Learning Performance Research Articles

The effects of embedded quizzes on self‐regulated processes and learning performance during a multimedia lesson

AbstractBackgroundPrevious research has demonstrated that quizzing can improve self‐regulation processes and learning performances. However, it remains unclear whether quizzes in multimedia material bring similar benefits, and whether interindividual differences such as working memory capacity (WMC) modulate quizzing effects.AimsThis study aimed to examine the effects of embedded quizzes (multiple quizzes within the learning sequence) with feedback during a multimedia learning sequence on self‐regulated processing (monitoring and control) and learning performances on repeated (material already tested) and new (untested) questions, and possible WMC moderation on performances.Sample and MethodsParticipants were 59 students who learned a multimedia neuroscience course module in either a no‐quiz or embedded quizzes condition. With regard to self‐regulated processing, control was assessed with eye‐tracking measures (fixation durations and transitions between texts and illustrations), and monitoring by comparing self‐reported judgements of learning with actual performances (calibration). At the end, students' WMC was measured, and learning performances were assessed with repeated and new questions.ResultsQuizzing improved learning performances on repeated and new questions, and enhanced monitoring (better calibration). Students with low WMC benefited more from quizzing for new questions than those with high WMC. However, contrary to our hypothesis, the embedded quizzes group allocated less time to the learning than the no‐quiz group.ConclusionsEmbedded quizzes are relevant tools to help students recalibrate their monitoring process and foster more effective self‐regulated learning. Results confirm that the use of quizzes can be recommended to improve the learning of multimedia material, especially among students with low WMC.

Antimicrobial stewardship to reduce overtreatment of asymptomatic bacteriuria in critical access hospitals: measuring a quality improvement intervention.

Asymptomatic bacteriuria (ASB) treatment is a common form of antibiotic overuse and diagnostic error. Antibiotic stewardship using the inappropriate diagnosis of urinary tract infection (ID-UTI) measure has reduced ASB treatment in diverse hospitals. However, critical access hospitals (CAHs) have differing resources that could impede stewardship. We aimed to determine if stewardship including the ID-UTI measure could reduce ASB treatment in CAHs. From October 2022 to July 2023, ten CAHs participated in an Intensive Quality Improvement Cohort (IQIC) program including 3 interventions to reduce ASB treatment: 1) learning labs (ie, didactics with shared learning), 2) mentoring, and 3) data-driven performance reports including hospital peer comparison based on the ID-UTI measure. To assess effectiveness of the IQIC program, change in the ID-UTI measure (ie, percentage of patients treated for a UTI who had ASB) was compared to two non-equivalent control outcomes (antibiotic duration and unjustified fluoroquinolone use). Ten CAHs abstracted a total of 608 positive urine culture cases. Over the cohort period, the percentage of patients treated for a UTI who had ASB declined (aOR per month = 0.935, 95% CI: 0.873, 1.001, P = 0.055) from 28.4% (range across hospitals, 0%-63%) in the first to 18.6% (range, 0%-33%) in the final month. In contrast, antibiotic duration and unjustified fluoroquinolone use were unchanged (P = 0.768 and 0.567, respectively). The IQIC intervention, including learning labs, mentoring, and performance reports using the ID-UTI measure, was associated with a non-significant decrease in treatment of ASB, while control outcomes (duration and unjustified fluoroquinolone use) did not change.

The Effects of Social Cues in Self-Produced Micro Video Lectures on Parasocial Interaction, Motivational Interest, Extraneous Cognitive Load, and Concept Learning Performance: an Empirical Study in a Junior High School

Background and Aim: Videos are popular in the dissemination of knowledge at scale, which can be seen in various video-based platforms and educational disruption. However, inconsistent results were reported in studies. This research is designed to determine the effectiveness of social cues in self-produced micro video lectures on parasocial interaction, motivational interest, extraneous cognitive load, and concept learning performance and the relationship among them. Materials and Methods: Self-produced videos for learning concepts of morality and the rule of law in junior high schools were used as the treatment. Tests were used to determine concept learning performance and Likert scales were utilized to collect information on the other three variables. 248 students from the eighth grade in a public school were involved in the sample. One-way ANOVA was used to analyze differences among groups, and path analysis as well as was adopted to determine the relationship between variables. Results: There was no significant difference among the four groups on each variable. Parasocial interaction has a significant positive impact on motivational interest, motivational interest has a negative significant impact on extraneous cognitive load, and extraneous cognitive load has a negative significant impact on concept learning performance. Conclusion: The effects of the increment of types of social cues can be ignored in 6-minute self-produced video lectures. Students’ parasocial interaction can influence concept learning performance via the mediation of the cognition process and motivational factors. The extraneous cognitive load decreases with the stimulation of the motivational interest.

Impact of Harmonization on MRI Radiomics Feature Variability Across Preprocessing Methods for Parkinson's Disease Motor Subtype Classification.

This study aimed to assess the reproducibility of MRI-derived radiomic features across multiple preprocessing methods for classifying Parkinson's disease (PD) motor subtypes and to evaluate the impact of ComBat harmonization on feature stability and machine learning performance. T1-weighted MRI scans from 140 PD patients (70 tremor-dominant and 70 postural instability gait difficulty) and 70 healthy controls were obtained from the Parkinson's Progression Markers Initiative (PPMI) database, acquired using different scanner models. Radiomic features were extracted from 16 brain regions using various preprocessing pipelines. ComBat harmonization was applied using a combined batch variable incorporating both scanner models and preprocessing methods. Intraclass correlation coefficients (ICC) and Kruskal-Wallis tests assessed feature reproducibility before and after harmonization. Feature selection was performed using Linear Support Vector Classifier with L1 regularization. Support vector machine classifiers were used for PD subtype classification. ComBat harmonization significantly improved feature reproducibility across all feature groups. The percentage of features showing excellent robustness (ICC ≥ 0.90) increased from 40.2 to 56.3% after harmonization. First-order statistic features showed the highest robustness, with 71.11% demonstrating excellent ICC after harmonization. The proportion of features significantly affected by preprocessing methods was reduced following harmonization. Classification accuracy improved dramatically, from a range of 34-75% before harmonization to 89-96% after harmonization across all preprocessing methods. AUC values similarly increased from 0.28-0.87 to 0.95-0.99 after harmonization. ComBat harmonization significantly enhanced the reproducibility of radiomic features across preprocessing methods and improved PD motor subtype classification performance. This study highlights the importance of harmonization in radiomics research for PD and suggests potential clinical applications in personalized treatment planning.

Proto-DS: A Self-Supervised Learning-Based Nondestructive Testing Approach for Food Adulteration with Imbalanced Hyperspectral Data

Conventional food fraud detection using hyperspectral imaging (HSI) relies on the discriminative power of machine learning. However, these approaches often assume a balanced class distribution in an ideal laboratory environment, which is impractical in real-world scenarios with diverse label distributions. This results in suboptimal performance when less frequent classes are overshadowed by the majority class during training. Thus, the critical research challenge emerges of how to develop an effective classifier on a small-scale imbalanced dataset without significant bias from the dominant class. In this paper, we propose a novel nondestructive detection approach, which we call the Dice Loss Improved Self-Supervised Learning-Based Prototypical Network (Proto-DS), designed to address this imbalanced learning challenge. The proposed amalgamation mitigates the label bias on the most frequent class, further improving robustness. We validate our proposed method on three collected hyperspectral food image datasets with varying degrees of data imbalance: Citri Reticulatae Pericarpium (Chenpi), Chinese herbs, and coffee beans. Comparisons with state-of-the-art imbalanced learning techniques, including the Synthetic Minority Oversampling Technique (SMOTE) and class-importance reweighting, reveal our method’s superiority. Notably, our experiments demonstrate that Proto-DS consistently outperforms conventional approaches, achieving the best average balanced accuracy of 88.18% across various training sample sizes, whereas the Logistic Model Tree (LMT), Multi-Layer Perceptron (MLP), and Convolutional Neural Network (CNN) approaches attain only 59.42%, 60.38%, and 66.34%, respectively. Overall, self-supervised learning is key to improving imbalanced learning performance and outperforms related approaches, while both prototypical networks and the Dice loss can further enhance classification performance. Intriguingly, self-supervised learning can provide complementary information to existing imbalanced learning approaches. Combining these approaches may serve as a potential solution for building effective models with limited training data.

Factors Impacting Students’ Learning Performance in Higher Education in Chengdu, China

Background and Aim: The effectiveness of education has always been a focus of social attention. Previous research has mostly focused on the level of education in primary and secondary schools, and some studies have shown that college students may have different motivations and learning concepts. Therefore, this study aims to take Xihua University in Chengdu, China as an example to systematically examine various factors that affect the academic performance of university students. Through quantitative analysis, this study investigated the impact of five independent variables (learning attitude, learning anxiety, learning motivation, teacher ability, and student participation) on the dependent variable (academic performance), especially in higher education. Moreover, it aims to identify significant differences between variables. Methodology: Research Design, Data, and Methods: This study used the Project Objective Consistency Index (IOC) to test the validity of the questionnaire, and Cronbach's Alpha was used as reliability in the pilot study (n=30). They used the multiple linear regression method to analyze the effective responses of 80 students at Xihua University and verify the significant relationship between variables. Afterward, a group of 30 students underwent a 14-week planned study. Then, the quantitative results after and before the scheduled study were compared in paired sample t-tests. Results: In multiple linear regression, studies have shown that learning attitude, learning anxiety, learning motivation, teacher ability, and student participation have a significant impact on student academic performance. Finally, the comparison results of paired sample t-tests indicate substantial differences in student academic performance between the post-strategic and pre-strategic plan stages. Conclusion: This research endeavors to foster students' learning performance in the context of Chengdu, China.

Modality Matters: Evidence for the Benefits of Speech-Based Adaptive Retrieval Practice in Learners with Dyslexia.

Retrieval practice-the process of actively calling information to mind rather than passively studying materials-has been proven to be a highly effective learning strategy. However, only recently, researchers have started to examine differences between learners in terms of the optimal conditions of retrieval practice in applied educational settings. In this study (N=118), we focus on learners with dyslexia. We compare their performance to the performance of typical learners in an adaptive retrieval practice task using both typing-based and speech-based response conditions. We find that typical learners outperform learners with dyslexia when they are asked to respond by typing, but that this difference disappears when learners respond by speech. Using a mathematical model to decompose response times, we demonstrate that this typing-specific disadvantage in learners with dyslexia is mainly a consequence of processing delays, rather than poorer memory performance. These findings contribute to a better understanding of the mechanisms underlying declarative learning in dyslexia, and they can be used to tailor educational technology toward the needs of neurodiverse learners.

Machine learning and deep learning models for preoperative detection of lymph node metastasis in colorectal cancer: a systematic review and meta-analysis.

To evaluate the diagnostic performance of Machine Learning (ML) and Deep Learning (DL) models for predicting preoperative Lymph Node Metastasis (LNM) in Colorectal Cancer (CRC) patients. A systematic review and meta-analysis were conducted following PRISMA-DTA and AMSTAR-2 guidelines. We searched PubMed, Web of Science, Embase, and Cochrane Library databases until February 16, 2024. Study quality and risk of bias were assessed using the QUADAS-2 tool. Data were analyzed using STATA v18, applying random-effects models to all analyses. Twelve studies involving 8321 patients were included, with most published in 2021-2024 (9/12). The pooled AUC of ML models for predicting LNM in CRC patients was 0.87 (95% CI: 0.82-0.91, I2:86.17) with a sensitivity of 78% (95% CI: 69-87%) and a specificity of 77% (95% CI: 64%-90%). In addition, when assessing the AUC reported by radiologists, both junior and senior radiologists had similar performance, significantly lower than the ML models. (P < 0.001). Subgroup analysis revealed higher AUCs in prospective studies (0.95, 95% CI: 0.87-1) compared to retrospective studies (0.85, 95% CI: 0.81-0.89) (P = 0.03). Studies without external validation exhibited significantly higher AUCs than those with external validation (P < 0.01). While there was no significant difference in AUC and sensitivity between the T1-T2 and T2-T4 stages, specificity was significantly higher in the T2-T4 stages than the low stages of T1 and T2 (95%, 95% CI: 92-98% vs. 61%, 95% CI: 44-78%; P < 0.01). ML models demonstrate strong potential for preoperative LNM staging and treatment planning in CRC, potentially reducing the need for additional surgeries and related health and financial burdens. Further prospective multicenter studies, with standardized reporting of algorithms, modality parameters, and LNM staging, are needed to validate these findings.

Is social interaction a catalyst for digital environments? Exploring affordances of synchronous delivery in online CLIL

ABSTRACT Content and Language Integrated Learning (CLIL) has emerged as a pedagogical approach integrating foreign language learning with subject content. Nevertheless, its foray into the online environment, where courses are commonly delivered in two modes: asynchronous and synchronous, remains unexplored. By integrating synchronous delivery into an asynchronous-only online CLIL course, this study aims to probe how synchronous delivery affects students’ learning performances (learning behaviors and outcomes), and investigate learners’ perspectives on synchronous delivery as affordances for enhancing these performances. The study involved 969 learners and two teachers, with data collected from six sources for triangulation: tracking data from social media chat logs, website learning records, course scores, teachers’ reflective narratives, observation of synchronous learning, and student interviews. The findings unveil the effectiveness of synchronous delivery in promoting learning performances, evident in students’ proactive time management and interactive behaviors, as well as elevated learning outcomes. Three affordances – pedagogical, technological, and social – are offered by synchronous delivery for promoting learning performances. Particularly noteworthy are the social affordances, which significantly enhance learning behaviors and outcomes by fostering efficient interactions, cordial interpersonal relationships, and a learning community.

Leadership Role of the School Principal in Developing and Improving Learner Performance in Eswatini Schools

This paper reports the findings of a study on the leadership role of school principals in developing and improving learner performance in schools. In this study, principals are the most influential educational leaders whose leadership role is inseparably linked to student performance. This study examines the basics of good leadership in line with student performance and offers practical suggestions that school leaders can employ. A qualitative case study embedded within the interpretative paradigm was conducted with five purposively sampled high schools and conveniently sampled principals in the Manzini region of Eswatini. Data were generated through semi-structured interviews. The findings showed that the principal is key to ensuring learner performance; thus, he is referred to as an instructional leader. The strategies principals employ to develop and improve learner performance are establishing a clear context relevant to school goals and focusing on those goals. Additionally, setting high expectations for student achievement and providing adequate resources and appropriate staff development programmes for teachers positively contribute to improving learner performance. Based on these findings, the study recommends that school principals have a knowledge-sharing platform on leadership practices and strategies for developing and improving learner performance.

Demographic Factors Associated with Students' Belief and Attitude towards Physics and Learning: A Cross-Sectional Study from Indonesia

Students' Belief and Attitude (BA) towards physics and learning are believed to influence student performance in learning. Identification and exploration of students' BA are crucial aspects in supporting the quality of physics learning. The Colorado Learning Attitudes about Sciences Survey (CLASS) is an instrument that is widely used to identify and explore students' beliefs and attitudes towards physics and learning. However, few studies have investigated the relationship between each categories of CLASS and the impact of demographic characteristics (gender, study program, school origin, interest in becoming a teacher, and length of study) on student beliefs. This study aims to fill in this gap. Using CLASS to measure these beliefs and attitudes among 439 physics students and prospective physics teachers from 11 universities. The survey data was analyzed with the Pearson Correlation, Mann-Whitney U, and Kruskal-Wallis tests. Overall, the results obtained show that all belief and attitude categories are significantly interconnected. Apart from that, there is also a positive and significant relationship between categories. This indicates that the three categories have demonstrated to be capable of constructing BA as a whole. However, no significant differences were discovered in the association between students' belief and attitudes towards physics and their learning based on demographic characteristics.

Common Mistakes in Pronunciation among Students in The English for Specific Purposes Department, Hanoi University

When learning English as a second language in a nation with a wide variety of races and cultures, the influence of the native tongue is nearly inevitable. Both the performance and communicative competence of learners can be negatively impacted by this influence. This paper is written for students in the English for Specific Purposes Department (ESPD), Hanoi University, who want to improve their learning of English Pronunciation. This mixed-methods study combine quantitative and qualitative methodologies. It focuses on some common pronunciation errors that EPSD students made when speaking English by analysing 120 recordings for the students’ mid-term test from their second semester in the academic year 2023-2024. Some common pronunciation errors are found, including consonant endings and vowel sounds. The study also shares some strategies for learners to improve their English pronunciation. It will be a helpful reference source for not only lecturers and learners at ESPD, Hanoi University, but other educational institutions across Vietnam.

Investigating the Effect of Role-Play Activity With GenAI Agent on EFL Students’ Speaking Performance

Role-play activities are considered a useful instructional design in enhancing the speaking performance of foreign language learners. However, in the traditional classroom context, learners may not readily have access to interlocutors for role-play activities. In this study, we proposed a designed method that integrated the GenAI agent into role-play activities and conducted a quasi-experiment at a Chinese university. A total of 53 Chinese students in an English course were assigned to an experimental group ( n = 29) that engaged in role-playing with a GenAI agent and a control group ( n = 24) that engaged in role-playing with their class peers. The results showed role-play activities with GenAI could facilitate students’ speaking performance, and one-way ANCOVA showed there was no significant difference in speaking performance improvement when compared with traditional role-play activities with peers. Moreover, the paired-sample Wilcoxon test showed that the treatment students had significant intrinsic motivation and self-efficacy improvement, which were higher than the control group. The findings provide important implications for the design and implementation of GenAI-integrated role-play activities for foreign language speaking learning.

An Effective Ensemble Approach for Preventing and Detecting Phishing Attacks in Textual Form

Phishing email assaults have been a prevalent cybercriminal tactic for many decades. Various detectors have been suggested over time that rely on textual information. However, to address the growing prevalence of phishing emails, more sophisticated techniques are required to use all aspects of emails to improve the detection capabilities of machine learning classifiers. This paper presents a novel approach to detecting phishing emails. The proposed methodology combines ensemble learning techniques with various variables, such as word frequency, the presence of specific keywords or phrases, and email length, to improve detection accuracy. We provide two approaches for the planned task; The first technique employs ensemble learning soft voting, while the second employs weighted ensemble learning. Both strategies use distinct machine learning algorithms to concurrently process the characteristics, reducing their complexity and enhancing the model’s performance. An extensive assessment and analysis are conducted, considering unique criteria designed to minimize biased and inaccurate findings. Our empirical experiments demonstrates that using ensemble learning to merge attributes in the evolution of phishing emails showcases the competitive performance of ensemble learning over other machine learning algorithms. This superiority is underscored by achieving an F1-score of 0.90 in the weighted ensemble method and 0.85 in the soft voting method, showcasing the effectiveness of this approach.

Enhancing learner autonomy and performance: A comprehensive exploration of Learning to Learn (L2L) facilitation.

Enhancing learner autonomy and performance: A comprehensive exploration of Learning to Learn (L2L) facilitation.

Curiosity-Driven Multi-Level Intrinsic Reward DQN for Enhanced Exploration in Reinforcement Learning

Abstract. Reinforcement learning (RL) has shown great potential in solving complex decision-making tasks. However, efficient exploration remains a significant challenge, particularly in environments with sparse or deceptive rewards. This paper introduces DQN-Mult-Cur, a novel reinforcement learning algorithm that enhances exploration by integrating curiosity-driven and multi-level intrinsic rewards within the Deep Q-Network (DQN) framework. The proposed method addresses the limitations of conventional exploration strategies by incentivizing agents to explore novel and meaningful states, thereby improving learning efficiency and performance. Extensive experiments across three standard environmentsCartPole-v1, MountainCar-v0 and Acrobot-v1demonstrate that DQN-Mult-Cur outperforms traditional DQN variants, achieving faster convergence, higher rewards, and greater stability. An ablation study further highlights the importance of each intrinsic reward component, confirming the robustness of the proposed approach. The results suggest that DQN-Mult-Cur offers a comprehensive solution to the exploration-exploitation trade-off in reinforcement learning, making it applicable to a wide range of challenging environments.

The impact of virtual laboratories on the academic performance of grade 12 learners in impulse and momentum

Virtual laboratories (labs) have evolved as an important tool in the sciences, delivering interactive and engaging platforms for learners to conduct experiments and comprehend complex concepts without the limitations of physical resources. They are an emerging teaching tool that offers interactive and immersive experiences that improve comprehension of complicated scientific concepts. This study assesses the impact of virtual labs on the academic achievement of Grade 12 learners on the topic of impulse and momentum in physics. It uses a quasi-experimental methodology to compare the performance of learners who used virtual labs to those who learned in the prevailing theoretical settings. A pre-test and post-test were used to assess learners’ academic success, and surveys were used to gather learner feedback on the learning tool. Preliminary data analysis indicates a statistically significant improvement in the test scores of learners using virtual labs, corroborating earlier studies that highlight the advantages of technologically enhanced learning environments. Furthermore, qualitative feedback indicates that learners found virtual labs more interesting and accessible, resulting in a more conducive learning environment. This work adds to the expanding body of evidence supporting the use of virtual labs in science teaching, especially in resource-constrained contexts where traditional laboratory equipment is limited or non-existent. The implications for curriculum design and educational policy are examined, with a focus on the importance of investing in digital infrastructure to support innovative teaching approaches.

Physical Facility Availability and Students’ Academic Performance in Public Secondary Schools in Trans Nzoia East Sub-County, Kenya

Education is essential for individual growth and national progress, but its success largely depends on the institutions that sustain it. Identifying how these institutional factors contribute to or hinder academic success is vital for overcoming challenges in the education sector and promoting academic achievement. This study seeks to examine how physical facility availability influences academic performance of learners in public secondary schools in Trans Nzoia East Sub-county, Kenya. The study set out to ascertain the following effects: the influence of science laboratories and computer laboratories in the academic performance of students in public secondary schools in Kenya's Trans Nzoia East Sub-County; assess the influence of library resources on the academic performance of students in public secondary schools in Kenya's Trans Nzoia East Sub-County; to evaluate the effects of dormitories on the academic performance of students in public secondary schools in Kenya's Trans Nzoia East Sub-County; and the impact of sanitation facilities and open fields on the academic performance of students in public secondary schools in Kenya's Trans Nzoia East Sub-County. Based on the principle of the Education Production Function, the study was conducted utilized a mixed-methods, descriptive approach. The research targeted a population of 4574 persons consisting of 73 principals, 1,280 serving instructors and 3221 senior students in government secondary institutions. Participants were selected using random sampling to appoint the sample of principals, teachers and students. A sample of 354 individuals were identified using the Krejcie &amp; Morgan table (1970) of choosing the appropriate number of sample. The sample consisted of 6 principals, 99 serving teachers and 249 form four students. To collect data, the study used interviews, questionnaires, document analysis and observation. The instruments were piloted in three public secondary schools in Trans Nzoia West region. Validity testing was done by the researcher being assisted by professionals who expertise in various aspects of validy. In reliability, test-retest method was administered and determined by the Cronbach alpha coefficient of a 0.7 threshold. . The analysis of quantitative data employed descriptive and inferential statistical techniques, with the aid of SPSS version 26 while qualitative data were categorized into themes corresponding to the study’s key objectives. The key findings indicated a statistically significant association between physical facilities and the academic performance of students. A Pearson correlation analysis demonstrated a significant positive relationship(0.968)between the availability of physical facilities and students’ academic performance indicating that students in schools with better facilities achieved higher scores academically. These results confirm that the presence of adequate physical facilities support better learning condition and hence improve students outcomes. It is recommended that educational policymakers invest in enhancing physical facilities to improve academic outcomes for students in public secondary schools.

Effects of subject-specific professional knowledge and skills of physical education teachers on students’ learning progress

AbstractAlthough the subject-specific professional knowledge of teachers is considered to be of central importance with regard to the learning performance of their students, there is only sparse empirical evidence on this issue. Moreover, current studies are primarily limited to the subject of mathematics. The aim of this article is to investigate the significance of subject-specific professional knowledge and skills of physical education (PE) teachers on the practical learning gains of their students based on a panel study. In order to measure both non-contextual knowledge and contextual skills on the teachers’ side, paper–pencil tests as well as video and text vignettes were used as test instruments. On the students’ side, practical sport skills were assessed in a pretest–posttest design to evaluate competence growth with regard to complex sports performance. An effect chain with mediation processes was postulated. Multilevel modeling analyzing 1669 students who were taught by 74 teachers could prove significant performance improvements for the students, but none of the four differentiated knowledge dimensions (content knowledge, non-contextual; pedagogical content knowledge, non-contextual; content knowledge, contextual; pedagogical content knowledge, contextual) of the teachers could explain these learning gains. Thus, the results are in line with various studies in mathematics and suggest that direct effects of teachers’ professional knowledge on students’ learning gains seem to be lost. The results therefore suggest that while basic and broad professional knowledge is required for instructional success, specific expertise in certain competence areas does not lead to additional learning success among students. Further research within the EPiC-PE study is needed to provide a more comprehensive picture of the relationships between professional knowledge and skills, classroom behavior, instructional processes, and learning outcomes.

Bias and Its Consequences : A Study of Machine Learning Performance

This paper addresses the concern about bias affecting the results of machine learning models. For this purpose, it uses the Adult Income dataset from OpenML for income classification. The conditions for bias are induced by underrepresenting people that earn &lt;= $50K in training data, thus checking the behavior of different models when encountering such a skewed distribution. Key metrics, namely accuracy and specificity (True Negative Rate), were analyzed for unbiased and biased training scenarios. The results show that Naive Bayes and Random Forest models were resistant to bias, but others, including SVM and Logistic Regression, suffered major performance drops. This study throws light on the robustness of different classifiers when exposed to biased data, requiring further bias mitigation strategies in real-world applications. This paper actually examines critically how bias in training data can significantly affect the performance of prediction, fairness, and model selection in income classification tasks.