integration in pedagogy from a theoretical standpoint but also provides practical recommendations for refining educational curricula and instructional strategies that promote effective AI integration.</div></div>","PeriodicalId":34469,"journal":{"name":"Computers and Education Artificial Intelligence","volume":"7 ","pages":"Article 100307"},"PeriodicalIF":0.0,"publicationDate":"2024-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142356973","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Fostering student competencies and perceptions through artificial intelligence of things educational platform 通过人工智能物联网教育平台培养学生的能力和感知力

Q1 Social Sciences

Computers and Education Artificial Intelligence

Pub Date : 2024-09-24 DOI: 10.1016/j.caeai.2024.100308

Sasithorn Chookaew , Pornchai Kitcharoen , Suppachai Howimanporn , Patcharin Panjaburee

The growing demand for artificial intelligence (AI) skills across various sectors has enhanced AI-focused careers and shaped academic exploration in educational institutions. These institutions have been actively developing teaching methods that enhance practical AI applications, particularly through integrating AI with the Internet of Things (IoT), leading to the emergence of the Artificial Intelligence of Things (AIoT). This convergence promises significant advancements in AI education, addressing gaps in structured learning methods for AIoT. This study explored AIoT's application in Smart Farming (SF) and its potential to enrich AI education and sectoral advancements. The AIoT platform was designed for SF simulations, integrating environmental sensing, AI processing, and user-friendly outputs. This platform was implemented with 40 first-year computer science university students in Thailand using a one-group pre-posttest design. This approach transformed theoretical AI concepts into experiential learning through interactive activities, demonstrating AIoT's capability to increase AI conceptual understanding, trigger AI competencies, and promote positive learning perceptions. Therefore, this study presented the results as indicative of the AIoT platform's potential benefits, emphasizing the need for further robust experimental research. This study contributes to educational technology discussions by suggesting improvements in AIoT platform effectiveness and highlighting areas for future investigation.

各行各业对人工智能（AI）技能的需求不断增长，促进了以人工智能为重点的职业发展，也影响了教育机构的学术探索。这些机构一直在积极开发能提高人工智能实际应用水平的教学方法，特别是通过将人工智能与物联网（IoT）相结合，促成了人工智能物联网（AIoT）的出现。这种融合有望极大地推动人工智能教育的发展，弥补人工智能物联网结构化学习方法的不足。本研究探讨了 AIoT 在智能农业（SF）中的应用及其丰富人工智能教育和行业进步的潜力。AIoT 平台专为智能农业模拟而设计，集成了环境传感、人工智能处理和用户友好的输出。泰国 40 名计算机科学专业的一年级学生采用单组前后测试设计实施了该平台。这种方法通过互动活动将人工智能理论概念转化为体验式学习，证明了人工智能物联网能够增强对人工智能概念的理解、激发人工智能能力并促进积极的学习感知。因此，本研究提出的结果表明，AIoT 平台具有潜在的优势，强调了进一步开展扎实实验研究的必要性。本研究为教育技术讨论提供了改进人工智能物联网平台有效性的建议，并强调了未来调查的领域。

{"title":"Fostering student competencies and perceptions through artificial intelligence of things educational platform","authors":"Sasithorn Chookaew , Pornchai Kitcharoen , Suppachai Howimanporn , Patcharin Panjaburee","doi":"10.1016/j.caeai.2024.100308","DOIUrl":"10.1016/j.caeai.2024.100308","url":null,"abstract":"<div><div>The growing demand for artificial intelligence (AI) skills across various sectors has enhanced AI-focused careers and shaped academic exploration in educational institutions. These institutions have been actively developing teaching methods that enhance practical AI applications, particularly through integrating AI with the Internet of Things (IoT), leading to the emergence of the Artificial Intelligence of Things (AIoT). This convergence promises significant advancements in AI education, addressing gaps in structured learning methods for AIoT. This study explored AIoT's application in Smart Farming (SF) and its potential to enrich AI education and sectoral advancements. The AIoT platform was designed for SF simulations, integrating environmental sensing, AI processing, and user-friendly outputs. This platform was implemented with 40 first-year computer science university students in Thailand using a one-group pre-posttest design. This approach transformed theoretical AI concepts into experiential learning through interactive activities, demonstrating AIoT's capability to increase AI conceptual understanding, trigger AI competencies, and promote positive learning perceptions. Therefore, this study presented the results as indicative of the AIoT platform's potential benefits, emphasizing the need for further robust experimental research. This study contributes to educational technology discussions by suggesting improvements in AIoT platform effectiveness and highlighting areas for future investigation.</div></div>","PeriodicalId":34469,"journal":{"name":"Computers and Education Artificial Intelligence","volume":"7 ","pages":"Article 100308"},"PeriodicalIF":0.0,"publicationDate":"2024-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142327804","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Which features of AI-based tools are important for students? A choice-based conjoint analysis 人工智能工具的哪些功能对学生很重要？基于选择的联合分析

Q1 Social Sciences

Computers and Education Artificial Intelligence

Pub Date : 2024-09-24 DOI: 10.1016/j.caeai.2024.100311

Jörg von Garrel, Jana Mayer

AI-based language tools such as ChatGPT have the potential to fundamentally change studying and teaching at universities. Since there have been few empirical studies on the use of AI systems by students, this article aims to analyze the use of AI in higher education. In particular, it focuses on identifying features that are important to students when using AI during their studies. For this purpose, a choice-based conjoint experiment was conducted through a survey involving over 6300 participants from German universities. The results show that students attach particular importance to the degree of scientific rigor. The optimal package for an AI-based tool for studying includes the citation of reliable or truthful sources, the detection and correction of errors in the input, a comprehensive and detailed formulation of the output, no AI-caused hallucinations and transparency of the database. These characteristics differ only slightly in group-specific analyses.

基于人工智能的语言工具（如 ChatGPT）有可能从根本上改变大学的学习和教学。由于有关学生使用人工智能系统的实证研究很少，本文旨在分析人工智能在高等教育中的使用情况。特别是，本文侧重于识别学生在学习过程中使用人工智能时的重要特征。为此，我们通过一项涉及德国高校 6300 多名参与者的调查，开展了一项基于选择的联合实验。结果显示，学生尤其重视科学的严谨性。人工智能学习工具的最佳组合包括：引用可靠或真实的资料来源、检测并纠正输入中的错误、全面而详细的输出表述、没有人工智能造成的幻觉以及数据库的透明度。这些特点在具体组别的分析中仅略有不同。

引用次数: 0

How to responsibly deploy a predictive modelling dashboard for study advisors? A use case illustrating various stakeholder perspectives 如何负责任地为研究顾问部署预测建模仪表板？一个说明不同利益相关者观点的使用案例

Q1 Social Sciences

Computers and Education Artificial Intelligence

Pub Date : 2024-09-20 DOI: 10.1016/j.caeai.2024.100304

Anouschka van Leeuwen, Marije Goudriaan, Ünal Aksu

Most higher education institutions employ study advisors to support their students. To adequately perform their task, study advisors have access to study information about their students. Using AI techniques to analyze that information and to predict if a student might be at risk of study delay could be a valuable tool in study advisors' practice. In this paper, we present a use case of how such a tool was developed (in the form of a dashboard) and which steps and considerations played a role in the responsible deployment of the tool. Three aspects are described: first, we present the timeline of the case study and zoom in on how the macro-level of the institution (where the groundwork is laid to facilitate AI-systems in education) and the micro-level of the implementation of the system influenced each other. Second, we describe which stakeholders were involved and what their ethical considerations were concerning data management, algorithms, and pedagogy. Third, we describe the initial evaluation of the dashboard in terms of study advisors’ experiences and provide suggestions on how to stimulate the responsible and useful implementation of a predictive modelling tool.

大多数高等教育机构都聘请学习顾问为学生提供支持。为了充分完成任务，学习顾问需要获取学生的学习信息。使用人工智能技术来分析这些信息，并预测学生是否有学习拖延的风险，可以成为学习顾问实践中的一个有价值的工具。在本文中，我们将介绍一个使用案例，说明如何开发这样一个工具（以仪表板的形式），以及在负责任地部署该工具的过程中，哪些步骤和考虑因素发挥了作用。本文从三个方面进行了阐述：首先，我们介绍了案例研究的时间轴，并放大了机构的宏观层面（为促进教育领域的人工智能系统奠定了基础）和系统实施的微观层面是如何相互影响的。其次，我们介绍了哪些利益相关者参与其中，以及他们在数据管理、算法和教学法方面的伦理考虑。第三，我们从学习顾问的经验角度描述了对仪表板的初步评估，并就如何促进负责任地、有用地实施预测建模工具提出了建议。

{"title":"How to responsibly deploy a predictive modelling dashboard for study advisors? A use case illustrating various stakeholder perspectives","authors":"Anouschka van Leeuwen, Marije Goudriaan, Ünal Aksu","doi":"10.1016/j.caeai.2024.100304","DOIUrl":"10.1016/j.caeai.2024.100304","url":null,"abstract":"<div><div>Most higher education institutions employ study advisors to support their students. To adequately perform their task, study advisors have access to study information about their students. Using AI techniques to analyze that information and to predict if a student might be at risk of study delay could be a valuable tool in study advisors' practice. In this paper, we present a use case of how such a tool was developed (in the form of a dashboard) and which steps and considerations played a role in the responsible deployment of the tool. Three aspects are described: first, we present the timeline of the case study and zoom in on how the macro-level of the <em>institution</em> (where the groundwork is laid to facilitate AI-systems in education) and the micro-level of the <em>implementation</em> of the system influenced each other. Second, we describe which stakeholders were involved and what their ethical considerations were concerning data management, algorithms, and pedagogy. Third, we describe the initial evaluation of the dashboard in terms of study advisors’ experiences and provide suggestions on how to stimulate the responsible and useful implementation of a predictive modelling tool.</div></div>","PeriodicalId":34469,"journal":{"name":"Computers and Education Artificial Intelligence","volume":"7 ","pages":"Article 100304"},"PeriodicalIF":0.0,"publicationDate":"2024-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666920X24001073/pdfft?md5=84e1d2eaf7c82fb91184236266a2bfd7&pid=1-s2.0-S2666920X24001073-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142315031","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Navigating the ethical terrain of AI in education: A systematic review on framing responsible human-centered AI practices 探索教育领域人工智能的道德领域：关于制定负责任的、以人为本的人工智能实践的系统综述

Q1 Social Sciences

Computers and Education Artificial Intelligence

Pub Date : 2024-09-19 DOI: 10.1016/j.caeai.2024.100306

Yao Fu , Zhenjie Weng

With the rapid development of artificial intelligence (AI) in recent years, there has been an increasing number of studies on integrating AI in various educational contexts, ranging from early childhood to higher education. Although systematic reviews have widely reported the effects of AI on teaching and learning, limited reviews have examined and defined responsible AI in education (AIED). To fill this gap, we conducted a convergent systematic mixed studies review to analyze key themes emerging from primary research. Following the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines, we searched Scopus and Web of Science and identified 40 empirical studies that satisfied our inclusion criteria. Specifically, we used four criteria for the screening process: (1) the study's full text was available in English; (2) the study was published before April 10th, 2024 in peer-reviewed journals or conference proceedings; (3) the study was primary research that collected original data and applied qualitative, quantitative, or mixed-methods as the study methodology; and (4) the study had a clear focus on ethical and/or responsible AI in one or multiple educational context(s). Our findings identified essential stakeholders and characteristics of responsible AI in K-20 educational contexts and expanded understanding of responsible human-centered AI (HCAI). We unveiled characteristics vital to HCAI, encompassing Fairness and Equity, Privacy and Security, Non-maleficence and Beneficence, Agency and Autonomy, and Transparency and Intelligibility. In addition, we provided suggestions on how to achieve responsible HCAI via collaborative efforts of stakeholders, including roles of users (e.g., students and educators), developers, researchers, and policy and decision-makers.

近年来，随着人工智能（AI）的快速发展，有关将人工智能融入从幼儿教育到高等教育等各种教育环境的研究越来越多。尽管系统性综述广泛报道了人工智能对教学和学习的影响，但对负责任的人工智能教育（AIED）进行研究和定义的综述却很有限。为了填补这一空白，我们开展了一项趋同的系统性混合研究综述，以分析主要研究中出现的关键主题。根据《系统性综述和元分析首选报告项目》（PRISMA）指南，我们搜索了 Scopus 和 Web of Science，确定了 40 项符合纳入标准的实证研究。具体来说，我们在筛选过程中使用了四项标准：（1）研究报告的全文为英文；（2）研究报告于 2024 年 4 月 10 日之前发表在同行评审期刊或会议论文集上；（3）研究报告为收集原始数据的初步研究，并将定性、定量或混合方法作为研究方法；以及（4）研究报告明确关注一种或多种教育环境中的道德和/或负责任的人工智能。我们的研究结果确定了 K-20 教育环境中负责任的人工智能的重要利益相关者和特征，并拓展了对负责任的以人为本的人工智能（HCAI）的理解。我们揭示了对 HCAI 至关重要的特征，包括公平与公正、隐私与安全、非恶意与有利、代理与自主以及透明与智能。此外，我们还就如何通过利益相关者的共同努力实现负责任的 HCAI 提出了建议，包括用户（如学生和教育工作者）、开发人员、研究人员以及政策和决策者的角色。

{"title":"Navigating the ethical terrain of AI in education: A systematic review on framing responsible human-centered AI practices","authors":"Yao Fu , Zhenjie Weng","doi":"10.1016/j.caeai.2024.100306","DOIUrl":"10.1016/j.caeai.2024.100306","url":null,"abstract":"<div><div>With the rapid development of artificial intelligence (AI) in recent years, there has been an increasing number of studies on integrating AI in various educational contexts, ranging from early childhood to higher education. Although systematic reviews have widely reported the effects of AI on teaching and learning, limited reviews have examined and defined responsible AI in education (AIED). To fill this gap, we conducted a convergent systematic mixed studies review to analyze key themes emerging from primary research. Following the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines, we searched Scopus and Web of Science and identified 40 empirical studies that satisfied our inclusion criteria. Specifically, we used four criteria for the screening process: (1) the study's full text was available in English; (2) the study was published before April 10th, 2024 in peer-reviewed journals or conference proceedings; (3) the study was primary research that collected original data and applied qualitative, quantitative, or mixed-methods as the study methodology; and (4) the study had a clear focus on ethical and/or responsible AI in one or multiple educational context(s). Our findings identified essential stakeholders and characteristics of responsible AI in K-20 educational contexts and expanded understanding of responsible human-centered AI (HCAI). We unveiled characteristics vital to HCAI, encompassing Fairness and Equity, Privacy and Security, Non-maleficence and Beneficence, Agency and Autonomy, and Transparency and Intelligibility. In addition, we provided suggestions on how to achieve responsible HCAI via collaborative efforts of stakeholders, including roles of users (e.g., students and educators), developers, researchers, and policy and decision-makers.</div></div>","PeriodicalId":34469,"journal":{"name":"Computers and Education Artificial Intelligence","volume":"7 ","pages":"Article 100306"},"PeriodicalIF":0.0,"publicationDate":"2024-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666920X24001097/pdfft?md5=c9dd430227b9715d8d56e7a8bbfc0e2a&pid=1-s2.0-S2666920X24001097-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142314947","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

AI-based prediction of academic success: Support for many, disadvantage for some? 基于人工智能的学业成功预测：对许多人有利，对某些人不利？

Q1 Social Sciences

Computers and Education Artificial Intelligence

Pub Date : 2024-09-19 DOI: 10.1016/j.caeai.2024.100303

Lisa Herrmann, Jonas Weigert

The use of computational tools to predict academic success has become increasingly popular. Machine learning algorithms, trained on past study histories, have been shown to provide valid predictions. However, knowing about biases and unfairness in algorithms, one should take a closer look at these predictions. This paper explores the extent to which the predictive accuracy of academic success varies between specific groups of students, focusing on traditional and non-traditional students (NTS), who have not acquired a higher education entrance qualification at school. In a case study the study compares several popular algorithms and their prediction quality, and investigates whether misclassified NTS show positive or negative biases. Results revealed that the accuracy of predicting academic success for NTS was significantly lower than when considering all students as a whole. The direction of the distortion cannot be determined exactly due to small case numbers. The study emphasizes that the possibility of bias always has to be considered when predicting study success, and the use of such tools must ensure there are no undesirable biases that could affect certain students.

使用计算工具来预测学业成绩已变得越来越流行。根据过去的学习历史进行训练的机器学习算法已被证明可以提供有效的预测。然而，由于算法中存在偏差和不公平现象，我们应该对这些预测进行更仔细的研究。本文探讨了学业成功的预测准确性在特定学生群体中的差异程度，重点关注在学校未获得高等教育入学资格的传统学生和非传统学生（NTS）。在案例研究中，该研究比较了几种流行的算法及其预测质量，并调查了被误分类的非传统学生是否表现出积极或消极的偏差。结果显示，预测非应届毕业生学业成功的准确率明显低于整体考虑所有学生时的准确率。由于案例数量较少，无法准确确定失真的方向。该研究强调，在预测学习成功时，必须始终考虑到存在偏差的可能性，而且在使用此类工具时必须确保不存在可能影响某些学生的不良偏差。

{"title":"AI-based prediction of academic success: Support for many, disadvantage for some?","authors":"Lisa Herrmann, Jonas Weigert","doi":"10.1016/j.caeai.2024.100303","DOIUrl":"10.1016/j.caeai.2024.100303","url":null,"abstract":"<div><div>The use of computational tools to predict academic success has become increasingly popular. Machine learning algorithms, trained on past study histories, have been shown to provide valid predictions. However, knowing about biases and unfairness in algorithms, one should take a closer look at these predictions. This paper explores the extent to which the predictive accuracy of academic success varies between specific groups of students, focusing on traditional and non-traditional students (NTS), who have not acquired a higher education entrance qualification at school. In a case study the study compares several popular algorithms and their prediction quality, and investigates whether misclassified NTS show positive or negative biases. Results revealed that the accuracy of predicting academic success for NTS was significantly lower than when considering all students as a whole. The direction of the distortion cannot be determined exactly due to small case numbers. The study emphasizes that the possibility of bias always has to be considered when predicting study success, and the use of such tools must ensure there are no undesirable biases that could affect certain students.</div></div>","PeriodicalId":34469,"journal":{"name":"Computers and Education Artificial Intelligence","volume":"7 ","pages":"Article 100303"},"PeriodicalIF":0.0,"publicationDate":"2024-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666920X24001061/pdfft?md5=586423a6b9f5f5318134629961734dad&pid=1-s2.0-S2666920X24001061-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142310457","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0