Article

Applying learning analytics for the early prediction of students' academic performance in blended learning

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

Blended learning combines online digital resources with traditional classroom activities and enables students to attain higher learning performance through well-defined interactive strategies involving online and traditional learning activities. Learning analytics is a conceptual framework and as a part of our Precision education used to analyze and predict students' performance and provide timely interventions based on student learning profiles. This study applied learning analytics and educational big data approaches for the early prediction of students' final academic performance in a blended Calculus course. Real data with 21 variables were collected from the proposed course, consisting of video-viewing behaviors, out-of-class practice behaviors, homework and quiz scores, and after-school tutoring. This study applied principal component regression to predict students' final academic performance. The experimental results show that students' final academic performance could be predicted when only one-third of the semester had elapsed. In addition, we identified seven critical factors that affect students' academic performance, consisting of four online factors and three traditional factors. The results showed that the blended data set combining online and traditional critical factors had the highest predictive performance. © 2018, International Forum of Educational Technology and Society.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... Through the online learning environments, students' learning logs can be recorded, and a large number of learning characteristics can be obtained. Previous studies applied machine learning algorithms to analyze the system logs to predict students' learning performance (Romero et al., 2013;Lu et al., 2018). By predicting the performance of at-risk students, teachers are able to intervene with students at an early stage to improve their chances of getting credit for the course (Lu et al., 2018;Huang et al., 2020). ...
... Previous studies applied machine learning algorithms to analyze the system logs to predict students' learning performance (Romero et al., 2013;Lu et al., 2018). By predicting the performance of at-risk students, teachers are able to intervene with students at an early stage to improve their chances of getting credit for the course (Lu et al., 2018;Huang et al., 2020). ...
... We found that this study was laborintensive due to manually collecting information on classroom activities. Previous studies (Jong et al., 2007;Ming and Ming, 2012;Lu et al., 2018;Huang et al., 2020) have found that online programming systems record useful student manipulation data. They used the data to predict whether students would have learning difficulties so that they could remedy them in a timely manner. ...
Article
Full-text available
To understand students’ learning behaviors, this study uses machine learning technologies to analyze the data of interactive learning environments, and then predicts students’ learning outcomes. This study adopted a variety of machine learning classification methods, quizzes, and programming system logs, found that students’ learning characteristics were correlated with their learning performance when they encountered similar programming practice. In this study, we used random forest (RF), support vector machine (SVM), logistic regression (LR), and neural network (NN) algorithms to predict whether students would submit on time for the course. Among them, the NN algorithm showed the best prediction results. Education-related data can be predicted by machine learning techniques, and different machine learning models with different hyperparameters can be used to obtain better results.
... Analysis of video viewing behaviors based on clickstream interactions such as pause, forward, and backward can reveal insights into students' preferences and learning styles (De Boer et al., 2011;Dissanayake et al., 2018). Understanding these behaviors can aid the design of learning environments (Li et al., 2015), and for implementing timely interventions aimed at improving students' learning performance (Kleftodimos & Evangelidis, 2014;Lu et al., 2018). ...
... Developing prediction models or determining predictors of learning performance is shown to help identify students who are at risk of academic failure or drop-out (Mubarak et al., 2021). These predictions can provide insights regarding the timing of interventions (Lu et al., 2018). Numerous studies have focused on predicting students' learning or test performance (Brinton et al., 2015;Guo et al., 2015;Hussain et al., 2022;Soni et al., 2018), with most having used navigational or demographic data (Gardner & Brooks, 2018;Shahiri et al., 2015). ...
... However, aggregating numerous output measurements into one single average measure may be said to go somewhat against the nature of employing repeated measures. Lu et al. (2018) investigated several features such as out-of-class practice behaviors, homework, and quiz scores, as well as video clickstream behaviors to predict students' learning performance for a blended course on calculus. The authors revealed seven critical factors, including the number of backward clicks per week and the number of plays per week in their study. ...
Article
Full-text available
Video clickstream behaviors such as pause, forward, and backward offer great potential for educational data mining and learning analytics since students exhibit a significant amount of these behaviors in online courses. The purpose of this study is to investigate the predictive relationship between video clickstream behaviors and students’ test performance with two consecutive experiments. The first experiment was performed as an exploratory study with 22 university students using a single test performance measure and basic statistical techniques. The second experiment was performed as a conclusive study with 16 students using repeated measures and comprehensive data mining techniques. The findings show that a positive correlation exists between the total number of clicks and students’ test performance. Those students who performed a high number of clicks, slow backward speed or doing backwards or pauses achieved better test performance than those who performed a lower number of clicks, or who used fast-backward or fast-forward. In addition, students’ test performance could be predicted using video clickstream data with a good level of accuracy (Root Mean Squared Error Percentage (%RMSE) ranged between 15 and 20). Furthermore, the mean of backward speed, number of pauses, and number/percentage of backwards were found to be the most important indicators in predicting students’ test performance. These findings may help educators or researchers identify students who are at risk of failure. Finally, the study provides design suggestions based on the findings for the preparation of video-based lectures.
... Many authors have identified that LMS usage data can be utilized to predict student grades (Conijn et al., 2017;Elbadrawy et al., 2015;Jo et al., 2015;Lu et al., 2018;Yang et al., 2017). Fewer researchers have explored whether LA interventions can act like nudges (Thaler & Sunstein, 2009) to change student behaviour in a way that is believed to positively impact learning outcomes (Hellings & Haelermans, 2020;Lim et al., 2019). ...
... When looking for which student behaviour to target for change, we need to identify which behaviour is most likely to impact student grades. While there are studies, mentioned above, that identify which LMS activity most strongly correlates with grades, other researchers have found that no LMS activity predicts future grades better than earlier grades (Conijn et al., 2017;Elbadrawy et al., 2015;Jayaprakash et al., 2014;Lu et al., 2018). ...
... One of the promises of LA is to identify which students are in need of additional support before it is too late. This is why many papers have focused on answering the question, which LMS usage indicator can best predict future student grades (Conijn et al., 2017;Elbadrawy et al., 2015;Jo et al., 2015;Lu et al., 2018;Yang et al., 2017). However, at least in this research, LMS usage data was not as good a predictor of future student performance as earlier student grades. ...
Article
Full-text available
This study seeks to explore if student behaviour can be changed using social modelling, specifically to increase usage of a learning management system (LMS), and whether any such increased LMS usage leads to higher student grades. After years of research into learning analytics, exploring which indicator can best predict student performance, with hopes of using that insight to improve student outcomes, there remain very few empirical studies which are randomized controlled trials, which is necessary to identify causation, and none that take place in a blended learning environment in the Global South. As learning analytics is a subject area for improving the learning of students worldwide, it is time to include more than just the Global North. In this experiment, 309 first year undergraduate participants were randomly assigned to control and treatment groups. Each member in the treatment group was sent a weekly email containing a link to an online dashboard showing the student’s performance compared against other students in the same cohort. Students in the treatment group did increase their use of the LMS but that increased usage did not translate into higher grades implying that the most important learning behaviours are not captured by the LMS, at least not in this study. Also of interest were that female students showed higher levels of engagement with the online dashboard and that the best predictor of a student’s grade in the second half of the semester was the student’s grade in the first half, supporting existing literature.
... Most of the recent studies employed students' online behavioral data in the prediction since most of the experiments were conducted in an online learning environment or a blended setting. Some of the examples are presented in (Abdullah et al., 2021;Cerezo et al., 2017;Lu et al., 2018;Mansouri et al., 2021;Shayan & van Zaanen, 2019), authors used students' behaviors logs containing their interactions with the online content to assess their performance, while Ayouni et al. (2021) and Hussain et al. (2018) used it to measure students' levels of engagement. (Chen & Cui, 2020;Chen et al., 2020;Dass et al., 2021;Macarini et al., 2019) utilized features related to time spent on online materials to identify at-risk of failing students or the ones who are most likely to dropout. ...
... The MLR used to forecast students' academic performance achieved an optimal pMSE of 198.62 and a pMAPC value of 0.81, it accurately predicted the academic scores of 8 out of 10 students. Meanwhile, Lu et al. (2018) employed principal component regression that obtained pMSE and pMAPC values of 159.17 and 0.82, respectively. ...
Article
Full-text available
The last few years have witnessed an upsurge in the number of studies using Machine and Deep learning models to predict vital academic outcomes based on different kinds and sources of student-related data, with the goal of improving the learning process from all perspectives. This has led to the emergence of predictive modelling as a core practice in Learning Analytics and Educational Data Mining. The aim of this study is to review the most recent research body related to Predictive Analytics in Higher Education. Articles published during the last decade between 2012 and 2022 were systematically reviewed following PRISMA guidelines. We identified the outcomes frequently predicted in the literature as well as the learning features employed in the prediction and investigated their relationship. We also deeply analyzed the process of predictive modelling, including data collection sources and types, data preprocessing methods, Machine Learning models and their categorization, and key performance metrics. Lastly, we discussed the relevant gaps in the current literature and the future research directions in this area. This study is expected to serve as a comprehensive and up-to-date reference for interested researchers intended to quickly grasp the current progress in the Predictive Learning Analytics field. The review results can also inform educational stakeholders and decision-makers about future prospects and potential opportunities.
... The target of student performance prediction in the offline education environment can be divided into single-course scores or grades [16,27], overall scores [33,34], comprehensive ranking [17,40], whether a specific student has failing subjects in a term [10], etc. The methods used include deep learning algorithms, random forests (RF), Gaussian regression and SVM, etc. Veeramanickam et al. [33] propose the cumulative dragonfly-based neural network (CDF-NN) algorithm that uses the dragonfly algorithm to train neural networks and predict student performance in the eighth term based on student scores in the first to seventh terms (e.g., experiment scores, test scores and attendance scores). ...
... The methods used include deep learning algorithms, random forests (RF), Gaussian regression and SVM, etc. Veeramanickam et al. [33] propose the cumulative dragonfly-based neural network (CDF-NN) algorithm that uses the dragonfly algorithm to train neural networks and predict student performance in the eighth term based on student scores in the first to seventh terms (e.g., experiment scores, test scores and attendance scores). Lu et al. [16] collect online teaching video viewing records of students and offline data about extracurricular practice, homework, test scores, and after-school tutoring, and use principal component regression to predict student course scores. Zong et al. [40] propose the tri-branch convolutional neural network (CNN) model to predict the comprehensive ranking of students based on student campus card data and historical performance data. ...
Article
Full-text available
Performance prediction is an important research facet of educational data mining. Most models extract student behavior features from campus card data for prediction. However, most of these methods have coarse time granularity, difficulty in extracting useful high-order behavior combination features, dependence on 6 historical achievements, etc. To solve these problems, this paper utilizes prediction of grade point average (GPA prediction) and whether a specific student has failing subjects (failing prediction) in a term as the goal of performance prediction and proposes a comprehensive performance prediction model of college students based on behavior features. First, a method for representing campus card data based on behavior flow is introduced to retain higher time accuracy. Second, a method for extracting student behavior features based on multi-head self-attention mechanism is proposed to automatically select more important high-order behavior combination features. Finally, a performance prediction model based on student behavior feature mode difference is proposed to improve the model’s prediction accuracy and increases the model’s robustness for students with significant changes in performance. The performance of the model is verified on actual data collected by the teaching monitoring big data platform of Xi’an Jiaotong University. The results show that the model’s prediction performance is better than the comparison algorithms on both the failing prediction and GPA prediction.
... In the field of learning analytics, there is a plethora of work undertaken where researchers have used clustering and prediction techniques to: a) identify learner profiles; b) discover what constructs (operationalised as clustering features) relate to academic performance, and finally; c) apply contemporary pedagogical approaches to assist or cater to learner needs (Castejón et al., 2016;Corrin et al., 2017;Khalil & Ebner, 2017;Lu et al., 2018;Marques & Belo, 2011;Rogiers et al., 2019;Strang, 2016;Südkamp et al., 2018;Tan et al., 2018;Watson et al., 2017;Yukselturk & Top, 2013). This section reviews the key techniques used to identify learner profiles in the published literature. ...
... It has been successfully used to identify patterns of learner interactions with online learning platforms, such as groups characterised by their level of engagement with the learning environment (Khalil & Ebner, 2017;Rodrigues et al., 2016;Tan et al., 2018). In this particular domain, there is a consensus that hard clustering techniques are proven to be compelling in producing results from profiling learners in online learning contexts (Khalil & Ebner, 2017;Lu et al., 2018;Mojarad et al., 2018, pp. 130-139;Pardo et al., 2017;Peach et al., 2019). ...
Article
Full-text available
The growth and uptake of educational technology has significantly reshaped the delivery of distance and online learning. With an unprecedented number of learners engaging with online modes of education, there is a growing need to understand the underlying student enrolment motivations, goals and learning behaviours evolving from a highly diverse student population. Research in learning analytics has advanced the use of digital data to understand student learning processes. However, there remains a limited understanding of how non-traditional learner characteristics, needs and motivational factors influence their learning behaviour and engagement strategies. Survey data from 232 students enrolled in fully online degree programs at a large public research university in Australia has been examined and used to represent 1687 students that have not completed the survey. To characterise the larger population of students, we combined their demographics, digital trace data, and course performance to provide richer insights of non-traditional learners in online learning. Data science approaches are applied, including an unsupervised machine learning technique that revealed the results of six unique learner profiles, clearly differentiated by their motivation, demographic, engagement and performance. While the findings show that each learner profile faces unique study challenges, there are also unique opportunities associated with each profile that could be utilised to improve their learning outcomes. The practical implications of the study on teaching practices are further discussed.
... Considerando lo anterior, y aun centrándonos exclusivamente en la combinación de docencia presencial y no presencial en un mismo programa educativo, aventurarse en delimitar las ventajas e inconvenientes de escenarios y propuestas tan diversas resulta altamente arriesgado. No obstante, tal y como sugeríamos al inicio, es habitual atribuir a las modalidades híbridas las ventajas que, habitualmente, se han asociado al e-learning o la docencia en línea o a distancia (Bocconi y Trenti, 2014; Lu et al., 2018;Oakley, 2016): ...
... According to Lu et al. (2018) and Mcquiggan et al. (2007), methods of EDM are frequently utilized to get an early indication of those who are not performing well and are at risk pupils in order to maximize the results. The early identification of students who are considered to be at risk helps higher education institutions to focus greater attention on the performance of these students, which eventually contributes to the achievement of better results. ...
Article
Full-text available
The goal of this study is to come up with a way to predict how well first-year college students in a professional program do in their classes (BCA). Tracking students' academic progress is an important area for ensuring the optimal growth of their analytical and logical skills. Being able to predict a student's academic success in the years immediately after graduation is useful for many different groups, including the government, legislators, and educators. An ensemble model is made for this task using a decision tree, a gradient boost algorithm, and some Naive Bayes techniques. This model gives the most accurate and reliable results. A questionnaire was created to find the factors that affect students' academic, social, behavioral, and demographic performance in school. Then, based on how well each of the three approaches performed, an ensemble model was created. The quality of the outcomes from the suggested ensemble model was evaluated using a 10-fold cross-validation technique. The output of an ensemble model allows for accurate and efficient prediction of student performance, and can help pinpoint students who are at risk of failing or dropping out of school. In order to create the current model, we employ both classification and regression techniques. With the current data set, the model achieves 99.1% accuracy in determining the important factors influencing students' academic success. As the suggested methodology allows for early identification of students who are at danger, it can also offer preventative and remedial strategies to boost students' overall academic performance.
... The general perception about BL is that it is an engagement of passive knowledge and a large number of students can be accommodated and learn through different mediums like online and face-to-face learning environments (Oakley, 2016). Through blended learning tendency of learning among students is also showing higher positive trends and thus increasing their The Blended Teaching and Page | 181 skills and abilities (Lu et al., 2018). BL also enhances the skills of communication, creativity, learning through different learning environment and above all they get to know the technology better and can use it for variety of purposes. ...
Article
Full-text available
Purpose – The aim of this study is to bring about a comprehensive overview of innovative tends (technologies) employ for IIoTs adoptions for achieving high-quality deployment. Methodology/Approach – This study initially identified 702 different articles from reputable research databases namely Science Direct, Emerald Insight, Wiley Online Library, Google Scholar, IEEE Access and Z Online Library from the years 2018 to June, 2022. A total of 32 articles were selected after undergoing a screening of its titles, keywords, abstracts and contents inclusion and exclusion for rigorous analysis. The data extracted were analysed and presented in form of descriptive statistics (tables, frequency, figures and charts) using Microsoft Excel software. Findings –The results carefully analyse how these trends influence the adoption of IIoTs in the industries. Novelty/value – Being the most growing innovation for industrial internet of things using industry 4.0 technologies have gradually changed the way traditional industries operate.
... The drawback of predicting the students' performance at the end of the semester is that students are not motivated in their current semester, which can result in students' early dropout. Few studies have been conducted that try to predict students' performance right from the start of the course length [13], [14]. Subsequently, the earliest possible intervention is possible, which can encourage students to stay on the right path. ...
Article
Full-text available
In this research study, we propose an Explainable Artificial Intelligence (XAI) model that provides the earliest possible global and local interpretation of students’ performance at various stages of course length. Global and local interpretation is provided in such a way that the prediction accuracy of a single local observation is close to the model’s overall prediction accuracy. For the earliest possible understanding of student performance, local and global interpretation is provided at 20%, 40%, 60%, 80%, and 100% of course length. Machine Learning (ML) and Deep Learning (DL) which are subfields of Artificial Intelligence (AI) have recently emerged to assist all educational institution’s in predicting the performance, engagement, and dropout rate of online students. Unfortunately, traditional ML and DL techniques lack in providing data analysis results in an understandable human way. Explainable AI (XAI), a new branch of AI, can be used in educational settings, specifically in VLEs, to provide the instructor with the study performance results of thousands or even millions of online students in a human-understandable way. Thus, unlike black box approaches such as traditional ML and DL techniques, XAI can help instructors to interpret the strengths and weaknesses of an individual student, providing them with timely personalized feedback and guidance. Various traditional and various ensemble ML algorithms were trained on demographic, clickstream, and assessment features to determine which algorithm gives the best performance result. The best-performing ML algorithm was ultimately selected and provided to the XAI model as an input for local and global interpretation of students’ study behavior at various percentages of course length. We have used various XAI tools to give students’ performance reports to instructors, in an explicable human way, at different stages of course length. The intermediate data analysis and performance reports will help instructors and all key stakeholders in decision-making and optimally facilitate online students.
... Blended and e-learning have therefore gained popularity in healthcare education (Jowsey et al, 2020). Blended (combined online and classroom based) learning (Garrison and Kanuka, 2004), is particularly attractive to institutions and policy-makers due to its positive impact on student motivation and performance, accessibility and convenience (Bramer, 2020;Lu et al., 2018). ...
Article
Full-text available
Background The global nursing and midwifery skills shortage and need for an expanded nursing workforce that is fit for contemporary care delivery is widely acknowledged. The immense pressure the profession was already under because of austerity, staff shortages and increasingly complex healthcare needs has been worsened by the Covid-19 pandemic. The UK is extending and evaluating the use of blended learning programmes for pre-registration nursing and midwifery students to help address these issues. This study sought to explore relevant nursing and midwifery experiences from outside the UK to help inform future health professional education policy here and elsewhere. Aim To explore international experiences of using blended learning programmes to prepare nursing and midwifery students for initial professional registration and the implications for workforce development. Design Cross-sectional, sequential, mixed methods study Participants /settings: Nursing/nurse education leaders from across International Council of Nurses regions Methods Exploratory online survey and follow-up case studies. Participants’ knowledge and experiences of blended learning were examined along with any perceived benefits for workforce development and successful strategies for addressing the challenges blended learning presents in this context. Case studies were developed from survey responses and follow up telephone calls to provide more detailed information about reported successes. Findings Participants reported flexibility, cost effectiveness, increased student/tutor and student/student communication and interaction as benefits of blended learning. Challenges included the design and use of interactive learning resources, appropriate preparation and support for staff and students, the potential of blended learning to exacerbate otherwise hidden disadvantage and the need for multi-stakeholder cost/benefit evaluation. Conclusions Blended learning is used globally in the pre-registration education of nurses, midwives and other healthcare professionals. These findings broadly mirror the literature regarding the benefits blended learning offers healthcare students, staff and organisations and the strategies employed to mitigate risk. As the deployment of blended learning nursing and midwifery programmes expands, further work is needed to address gaps in the current evidence base regarding the practice and impact of this approach. These concern adequate preparation and support of students and staff, ensuring access to appropriate equipment and connectivity, exploration of student perceptions that online learning is of lesser value and comprehensive multi-stakeholder, exploratory evaluation to uncover any hidden factors and impact. Tweetable abstract Blended learning plays an effective part in the education of pre-registration nursing and midwifery students to address global workforce shortages but further work is needed to address gaps in the current evidence base regarding the practice and impact of this approach.
... Therefore, it is of the utmost importance to develop techniques which can identify reasons and forecast students' projected performance. To achieve this, a number of studies [5][6][7][8][9][10] have been conducted in the recent past to explore the e-learning domain. ...
Article
Full-text available
Virtual learning environment (VLE) is vital in the current age and is being extensively used around the world for knowledge sharing. VLE is helping the distance-learning process, however, it is a challenge to keep students engaged all the time as compared to face-to-face lectures. Students do not participate actively in academic activities, which affects their learning curves. This study proposes the solution of analyzing students’ engagement and predicting their academic performance using a random forest classifier in conjunction with the SMOTE data-balancing technique. The Open University Learning Analytics Dataset (OULAD) was used in the study to simulate the teaching–learning environment. Data from six different time periods was noted to create students’ profiles comprised of assessments scores and engagements. This helped to identify early weak points and preempted the students performance for improvement through profiling. The proposed methodology demonstrated 5% enhanced performance with SMOTE data balancing as opposed to without using it. Similarly, the AUC under the ROC curve is 0.96, which shows the significance of the proposed model.
... Na [12] proposed a framework for predicting students' learning performance based on a behavioral model and described their behavior characteristics and added context information to the collaborative filtering algorithm, including student knowledge point mastery and class knowledge points, and students' mastery is predicted according to the learning path excavated. Lu, et al. [13] applied learning analytics and educational big data approaches, including proposed course, consisting of video viewing behaviors, out-of-class practice behaviors, homework and quiz scores, and after-school tutoring, for the early prediction of students' final academic performance by principal component regression in a blended course. Moises et al. [14] used machine learning to create models for the early prediction of students' performance in solving LMS assignments, by just analyzing the LMS log files generated up to the moment of prediction. ...
Article
Full-text available
In online learning, students’ learning data such as time and logs are commonly used to predict the student’s learning performance. In a hybrid context, learning activities occur both online and offline. Thus, how to integrate online and offline learning data effectively for an accurate learning performance prediction becomes very challenging. This paper proposes a “prediction and alert” model for students’ learning performance in a hybrid learning context. The model is developed and evaluated through analyzing the 16-week (one semester) attributes of English learning data of 50 students in the eighth grade. Six significant variables were determined as learning performance attributes, namely, qualified rate, excellent rate, scores, number of practice sessions, practice time, and completion. The proposed model was put into actual practice through four months of application and modification, in which a sample of 50 middle school students participated. The model shows the feasibility and effectiveness of data analysis for hybrid learning. It can support students’ continuous online and offline learning more effectively.
... Nosseir and Fath [26] developed a mobile application for prediction of student performance combining Fuzzy Logic and Artificial Neural Networks. Among other AI methods that have been investigated to develop accurate models to forecast student retention are Decision Tree, Artificial Neural Networks Naive Bayesian, Support Vector Machine, and Genetic Algorithm [18,[27][28][29][30][31][32][33][34][35][36]. ...
Article
Full-text available
Understanding the key factors that play an important role in students’ performance can assist improvements in the teaching-learning process. As an alternative, artificial intelligence (AI) methods have enormous potential, facilitating a new trend in education. Despite the advances, there is an open debate on the most suitable model for machine learning applied to forecast student performance patterns. This paper addresses this gap, where a comparative analysis between AI methods was performed. As a research hypothesis, a fuzzy inference system (FIS) should provide the best accuracy in this forecast task, due to its ability to deal with uncertainties. To do so, this paper introduces a model proposal based on AI using a FIS. An online survey was carried to collect data. Filling out a self-report, respondents declare how often they use some learning strategies. In addition, we also used historical records of students’ grades and retention from the last 5 years before the COVID pandemic. Firstly, two experimental groups were composed of students with failing and passing grades, compared by the Mann-Whitney test. Secondly, an association between the ‘frequency of using learning strategies’ and ‘occurrence of failing grades’ was quantified using a logistic regression model. Then, a discriminant analysis was performed to build an Index of Student Performance Expectation (SPE). Considering the learning strategies with greater discriminating power, the fuzzy AI-based model was built using the database of historical records. The learning strategies with the most significant effect on students’ performance were lesson review (34.6%), bibliography reading (25.6%), class attendance (23.5%), and emotion control (16.3%). The fuzzy AI-based model proposal outperformed other AI methods, achieving 94.0% accuracy during training and a generalization capacity of 91.9% over the testing dataset. As a practical implication, the SPE index can be applied as a tool to support students’ planning in relation to the use of learning strategies. In turn, the AI model based on fuzzy can assist professors in identifying students at higher risk of retention, enabling preventive interventions
... There are a multitude of published studies describing complex algorithms and models that predict (with variable levels of accuracy) the likelihood of a student failing their subjects or leaving University (Jayaprakash et al 2014;Wolff et al 2013;Lacave, Molina & Cruz-Lemus 2018;Lu et al 2018;Tempelaar et al 2018). However, in this pilot project we took a "back to basics" approach through building solid relationships with teaching academics and divisional support staff to devise a program of proactive, pre-census support and contact at the subject level. ...
Article
Full-text available
Student progression, attrition and completion are key metrics for all Universities. Subject level progress rates can be a useful lead indicator of these metrics; thus allowing quality improvement activities to be piloted and evaluated in shorter timeframes. The aim of this pilot study was to examine the effect of a series of pre-census interventions on subject progress rates. In 2018, twenty five subjects from the Faculty of Science at Charles Sturt University were selected to participate in the pilot based on a history of poor subject progress rates. Approximately 700 disengaged students were identified via learning analytics and emailed in week 2. In weeks 3 and 4 a range of “triggers” were used to identify 425 students who remained disengaged and contact was made via phone and/or email. Overall average subject progress rates increased from 66% in 2017 to 80% in 2018 (p<0.05) with a significant reduction in the average number of fail and fail withdrawn grades (from 42 to 26 per subject, p<0.05). Overall, we demonstrated that significant improvements in subject progress rates can be achieved when academics and Divisional support staff work collaboratively to identify, contact and support disengaged students prior to the HECS census date.
... As a result, students are encouraged to connect socially with professors and classmates in a BL setting while also having access to lexibility in time and location that is impossible in face-to-face training (Eryilmaz, 2015). Lu et al. (2018) claim that BL is now well-liked by institutions due to its positive impacts on pupil motivation and performance. BL that incorporates online learning has many bene its, including encouraging online research, establishing connections between practitioners and the global community, instilling self-discipline, giving practitioners access to the enormous and reliable knowledge sources they need for work and school, and (Paul, 2021). ...
... The general perception about BL is that it is an engagement of passive knowledge and a large number of students can be accommodated and learn through different mediums like online and face-to-face learning environments (Oakley, 2016). Through blended learning tendency of learning among students is also showing higher positive trends and thus increasing their The Blended Teaching and Page | 181 skills and abilities (Lu et al., 2018). BL also enhances the skills of communication, creativity, learning through different learning environment and above all they get to know the technology better and can use it for variety of purposes. ...
Chapter
Full-text available
The purpose of this project is to determine how information and communication technology (ICT) may be used as a transformational tool to re-engineer educational systems and change processes using synchronous and asynchronous technologies for e-learning. The study also compared the characteristics and functions of synchronous and asynchronous technologies for e-learning systems that will work with the present COVID-19 pandemic related situations. At first, 88 articles were downloaded, and 34 of those articles relating to the topic were evaluated based on title and content criteria. Research databases including Sciences Direct, Emerald Insight, Google Scholar, and Research Gate were used to find these appropriate publications. Both synchronous and asynchronous technologies are highly helpful in the current circumstances, according to the study's conclusions. In addition, this study offers some suggestions on how the parents' financial condition may influence the adoption and use of ICT in the educational system. In line with this, the research also identifies which of these two technologies will best fit their capabilities. Email, web 2.0 technologies like Facebook, WhatsApp, Twitter, YouTube, and Telegram, including audio/video conferencing, discussion boards, and many more, all seem to be generally relatively straightforward to use, accessible for free, and may be utilised as platforms.
... The application viewpoint focuses on the uses and practical applications of LA. Many researchers in the field of LA studied the applications of LA and their studies highlighted specific purposes of LA, such as prediction of learning performance and retention (Hicks, 2018;Lu et al., 2018;Marbouti, Diefes-Dux, & Madhavan, 2016;Yu et al., 2018) or understanding of learners' behaviors (Berland, Martin, Benton, Smith, & Davis, 2013;Martín-Monje, Castrillo, & Mañana-Rodríguez, 2018;Ruipérez-Valiente, Muñoz-Merino, Leony, & Kloos, 2015), rather than discussing classification of applications of LA. There is not enough research on classification frameworks for LA seen from the application viewpoint. ...
... However, each case study has its own characteristics and nature, so different techniques can be selected as the best option to predict students' behavior. Lu et al. [49] applied principal component regression in the bended learning system to predict students' final performance. Xu et al. [50] used K-nearest neighbors (KNNs) and gradient boosting decision tree (GBDT) in the blended learning system to the predict of student performance and the possibility of early intervention. ...
Article
Full-text available
University electronic learning (e-learning) has witnessed phenomenal growth, especially in 2020, due to the COVID-19 pandemic. This type of education is significant because it ensures that all students receive the required learning. The statistical evaluations are limited in providing good predictions of the university’s e-learning quality. That is forcing many universities to go to online and blended learning environments. This paper presents an approach of statistical analysis to identify the most common factors that affect the students’ performance and then use artificial neural networks (ANNs) to predict students’ performance within the blended learning environment of Saudi Electronic University (SEU). Accordingly, this dissertation generated a dataset from SEU’s Blackboard learning management system. The student’s performance can be tested using a set of factors: the studying (face-to-face or virtual), percentage of attending live lectures, midterm exam scores, and percentage of solved assessments. The results showed that the four factors are responsible for academic performance. After that, we proposed a new ANN model to predict the students’ performance depending on the four factors. Firefly Algorithm (FFA) was used for training the ANNs. The proposed model’s performance will be evaluated through different statistical tests, such as error functions, statistical hypothesis tests, and ANOVA tests.
... There seems to be a consensus among researchers that blended learning methods have potential to impact the learning process and outcomes (Bernard et al., 2014;Lu et al., 2018;Peng & Fu, 2021). Blended learning was found to have a positive effect on better academic achievement (Evensen et al., 2020;Grønlien et al., 2021;Shang & Liu, 2018), to be more effective in terms of declarative and procedural knowledge instruction (Aguado et al., 2011), to be more self-regulated (Joe et al., 2017), and more motivated . ...
Article
Full-text available
The research aimed to make a quasi-experimental study of College English in a Chinese blended learning community environment. A mixed method was adopted to collect data about students’ academic achievement and learning satisfaction. Academic achievement consisted of obtrusive data from the final exam, and learning satisfaction data came from a unified standard questionnaire organized by the university. Utilizing quasi-experimental design, results showed that the experimental class (C1) adopting the framework of blended learning community (BLC) earned a higher level of academic achievement and satisfaction. It is concluded in this research that the BLC framework was shown to have a positive effect on students’ academic achievement and learning satisfaction than face-to-face (FTF) learning in the context of English learning in higher education. Pedagogical implications and future research recommendations were also included.
... Students' academic performance is one of the main priorities for educators in determining educational success at all educational levels (Tan et al., 2019). Predicting students' performance enables early interventions and support to be devised and implemented by educators to enhance their performance, particularly for students who are at risk of failure (Lu et al., 2018;Wakelam et al., 2020). Despite the increasingly growing popularity of DM in various educational settings, EDM studies are still rather minimal and are not discussed in-depth, particularly in developing countries such as Malaysia (Jamil et al., 2018;Shahiri & Husain, 2015). ...
Article
Full-text available
This study attempts to predict secondary school students’ performance in English and Mathematics subjects using data mining (DM) techniques. It aims to provide insights into predictors of students’ performance in English and Mathematics, characteristics of students with different levels of performance, the most effective DM technique for students’ performance prediction, and the relationship between these two subjects. The study employed the archival data of students who were 16 years old in 2019 and sat for the Malaysian Certificate of Examination (MCE) in 2021. The learning of English and Mathematics is a concern in many countries. Three main factors, namely students’ past academic performance, demographics, and psychological attributes were scrutinized to identify their impact on the prediction. This study utilized the Orange software for the DM process. It employed Decision Tree (DT) rules to determine the characteristics of students with low, moderate, and high performance in English and Mathematics subjects. DT and Naïve Bayes (NB) techniques show the best predictive performance for English and Mathematics subjects, respectively. Such characteristics and predictions may cue appropriate interventions to improve students’ performance in these subjects. This study revealed students’ past academic performance as the most critical predictor, as well as a few demographics and psychological attributes. By examining top predictors derived using four different classifier types, this study found that students’ past Mathematics performance predicts their MCE English performance and students’ past English performance predicts their MCE Mathematics performance. This finding shows students’ performances in both subjects are interrelated.
... Sandoval et al. [130] use background and academic records of students to predict a final score with implementation of linear regression, robust linear regression and random forest algorithms. In [131] authors apply principal component regression to features related to internal assessment and video viewing to predict final academic performance. ...
Thesis
Educational institutions seek to design effective mechanisms that improve academic results, enhance the learning process, and avoid dropout. The performance analysis and performance prediction of students in their studies may show drawbacks in the educational formations and detect students with learning problems. This induces the task of developing techniques and data-based models which aim to enhance teaching and learning. Classical models usually ignore the students-outliers with uncommon and inconsistent characteristics although they may show significant information to domain experts and affect the prediction models. The outliers in education are barely explored and their impact on the prediction models has not been studied yet in the literature. Thus, the thesis aims to investigate the outliers in educational data and extend the existing knowledge about them. The thesis presents three case studies of outlier detection for different educational contexts and ways of data representation (numerical dataset for the German University, numerical dataset for the Russian University, sequential dataset for French nurse schools). For each case, the data preprocessing approach is proposed regarding the dataset peculiarities. The prepared data has been used to detect outliers in conditions of unknown ground truth. The characteristics of detected outliers have been explored and analysed, which allowed extending the comprehension of students' behaviour in a learning process. One of the main tasks in the educational domain is to develop essential tools which will help to improve academic results and reduce attrition. Thus, plenty of studies aim to build models of performance prediction which can detect students with learning problems that need special help. The second goal of the thesis is to study the impact of outliers on prediction models. The two most common prediction tasks in the educational field have been considered: (i) dropout prediction, (ii) the final score prediction. The prediction models have been compared in terms of different prediction algorithms and the presence of outliers in the training data. This thesis opens new avenues to investigate the students' performance in educational environments. The understanding of outliers and the reasons for their appearance can help domain experts to extract valuable information from the data. Outlier detection might be a part of the pipeline in the early warning systems of detecting students with a high risk of dropouts. Furthermore, the behavioral tendencies of outliers can serve as a basis for providing recommendations for students in their studies or making decisions about improving the educational process.
... Most of the early studies focused on pedagogy and psychology, trying to explore the key factors affecting students' academic performance, such as personality composition, learning motivation, family environment, etc. [3]. is kind of research is mainly based on the self-assessment reports provided by some students, which has great defects in sample size and timeliness, and the conclusions are also susceptible to the influence of the subjective consciousness of the interviewed individuals. ...
Article
Full-text available
The prediction and analysis of student achievement aim to realize personalized guidance for students and improve student achievement and teacher’s teaching achievement. Student achievement is affected by many factors such as family environment, learning conditions, and individual performance. Traditional prediction methods often ignore that different factors have different effects on the same student’s score, and different students have different effects on the same factor, so the model constructed cannot realize personalized analysis and guidance for students. Therefore, this paper proposes a prediction model based on the analytic hierarchy process and genetic algorithm. Firstly, according to the relationship among different levels, the analytic hierarchy process (AHP) model is established. Then, a k-means clustering algorithm is used to process the experimental data. Secondly, in order to get rid of the negative impact of the randomness of the initial threshold and weight on model prediction accuracy, which leads to the prediction result falling into a local minimum, a genetic algorithm is proposed to find the optimal initial threshold and weight of model first. Finally, a prediction model based on the BP neural network is established to predict students’ scores, which proves that the prediction effect is good. The experiment was conducted with English major students in a university as the research object. Experimental results show that compared with traditional data mining methods, the proposed method has better prediction accuracy.
... A student that exhibits low progress or is likely to leave the university due to issues and other aspects are identified as ''at-risk'' students [14]. Prediction of these students as early as possible can bring many benefits and helpful tools to institutions and students, such as an early warning system [10], [15], [16] generated to resolve issues that might lead students to defer or discontinue their studies. Additionally, predicting the successful students can help educators identify the characteristics of an outstanding student, and these traits can be established as a set of principles that can be provided for students when they commence in a university. ...
Article
Full-text available
Student retention is an essential measurement metric in education, indicated by retention rates, which are accumulated as students re-enroll from one academic year to the next. High retention rates can be obtained if institutions aim to provide appropriate support and teaching methods among the various practices to prevent students from deferring their studies. To address this pressing challenge faced by educational institutions, the underlying factors and the methodological aspects of building robust predictive models are reviewed and scrutinized. Educational Data Mining (EDM) and Learning Analytics (LA) have been widely adopted for knowledge discovery from educational data sources, improving the teaching practice, and identifying at-risk students. Various predictive techniques are applied in LA, such as Machine Learning (ML), Statistical Analysis, and Deep Learning (DL). To gain an in-depth review of these techniques, academic publications have been reviewed to highlight their potential to resolve Student Retention issues in education. Additionally, the paper presents a taxonomy of ML approaches and a comprehensive review of the success factors and the features that are not indicative of student performance in three different learning environments: Traditional Learning, Blended Learning, and Online Learning. The survey reveals that supervised ML and DL techniques are broadly applied in Student Retention. However, the application of ensemble and unsupervised learning clustering techniques supporting the heterogenous and homogenous groups of students is generally lacking. Moreover, static and traditional features are commonly used in student performance, ignoring vital factors such as educators-related, cognitive, and personal data. Furthermore, the paper highlights open challenges for future research directions.
... Blended learning, also known as hybrid learning or mixed-mode education, is an instructional approach that combines the use of one or two different learning methodologies with the more conventional model of instruction in a classroom setting (Graham, 2006;Lee et al., 2017;Thai et al., 2017;Vasyura et al., 2020). Improving data analysis and computation skills has contributed to the popularity of the blended learning instructional style (Lu et al., 2018). Integration of face-to-face learning experiences in the classroom with online learning experiences in a thoughtful manner (Garrison and Kanuka, 2004). ...
Article
Full-text available
Blended learning is gaining popularity because it has shown to be a successful method for accommodating an increasingly varied student body while enhancing the learning environment by incorporating online teaching materials. Higher education research on blended learning contributes to the blended learning literature. The ideas for future researchers are a vital component of research-based research articles. This study aims to consolidate the recommendations made for future studies. Research articles published in Scope-indexed journals over the past 5 years were analyzed in this context. Each cited passage from the research was read and coded independently in this analysis. After a period of time, the codes were merged into categories and themes. In the results section, direct citations were used to support the codes. The number of publications increased starting in 2017 and continuing through 2020. In the year 2020, most articles were published. Approximately half of the publications provide recommendations for future research. The researchers’ recommendations were gathered under the titles “Research Content” and “Replication and Method” the researchers’ recommendations were gathered.
... However, this comes at the cost of an increased likelihood of biases affecting the processing of students' information for assessment and decision-making [16,17], including personal preferences for the parameters to take into account and for the type and amount of data that are necessary to generate sensible predictions. Consider that, as the semester progresses, the amount of information about students that is available to an educator accumulates, but its utility decreases as remedial actions become harder to implement and their success becomes more uncertain [18,19]. Earlier predictions are unquestionably more valuable than later predictions, but at the beginning of a course, very little information is available to the educator, making predictions about a student's difficulties even more uncertain (e.g., is initial poor performance symptomatic of a momentary hurdle, perhaps linked to the idiosyncrasies of an assignment, or a reliable indication of serious issues?). ...
Article
Full-text available
The extent to which grades in the first few weeks of a course can predict overall performance can be quite valuable in identifying at-risk students, informing interventions for such students, and offering valuable feedback to educators on the impact of instruction on learning. Yet, research on the validity of such predictions that are made by machine learning algorithms is scarce at best. The present research examined two interrelated questions: To what extent can educators rely on early performance to predict students’ poor course grades at the end of the semester? Are predictions sensitive to the mode of instruction adopted (online versus face-to-face) and the course taught by the educator? In our research, we selected a sample of courses that were representative of the general education curriculum to ensure the inclusion of students from a variety of academic majors. The grades on the first test and assignment (early formative assessment measures) were used to identify students whose course performance at the end of the semester would be considered poor. Overall, the predictive validity of the early assessment measures was found to be meager, particularly so for online courses. However, exceptions were uncovered, each reflecting a particular combination of instructional mode and course. These findings suggest that changes to some of the currently used formative assessment measures are warranted to enhance their sensitivity to course demands and thus their usefulness to both students and instructors as feedback tools. The feasibility of a grade prediction application in general education courses, which critically depends on the accuracy of such tools, is discussed, including the challenges and potential benefits.
... Iqbal et al. (2017) used a restricted Boltzmann machines (RBM) model to predict the grade of students who have been admitted to a bachelor's degree program. Lu et al. (2018) used a principal component regression model to predict students' academic performance. Khan et al. (2019) used the decision tree algorithm to identify students' final grades in an introductory programming course. ...
Article
Full-text available
Educational data mining (DEM) provides valuable educational information by applying data mining tools and techniques to analyze data at educational institutions. In this paper, tree-based machine learning algorithms are used to predict students’ overall academic performance in their bachelor’s program. The transcript data of the students in the same department in a Chinese university were collected. All the courses in the bachelor’s program were then divided into six typical categories, and the mean GPAs of each category were taken as primary input features for prediction. Three tree-based machine learning models were established, i.e. decision tree (DT), Gradient boosting decision tree (GBDT) and random forest (RF). Results show that we can successfully identify more than 80% of the students at low-performance risk using the RF model at the end of the second semester, which is meaningful because the global quality of teaching and learning of the department can be improved by taking targeted measures in time according to the machine learning model. Feature importance and the structure of decision tree were also analyzed to extract knowledge that is valuable for both students and teachers. The results of this case study can be used as a reference for other engineering departments in China.
... Over time, several emerging technologies create more flexible and diverse forms of applied blended learning and may improve its effectiveness. Examples include augmented reality (Dunleavy et al., 2009), learning management systems (Dias & Diniz, 2013), learning analytics (Lu et al., 2018) and virtual reality (Nortvig et al., 2020). Cheung and Slavin (2012) analysed the effect of publication year on the effectiveness of educational technology, and they found that the effect differed significantly across periods. ...
Article
Full-text available
Background Blended learning programs in Kindergarten through Grade 12 (K‐12) classrooms are growing in popularity; however, previous studies assessing their effects have yielded inconsistent results. Further, their effects have not been completely quantitatively synthesized and evaluated. Objectives The purpose of this study is to synthesize the overall effects of blended learning on K‐12 student performance, distinguish the most effective domains of learning outcomes, and examine the moderators of the overall effects. Methods For the purpose, this study conducted a meta‐analysis of 84 studies published between 2000 and 2020, and involved 30,377 K‐12 students. Results and Conclusions Results revealed that blended learning can significantly improve K‐12 students' overall performance [g = 0.65, p < 0.001, 95% CI = (0.54–0.77)], particularly in the cognitive domain [g = 0.74, p < 0.001, 95% CI = (0.61–0.88)). The testing of moderators indicates that the factors moderating the impact of blended learning on student performance in these studies included group activities, educational level, subject, knowledge type, instructor, sample size, intervention duration and region. Implications The results indicate that blended learning is an effective way to improve K‐12 students' performance compared to traditional face‐to‐face (F2F) learning. Additionally, these findings highlight valuable recommendations for future research and practices related to effective blended learning approaches in K‐12 settings.
... LA allows teachers to predict pupils' achievement (Hue et al., 2015) and improve pupils' mathematics achievement (Lu et al., 2018). Such prediction helps teachers take proactive steps to improve pupils' mastery of learning. ...
Article
Full-text available
Game-based learning has received increasing attention in recent years as it could help improve pupils’ motivation, self-efficacy, and achievement. Technological innovations like learning analytics (LA) and GBL offer pedagogical support for teachers. GBL could significantly support pupils’ learning as a learning approach compared to conventional approaches. Therefore, there is a need to elevate “ teachers’ level of knowledge on the impact of GBL. In the meantime, LA could be used to collect, analyze, and report data on the impact of GBL on pupils’ learning performance. In this light, GBL applications have been developed to facilitate the use of LA for teaching and learning. This paper describes the design of GBL with LA integration for teaching mathematics in primary schools. It documents the construction of the GBL and AL app, which is grounded on the Dick, Carey, and Carey Model and the theory of constructivism. In addition, the cognitive load theory was applied to ensure that the application accommodates pupils’ cognitive load. This study also validated the design of the GBL, and it was found to be relevant and engaging. Keywords: Game-based learning, mathematics, analytics, technology, education
... For this purpose, different approaches have been introduced. For example, Lu et al. (2018), or Panigutti et al. (2021 propose tools to examine fairness. Galhotra et al. (2017) on the other hand propose a testing algorithm, that generates test cases to measure fairness and discrimination. ...
... Other studies have also highlighted learning analytics to predict students' learning performance [15] and learning intervention [17]. For example, in Lu et al.'s study [18], learning analytics has been applied to analyze and predict students' performance by analyzing student learning profiles using the regression model to improve prediction performance. This study presents the possibility of group patterns and peer leader identification observed in peer learning for potential use in other peer learning contexts. ...
Article
Full-text available
Physical interaction in peer learning has been proven to improve students’ learning processes, which is pertinent in facilitating a fulfilling learning experience in learning theory. However, observation and interviews are often used to investigate peer group learning dynamics from a qualitative perspective. Hence, more data-driven analysis needs to be performed to investigate the physical interaction in peer learning. This paper complements existing works by proposing a framework for exploring students’ physical interaction in peer learning based on the graph analytics modeling approach focusing on both centrality and community detection, as well as visualization of the graph model for more than 50 students taking part in group discussions. The experiment was conducted during a mathematics tutorial class. The physical interactions among students were captured through an online Google form and represented in a graph model. Once the model and graph visualization were developed, findings from centrality analysis and community detection were conducted to identify peer leaders who can facilitate and teach their peers. Based on the results, it was found that five groups were formed during the physical interaction throughout the peer learning process, with at least one student showing the potential to become a peer leader in each group. This paper also highlights the potential of the graph analytics approach to explore peer learning group dynamics and interaction patterns among students to maximize their teaching and learning experience.
... Regarding the blended learning environment, another approach was applied in [7] using principal component regression to predict the student's final result. Their experiments show that the student's final grade could be expected when only one-third of the semester had elapsed, which is close to our approach as we can predict it when only five weeks out of fourteen passed. ...
Article
Full-text available
Recommending quizzes in e-Learning systems always represents a challenging task, as the quality of recommendations may have a high impact on the student’s progress. We propose a data analysis workflow based on building multiple stacks of models that use information from former students’ taken quizzes. The current implementation uses the RandomForest algorithm for building the models on a real-world dataset that has been obtained in a controlled environment. As preprocessing techniques, we have used normalization and discretization such that training data have been used for classification and regression tasks. At run-time, the models are queried for classifying the student and inferring an optimal quiz that is personalized for the student. We have evaluated the accuracy parametrized on the previous number of quizzes and found that a possible optimal timeframe for each class of students should be used and may provide more helpful quizzes.
... MOOCs, on the other hand, frequently record low completion rates and significant dropout rates (Sun, Ni, Zhao, Shen, & Wang, 2019). Several research have offered ways for predicting students' course success or failure (Er et al., 2019;Lu et al., 2018). A logistic regression model is one such method for prediction. ...
Article
Full-text available
The system architecture of big data in massive open online courses (BD-MOOCs System Architecture) is composed of six components. The first component was comprised of big data tools and technologies such as Hadoop, YARN, HDFS, Spark, Hive, Sqoop, and Flume. The second component was educational data science, which is composed of the following four parts: EDM, ERS, AA, and S/II. The third component was a description of three basic elements of a big data system: data capture, management, and analysis. The fourth component was that MOOCs were classified as cMOOCs, xMOOCs, quasi-MOOCs, hMOOCs, and other related. The fifth component included the steps of MOOC development: design, delivery, and assessment. Finally, MOOCs present educational data science challenges such as analyzing student interactions, estimating dropout risk, grading, and making recommendations. Overall, the BD-MOOCs system architecture design was suitable at the highest level.
... 13 different prediction techniques were applied to analyze the data collected over a 14-week semester with promising predictions being made as early as the 3rd week in the semester. Lu et al. (2018) investigated the prediction of a students' academic performance by applying principle component regression (PCR) to 21 indicators that were aggregated from the log data of a blended learning Calculus course. Prediction of the final academic performance could be made as early as one-third into the semester, and seven important factors in performance were identified, with the best prediction resulting from the use of blended data from both online and traditional classes. ...
Article
Full-text available
Digitized learning materials are a core part of modern education, and analysis of the use can offer insight into the learning behavior of high and low performing students. The topic of predicting student characteristics has gained a lot of attention in recent years, with applications ranging from affect to performance and at-risk student prediction. In this paper, we examine students reading behavior using a digital textbook system while taking an open-book test from the perspective of engagement and performance to identify the strategies that are used. We create models to predict the performance and engagement of learners before the start of the assessment and extract reading behavior characteristics employed before and after the start of the assessment in a higher education setting. It was found that strategies, such as: revising and previewing are indicators of how a learner will perform in an open ebook assessment. Low performing students take advantage of the open ebook policy of the assessment and employ a strategy of searching for information during the assessment. Also compared to performance, the prediction of overall engagement has a higher accuracy, and therefore could be more appropriate for identifying intervention candidates as an early-warning intervention system.
... A lot of researchers have done researches to elaborate its effectiveness from grade one to higher education in various disciplines (Marchalot et al., 2018;Zhang & Zhu, 2020) and proved to be one of the most dynamic learning strategies in various disciplines. Lu et al., (2018) suggested that blended learning strategy is endorsed by various colleges and universities in various disciplines because of its positive results on students' academic achievements and critical thinking skills. Cuesta, (2010) suggested that the key objective of blended learning strategy is to offer a platform for the learners according to their skills, styles and needs. ...
Article
Full-text available
Learning strategies have shifted from conventional information to communication technology-based learning since the beginning of the twenty-first century. A study of published articles on blended and traditional learning strategies was done to emphasise the value and significance of both learning strategies and to investigate their efficacy in promoting a safer learning environment in different educational levels. Thirty-six (36) research articles from various disciplines published in Web of Science and Scopus databases were chosen for review. According to the review of researches, blended learning demonstrated to be a more successful learning approach than traditional learning strategy in the majority of studies. Twenty-five (25) studies found a statistically more significant benefit in blended learning strategy for academic success, critical and creative abilities, and a safer learning environment in diverse disciplines, out of 36 published articles evaluated. Based on the findings of this study, it is strongly suggested that blended learning strategies be used to attain high academic and professional goals while also providing a safer learning environment in educational institutions and society.
Article
The flipped classroom approach is aimed at improving learning outcomes by promoting learning motivation and engagement. Recommendation systems can also be used to improve learning outcomes. With the rapid development of artificial intelligence (AI) technology, various systems have been developed to facilitate student learning. Accordingly, we applied AI-enabled personalized video recommendations to stimulate students' learning motivation and engagement during a systems programming course in a flipped classroom setting. We assigned students to control and experimental groups comprising 59 and 43 college students, respectively. The students in both groups received flipped classroom instruction, but only those in the experimental group received AI-enabled personalized video recommendations. We quantitatively measured students’ engagement based on their learning profiles in a learning management system. The results revealed that the AI-enabled personalized video recommendations could significantly improve the learning performance and engagement of students with a moderate motivation level.
Chapter
Educational Data Mining (EDM) is a discipline developed by focusing on improving independent and adaptive learning methods to find hidden education patterns. In this area, heterogeneous data is known to continue to develop in a big-data paradigm. Several specific data mining techniques are required to extract information with an adaptive value from the available educational data. Therefore, this study aims to present a grouping approach related to partitioning students into a different group or cluster based on the students’ behavior during lessons. Then, the architecture related to the e-learning system will be personalized to detect and provide suitable teaching methods and content according to each student's learning ability so that students can improve their quality and learning ability. The grouping methods that can be done in this educational data mining include K-Means, K-Medoids, Agglomerative Hierarchical Cluster Trees, Noise-Based Application Density-Based Spatial Clustering, and Fast Search and Density Peak Findings through Heat Diffusion (CFSFDP-HD) Shows the average compute time with different student count benchmarks: 600, 1200, 1800, 2400, 3000, 3600. Then, it has been found that the CFSFDP-HD method has strong results compared to other methods.KeywordsEducational data miningBig dataClusteringProfile learninge-Learning
Chapter
Currently, learning early warning mainly uses two methods, student classification and performance regression, both of which have some shortcomings. The granularity of student classification is not fine enough. The performance regression gives an absolute score value, and it cannot directly show the position of a student in the class. To overcome the above shortcomings, we will focus on a rare learning early warning method — ranking prediction. We propose a dual-student performance comparison model (DSPCM) to judge the ranking relationship between a pair of students. Then, we build the model using data including class quiz scores and online behavior times and find that these two sets of features improve the Spearman correlation coefficient for the ranking prediction by 0.2986 and 0.0713, respectively. We also compare the process proposed with the method of first using a regression model to predict scores and then ranking students. The result shows that the Spearman correlation coefficient of the former is 0.1125 higher than that of the latter. This reflects the advantage of the DSPCM in ranking prediction.KeywordsLearning early warningStudent ranking predictionClass quiz scoreOnline behavior time
Chapter
Blended teaching has the characteristics of small scale, strong controllability, definite learning tasks and consideration of both online and offline teaching. The quantitative evaluation indicators of learners' blended learning behavior enthusiasm and stability are proposed, and then used for learning behavior analysis and performance prediction. It analyzes the distribution, correlation, consistency and effectiveness of online and offline learning behavior indicators, and it is found that there is a high correlation between learning behavior indicators and the final grade. The prediction is carried on the data set composed of learning behavior indicators, students' basic information, online and offline learning data. The improved forest optimization algorithm is applied to select features. The naive Bayes, decision tree and random forest classifier are used to predict the final performance. The experiments show that the learning behavior indicators can effectively reduce the scale of feature set and improve the performance prediction effect.KeywordsLearning behavior analysisPerformance predictionBlended teachingData miningInformation entropyFeature engineering
Article
Full-text available
In this work, 29 features were defined and implemented to be automatically extracted and analysed in the context of NeuroK, a learning platform within the neurodidactics paradigm. Neurodidactics is an educational paradigm that addresses optimization of the learning and teaching process from the perspective of how the brain functions. In this context, the features extracted can be fed as input into various machine learning algorithms to predict the students’ performance. The proposed approach was tested with data from an international course with 698 students. Accuracies greater than 0.99 were obtained in predicting the students’ final performance. The best model was achieved with the Random Forest algorithm. It selected 7 relevant features, all with a clear interpretation in the learning process. These features are related to the principles of neurodidactics, and reflect the importance of a social learning and constructivist approach in this context. This work constitutes a first step in relating the tools of learning analytics to neurodidactics. The method, after its adaptation to capture relevant features corresponding to different contexts, could be implemented on other management learning platforms, and applied to other online courses with the aim of predicting the students’ performance, including real-time tracking of their progress and risk of dropout.
Article
This study examined learning processes in undergraduate online general chemistry courses. The study aimed to characterize learners according to their learning patterns and to identify indicators that predict students' success in an online environment. Specifically, we focused on the role of a central factor affecting success in online courses: self-regulated learning and learner engagement. To this end, we used a mixed methods approach that combines semi-structured interviews and statistical analysis. We applied two logistic regression models and a decision tree algorithm and found two parameters that can predict completion of the course: the submission status of an optional assignment and the students' cumulative video opening pattern (SCOP). Recommendations for institutions and lecturers regarding the benefits of implementing these models to identify self-regulated learning patterns in online courses and to design future effective interventions are discussed. Regarding students, we emphasize the importance of time management and how choices they make with respect to their learning process affect their potential for success.
Article
Full-text available
Massive Open Online Courses (MOOCs) have received much attention in higher education; however, evidence about MOOCs at the K-12 level is scarce. To shed light on the phenomenon, we use the i-MOOC that aims at fostering upper secondary level students’ information literacy. The i-MOOC is a blended MOOC developed and refined in a design research process; it meets established criteria for high-quality MOOCs. In 2020, 1032 upper secondary level students in German-speaking Switzerland took the i-MOOC; the sample comprises N = 167 students who voluntarily filled in a questionnaire. The students are mainly from high schools and vocational schools. Learning effects are captured with a performance test. Information literacy gains are significant and medium in size: d = 0.75. The technology acceptance of students is evaluated using the extended unified theory of acceptance and use of technology (UTAUT2). Student technology acceptance of K-12 MOOCs is primarily driven by hedonic motivation, i.e., perceived fun and entertainment. However, this type of motivation negatively predicts learning gains. Implications for teachers and educational decision makers are discussed.
Article
Learning analytics has drawn the attention of academics and administrators in recent years as a tool to better understand students' needs and to tailor appropriate and timely responses. While many value the potential of learning analytics, it is not without critics, especially with regards to ethical concerns surrounding the level and type of data gathered, and scepticism on data's ability to measure something useful and actionable. This paper gathers the thoughts of key stakeholders', including staff and students, and their expectations of learning analytics, their priorities for using student data, and how they should be supported to act on the data, with the aim of aiding institutions with their plans to implement learning analytics. For this analysis we explored stakeholders' awareness, concerns, priorities and support needs with respect to effectively accessing, interpreting and utilising learning data through the use of surveys and focus groups. These were adapted from the previously published SHEILA framework protocols with additional topics added relating to awareness, uses of data, and support. The focus groups were used to capture prevalent themes, followed by surveys to gain perspectives on these themes from a wider stakeholder audience. Overall, results suggest there are significant differences in the perspectives of each of the stakeholders. There is also a strong need for both additional training and ongoing support to manage and realise stakeholders' understanding and goals around learning analytics. Further research is necessary to explore the needs of other stakeholders not captured in this study especially around differently abled students.
Chapter
COVID-19 continues to overwhelm the education sectors globally posing threats to progress made towards inclusive and equity education in the previous years. Before COVID-19, continuous evaluation methods were systematically done manually in many learning institutions, now it is difficult to timely identify underperforming students, and students who are at risk of dropping out and provision of timely remedial and appropriate actions tailored for individual’s needs. To alleviate this situation, early prediction of students’ performance using deep learning techniques and data generated from smart learning environments becomes imperative. Therefore, this study aimed at providing a pioneering comprehensive review of hybrid and deep learning models to predict students’ performance in online learning environments . The study revealed that deep learning techniques extract hidden data to predict students’ performance, identify students at risk of dropping out, monitor students’ cognitive learning styles and unusual learning behaviours, emotional state of students to facilitate pedagogical content knowledge, instructional designs and appropriate action promptly. These models use various performance predictors such as course attributes, study time and duration, internal assessments, socio-economic, students’ legacy and learning environment. Furthermore, the study revealed that the psychological state of students was not taken into consideration, yet it impacts learning outcomes. However, the varying context of implementation could be the leading cause of differences in perspective to determine performance predictors that are reliable to predict student performance. Predicting students’ performance should be done prior, during and at the end of the course to ensure effective implementation of educational interventions. KeywordsDeep learningStudents’ academic performanceSmart online learning
Article
Problem Statement Advanced technology makes a positive and negative impact on the student’s academic performances. It makes a huge impact on the learners psychological factors. On the other hand, frequent distraction makes a negative impact on the students overall academic performances. Poor network connectivity is also responsible for creating obstacles in academic performance. Methodology Secondary data collection techniques have been used to justify the qualitative research method. A systematic review and thematic analysis process have been used in the findings and analysis section Main Findings The fellow researchers have detected that advanced technology helped to motivate students to improve their learning abilities. It helps to determine the learning goals for better future performances. It helps to improve students' soft skills for enrichment of their performances. Conclusion From this current research article, it has been stated that blended learning skills help to improve students’ academic performance. It meets the student's requirements by providing numerous resources to acquire knowledge.
ResearchGate has not been able to resolve any references for this publication.