ABSTRACT
Affected by the COVID-19 epidemic, the final examinations at many universities and the recruitment interviews of enterprises were forced to be transferred to online remote video invigilation, which undoubtedly improves the space and possibility of cheating. To solve these problems, this paper proposes an intelligent invigilation system based on the EfficientDet target detection network model combined with a centroid tracking algorithm. Experiments show that cheating behavior detection model proposed in this paper has good detection, tracking and recognition effects in remote testing scenarios. Taking the EfficientDet network as the detection target, the average detection accuracy of the network is 81%. Experiments with real online test videos show that the cheating behavior detection accuracy can reach 83.1%. In addition, to compensate for the shortage of image detection, we also design an audio detection module to carry out auxiliary detection and forensics. The audio detection module is used to continuously detect the environmental sound of the examination room, save suspicious sounds and provide evidence for judging cheating behavior.
ABSTRACT
INTRODUCTION: During the recent pandemic, e-learning and e-assessment methods have been implemented worldwide, providing opportunities for further implementation in the dental curriculum. This study aims to research the dental students' and dental faculty's perceptions of online exams with e-invigilation. MATERIALS AND METHODS: Online questionnaires were developed and delivered, after three semesters of online exams, to all students and faculty. Descriptive statistics were performed, and answers were grouped into Principal Components (PC) using Principal Component Analysis (PCA). Statistical significance was set at p < .05. RESULTS: Two-hundred and sixty dental students (83.7%) and 24 dental faculty members (63.1%) answered the online questionnaires. PCA of students' responses revealed 4 PC, 'University support to students', 'Comparison between online and face-to-face exams', 'Preparation for the online exams' and 'Attitudes towards the technology used for the online exams'. PCA of faculty responses revealed 5 PC: 'Comparison between online and face-to-face exams', 'University support to faculty', 'Faculty attitudes towards the exam procedures', 'Human factors associated with the exam procedures' and 'Exam invigilation'. The overall satisfaction was high for both staff and students (higher for students and female staff). Students with previous experience in online exams scored more positively than first-year students. University support, process-related stress and e-invigilation were highlighted. CONCLUSIONS: The overall satisfaction with the e-exams was high, despite the technical problems, time-consuming processes and related stress. University support (training, IT support and resources) and mock exams emerged as important elements of online exams, as was the e-invigilation, which students perceived as efficient and non-intrusive.
ABSTRACT
With online examinations prevailing, academic integrity has become a concern. This study is to explore perceptions of online tests among students. An anonymous online survey was conducted among 156 students enrolled in the health professional courses from the CUHK Faculty of Medicine in the academic year 2019-2020. The majority preferred traditional onsite examination (75%) over online format type (25%) because of a less technical requirement (86%), immediate support from invigilators for unanticipated situations (74%), and easiness of focusing (64%);however, some prefer online examinations due to convenience (66%). The pressure and anxiety towards the study do not have significant differences. The reasons for cheating include the desire to pass (42%), peer influence (42%), outstanding grades (38%), and ease of browsing other websites (31%). The application of the blackboard system, Respondus LockDown Browser, with Zoom invigilation minimises the chance of cheating.
ABSTRACT
A key role of universities is the credentialing of student learning by awarding degrees and diplomas. This requires universities to have confidence in the integrity of their assessment processes and in turn, external stakeholders to have the same confidence. This study investigates the following research question: 'Has COVID-19 had an impact on the assessment and invigilation of accounting courses in Australia and New Zealand and, if so, how?' This is a critical issue for accounting faculty in many countries as COVID-19 has forced a shift in the way assessments are administered - from face to face to online. The study involved a survey of accounting faculty in Australia and New Zealand and found changes occurred to how students were assessed because of COVID-19 and a variety of institutional responses to this. The paper makes recommendations for accounting educators, universities, and the professional accounting bodies.
ABSTRACT
Exams play a significant role in an educational institution in learning, and it predicts strength and weaknesses of a student. There are two ways to conduct an exam, i.e. physical or online. In both ways of examination, cheating has become a major problem for academic institutions. Students use various cheating methods, such as using gestures, cheat sheets, written notes on hands, dodging cameras, smartphones, dim lights, etc. Due to physical limitations, typical invigilation systems are not successful in preventing students from cheating. An automated and yet authentic method must be made to prevent and detect different types of cheating during exams. In this article, several invigilation systems have been discussed, including manual and automated systems;they comprise various means adopted for cheating and methods for their interception. Deep learning-based surveillance systems have proved to be more efficient and accurate than manual methods. A taxonomy and detail analysis of cheating and deep learning algorithms like YOLO, LSTM, ResNet, Faster-RCNN, and others is discussed. This study investigates and summarizes different contributions of researchers in an organized manner and shows that using a deep learning approach/method is an efficient and precise way to detect and prevent cheating during exams. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.
ABSTRACT
BACKGROUND: Medical education has historically relied on high stakes knowledge tests sat in examination centres with invigilators monitoring academic malpractice. The COVID-19 pandemic has made such examination formats impossible, and medical educators have explored the use of online assessments as a potential replacement. This shift has in turn led to fears that the change in format or academic malpractice might lead to considerably higher attainment scores on online assessment with no underlying improvement in student competence. METHOD: Here, we present an analysis of 8092 sittings of the Prescribing Safety Assessment (PSA), an assessment designed to test the prescribing skills of final year medical students in the UK. In-person assessments for the PSA were cancelled partway through the academic year 2020, with 6048 sittings delivered in an offline, traditionally invigilated format, and then 2044 sittings delivered in an online, webcam invigilated format. RESULTS: A comparison (able to detect very small effects) showed no attainment gap between online (M = 0.762, SD = 0.34) and offline (M = 0.761, SD = 0.34) performance. CONCLUSIONS: The finding suggests that the transition to online assessment does not affect student performance. The findings should increase confidence in the use of online testing in high-stakes assessment.