Behavioral Feedback Research Articles

Grey wolf algorithm for cost optimization of cloud computing repairable system with [formula omitted]-policy, discouragement and two-level Bernoulli feedback

With the increasing reliance on cloud computing as the foundational infrastructure for technological operations, the urgency and efficacy with which failed virtual machines can be successfully repaired earns significant weight. In this study, we have presented a queueing model that analyzes the repair process of failed virtual machines in cloud computing system. This model includes various important components, such as control N-policy, server unreliability, discouragement behavior of failed virtual machines (like balking and reneging), retention of reneged machines, redundancy, and two-level Bernoulli feedback. A steady-state performance analysis has been performed. The proposed model provides practical suggestions for cloud service providers to improve the reliability and availability of their services, which insights more than theoretical frameworks. This study is unique in the field of queueing models of cloud computing machining environment, as it incorporates parallel consideration of control N policies, balking, reneging, retention, server vacation, two-level Bernoulli feedback and unreliable server behavior. A thorough examination of the behavior of key performance indicators has been conducted and presented using detailed graphical representations. In addition, a cost function is formulated for the proposed model. Further, the Grey Wolf Optimization (GWO) and Particle Swarm Optimization (PSO) techniques are utilized to optimize the total cost function, thereby gaining valuable insights into the system’s behavior across various conditions as well as to compare the outcomes of GWO and PSO.

Development of Trustworthiness for Cloud Service Providers Using DBN-Based Trust Model in Cloud Computing Environment

In order to facilitate diverse computing resources and services, cloud computing has evolve into a promising paradigm on-demand over the internet. To access services, cloud users have to rely on third-party service providers. Choosing a suitable Cloud Service Provider (CSP) with a raise in available cloud services in order to deliver the service safely is considered a serious concern for users. Regrettably, there are various problems that minimize the growth of cloud computing, like privacy, security loss, and control. The security issue is regarded as the most important element that could avoid the evolution of cloud computing. In the cloud environment, to handle the user’s requests, trust measures play a significant role when choosing appropriate service providers. Therefore, trustworthiness evaluation of CSP prior to choosing it to facilitate the service has become a significant obligation in cloud environment. In this work, a trust model, Deep Behavioral Feedback Quality of Service and Statistics based trust (Deep BFQS-trust), is developed to calculate trustworthiness of CSP based on its feedback and behavior, QoS and statistics-based given by the users. Also, to calculate behavioral trust values, various QoS attributes are considered. In order to maintain and calculate feedback trust value for service provider, diverse parameters from service level agreement are utilized. By computing the collective trust, trustworthiness of cloud service provider is judged that is computed by these trust factors. Moreover, the weights of the collective trust are determined by employing Deep belief network (DBN) model. Finally, the proposed Deep BFQS-trust technique is compared with the existing approaches, and exhibits that the proposed model attained utmost trustworthiness and successful interaction with the values of 0.860 and 0.888, respectively.

Social behavior, academic achievement, language skills, and peer rejection in elementary school classes: the moderating role of teacher feedback

The study examines the influence of teacher feedback in the relationship between peer rejection and student level predictors of rejection. Feedback on academic performance and social behaviour recorded during a standardised lesson was analysed for each of 36 Grade 1–3 classes (N = 709). Student social behaviour, academic achievement, and language skills were assessed at the start of the school year. Peer rejection was assessed at both the beginning and end of the school year. Three types of feedback behaviour were identified: Teachers who gave most negative feedback on social behaviour, those who gave most positive feedback on academic performance, and those who gave less feedback, positive or negative. Results provide evidence for the moderating role of teacher behaviour in the relationship between student academic achievement and peer rejection. In classrooms with the highest proportion of positive feedback on academic performance, academic achievement predicted peer rejection and there was a decrease in peer rejection over an academic year.

Effect of Redox Mediators on Corrosion Behavior and Scanning Electrochemical Microscopy Response.

Scanning electrochemical microscopy (SECM) is widely used to measure local electrochemical reactivity of corroding surfaces. A major criticism of using SECM in feedback mode for corrosion studies is the requirement of an external redox mediator (RM) as it could react with the metal and affect the Nernst potential at the metal-solution interface. Consequently, it becomes challenging to differentiate the interference caused by the RM from the local reactivity of the metal. Herein, a multiscale electrochemical approach is presented to investigate the effect of RM choice on the corroding substrate. Two common RMs, ferrocenemethanol and hexaammineruthenium(III) chloride, were used to perform SECM over copper and aluminum. It was found during macroscale electrochemical measurements that Ru(NH)63+ acted as an oxidant and promoted corrosion. The SECM feedback behavior varied for copper depending on the RM used, suggesting that the corrosion reactions controlled the negative feedback mechanism, not the formation of an insulating passive film. The passivated aluminum surface consistently exhibited negative feedback, regardless of the RM used. SECM approach curves also displayed a distortion in the steady state current, which was caused by the deposition of substrate-generated species on the microelectrode. These deviations in feedback response were accounted for during analysis through incorporation into a finite element model to accurately extract the RM kinetic rate constants. The importance of understanding these processes is highlighted to avoid misinterpretation of passive behavior and advances toward a more quantitative use of SECM for corrosion studies.

Task-driven Risk-bounded Hierarchical Reinforcement Learning Based on Iterative Refinement

Deep Reinforcement Learning (DRL) has garnered substantial acclaim for its versatility and widespread applications across diverse domains. Aligned with human-like learning, DRL is grounded in the fundamental principle of learning from interaction, wherein agents dynamically adjust behavior based on environmental feedback in the form of rewards. This iterative trial-and-error process, mirroring human learning, underscores the importance of observation, experimentation, and feedback in shaping understanding and behavior. DRL agents, trained to navigate complex surroundings, refine their knowledge through hierarchical and abstract representations, empowered by deep neural networks. These representations enable efficient handling of long-horizon tasks and flexible adaptation to novel situations, akin to the human ability to construct mental models for comprehending complex concepts and predicting outcomes. Hence, abstract representation building emerges as a critical aspect in the learning processes of both artificial agents and human learners, particularly in long-horizon tasks. Furthermore, human decision-making, deeply rooted in evolutionary history, exhibits a remarkable capacity to balance the tradeoff between risk and cost across various domains. This cognitive process involves assessing potential negative consequences, evaluating factors such as the likelihood of adverse outcomes, severity of potential harm, and overall uncertainty. Humans intuitively gauge inherent risks and adeptly weigh associated costs, extending beyond monetary expenses to include time, effort, and opportunity costs. The nuanced ability of humans to consider the tradeoff between risk and cost highlights the complexity and adaptability of human decision-making, a skill lacking in typical DRL agents. Principles like these derived from human-like learning present an avenue for inspiring advancements in DRL, fostering the development of more adaptive and intelligent artificial agents. Motivated by these observations and focusing on practical challenges in robotics, our efforts target risk-aware stochastic sequential decision-making problem which is crucial for tasks with extended time frames and varied strategies. A novel integration of model-based conditional planning with DRL is proposed, inspired by hierarchical techniques. This approach breaks down complex tasks into manageable subtasks(motion primitives), ensuring safety constraints and informed decision-making. Unlike existing methods, our approach addresses motion primitive improvement iteratively, employing diverse prioritization functions to guide the search process effectively. This risk-bounded planning algorithm seamlessly integrates conditional planning and motion primitive learning, prioritizing computational efforts for enhanced efficiency within specified time limits.

Behavioral skills training for teaching safety skills to mental health service providers compared to training-as-usual: a pragmatic randomized control trial.

Violence in the healthcare workplace has been a global concern for over two decades, with a high prevalence of violence towards healthcare workers reported. Workplace violence has become a healthcare quality indicator and embedded in quality improvement initiatives of many healthcare organizations. The Centre for Addiction and Mental Health (CAMH), Canada's largest mental health hospital, provides all clinical staff with mandated staff safety training for self-protection and team-control skills. These skills are to be used as a last resort when a patient is at imminent risk of harm to self or others. The purpose of this study is to compare the effectiveness of two training methods of this mandated staff safety training for workplace violence in a large psychiatric hospital setting. Using a pragmatic randomized control trial design, this study compares two approaches to teaching safety skills CAMH's training-as-usual (TAU) using the 3D approach (description, demonstration and doing) and behavioural skills training (BST), from the field of applied behaviour analysis, using instruction, modeling, practice and feedback loop. Staff were assessed on three outcome measures (competency, mastery and confidence), across three time points: before training (baseline), immediately after training (post-training) and one month later (follow-up). This study was registered with the ISRCTN registry on 06/09/2023 (ISRCTN18133140). With a sample size of 99 new staff, results indicate that BST was significantly better than TAU in improving observed performance of self-protection and team-control skills. Both methods were associated with improved skills and confidence. However, there was a decrease in skill performance levels at the one-month follow-up for both methods, with BST remaining higher than TAU scores across all three time points. The impact of training improved staff confidence in both training methods and remained high across all three time points. The study findings suggest that BST is more effective than TAU in improving safety skills among healthcare workers. However, the retention of skills over time remains a concern, and therefore a single training session without on-the-job-feedback or booster sessions based on objective assessments of skill may not be sufficient. Further research is needed to confirm and expand upon these findings in different settings.

Towards the implications of positive teacher interpersonal behaviors for Spanish learners’ academic engagement: Voices from Chinese SFL learners

With the advent of positive psychology in the area of language education, more focus has been placed on the consequences of favorable teacher communication behaviors in language classes. Nonetheless, the function of language instructors' interpersonal behaviors in raising learners' engagement is somehow unknown. Furthermore, to the best of our knowledge, no research study has been carried out in Spanish language classes to explore the function of teacher communication behaviors in learners' engagement. To fill this lacuna, the current inquiry looked into the role of teachers' positive interpersonal factors in Chinese SFL learners' behavioral, cognitive, and emotional engagement. For this purpose, a random sample of 32 SFL learners was chosen to participate in our online interview sessions. The answers of SFL learners to the interview questions were thematically analyzed via MAXQDA software. The thematic analysis findings pointed to the value of teacher communication behaviors in improving SFL learners' engagement. The analysis outcomes also demonstrated the potential of 11 positive interpersonal behaviors (rapport, care, clarity, credibility, confirmation, immediacy, closeness, praise, feedback, respect, and stroke) in increasing Spanish language learners' academic engagement. The practical implications that may emerge from the present study's outcomes are finally discussed.

Integrating ecological feedbacks across scales and levels of organization

In ecosystems, species interact in various ways with other species, and with their local environment. In addition, ecosystems are coupled in space by diverse types of flows. From these links connecting different ecological entities can emerge circular pathways of indirect effects: feedback loops. This contributes to creating a nested set of ecological feedbacks operating at different organizational levels as well as spatial and temporal scales in ecological systems: organisms modifying and being affected by their local abiotic environment, demographic and behavioral feedbacks within populations and communities, and spatial feedbacks occurring at the landscape scale. Here, we review how ecological feedbacks vary in space and time, and discuss the emergent properties they generate such as species coexistence or the spatial heterogeneity and stability of ecological systems. With the aim of identifying similarities across scales, we identify the abiotic and biotic modulators that can change the sign and strength of feedback loops and show that these feedbacks can interact in space or time. Our review shows that despite acting at different scales and emerging from different processes, feedbacks generate similar macroscopic properties of ecological systems across levels of organization. Ultimately, our contribution emphasizes the need to integrate such feedbacks to improve our understanding of their joint effects on the dynamics, patterns, and stability of ecological systems.

Social tipping dynamics in the energy system

Abstract. This paper reviews evidence on how the fast growth in renewable energy technologies can trigger social tipping dynamics that potentially accelerate a system-wide energy transition. It does so by reviewing a variety of literature across several disciplines addressing socio-technical dimensions of energy transitions. The tipping dynamics in wind and solar power create potential for cascading effects to energy demand sectors, including household energy demand. These most likely start with shift actions and adoption of household-scale batteries and heat pumps. Key enablers are strong regulations incentivising reductions in demand and setting minimum efficiency levels for buildings and appliances. While there is evidence of spillovers to more environmentally friendly behaviour, the extent of these and the key leverage points to bring them about present a knowledge gap. Moreover, these behavioural feedback loops require strong additional policy support to “make them stick”. Understanding the economic and social tipping dynamics in a system can empower decision-makers, fostering realistic energy transition policies. This paper highlights energy communities as a promising niche for leveraging tipping dynamics. Ultimately, bridging the gap between these tipping dynamics and institutional reforms is crucial for unlocking the full potential of sustainable energy systems.

SPICE: the connection between cosmic reionization and stellar feedback in the first galaxies

ABSTRACT We present SPICE, a new suite of radiation-hydrodynamic, cosmological simulations targeting the epoch of reionization. The goal of these simulations is to systematically probe a variety of stellar feedback models, including ‘bursty’ and ‘smooth’ forms of supernova energy injection, as well as poorly explored physical scenarios such as hypernova explosions and radiation pressure on dust. We show that even subtle differences in the behaviour of supernova feedback drive profound differences in reionization histories, with burstier forms of feedback causing earlier reionization. However, we also find that some global galaxy properties, such as the dust-attenuated luminosity functions and star formation main sequence, remain degenerate between models. In particular, we show that stellar feedback and its strength determine the morphological mix of galaxies emerging by $z \, = \, 5$ and that the reionization history is inextricably connected to intrinsic properties such as galaxy kinematics and morphology. While star-forming, massive discs are prevalent if supernova feedback is ‘smooth’, ‘bursty’ feedback preferentially generates dispersion-dominated systems. Different modes of feedback produce different strengths of outflows, altering the interstellar/circumgalactic medium in different ways, and in turn strongly affecting the escape of Lyman continuum (LyC) photons. We establish a correlation between galaxy morphology and LyC escape fraction, revealing that dispersion-dominated systems have escape fractions 10–50 times higher than their rotation-dominated counterparts at all redshifts. At the same intrinsic luminosity, dispersion-dominated systems should thus preferentially generate large H ii regions as compared to their rotation-dominated counterparts. Since dispersion-dominated systems are more prevalent if stellar feedback is more explosive, reionization occurs earlier in our simulation with burstier feedback. We argue that statistical samples of post-reionization galaxy morphologies (using both stellar and gaseous components) probed with telescopes such as JWST, ALMA, and MUSE can constrain stellar feedback at z > 5 and models of cosmic reionization.

Piloting co-developed behaviour change interventions to reduce exposure to air pollution and improve self-reported asthma-related health.

Exposure to air pollution can exacerbate asthma with immediate and long-term health consequences. Behaviour changes can reduce exposure to air pollution, yet its 'invisible' nature often leaves individuals unaware of their exposure, complicating the identification of appropriate behaviour modifications. Moreover, making health behaviour changes can be challenging, necessitating additional support from healthcare professionals. This pilot study used personal exposure monitoring, data feedback, and co-developed behaviour change interventions with individuals with asthma, with the goal of reducing personal exposure to PM2.5 and subsequently improving asthma-related health. Twenty-eight participants conducted baseline exposure monitoring for one-week, simultaneously keeping asthma symptom and medication diaries (previously published in McCarron et al., 2023). Participants were then randomised into control (n = 8) or intervention (n = 9) groups. Intervention participants received PM2.5 exposure feedback and worked with researchers to co-develop behaviour change interventions based on a health behaviour change programme which they implemented during the follow-up monitoring week. Control group participants received no feedback or intervention during the study. All interventions focused on the home environment. Intervention group participants reduced their at-home exposure by an average of 5.7 µg/m³ over the monitoring week (-23.0 to +3.2 µg/m³), whereas the control group had a reduction of 4.7 µg/m³ (-15.6 to +0.4 µg/m³). Furthermore, intervention group participants experienced a 4.6% decrease in participant-hours with reported asthma symptoms, while the control group saw a 0.5% increase. Similarly, the intervention group's asthma-related quality of life improved compared to the control group. This pilot study investigated a novel behaviour change intervention, utilising personal exposure monitoring, data feedback, and co-developed interventions guided by a health behaviour change programme. The study aimed to reduce personal exposure to fine particulate matter (PM2.5) and improve self-reported asthma-related health. Conducting a randomised controlled trial with 28 participants, co-developed intervention successfully targeted exposure peaks within participants' home microenvironments, resulting in a reduction in at-home personal exposure to PM2.5 and improving self-reported asthma-related health. The study contributes valuable insights into the environmental exposure-health relationship and highlights the potential of the intervention for individual-level decision-making to protect human health.

A (dynamic) investigation of stereotypes, belief‐updating, and behavior

AbstractUsing a controlled experiment, we study the dynamic effects of feedback on decision‐making across verbal skills and math. Before feedback, men are more optimistic about their performance and more willing to compete than women, especially in math. While feedback shifts individuals' beliefs and behavior, we see substantial persistence of gender gaps 1 week later. This is particularly true among individuals who receive negative feedback. Our results are not well‐explained by motivated reasoning; in fact, negative feedback is more likely to be recalled than positive feedback. Overall, our results highlight the challenges involved in overcoming gender gaps in dynamic settings.

Construction of the "Full Path" of restorative effects on older adults' mental health in parks under seasonal differences: taking Changchun as an example.

With the aging and older adults' mental health problems in China, more attention has been paid to the restorative environment. As an important restorative environment in the city, the mental health restorative effect of park environment has been confirmed. However, further exploration is needed to determine whether winter parks have positive effects, their differences from non-winter parks, and the specific pathways of these effects. Therefore, this study constructed a "full path" for the restorative effects of older adults' mental health in parks under seasonal differences, including four components: perceived environment, affective feedback, behavioral feedback, and restorative effect, forming four pathways. Based on this, this study obtained 211 and 240 sample data in winter and non-winter parks, and verified the validity of various hypotheses and mediation paths using structural equation models. It found that: (1) overall restorative effects existed in different seasons; (2) in winter, perceived environmental assessment was not a direct antecedent of restorative effects, and affective feedback and Moderate and Vigorous Physical Activity (MVPA) feedback were important mediating factors, and the chain mediated pathway existed; (3) in non-winter, both direct, indirect and chain mediated effects existed, and affective feedback and Low Physical Activity (LPA) feedback were important mediating factors. Based on this, this study divided parks into "affective inducing" and "behavioral promoting" types, and proposed corresponding planning priorities to positively intervene in planning and design practices.

Not so simple: evaluating consequences validity evidence for a workplace-based assessment in surgery

PurposeWorkplace-based assessments (WBAs) of trainee operative skills are widely used in surgical education as formative assessments to facilitate feedback for learning, but the evidence to support this purpose is mixed. Further evaluation of the consequences of assessment use and score interpretation is needed to understand if there is alignment between the intended and actual impacts of assessment. This study examines consequences validity evidence for an operative WBA, exploring whether WBA use is consistent with the goals of formative assessment for learning.MethodsEight residents and 9 faculty within the Department of Otolaryngology—Head and Neck Surgery at a tertiary institution completed semi-structured interviews after participating in a pilot of a surgical WBA, the System for Improving and Measuring Procedural Learning in the OR (SIMPL OR). Residents received feedback from attendings via both scores (performance and autonomy ratings) and recorded dictations. Interview questions explored faculty and resident perceptions of feedback behaviors and perceived impacts on their teaching or learning practices. Three researchers analyzed transcripts using directed qualitative content analysis to generate themes and evaluated how the perceived impacts aligned with formative purposes for assessment and score use.ResultsBoth faculty and residents identified intended impacts of formative assessment, including (1) greater emphasis on feedback, (2) support for a postoperative feedback routine, and (3) facilitation of case-specific reflection. Residents also used score and verbal feedback for (1) calibrating case perceptions and (2) benchmarking performance to an external standard. The recorded dictations supported feedback by (1) providing context for ratings, (2) facilitating review of dictated feedback, and (3) prompting faculty for deliberate feedback. Unintended impacts included: (1) emotional discomfort during the assessment process, (2) increased feedback frequency but not diversity or quality, (3) inadequate support for feedback conversations, and (4) limited next steps for teaching or learning. Assessment usage declined over the pilot period.ConclusionsThe validity evidence gathered in this study suggests an operative WBA can be used for formative purposes to improve perceptions of feedback, but unintended consequences and implementation challenges limited ultimate impacts on teaching and learning. User perspectives can add important elements to consequences validity evidence and should be further evaluated in different implementation settings to better understand how WBAs can achieve their formative goals.

Developmental leadership as a driver of employee feedback behaviors: the roles of learning goal orientation and leader high performance expectations

PurposeDrawing on goal orientation theory, this study intends to investigate whether, how and when developmental leadership influences employees’ feedback seeking behavior (FSB) and feedback avoidance behavior (FAB).Design/methodology/approachThe authors used a two-wave survey of 416 full-time employees in China. The hypotheses were tested with path analyses.FindingsDevelopmental leadership exerts a positive influence on employees’ FSB and a negative impact on FAB through learning goal orientation (LGO). Additionally, leaders’ high performance expectations (LHPE) not only strengthen the positive effects of developmental leadership on FSB but also intensify its weakening effects on FAB.Originality/valueThe findings enrich current understanding by associating developmental leadership with employees’ FSB/FAB, offering a new viewpoint on its positive impacts. This study also provides deeper insights into when the benefits of developmental leadership are intensified.

A sustainable and trusted solution for IoT-based federated learning using feedback behavior

Federated learning is defined as one of the solutions that use distributed clients to train and aggregate the target model without sharing the private information of clients in the network. Hence, federated learning may easily handle and tackle the generated information in a more convenient, feasible and secure manner, while involving the intelligent devices to perform testing and training at the local edges. However, the design of such federated learning solutions faces multiple difficulties due to the cost and complexity of information computation, along with a high energy consumption. Furthermore, the identification of resource-constrained devices which ensures significant power consumption along with privacy and security to the devices makes it critical to attain a sustainable federated learning environment. The aim of this paper is to integrate feedback-based and objective trust models to accurately identify the behavior of communicating devices in the network. The secure transmission of information forwarded by trusted IoT devices is computed by the objective model. In addition, the energy consumption required for measuring the authenticity of a device can be further diminished using a behavior and feedback trusted model. In addition, the past history behavior and feedback received from neighboring devices may also contribute to achieving an efficient and sustainable security mechanism with reduced communication steps and computational delay in the network. The proposed solution is evaluated to analyze the feasibility and performance in terms of energy consumption and resource utilization, while also considering various security metrics for comparison against existing methods. The proposed mechanism out-performed 87% improvement in terms of security metrics in comparison of existing approaches.

Comparing trained EFL peer reviewers’ feedback: From claim to reality

Comparing trained L2 writing student reviewers’ feedback behaviors as well as examining the extent to which their claims are aligned with their actual evaluation practices have received limited scholarly attention. Employing think-aloud protocols, one cause and one effect essays evaluated by five upper-intermediate L2 learners, and follow-up semi-structured interviews, this case study research aimed to explore trained L2 peer reviewers’ feedback behaviors and the matches and mismatches between their claims and evaluation practices. While the first and the second data source compared the participants’ actual feedback practices in terms of nature, type, and validity, the last source probed their claims on peer evaluation. The findings contribute to peer feedback research by emphasizing on the need for individual, customized, and constant peer review training sessions rather than general, all-purpose, and decontextualized instructions which can consequently improve peer feedback quality in L2 writing contexts.

Exploring the REIT architecture for requirements elicitation interview training with robotic and virtual tutors

Requirements elicitation interviews are a widely adopted technique where the interview success depends on the interviewer’s preparedness and communication skills. Students can enhance these skills through practice interviews. However, organizing practice interviews for many students presents scalability challenges, given the time and effort required to involve stakeholders in each session. To address this, we propose REIT, an extensible architecture for Requirements Elicitation Interview Training system leveraging technologies such as robots and voice systems. REIT has components to support both the interview phase, wherein students act as interviewers while the system assumes the role of an interviewee, and the feedback phase, during which the system assesses students’ performance and offers contextual and behavioral feedback to enhance their interviewing skills. We demonstrate the applicability of REIT through two implementations: RoREIT with a physical robotic agent and VoREIT with a virtual voice-only agent. We empirically evaluated both instances with a group of graduate students. The participants appreciated both systems. They demonstrated higher learning gain when trained with RoREIT, but they found VoREIT more engaging and easier to use. These findings indicate that each system has distinct benefits and drawbacks, suggesting that educators can customize REIT for various settings, considering preferences and available resources.

A scoping review of feedback in physical education: Conceptualisations and the role of teachers and students

Given the important role of feedback in student learning, this scoping review investigated how feedback has been conceptualised and examined, and how the teacher and student have been positioned in feedback processes in physical education (PE) research. Following recommended processes for conducting a scoping review, 110 papers were included, and data were extracted and synthesised to address the aim. Findings identified a lack of conceptual and theoretical underpinnings, and considerable variability in feedback terminology used in PE. Feedback was mainly positioned as being transmitted verbally from the teacher to the student to evaluate performances. Research methods mainly focused on the teacher's feedback behaviours, with less attention given to how the student receives, interprets, and uses feedback to advance their learning. Based on the findings, we recommend that future feedback research in PE is underpinned by theory and frameworks that acknowledge the active role of both the teacher and the student in the feedback process. Additionally, given the critical element of feedback effectiveness is not in the provision of feedback, but how it is interpreted and used, research methods that further consider how students engage with feedback opportunities are required. If we are to maximise feedback's learning potential, future PE research and practices should consider how students can be further activated in the feedback process.

A Robust Sequential Recommendation Model Based on Multiple Feedback Behavior Denoising and Trusted Neighbors

At present, most of the personalized sequential recommendations utilize users’ implicit positive feedback (such as clicks) to predict user behavior, ignoring the impact of implicit negative feedback and explicit feedback on the accuracy of recommendation results prediction. In this paper, we propose a robust sequence recommendation model based on multi feedback behavior denoising and trusted neighbors, which utilizes multiple feedback behavior data for feature denoising and considers trusted nearest neighbor information to improve model performance. Firstly, by learning the feature representations and interactions of various types of feedback, explicit feedback is used to map and purify implicit feedback with the same and different attributes, resulting in unbiased user performance. Then, we design a filter attention network to identify highly trusted neighbor information. Finally, we integrate pure user interest representations and trusted nearest neighbor representations to improve the accuracy and robustness of the model. The experimental results on two publicly available datasets show that the proposed sequential recommendation model can achieve superior results to baseline methods in both AUC and RelaImpr.