Search

Despite the growing interest in incorporating response time data into item response models, there has been a lack of research investigating how the effect of speed on the probability of a correct response varies across different groups (e.g., experimental conditions) for various items (i.e., differential response time item analysis). Furthermore, previous research has shown a complex relationship between response time and accuracy, necessitating a functional analysis to understand the patterns that manifest from this relationship. In this study, response time data are incorporated into an item response model for two purposes: (a) to examine how individuals' speed within an experimental condition affects their response accuracy on an item, and (b) to detect the differences in individuals' speed between conditions in the presence of within-condition effects. For these two purposes, by-variable smooth functions are employed to model differential and functional response time effects by experimental condition for each item. This model is illustrated using an empirical data set to describe the effect of individuals' speed on their reading comprehension ability in two experimental conditions of reading medium (paper vs. digital) by item. A simulation study showed that the recovery of parameters and by-variable smooth functions of response time was satisfactory, and that the type I error rate and power of the test for the by-variable smooth function of response time were acceptable in conditions similar to the empirical data set. In addition, the proposed method correctly identified the range of response time where between-condition differences in the effect of response time on the probability of a correct response were accurate.

It is unknown whether and to what extent cognitive individual differences may play different roles in paper versus computer-based second language (L2) writing. This exploratory study is a first attempt to explore this issue, focusing on the effects of working memory and language aptitude on the quality of paper versus computer-based L2 writing performance. Forty-two Spanish learners of L2 English performed a problem-solving task either digitally or on paper, took a working memory n-back test, and completed LLAMA tests to measure language aptitude. The quality of their L2 written texts was assessed in terms of complexity, accuracy and fluency (CAF) measures. The results indicated that the role of cognitive individual differences may vary depending on the writing environment.

This study compared the effects of computer-mediated (CM) versus pen-and-paper (P&P) writing on written accuracy and feedback processing in tasks written and rewritten collaboratively following a pedagogical treatment in two intact authentic classrooms. The study involved 32 secondary education low-proficiency English-as-a-foreign-language (EFL) learners writing two descriptive texts collaboratively and receiving in-class training in the identification and correction of grammatical, lexical, and mechanical errors. Participants were provided with unfocused direct error correction (EC). Error logs were used to facilitate noticing of teacher corrections (i.e., feedback processing). Dyads were required to rewrite their texts for evidence of feedback uptake. Results indicate that writing collaboratively on the computer with the availability of the Internet contributes to increased grammatical and lexical accuracy. No differences were found between writing environments regarding feedback processing or accuracy of rewritten texts.

The purpose of this study was to investigate the impact of mLearning or mobile device practice on the attention and accuracy of student's use of math concepts, specifically, telling time. A single subject, alternating treatment design was used to compare mLearning to paper and pencil practice in four 3rd grade male students. Results were mixed; two children were observed to be more on-task during the mLearning practice, and two children were observed to perform similarly across both conditions. Additionally, two children performed similarly on correctly completed problems across both conditions, and two children performed better using paper and pencil practice. All students completed more math problems during the paper and pencil practice.

Incorporating counterarguments can be challenging for many student-writers, including those for whom English is a second or other language. In this study, we present findings that may lead to improvements in students' understanding of the benefits of integrating counterarguments. In our study, expert readers were presented with excerpts of student academic writing in order to assess the degree to which counterarguments could be identified through student deployment of metadiscoursal features. The results suggest that student-writers deploy metadiscoursal features with insufficient frequency and consistency. The results further suggest that when metadiscoursal features are used in counterarguments, their deployment is largely restricted to the sentence-initial position. The study also considers the pedagogical and technological issues associated with the findings.

This study aimed to compare the accuracy of the test scores as results of Test of English Proficiency (TOEP) based on paper and pencil test (PPT) versus computer-based test (CBT). Using the participants' responses to the PPT documented from 2008-2010 and data of CBT TOEP documented in 2013-2014 on the sets of 1A, 2A, and 3A for the Listening and Reading section, the researcher estimated the reliability estimation results using classical test theory and the value of information function and on the item response theory on PPT are then compared with CBT, which has the greater reliability and the value of information functions is said to be more accurate. The study shows that with the classical test theory approach, the reliability coefficients between the scores of the results of PPT and those of CBT are almost the same, and using the item response theory, it was found that although the value of the information function on PPT and CBT are relatively similar in several subtests, there is a tendency for participants with the moderate ability that CBT is more accurate than PPT, and for the low and high ability of participants, PPT tends to be more accurate than CBT.

Effective nutrition policies require timely, accurate individual dietary consumption data; collection of such information has been hampered by cost and complexity of dietary surveys and lag in producing results. The objective of this work was to assess accuracy and cost-effectiveness of a streamlined, tablet-based dietary data collection platform for 24-hour individual dietary recalls (24HR) administered using INDDEX24 platform v. a pen-and-paper interview(PAPI) questionnaire, with weighed food record (WFR) as a benchmark. This cross-sectional comparative study included women 18-49 years old from rural Burkina Faso (n 116 INDDEX24; n 115 PAPI). A WFR was conducted; the following day, a 24HR was administered by different interviewers. Food consumption data were converted into nutrient intakes. Validity of 24HR estimates of nutrient and food group consumption was based on comparison with WFR using equivalence tests (group level) and percentages of participants within ranges of percentage error (individual level). Both modalities performed comparably estimating consumption of macro- and micronutrients, food groups and quantities (modalities' divergence from WFR not significantly different). Accuracy of both modalities was acceptable (equivalence to WFR significant at P < 0·05) at group level for macronutrients, less so for micronutrients and individual-level consumption (percentage within ±20 % for WFR, 17-45 % for macronutrients, 5-17 % for micronutrients). INDDEX24 was more cost-effective than PAPI based on superior accuracy of a composite nutrient intake measure (but not gram amount or item count) due to lower time and personnel costs. INDDEX24 for 24HR dietary surveys linked to dietary reference data shows comparable accuracy to PAPI at lower cost.

Tablet computers (tablets) are positioned to be powerful, innovative, effective, and motivating research and assessment tools. We addressed two questions critical for evaluating the appropriateness of using tablets to study number-line estimation, a skill associated with math achievement and argued to be central to numerical cognition. First, is performance with paper and pencil comparable with performance on a tablet? Second, is comparability affected by students' preference for one method of presentation? Thirty-two students in Grade 6 estimated targets on a number line; half estimated with paper and pencil and half with a tablet. For both presentation methods, students' performance was comparable. Students liked both presentation conditions equally but, when asked to choose, most students preferred the tablet. Preference did not influence comparability of results across presentation methods. Finally, students' reasons for their preferences were explored, along with implications for using tablet applications in research and educational assessment.

Many teachers like to introduce the Bohr atom toward the end of an introductory physics course. This is an excellent idea, given the historic importance of Bohr's 1913 work, which provided the bridge from Planck's quantized interaction of matter and radiation (1900) to the full theory of quantum mechanics (1925-28). Unfortunately, the version of the Bohr atom that appears in many textbooks and is no doubt often presented to students is more wrong than right and may leave both teachers and students wondering why, more than a hundred years later, it is still being taught. This "pedagogic" version postulates that an electron in a stationary state moves in a circular orbit with an angular momentum that is an integral multiple of h/2p (L = nh/2p = nh)--h for the lowest-energy state, 2h for the next state, and so on. This picture of the hydrogen atom is wrong in two senses. First it doesn't conform to our present understanding of the hydrogen atom. This, in itself, is not a reason to scrap it, for the historical development of quantum physics is certainly of interest. But second, it doesn't conform to the essence of what Bohr actually did. That "is" a reason not to teach about circular orbits and L = nh. Much excellent scholarship has been devoted to Bohr's 1913 work. This article is not intended to extend or enrich that scholarship. Rather, I want to draw the attention of teachers of introductory physics to the high points of Bohr's classic first 1913 paper on the hydrogen atom and to clarify what in that paper is right and what is wrong (as well as what is weird). As I have pointed out in an earlier short communication, there is more right and less wrong than is widely recognized. As to what is weird, well, it is interesting.

BackgroundCongo Red Dot Paper Test (CRDPT) appears to be a simple, cost-effective, non-invasive diagnostic tool for hypertensive disorders of pregnancy (HDP). The main objective of the study is to assess the effectiveness of CRDPT in detecting HDP.MethodsThis is a systemic review and meta-analysis of published studies on the effectiveness of CRDPT in the detection of HDP. The study was conducted in line with the PRISMA-DTA guidelines. The PICOS framework was used to search for relevant articles using Medline, PubMed, Google Scholar, Web of Science, and the Cochrane Library databases. The articles were screened against a set of inclusion and exclusion criteria and analysed using the Review Manager 5.4 software.ResultsA title, abstract and full article screening was conducted on 18,153 potential articles based on the inclusion and exclusion criteria. The screening yielded five articles for meta-analysis. The total number of normotensive pregnant women (n = 3,380) in the included studies was five times higher than the total number of women with pre-eclampsia (n = 535). A difference between the HDP and normotensive group was noted. This is indicated by a significantly decreased in the effectiveness of CRDPT in detecting HDP as compared to normotensive group [Risk Ratio (RR) = 6.32 (2.17, 18.43) p

Many schools use computer-based testing to measure students' progress for end-of-the-year and statewide assessments. There is little research to support whether computer-based testing accurately reflects student progress, particularly among students with learning, performance, and generalization difficulties. This article summarizes an investigation of a reading strategy on comprehension of expository and narrative text; comprehension was assessed using both paper- and computer-based tasks. Results showed that students increased their learning strategy use and reading comprehension accuracy when using the strategy on computer-based assessments with expository text. Also students generalized their strategy use to paper-pencil assessments with narrative text. Reading comprehension of narrative passages also increased. The results suggest the value of teaching a reading strategy to students who struggle with both paper-pencil and computer-based assessments of comprehension.

The practice of Corrective Feedback (CF), which is situated within a Second Language Acquisition (SLA) Paradigm, is currently positioned towards an accuracy-orientated delivery based on native speaker norms. This is despite the recognition in different areas of linguistic research that there is considerable variation in the way that English is spoken around the world. This paper argues that the epistemological assumptions and methodological approaches to investigate CF within an SLA paradigm have various underlying weaknesses that undermine research findings. These findings purport to provide support for an accuracy-orientated CF in the English classroom. However, it is suggested in this paper that a meaning-orientated CF would be more reflective and beneficial for students given the transformative changes that have occurred to English over the past 30 years. This perspective is discussed in relation to one teacher's approach to CF who participated in a larger project which examined CF conducted in a British-Sino University.

The Artificial Intelligence Recommender System has emerged as a significant research interest. It aims at helping users find things online by offering recommendations that closely fit their interests. Recommenders for research papers have appeared over the last decade to make it easier to find publications associated with the field of researchers' interests. However, due to several issues, such as copyright constraints, these methodologies assume that the recommended articles' contents are entirely openly accessible, which is not necessarily the case. This work demonstrates an efficient model, known as RPRSCA: Research Paper Recommendation System Using Effective Collaborative Approach, to address these uncertain systems for the recommendation of quality research papers. We make use of contextual metadata that are publicly available to gather hidden relationships between research papers in order to personalize recommendations by exploiting the advantages of collaborative filtering. The proposed system, RPRSCA, is unique and gives personalized recommendations irrespective of the research subject. Thus, a novel collaborative approach is proposed that provides better performance. Using a publicly available dataset, we found that our proposed method outperformed previous uncertain methods in terms of overall performance and the capacity to return relevant, valuable, and quality publications at the top of the recommendation list. Furthermore, our proposed strategy includes personalized suggestions and customer expertise, in addition to addressing multi-disciplinary concerns. [ABSTRACT FROM AUTHOR]

An analytic matrix comprised of multiple media literacy teaching and learning principles is conceptualized to examine a model of news literacy developed by journalism educators at Stony Brook University. The multidimensional analysis indicates that news literacy instructors focus on teaching students how to question and assess the veracity of news texts, and their approach favors cognitive skill development over other ways people make meaning of media messages. Based on these findings, a cognitive theory of news literacy is proposed as a means to situate the journalistic methods and mindsets that informed the Stony Brook curriculum within the parameters of established media literacy education theories and practices.

Recent advancements in generative artificial intelligence (GAI) have sparked debates and research on its transformative potential in education. This study explored how a specific group of educators partner with GAI tools, particularly ChatGPT, to complement and enhance their teaching. Within an interpretative paradigm, the study used an exploratory case study design to investigate how 13 educators enrolled as students in a structured master's programme at an open distance learning university in South Africa use GAI tools in their teaching. Their posts on a discussion forum were used to collect data. Findings reveal that they actively employ GAI to streamline their teaching practices, specifically to personalized tutors, simplifying English texts and translations, assessments, lesson planning, and critical thinking tasks. The study also highlights the need to understand the limitations and boundaries of GAI, including concerns about accuracy, biases and reliability. Importantly, the absence of institutional guidelines on GAI raises questions about the necessity for clear guidelines and policies to ensure responsible and ethical integration in educational contexts. The paper concludes by emphasizing the transformative potential of GAI and the fact that it will increasingly act as a partner in teaching and influence modern teaching practices. Ultimately, GAI's foothold in education is undeniable, challenging educators and policymakers to navigate its implications while maximizing its potential to support effective teaching.

In calibration tasks students assess exemplar texts using criteria against which their own work will be assessed. Typically, these tasks are used in the context of training for peer assessment. Little research has been conducted on the benefits of calibration tasks, such as benchmarking, as learning opportunities in their own right. This paper examines a dataset from a long-running benchmarking task (~500 students per semester, for four semesters). We investigate the relationship of benchmarking performance to other student outcomes, including ability to self-assess accurately. We show that students who complete the benchmarking perform better, that there is a relationship between benchmarking performance and self-assessment performance, and that students appreciate the support for learning that benchmarking tasks provide. We discuss implications for teaching and learning flagging the potential of calibration tasks as an under-explored tool.

In the field of second/foreign language learning, Journal Writing (JW) has been used as a tool for self-expression, acknowledged for its multiple benefits with regard to various aspects of language learning. With the advancements of technology, however, few studies have addressed the potentials of internet-based platforms such as blogs in the practice of JW. The present study, therefore, has attempted to juxtapose the traditional paper-and-pencil mode of JW with Blog JW to explore the potential impacts on L2 learners’ writing skill. To this end, four elements of accuracy, fluency, lexical, as well as syntactic complexity were examined in the participants’ writing samples through relevant measures. To analyze the data, (M)ANCOVAs and (M)ANOVAs were conducted, the results of which are summarized and presented. The general conclusion of the study was that, JW, in particular, in the form of blogging has the potential of enhancing the quality of language learners’ writing.

Purpose: The purpose of this study was to investigate work-related adaptive performance from a longitudinal process perspective. This paper clustered specific behavioral patterns following the introduction of a change and related them to retentivity as an individual cognitive ability. In addition, this paper investigated whether the occurrence of adaptation errors varied depending on the type of change content. Design/methodology/approach: Data from 35 participants collected in the simulated manufacturing environment of a Research and Application Center Industry 4.0 (RACI) were analyzed. The participants were required to learn and train a manufacturing process in the RACI and through an online training program. At a second measurement point in the RACI, specific manufacturing steps were subject to change and participants had to adapt their task execution. Adaptive performance was evaluated by counting the adaptation errors. Findings: The participants showed one of the following behavioral patterns--(1) no adaptation errors, (2) few adaptation errors, (3) repeated adaptation errors regarding the same actions, or (4) many adaptation errors distributed over many different actions. The latter ones had a very low retentivity compared to the other groups. Most of the adaptation errors were made when new actions were added to the manufacturing process. Originality/value: Our study adds empirical research on adaptive performance and its underlying processes. It contributes to a detailed understanding of different behaviors in change situations and derives implications for organizational change management.

The emergence of ChatGPT as one of the most advanced chatbots and its ability to generate diverse data has given room for numerous discussions worldwide regarding its utility, particularly in advancing medical education and research. This study seeks to assess the performance of ChatGPT in medical biochemistry to evaluate its potential as an effective self-learning tool for medical students. This evaluation was carried out using the university examination question papers of both parts 1 and 2 of medical biochemistry which comprised theory and multiple choice questions (MCQs) accounting for a total of 100 in each part. The questions were used to interact with ChatGPT, and three raters independently reviewed and scored the answers to prevent bias in scoring. We conducted the inter-item correlation matrix and the interclass correlation between raters 1, 2, and 3. For MCQs, symmetric measures in the form of kappa value (a measure of agreement) were performed between raters 1, 2, and 3. ChatGPT generated relevant and appropriate answers to all questions along with explanations for MCQs. ChatGPT has "passed" the medical biochemistry university examination with an average score of 117 out of 200 (58%) in both papers. In Paper 1, ChatGPT has secured 60 ± 2.29 and 57 ± 4.36 in Paper 2. The kappa value for all the cross-analysis of Rater 1, Rater 2, and Rater 3 scores in MCQ was 1.000. The evaluation of ChatGPT as a self-learning tool in medical biochemistry has yielded important insights. While it is encouraging that ChatGPT has demonstrated proficiency in this area, the overall score of 58% indicates that there is work to be done. To unlock its full potential as a self-learning tool, ChatGPT must focus on generating not only accurate but also comprehensive and contextually relevant content.

As the cradle of cultivating talents, universities are facing great opportunities and challenges in their education. Among them, IPE (ideological and political education), as an important foundation for the future growth of university students, is of great significance. This paper discusses the relationship between IPE and psychological fitness education in university teaching. This paper expounds the necessity and feasibility of playing the role of psychological fitness education in IPECU (ideological and political education in colleges and universities). Based on this, this paper gives the strategy of infiltrating psychological fitness education into IPE. This paper combines NN (neural network) method to construct an assessment model of IPE quality. In this paper, MATLAB is used for simulation and comparative analysis. The final experiment shows that the RMSE of this algorithm is 0.512, MAE is 1.089, and the accuracy of the algorithm is 0.958.

This paper opens with a family anecdote in which my future mother in-law, when asked what wise advice she would offer undergraduate university students, replied, "I would tell them it's not going to be okay." Can we learn to keep engaging with the world despite its inevitable disappointments? I propose that Stoic philosophy, by "orienting" our "attention" and "courage," can help us navigate the troubled post-COVID world we share. To help make this more concrete, I describe a critical moment I observed in which a maskless shopper insulted fellow patrons in a grocery store for wearing a mask. I then develop the Stoic themes of acknowledgement (a commitment to the facts) and affinity (reaching out to others to build community). In the conclusion I return to the "maskless shopper" incident to consider how my two Stoic themes might help open a dialogue with this person. After discussing the limitations to such an undertaking, given the surge in populism over the last decade, I conclude with the appropriately tough-minded Stoic proposition that despite the obstacles, we must keep trying.

Predictive student models are increasingly used in learning environments due to their ability to enhance educational outcomes and support stakeholders in making informed decisions. However, predictive models can be biased and produce unfair outcomes, leading to potential discrimination against certain individuals and harmful long-term implications. This has prompted research on fairness metrics meant to capture and quantify such biases. Nonetheless, current metrics primarily focus on predictive performance comparisons between groups, without considering the behavior of the models or the severity of the biases in the outcomes. To address this gap, we proposed a novel metric in a previous work (Verger et al., 2023) named "Model Absolute Density Distance" (MADD), measuring algorithmic unfairness as the difference of the probability distributions of the model's outcomes. In this paper, we extended our previous work with two major additions. Firstly, we provided theoretical and practical considerations on a hyperparameter of MADD, named "bandwidth," useful for optimal measurement of fairness with this metric. Secondly, we demonstrated how MADD can be used not only to measure unfairness but also to mitigate it through postprocessing of the model's outcomes while preserving its accuracy. We experimented with our approach on the same task of predicting student success in online courses as our previous work, and obtained successful results. To facilitate replication and future usages of MADD in different contexts, we developed an open-source Python package called maddlib (https://pypi.org/project/maddlib/). Altogether, our work contributes to advancing the research on fair student models in education.

Although an accurate measure of vocabulary size is integral to understanding the proficiency of language learners, the validity of multiple-choice (M/C) vocabulary tests to determine this has been questioned due to users guessing correct answers which inflates scores. In this paper the nature of guessing and partial knowledge used when taking the Vocabulary Size Test (VST) is examined. The analysis evaluates the thought processes of test takers through think-aloud protocols and self-reports. This provides a taxonomy of seven types of guesses used by learners while taking the VST and measures the frequency of their occurrence. Based on qualitative coding, guesses are investigated to determine if they exhibit partial word knowledge which is relevant to the test construct. The findings suggest that both guesses resulting from the use of partial knowledge and random guesses are included in estimates and this has a detrimental effect on test accuracy. The paper concludes that the VST does not provide an accurate measure of the vocabulary necessary for reading due to guessing and the meaning-recognition format. It also suggests that the role of partial knowledge should be considered when producing more sophisticated vocabulary tests in the future.

This paper discusses the factors affecting the behaviours for coping with fake news among young people. The data were collected from a survey conducted in late 2019, which sampled 2112 secondary school students from 21 partnering schools. This study aims to understand the opinions and behaviours of teenagers towards disinformation when fake news was prevalent during the anti-extradition bill protests in Hong Kong. It finds that awareness of the problem alone had limited influence in facilitating coping strategies. Civic awareness and interaction with social media were useful predictors of internal and external coping behaviours, respectively. Confidence about one's ability to detect fake news was a crucial factor, yet a concern for the value of truth stood out as the strongest predictor of fake news coping behaviours.

Though a substantial amount of research exists on imputing missing scores in educational assessments, there is little research on cases where responses or scores to an item are missing for all test takers. In this paper, we tackled the problem of imputing missing scores for tests for which the responses to an item are missing for all test takers. We considered three missing-data imputation methods--the median method, the item response theory (IRT) method, and the two-way method--for imputing scores. We compared the performance of these three imputation methods with respect to their accuracy in estimating scaled scores and test reliability for the aforementioned problem. Real data were used in the comparison. All three methods performed well in imputing scaled scores with negligible imputation error: The IRT method and the median method provided slightly more accurate scaled scores. The two-way method provided the most accurate reliability estimates. Recommendations for practice are provided.

Background The Medical Certificate of Stillbirth (MCS) records data about a baby's death after 24 weeks of gestation but before birth. Major errors that could alter interpretation of the MCS were widespread in two UK-based regional studies. Methods A multicentre evaluation was conducted, examining MCS issued 1 January 2018 to 31 December 2018 in 76 UK obstetric units. A systematic case-note review of stillbirths was conducted by Obstetric and Gynaecology trainees, generating individual 'ideal MCSs' and comparing these to the actual MCS issued. Anonymized central data analysis described rates and types of error, agreement and factors associated with major errors. Results There were 1120 MCSs suitable for assessment, with 126 additional submitted data sets unsuitable for accuracy analysis (total 1246 cases). Gestational age demonstrated 'substantial' agreement [K = 0.73 (95% CI 0.70–0.76)]. Primary cause of death (COD) showed 'fair' agreement [K = 0.26 (95% CI 0.24–0.29)]. Major errors [696/1120; 62.1% (95% CI 59.3–64.9%)] included certificates issued for fetal demise at <24 weeks' gestation [23/696; 3.3% (95% CI 2.2–4.9%)] or neonatal death [2/696; 0.3% (95% CI 0.1–1.1%)] or incorrect primary COD [667/696; 95.8% (95% CI 94.1–97.1%)]. Of 540/1246 [43.3% (95% CI 40.6–46.1%)] 'unexplained' stillbirths, only 119/540 [22.0% (95% CI 18.8–25.7%)] remained unexplained; the majority were redesignated as either fetal growth restriction [FGR: 195/540; 36.1% (95% CI 32.2–40.3%)] or placental insufficiency [184/540; 34.1% (95% CI 30.2–38.2)]. Overall, FGR [306/1246; 24.6% (95% CI 22.3–27.0%)] was the leading primary COD after review, yet only 53/306 [17.3% (95% CI 13.5–22.1%)] FGR cases were originally attributed correctly. Conclusion This study demonstrates widespread major errors in MCS completion across the UK. MCS should only be completed following structured case-note review, with particular attention on the fetal growth trajectory. [ABSTRACT FROM AUTHOR]

Abstract Citation prediction of scholarly papers is of great significance in guiding funding allocations, recruitment decisions, and rewards. However, little is known about how citation patterns evolve over time. By exploring the inherent involution property in scholarly paper citation, we introduce the Paper Potential Index (PPI) model based on four factors: inherent quality of scholarly paper, scholarly paper impact decaying over time, early citations, and early citers' impact. In addition, by analyzing factors that drive citation growth, we propose a multi-feature model for impact prediction. Experimental results demonstrate that the two models improve the accuracy in predicting scholarly paper citations. Compared to the multi-feature model, the PPI model yields superior predictive performance in terms of range-normalized RMSE. The PPI model better interprets the changes in citation, without the need to adjust parameters. Compared to the PPI model, the multi-feature model performs better prediction in terms of Mean Absolute Percentage Error and Accuracy; however, their predictive performance is more dependent on the parameter adjustment. [ABSTRACT FROM AUTHOR]

This research was triggered by the identified need in literature for large-scale studies about the kinds of designs that teachers create for mobile learning (m-learning). These studies require analyses of large datasets of learning designs. The common approach followed by researchers when analyzing designs has been to manually classify them following high-level pedagogically guided coding strategies, which demands extensive work. Therefore, the first goal of this paper is to explore the use of supervised machine learning (SML) to automatically classify the textual content of m-learning designs using pedagogically relevant classifications, such as the cognitive level demanded by students to carry out specific designed tasks, the phases of inquiry learning represented in the designs, or the role that the situated environment has in the designs. Because not all SML models are transparent, but researchers often need to understand their behaviour, the second goal of this paper is to consider the trade-off between models' performance and interpretability in the context of design analytics for m-learning. To achieve these goals, we compiled a dataset of designs deployed using two tools, Avastusrada and Smartzoos. With this dataset, we trained and compared different models and feature extraction techniques. We further optimized and compared the best performing and most interpretable algorithms (EstBERT and Logistic Regression) to consider the second goal with an illustrative case. We found that SML can reliably classify designs with accuracy > 0.86 and Cohen's kappa > 0.69.

This paper examines the differences correlation between instructor- and peer-based participation scores and oral proficiency in an EFL course focusing on oral communication. It finds that there is very little correlation between instructor-based and peer-based participation scores (r = 0.187, p = 0.053) and that the former is more associated with improvement in fluency (r = 0.266, p = 0.005), but that the latter is strongly associated with pre-class levels of fluency measures such as trimmed speech rate (r = 0.364, p < 0.001). However, neither measure of participation was particularly correlated with post-class levels of fluency. Therefore, some caution is required by instructors when using in-class participation scores. While it is important to encourage students to improve by participating in classes, the methods of assigning participation scores examined in this paper may not be appropriate as a part of the composite final grade in the class as they do not correlate well to final skill level.

Automatic grading models are valued for the time and effort saved during the instruction of large student bodies. Especially with the increasing digitization of education and interest in large-scale standardized testing, the popularity of automatic grading has risen to the point where commercial solutions are widely available and used. However, for short answer formats, automatic grading is challenging due to natural language ambiguity and versatility. While automatic short answer grading models are beginning to compare to human performance on some datasets, their robustness, especially to adversarially manipulated data, is questionable. Exploitable vulnerabilities in grading models can have far-reaching consequences ranging from cheating students receiving undeserved credit to undermining automatic grading altogether--even when most predictions are valid. In this paper, we devise a black-box adversarial attack tailored to the educational short answer grading scenario to investigate the grading models' robustness. In our attack, we insert adjectives and adverbs into natural places of incorrect student answers, fooling the model into predicting them as correct. We observed a loss of prediction accuracy between 10 and 22 percentage points using the state-of-the-art models BERT and T5. While our attack made answers appear less natural to humans in our experiments, it did not significantly increase the graders' suspicions of cheating. Based on our experiments, we provide recommendations for utilizing automatic grading systems more safely in practice.

A 7-month mathematics proficiency program was conducted in a primary Australian Indigenous community school. This paper focuses on outlining the specific methodologies employed to explore how students' mathematical proficiency changed throughout the implementation of the program in Years 2 to 4 (~ 7 to 9 years old). A mixed methods research design was utilised, and findings will be presented to evidence how the combination of standardised tests, diagnostic tests, and Newman interviews were useful in capturing and making visible young Indigenous student's mathematical proficiency. Whilst standardised tests provided a useful and comparable measure of student achievement, diagnostic tests and Newman interviews gave space for Indigenous student voice and demonstrated their strengths and areas for improvement in relation to their conceptual understanding, procedural fluency, and strategic competence. From these findings, recommendations concerning the adjustment of data collection procedures for young students in this setting are presented. The findings question the accuracy of standardised tests in revealing young students' proficiency, and this has implications for the extent to which standardised tests are relied upon to inform educational reform particularly for Indigenous students. Striving for equitable educational outcomes is an important endeavour in Australia, and such undertakings must be driven by meaningful and accurate evidence of students' proficiency in mathematics.

Categorical growth models describe examinee growth in terms of performance-level category transitions, which implies that some percentage of examinees will be misclassified. This paper introduces a new procedure for estimating the classification accuracy of categorical growth models, based on Rudner's classification accuracy index for item response theory-based assessments. Results of a simulation study are presented to provide evidence for the accuracy and validity of the approach. Also, an empirical example is presented to demonstrate the approach using data from the Indiana Student Performance Readiness and Observation of Understanding Tool growth model, which classifies examinees into growth categories used by the Office of Special Education Programs to monitor the progress of preschool children who receive special education services.

The recent development of AI Chatbot -- specifically ChatGPT -- has gained dramatic attention from users as evident by ongoing discussion among the education fraternity. We argue that prior to making any conclusion, it is important to understand how ChatGPT is being used in higher education across the globe. This paper makes a significant contribution by systematically reviewing the global literature on the use of ChatGPT in higher education using PRISMA guidelines. We included 69 studies in the analysis based on inclusion and exclusion criteria. We presented the scope of published literature in three aspects: (i) contextual, (ii) methodological, and (iii) disciplinary. Most of the studies have been carried out in HICs (n = 53; 77%) representing the field of higher education (n = 37; 54%) without specifying the discipline, while only a few studies were based on empirical data (n = 19; 27%). The findings based on included studies reveal that ChatGPT serves as a convenient tool to assist teachers, students, and researchers in various tasks. While the specific uses vary, the underlying motivation remains consistent: seeking personal benefits and reducing academic burdens. Teachers use it for personal and professional learning and resource generation while students use it as personal tutors for various learning purposes. However, concerns related to accuracy, reliability, academic integrity, and potential negative effects on cognitive and social development were consistently highlighted in many studies. To address these concerns, we have proposed a comprehensive framework for universities along with directions for future research in higher education as an optimal response.

Researchers in second language (L2) and education domain use different statistical methods to assess their constructs of interest. Many L2 constructs emerge from elements/parts, i.e., the elements "define" and "form" the construct and not the other way around. These constructs are referred to as emergent variables (also called components, formative constructs, and composite constructs). Because emergent variables are composed of elements/parts, they should be assessed through confirmatory composite analysis (CCA). Elements of emergent variables represent unique facets of the construct. Thus, such constructs cannot be properly assessed by confirmatory factor analysis (CFA) because CFA and its underlying common factor model regard these elements to be similar and interchangeable. Conversely, the elements of an emergent variable uniquely define and form the construct, i.e., they are not similar or interchangeable. Thus, CCA is the preferred approach to empirically validate emergent variables such as language skills L2 students' behavioral engagement and language learning strategies. CCA is based on the composite model, which captures the characteristics of emergent variables more accurately. Aside from the difference in the underlying model, CCA consists of the same steps as CFA, i.e., model specification, model identification, model estimation, and model assessment. In this paper, we explain these steps. and present an illustrative example using publicly available data. In doing so, we show how CCA can be conducted using graphical software packages such as Amos, and we provide the code necessary to conduct CCA in the R package lavaan.

Recent investigations on how people derive meaning from language have focused on task-dependent shifts between two cognitive systems. The symbolic (amodal) system represents meaning as the statistical relationships between words. The embodied (modal) system represents meaning through neurocognitive simulation of perceptual or sensorimotor systems associated with a word's referent. A primary finding of literature in this field is that the embodied system is only dominant when a task necessitates it, but in certain paradigms, this has only been demonstrated using nouns and adjectives. The purpose of this paper is to study whether similar effects hold with verbs. Experiment 1 evaluated a novel task in which participants rated a selection of verbs on their implied vertical movement. Ratings correlated well with distributional semantic models, establishing convergent validity, though some variance was unexplained by language statistics alone. Experiment 2 replicated previous noun-based location-cue congruency experimental paradigms with verbs and showed that the ratings obtained in Experiment 1 predicted reaction times more strongly than language statistics. Experiment 3 modified the location-cue paradigm by adding movement to create an animated, temporally decoupled, movement-verb judgment task designed to examine the relative influence of symbolic and embodied processing for verbs. Results were generally consistent with linguistic shortcut hypotheses of symbolic-embodied integrated language processing; location-cue congruence elicited processing facilitation in some conditions, and perceptual information accounted for reaction times and accuracy better than language statistics alone. These studies demonstrate novel ways in which embodied and linguistic information can be examined while using verbs as stimuli.

There is no need or point to testing of knowledge, attributes, traits, behaviours or abilities of an individual if information obtained from the test is inaccurate. However, by and large, it seems the estimation of psychometric properties of test items in classroomshas been completely ignored otherwise dying slowly in most testing environments. In the quest to obtain sound and efficient test results, it is imperative that assessorsrely on some psychometric properties to make informed classrooms decisions. These psychometric properties can be estimated using KuderRichardson20 Formula. In this study, 30 multiple-choice items were administered and used for the study. The strength of each item was analysed by looking at their difficulty level and how theydiscriminated among the students. Reliability tests were also conducted in addition to the item analysis to observe the quality of the test as a whole. With lucid prose, KR-20 was used to estimate the psychometric properties of 30 set integrated science test items (which werescored dichotomously)to serve as a primer for assessorsin higher institutions. The procedure produced coefficient value of 0.6915which is approximately 0.7 implying that the reliability of the test was high.The procedure we used to arrive at the obtained coefficient is extensively outlined in the paper. We concluded thatthe suggested procedure (KR-20) for estimating psychometric properties may have a paradigm shift in classroom testing situations where it will communicate to teachers on the efficiency and process of teachermade tests. In essence, this could enhance the quest of obtaining the real knowledge, attributes, traits, behaviours or abilities of students by using test items that are reliable and dependable.

Presentation skills are considered as the most vital proficiencies required in higher education. This research paper examined the difficulties of oral presentation EFL learners encountered, strategies used to make oral presentation well and the perceived role of lecturer in improving EFL learners' oral presentation skills. With regard to achieving these objectives, questionnaires distributed to 150 EFL learners, and face-to-face interviews were conducted. The quantitative data were analyzed through SPSS by conducting frequencies and percentages, independent sample t-test and ANOVA tests. The results of the study revealed that learners had a problem with oral fluency, accuracy and pronunciation during oral presentation, feel frightened when a lot of people are watching them, afraid of being assessed by their classmates in front of the class, having low self-confidence. Furthermore, considering a crucial role of lecturer in improving oral presentation skills, this study also discovered that lecturer has to show learners videos of good speakers for enhancing their presentation skills and lecturer has to give learners freedom to choose their own topics that lower their anxiety. It is hoped that the ideas discussed will provide the learners and lecturers with basic guidelines of how to improve the learners' oral presentation and assist learners to work their way out of anxiety.

Language proficiency is multi-componential in nature and, according to many SLA researchers and L2 practitioners, is best captured by the concepts of complexity, accuracy, and fluency (CAF). Two of the three studies reported in this paper adopted the Cognition Hypothesis (Robinson, 2003) as the means to address the question of whether increasing task complexity positively impacts learning and performance. The Cognition Hypothesis claims that increasing cognitive task complexity along specific dimensions has the potential to promote "greater analysis, modifications, and restructuring of the IL (interlanguage) with consequent performance effects" (Robinson, 2001a, p. 302). The studies that invoke the Cognition Hypothesis either refer to Robinson's Triadic Framework (Robinson, 2001a) and the Cognition Hypothesis to structure their research and explain the findings, or they take a more radical stand to argue that the premises underlying the Cognition Hypothesis and the claims derived from it are problematic. This forum begins with a literature review on CAF. It then reports on metric indices employed in the studies to measure the phenomenon of linguistic complexity and the results pertaining to grammatical and lexical complexity. Next, it describes the status of theory in empirical research and the challenges research on CAF faces. The closing presents a discussion on how theory impacted the design of the empirical studies sampled for this paper.

Purpose: The purpose of this study was to examine if reading purpose (study or entertainment) varied the effect of reading medium on comprehension and accuracy of perceptions of comprehension. A secondary purpose was to examine how mind wandering relates to reading medium and reading purpose. An unanticipated purpose was examining the role of emergency remote COVID-19 instruction on reading comprehension, mind wandering, as well as both preference for and use of screens and paper. Methods: In this study, undergraduate students (N = 133) were randomly assigned to reading purposes of study or entertainment as well as reading from paper (from a book) or screens (from an iPad). Results: Neither reading medium nor purpose had reliable differences in performance on the text comprehension assessment, metacomprehension, or mind wandering. Reading from one's preferred medium appeared to be related to more accurate metacomprehension. Exploratory analyses indicated less studying from paper and more task-unrelated thoughts while reading from screens after the COVID-19 learning changes. Conclusions: Reading medium may possibly have less effect on comprehension when readers have purposes for reading and the haptic experience of reading is similar by medium. Medium effects on metacomprehension and mind wandering may vary depending on characteristics of the reader.

This paper reports an experiment that investigates interpretive distinctions between two different expressions of generalization in Spanish. In particular, our aim was to find out when the distinction between generic statements (GS) such as "Tigers have stripes" and universally quantified statements (UQS) such as "All tigers have stripes" was acquired in Spanish-speaking children of two different age groups (4/5-year-olds and 8/9-year-olds), and then compare these results with those of adults. The starting point of this research was the semantic distinction between GS and UQS in that the former admit exceptions, unlike the latter. On the other hand, several authors have observed a Generic overgeneralization effect (GOG) consisting in allowing for UQS to be felicitous in the face of exceptions, thus proposing that this "error" stems from GS being defaults (simpler, more easily learned and processed). In the current paper we aimed to test the "Generics as Default" (GaD) hypothesis by comparing GS and UQS in three different age ranges. Our data show that, overall, the accuracy of GS is greater than the accuracy of UQS. Moreover, we also confirm a hypothesized interaction between age and NP type (GS vs UQS). Further, we present several data points that are not predicted by the GaD, including an observed decline in the accuracy of GS in the older group of children as well as in adults, and that children fail at rejecting statements that are not considered to be true generalizations.

Robust indicators are important for identifying disadvantaged pupils in education, and for ensuring that they are rightly receiving relevant state-funded assistance. This paper compares the quality and completeness of data from England on student eligibility for free school meals (FSM) based on an administrative census, with more all-encompassing household income measures, from a smaller sample of young people. The first measure comes from the National Pupil Database (NPD), and the second from Next Steps (NS). The two datasets are linked at the individual student level. In this restricted group, FSM data is more complete (97%) than household income (47%). The bias created by missing data on income in NS calls into question its more general usefulness for analysts. FSM cannot be read neatly from income, such as referring to an income below a certain level, and vice versa. Many reportedly low-income children are not listed as FSM-eligible. However, the two values are linked, while each also provides unique information. Both measures predict attainment at school, to some extent. The paper concludes that FSM is the more practical measure at present, but also considers how access to limited income data could be made more widespread while maintaining individual data rights.

Misinformation is a global pandemic, fueled by the sophistication of the human intellect, algorithmic systems among other factors. Enhanced by the proliferation of algorithms optimized for engagement and reactions on social media, misinformation has ignited or hampered sociopolitical participation and movements and dissuaded citizens from being vaccinated, for example. Observations have shown that efforts to contain misinformation have largely been tech-based, with ubiquitous impressions that it can be coded into extinction and/or fact-checked with automated tools. This paper, therefore, contributes to the debate that there are mechanisms that should be explored beyond algorithms and fact-checking in the fight against misinformation. The paper adopted an integrative literature review approach, using purposive selection of 22 full texts from Google Scholar, JStor and other sources as captured in Table 1. The PRISMA flow diagram was used to show the search process. Findings from the literature reviewed showed that algorithms and fact-checking have made significant impacts in identifying, verifying and correcting misinformation. Nonetheless, they have drawbacks that should be complemented with information literacy programs/services and information ethics. The study suggests that information literacy and information ethics be made integral parts of educational modules and awareness should be increased about nonalgorithmic approaches to solving misinformation problems in order to proactively build a more informed public.

In order to study the big data of college students' employment, this paper takes the big data of college students' employment as the premise, analyzes the current employment data by establishing a DBN model, and puts forward relevant management measures, aiming to provide scientific basis for the management of graduates' employment data. The results are as follows: By comparing the application evaluation of linear regression method, BP neural network and DBN model, this paper finds that DBN model has better accuracy and lower error and has better advantages in the application of college students' employment data management characteristics. In addition, the development of social economy and the number of college graduates are the key factors for the employment rate of college students. Therefore, this paper suggests that through the use of big data technology, college will build a data platform for college students' employment management and provide a carrier for college students to obtain professional information and employment management information.

This paper aims to analysis an English language textbook used in "middle school" (high school) in Saudi Arabia in-depth, to ascertain the type of pedagogy pushed by the teacher's book. The criteria of analysis are adapted from Communicative Language Teaching (CLT). These criteria are used to analyses the textbook in terms of how communicative it is and personalised, how authentic, how meaningful; its degree of linguistic complexity; whether it is scaffolded, accuracy or fluency oriented; and how much of all each of these categories is in the textbook activities. The methodology of this paper is qualitative and presents only one aspect, textbook analysis, of a broader case study which forms part of a Ph.D. project that explores the factors behind the failure of implementation of CLT in Saudi Arabia. The results of this analysis show that in the textbook there is a careful balance between accuracy and fluency. There are some person, alised and meaningful activities, however most of the activities lack of authenticity which limits freedom for the students in using use the target language.