key: cord-343764-xan5g88w
authors: Butler-Henderson, Kerryn; Crawford, Joseph
title: A systematic review of online examinations: A pedagogical innovation for scalable authentication and integrity
date: 2020-09-22
journal: Comput Educ
DOI: 10.1016/j.compedu.2020.104024
sha: 
doc_id: 343764
cord_uid: xan5g88w

Digitization and automation across all industries has resulted in improvements in efficiencies and effectiveness to systems and process, and the higher education sector is not immune. Online learning, e-learning, electronic teaching tools, and digital assessments are not innovations. However, there has been limited implementation of online invigilated examinations in many countries. This paper provides a brief background on online examinations, followed by the results of a systematic review on the topic to explore the challenges and opportunities. We follow on with an explication of results from thirty-six papers, exploring nine key themes: student perceptions, student performance, anxiety, cheating, staff perceptions, authentication and security, interface design, and technology issues. While the literature on online examinations is growing, there is still a dearth of discussion at the pedagogical and governance levels.

Learning and teaching is transforming away from the conventional lecture theatre designed to seat 100 to 10,000 passive students towards more active learning environments. In our current climate, this is exacerbated by COVID-19 responses (Crawford et al., 2020) , where thousands of students are involved in online adaptions of face-to-face examinations (e.g. online Zoom rooms with all microphones and videos locked on). This evolution has grown from the need to recognize that students now rarely study exclusively and have commitments that conflict with their University life (e.g. work, family, social obligations). Students have more diverse digitally capability (Margaryan et al., 2011) and higher age and gender diversity (Eagly & Sczesny, 2009; Schwalb & Sedlacek, 1990) . Continual change of the demographic and profile of students creates a challenge for scholars seeking to develop a student experience that demonstrates quality and maintains financial and academic viability (Gross et al., 2013; Hainline et al., 2010) .

Universities are developing extensive online offerings to grow their international loads and facilitate the massification of higher learning. These protocols, informed by growing policy targets to educate a larger quantity of graduates (e.g. Kemp, 1999; Reiko, 2001) , have challenged traditional university models of fully on-campus student attendance. The development of online examination software has offered a systematic and technological alternative to the end-of-course summative examination designed for final authentication and testing of student knowledge retention, application, and extension. As a result of the COVID-19 pandemic, the initial response in higher education across many countries was to postpone examinations (Crawford et al 2020).

However, as the pandemic continued, the need to move to either an online examination format or alternative assessment became more urgent.

This paper is a timely exploration of the contemporary literature related to online examinations in the university setting, with the hopes to consolidate information on this relatively new pedagogy in higher education. This paper begins with a brief background of traditional examinations, as the assumptions applied in many online examination environments build on the techniques and assumptions of the traditional face-to-face gymnasium-housed invigilated examinations. This is followed by a summary of the systematic review method, including search strategy, procedure, quality review, analysis, and summary of the sample.

Print-based educational examinations designed to test knowledge have existed for hundreds of years. The New York State Education Department has "the oldest educational testing service in the United States" and has been delivering entrance examinations since 1865 (Johnson, 2009, p. 1; NYSED, 2012) . In pre-Revolution Russia, it was not possible to obtain a diploma to enter university without passing a high-stakes graduation examinations (Karp, 2007) .

These high school examinations assessed and assured learning of students in rigid and highsecurity conditions. Under traditional classroom conditions, these were likely a reasonable practice to validate knowledge. The discussion of authenticating learning was not a consideration at this stage, as students were face to face only. For many high school jurisdictions, these are designed to strengthen the accountability of teachers and assess student performance (Mueller & Colley, 2015) .

In tertiary education, the use of an end-of-course summative examination as a form of validating knowledge has been informed significantly by accreditation bodies and streamlined financially viable assessment options. The American Bar Association has required a final course examination to remain accredited (Sheppard, 1996) . Law examinations typically contained brief didactic questions focused on assessing rote memory through to problem-based assessment to J o u r n a l P r e -p r o o f evaluate students' ability to apply knowledge (Sheppard, 1996) . In accredited courses, there are significant parallels. Alternatives to traditional gymnasium-sized classroom paper-and-pencil invigilated examinations have been developed with educators recognizing the limitations associated with single-point summative examinations (Butt, 2018) .

The objective structured clinical examinations (OSCE) incorporate multiple workstations with students performing specific practical tasks from physical examinations on mannequins to short-answer written responses to scenarios (Turner & Dankoski, 2008) . The OSCE has parallels with the patient simulation examination used in some medical schools (Botezatu et al., 2010) .

Portfolios assess and demonstrate learning over a whole course and for extracurricular learning (Wasley, 2008) .

The inclusion of online examinations, e-examinations, and bring-your-own-device models have offered alternatives to the large-scale examination rooms with paper-and-pencil invigilated examinations. Each of these offer new opportunities for the inclusion of innovative pedagogies and assessment where examinations are considered necessary. Further, some research indicates online examinations are able to discern a true pass from a true fail with a high level of accuracy (Ardid et al., 2015) , yet there is no systematic consolidation of the literature. We believe this timely review is critical for the progression of the field in first stepping back and consolidating the existing practices to support dissemination and further innovation. The pursuit of such systems may be to provide formative feedback and to assess learning outcomes, but a dominant rationale for final examinations is to authenticate learning. That is, to ensure the student whose name is on the student register, is the student who is completing the assessed work.

The development of digitalized examination pilot studies and case studies are becoming an expected norm with universities developing responses to a growing online curriculum offering J o u r n a l P r e -p r o o f (e.g. Al-Hakeem & Abdulrahman, 2017; Alzu'bi, 2015; Anderson et al., 2005; Fluck et al., 2009; Fluck et al., 2017; Fluck, 2019; Seow & Soong, 2014; Sindre & Vegendla, 2015; Steel et al., 2019; Wibowo et al., 2016) .

As many scholars highlight, cheating is a common component of the contemporary student experience (Jordan, 2001; Rettinger & Kramer, 2009 ) despite that it should not be. Some are theorizing responses to the inevitability of cheating from developing student capacity for integrity (Crawford, 2015; Wright, 2011) to enhancing detection of cheating (Dawson & Sutherland-Smith, 2018 and legislation to ban contract cheating (Amigud & Dawson, 2020) . We see value in the pursuit of methods that can support integrity in student assessment, including during rapid changes to the curriculum. The objective of this paper is to summarize the current evidence on online examination methods, and scholarly responses to authentication of learning and the mitigation of cheating, within the confines of assessment that enables learning and student wellbeing. We scope out preparation for examinations (e.g. Nguyen & Henderson, 2020) to enable focus on the online exam setting specifically.

To address the objective of this paper, a systematic literature review was undertaken, following the PRISMA approach for article selection (Moher et al., 2009) . The keyword string was developed incorporating the U.S. National Library of Medicine (2019) MeSH (Medical Subject Headings) terms: [("online" OR "electronic" OR "digital") AND ("exam*" OR "test") AND ("university" OR "educat*" OR "teach" OR "school" OR "college")]. The following databases dissertations) are not peer-reviewed, and we aimed to identify rigorous best practice literature, we excluded these. We subsequently conducted a general search in Google Scholar and found no additional results. All records returned from the search were extracted and imported into the Covidence® online software by the first author.

The online Covidence® software facilitated article selection following the PRISMA approach.

Each of the 1,906 titles and abstracts were double-screened by the authors based on the eligibility criteria. We also excluded non-higher education examinations, given the context around student demographics is often considerably different than vocational education, primary and high schools. Where there was discordance between the authors on a title or abstract inclusion or exclusion, consensus discussions were undertaken. The screening reduced the volume of papers significantly because numerous papers related to a different education context or involved online or digital forms of medical examinations. Next, the full-text for selected abstracts were doublereviewed, with discordance managed through a consensus discussion. The papers selected following the double full-text review were accepted for this review. Each accepted paper was reviewed for quality using the MMAT system (Hong et al., 2018) and the scores were calculated J o u r n a l P r e -p r o o f as high, medium, or low quality based on the matrix (Hong et al., 2018) . A summary of this assessment is presented in Table 1 .

Following the process described by Braun and Clarke (2006) , an inductive thematic approach was undertaken to identify common themes identified in each article. This process involves six stages: data familiarization, data coding, theme searching, theme review, defining themes, and naming themes. Familiarization with the literature was achieved during the screening, full-text, and quality review process by triple exposure to works. The named authors then inductively coded half the manuscripts each. The research team consolidated the data together to identify themes. Upon final agreement of themes and their definitions, the write-up was split among the team with subsequent review and revision of ideas in themes through independent and collaborative writing and reviewing (Creswell & Miller, 2000; Lincoln & Guba, 1985) . This resulted in nine final themes, each discussed in-depth during the discussion.

There were thirty-six (36) articles identified that met the eligibility criteria and were selected following the PRISMA approach, as shown in Figure 1 . 

The selected articles are from a wide range of discipline areas and countries. Table 1 summarizes the characteristics of the selected articles. The United States of America held a vast majority (14, J o u r n a l P r e -p r o o f 38.9%) of the publications on online examinations, followed by Saudi Arabia (4, 11.1%), China (2, 5.6%), and Australia (2, 5.6%). When aggregated at the region-level, there was an equality of papers from North America and Asia (14, 38.9% each), with Europe (6, 16.7%) and Oceania (2, 5.6%) least represented in the selection of articles. There has been considerable growth in publications in the past five years, concerning online examinations. Publications between the years 2009 and 2015 represented a third (12, 33.3%) of the total number of selected papers. The majority (24, 66.7%) of papers were published in the last three years. Papers that described a system but did not include empirical evidence scored a low-quality rank as they did not meet many of the criteria that relate to the evaluation of a system.

When examining the types of papers, the majority (30, 83.3%) were empirical research, with the remainder commentary papers (6, 16.7%). Of the empirical research papers, three-quarters of the paper reported a quantitative study design (32, 88.9%) compared to two (5.6%) qualitative study designs and two (5.6%) that used a mixed method. For quantitative studies, there was a range between nine and 1,800 student participants (x̄ = 291.62) across 26 studies, and a range between two and 85 staff participants (x̄ = 30.67) in one study. The most common quantitative methods were self-administered surveys and analysis of numerical examination student grades (38% each).

Qualitative and mixed methods studies only adopted interviews (6%). Only one qualitative study reported a sample of students (n = 4), with two qualitative studies reporting a sample of staff (n = 2, n = 5). Students perceive the online examinations saves time (75.0% of students surveyed) and is more economical (87.5%) than paper examinations (Attia, 2014) . It provides greater flexibility for completing examinations (Schmidt, Ralph & Buskirk, 2009 ) with faster access to remote student papers (87.5%) and students trust the result of online over paper-based examinations (78.1%: Attia, 2014) . The majority of students (59.4%: Attia, 2014; 55.5%: Pagram et al., 2018) perceive that the online examination environment makes it easier to cheat. More than half (56.25%) of students believe that a lack of information communication and technology (ICT) skill do not adversely affect performance in online examinations (Attia 2014) . Nearly a quarter (23%) of students reported (Abdel Karim & Shukur, 2016 ) the most preferred font face (type) was Arial, a font also recommended by Vision Australia (2014) in their guidelines for online and print inclusive design and legibility considerations. Nearly all (87%) students preferred black text color on a white background color (87%). With regards to onscreen time counters, a countdown counter was the most preferred option (42%) compared to a traditional analogue J o u r n a l P r e -p r o o f clock (30%) or an ascending counter (22%). Many systems allow students to set their preferred remaining time reminder or alert, including 15 minutes remaining (35% students preferred), 5 minutes remaining (26%), mid-examination (15%) or 30 minutes remaining (13%).

Several studies in the sample referred to a lack of score variation between the results of examination across different administration methods. For example, student performance did not have significant difference in final examination scores across online and traditional examination modalities (Gold & Mozes-Carmel, 2017) . This is reinforced by a test of validity and reliability of computer-based and paper-based assessment that demonstrated no significant difference (Oz & Ozturan, 2018) , and equality of grades identified across the two modalities (Stowell & Bennett, 2010) .

When considering student perceptions, of the studies documented in our sample, there tended to be favorable ratings of online examinations. In a small sample of 34 postgraduate students, the respondents had positive perceptions towards online learning assessments (67.4%).

The students also believed it contributed to improved learning and feedback (67.4%), and 77 percent had favorable attitudes towards online assessment (Attia, 2014) . In a pre-examination survey, students indicated they preferred to type than to write, felt more confident about the examination, and had limited issues with software and hardware (Pagram, 2018) . With the same sample in a post-examination survey, within the design and technology examination, students felt the software and hardware were simple to use, yet many students did not feel at ease from their use of an e-examination .

J o u r n a l P r e -p r o o f Rios and Liu (2017) compared proctored and non-proctored online examinations across several aspects, including test-taking behavior. Their study did not identify any difference in the test-taking behavior of students between the two environments. There was no significant difference between omitted items and not-reached items. Furthermore, with regards to rapid guessing, there was no significant difference. A negligible difference existed for students aged older than thirty-five years, yet gender was a nonsignificant factor.

Scholars have an increasing awareness of the role that test anxiety has in reducing student success in online learning environments (Kolski & Weible, 2018) . The manuscripts identified by the literature scan, identified inconsistencies of results for the effect that examination modalities have on student test anxiety. A study of 69 psychology undergraduates identified that students who typically experienced high anxiety in traditional test environments had lower anxiety levels when completing an online examination (Stowell & Bennett, 2010) . In a quasi-experimental study (n = 38 nursing students), when baseline anxiety is controlled, students in computer-based examinations had higher degrees of test anxiety.

In 34 postgraduate student interviews, only three opposed online assessment based on perceived lack of technical skill (e.g. typing; Attia, 2014) . Around two-thirds of participants identified some form of fear-based on internet disconnection, electricity, slow typing, or family disturbances at home. A 37 participant Community College study used proximal indicators (e.g. lip licking and biting, furrowed eyebrows, and seat squirming) to assess the rate of test anxiety in webcam-based examination proctoring (Kolski & Weible, 2018) . Teacher strategies to reduce anxiety in their students include enabling students to consider, review, and acknowledge their J o u r n a l P r e -p r o o f anxieties (Kolski & Weible, 2018) . Responses such as students writing of their anxiety, or responding to multiple-choice questionnaire on test anxiety, reduced anxiety. Students in the test group and provided anxiety items or expressive writing exercises, performed better (Kumar, 2014).

Cheating was the most prevalent area among all the themes identified. Cheating in asynchronous, objective, and online assessments is argued by some to be at unconscionable levels (Sullivan, 2016) . In one survey, 73.6 percent of students felt it was easier to cheat on online examinations than regular examinations (Aisyah et al., 2018) . This is perhaps because students are monitored in paper and pencil examinations, compared to online examinations where greater control of variables is required to mitigate cheating. Some instructors have used randomized examination batteries to minimize cheating potential through peer-to-peer sharing (Schmidt et al., 2009 ).

Scholars identify various methods for mitigating cheating. Identifying the test taker, preventing examination theft, unauthorized use of textbook/notes, preparing a set-up for online examination, unauthorized student access to a test bank, preventing the use of devices (e.g. phone, Bluetooth, and calculators), limiting access to other people during the examination, equitable access to equipment, identifying computer crashes, inconsistency of method for proctoring (Hearn Moore et al., 2017) . In another, the issue for solving cheating is social as well as technological. While technology is considered the current norm for reducing cheating, these tools have been mostly ineffective (Sullivan, 2016) . Access to multiple question banks through effective quiz design and delivery is a mechanism to reduce the propensity to cheat, by reducing the stakes through multiple delivery attempts (Sullivan, 2016) . Question and answer J o u r n a l P r e -p r o o f randomization, continuous question development, multiple examination versions, open book options, time stamps, and diversity in question formats, sequences, types, and frequency are used to manage the perception and potential for cheating. In the study with MBA students, perception of the ability to cheat seemed to be critical for the development of a safe online examination environment (Sullivan, 2016) .

Dawson (2016) in a review of bring-your-own-device examinations including:

• Copying contents of USB to a hard drive to make a copy of the digital examination available to others,

• Use of a virtual machine to maintain access to standard applications on their device,

• USB keyboard hacks to allow easy access to other documents (e.g. personal notes),

• Modifying software to maintain complete control of their own device, and

• A cold boot attack to maintain a copy of the examination.

The research on cheating has focused mainly on technical challenges (e.g. hardware to support cheating), rather than ethical and social issues (e.g. behavioral development to curb future cheating behaviors. The latter has been researched in more depth in traditional assessment methods (e.g. Wright, 2015) . In a study on Massive Open Online Courses (MOOCs), motivations for students to engage in optional learning stemmed from knowledge, work, convenience, and personal interest (Shapiro et al., 2017) . This provides possible opportunities for future research to consider behavioral elements for responding to cheating, rather than institutional punitive arrangements.

J o u r n a l P r e -p r o o f Schmidt, Ralph and Buskirk (2009) also examined the perceptions of academics with regards to online examination. Academics reported that their biggest concern with using online examinations is the potential for cheating. There was a perception that students may get assistance during an examination. The reliability of the technology is the second more critical concern of academic staff. This includes concerns about internet connectivity as well as computer or software issues. The third concern is related to ease of use, both for the academic and for students. Academics want a system that is easy and quick to create, manage and mark examinations, and students can use with proficient ICT skills (Schmidt et al., 2009) . Furthermore, staff reported in a different study that marking digital work was easier and preferred it over paper examinations because of the reduction in paper (Pagram et al., 2018) . They believe preference should be given to using university machines instead of the student using their computer, mainly due to issues around operating system compatibility and data loss.

Authentication was recognized as a significant issue for examination. Some scholars indicate that the primary reason for requiring physical attendance to proctored examinations is to validate and authenticate the student taking the assessment (Chao et al., 2012) . Importantly, the validity of online proctored examination administration procedures is argued as lower than proctored oncampus examinations (Rios & Liu, 2017) . Most responses to online examinations use bringyour-own-device models where laptops are brought to traditional lecture theatres, use of software on personal devices in any location desired, or use of prescribed devices in a classroom setting.

The primary goal of each is to balance the authentication of students and maintain the integrity and value of achieving learning outcomes.

In a review of current authentication options (AbuMansoor, 2017), the use of fingerprint reading, streaming media, and follow-up identifications were used to authenticate small cohorts of students. Some learning management systems (LMS) have developed subsidiary products (e.g.

Weaver within Moodle) to support authentication processes. Some biometric software uses different levels to authenticate keystrokes for motor controls, stylometry for linguistics, application behavior for semantics, capture to physical or behavioral samples, extraction of Some scholars recommend online synchronous assessments as an alternative to traditional proctored examinations while maintaining the ability to manually authenticate (Chao et al., 2012) . In these assessments: quizzes are designed to test factual knowledge, practice for procedural, essay for conceptual, and oral for metacognitive knowledge. A 'cyber face-to-face' element is required to enable the validation of students.

The interface of a system will impact on whether a student perceives the environment to be an enabler or barrier for online examinations. Abdel Karim and Shukur (2016) summarized the potential interface design features that emerged from a systematic review of the literature on this topic, as shown in Table 2 . The incorporation of navigation tools has also been identified by students and staff as an essential design feature (Rios & Liu, 2017) , as is an auto-save functionality (Pagram et al., 2018). 

None of the studies that included technological problems in its design reported any issues (Böhmer et al., 2018; Matthíasdóttir & Arnalds, 2016; Schmidt et al., 2009) . One study stated J o u r n a l P r e -p r o o f that 5 percent of students reported some problem ranging from a slow system through to the system not working well with the computer operating system, however, the authors stated no technical problems that resulted in the inability to complete the examination were reported (Matthíasdóttir & Arnalds, 2016) . In a separate study, students reported that they would prefer to use university technology to complete the examination due to distrust of the system working with their home computer or laptop operating system or the fear of losing data during the examination (Pagram et al., 2018) . While the study did not report any problems loading on desktop machines, some student laptops from their workplace had firewalls, and as such had to load the system from a USB.

This systematic literature review sought to assess the current state of literature concerning online examinations and its equivalents. For most students, online learning environments created a system more supportive of their wellbeing, personal lives, and learning performance. Staff preferred online examinations for their workload implications and ease of completion, and basic evaluation of print-based examination logistics could identify some substantial ongoing cost savings. Not all staff and students preferred the idea of online test environments, yet studies that considered age and gender identified only negligible differences (Rios & Liu, 2017) .

While the literature on online examinations is growing, there is still a dearth of discussion at the pedagogical and governance levels. Our review and new familiarity with papers led us to point researchers in two principal directions: accreditation and authenticity. We acknowledge that there are many possible pathways to consider, with reference to the consistency of application, the validity and reliability of online examinations, and whether online examinations enable better measurement and greater student success. There are also opportunities to synthesize J o u r n a l P r e -p r o o f online examination literature with other innovative digital pedagogical devices. For example, immersive learning environments (Herrington et al., 2007) , mobile technologies (Jahnke & Liebscher, 2020) ; social media (Giannikas, 2020), and web 2.0 technologies (Bennett et al., 2012) . The literature examined acknowledges key elements of the underlying needs for online examinations from student, academic, and technical perspectives. This has included the need for online examinations need to accessible, need to be able to distinguish a true pass from a true fail, secure, minimize opportunities for cheating, accurately authenticates the student, reduce marking time, and designed to be agile in software or technological failure.

We turn attention now to areas of need in future research, and focus on accreditation and authenticity over these alternates given there is a real need for more research prior to synthesis of knowledge on the latter pathways.

The influence of external accreditation bodies was named frequently and ominously among the sample group, but lacked clarity surrounding exact parameters and expectations. Rios (2017, p. 231) identified a specific measure was used "for accreditation purposes". Hylton et al. (2016, p. 54) specified that the US Department of Education requires "appropriate procedures or technology are implemented" to authentic distance students. Gehringer and Peddycord (2013) empirically found that online/open-web examinations provided more significant data for accreditation. Underlying university decisions to use face-to-face invigilated examination settings is to enable authentication of learning -a requirement of many governing bodies globally. The continual refinement of rules has enabled a degree of assurance that students are who they say they are.

Nevertheless, sophisticated networks have been established globally to support direct student cheating from completing quick assessments and calculators with secret search engine capability through to full completion of a course inclusive of attending on-campus invigilated examinations. The authentication process in invigilated examinations does not typically account for distance students who have a forged student identification card to enable a contract service to complete their examinations. Under the requirement assure authentication of learning, invigilated examinations will require revision to meet contemporary environments. The inclusion of a broader range of big data from keystroke patterns, linguistics analysis, and whole-of-student analytics over a student lifecycle is necessary to identify areas of risk from the institutional perspective. Where a student has a significantly different method of typing or sentence structure, it is necessary to review.

An experimental study on the detection of cheating in a psychology unit found teachers could detect cheating 62 percent of the time (Dawson & Sutherland-Smith, 2017) . Automated algorithms could be used to support the pre-identification of this process, given lecturers and professors are unlikely to be explicitly coding for cheating propensity when grading multiple hundreds of papers on the same topic. Future scholars should be considering the innate differences that exist among test-taking behaviors that could be codified to create pattern recognition software. Even in traditional invigilated examinations, the use of linguistics and handwriting evaluations could be used for cheating identification.

The literature identified in the sample discussed with limited depth the role of authentic assessment in examinations. The evolution of pedagogy and teaching principles (e.g. J o u r n a l P r e -p r o o f constructive alignment; Biggs, 1996) have paved the way for revised approaches to assessment and student learning. In the case of invigilated examinations, universities have been far slower to progress innovative solutions despite growing evidence that students prefer the flexibility and opportunities afforded by digitalizing exams. University commitments to the development of authentic assessment environments will require a radical revision of current examination practice to incorporate real-life learning processes and unstructured problem-solving (Williams & Wong, 2009 ). While traditional examinations may be influenced by financial efficacy, accreditation, and authentication pressures, there are upward pressures from student demand, student success, and student wellbeing to create more authentic learning opportunities.

The online examination setting offers greater connectivity to the kinds of environments graduates will be expected to engage in on a regular basis. The development of time management skills to plan times to complete a fixed time examination is reflected in the business student's need to pitch and present at certain times of the day to corporate stakeholders, or a dentist maintaining a specific time allotment for the extraction of a tooth. The completion of a selfregulated task online with tangible performance outcomes is reflected in many roles from lawyer briefs on time-sensitive court cases to high school teacher completions of student reports at the end of a calendar year. Future practitioner implementation and evaluation should be focused on embedding authenticity into the examination setting, and future researchers should seek to understand better the parameters by which online examinations can create authentic learning experiences for students. In some cases, the inclusion of examinations may not be appropriate; and in these cases, they should be progressively extracted from the curriculum.

J o u r n a l P r e -p r o o f

As institutions begin to provide higher learning flexibility to students with digital and blended offerings, there is scholarly need to consider the efficacy of the examination environment associated with these settings. Home computers and high-speed internet are becoming commonplace (Rainie & Horrigan, 2005) , recognizing that such an assumption has implications for student equity. As Warschauer (2007, p. 41) puts it, "the future of learning is digital". Our ability as educators will be in seeking to understand how we can create high impact learning opportunities while responding to an era of digitalization. Research considering digital fluency in students will be pivotal (Crawford & Butler-Henderson, 2020) . Important too, is the scholarly imperative to examine the implementation barriers and successes associated with online examinations in higher education institutions given the lack of clear cross-institutional case studies. There is also a symbiotic question that requires addressing by scholars in our field, beginning with understanding how online examinations can enable higher education, and likewise how higher education can shape and inform the implementation and delivery of online examinations.

This study adopted a rigorous PRISMA method for preliminary identification of papers for inclusion, the MMAT protocol for identifying the quality of papers, and an inductive thematic analysis for analyzing papers included. These processes respond directly to limitations of subjectivity and assurance of breadth and depth of literature. However, the systematic literature review method limits the papers included by the search criteria used. While we opted for a broad set of terms, it is possible we missed papers that would typically have been identified in other 

This systematic literature review considered the contemporary literature on online examinations and their equivalents. We discussed student, staff, and technological research as it was identified in our sample. The dominant focus of the literature is still oriented on preliminary evaluations of implementation. These include what processes changed at a technological level, and how students and staff rated their preferences. There were some early attempts to explore the effect of J o u r n a l P r e -p r o o f online examinations on student wellbeing and student performance, along with how the changes affect the ability for staff to achieve.

Higher education needs this succinct summary of the literature on online examinations to understand the barriers and how they can be overcome, encouraging greater uptake of online examinations in tertiary education. One of the largest barriers is perceptions of using online examinations. Once students have experienced online examinations, there is a preference for this format due to its ease of use. The literature reported student performance did not have significant difference in final examination scores across online and traditional examination modalities.

Student anxiety decreased once they had used the online examination software. This information needs to be provided to students to change students' perceptions and decrease anxiety when implementing an online examination system. Similarly, the information summarized in this paper needs to be provided to staff, such as the data related to cheating, reliability of the technology, ease of use, and reduction in time for establishing and marking examinations. When selecting a system, institutions should seek one that includes biometrics with a high level of precision, such as user authentication, and movement, sound, and keystroke monitoring (reporting deviations so the recording can be reviewed). These features reduce the need for online examinations to be invigilated. Other system features should include locking the system or browser, cloud-based technology so local updates are not required, and an interface design that makes using the online examination intuitive. Institutions should also consider how it will address technological failures and digital disparities, such as literacy and access to technology.

We recognize the need for substantially more evidence surrounding the postimplementation stages of online examinations. The current use of online examinations across disciplines, institutions, and countries needs to be examined to understand the successes and gaps.

Beyond questions of 'do students prefer online or on-campus exams', serious questions of how student mental wellbeing, employability, and achievement of learning outcomes can be improved as a result of an online examination pedagogy is critical. In conjunction is the need to break down the facets and types of digitally enhanced examinations (e.g. online, e-examination , BYOD examinations, and similar) and compare each of these for their respective efficacy in enabling student success against institutional implications. While this paper was only able to capture the literature that does exist, we believe the next stage of literature needs to consider broader implications than immediate student perceptions toward the achievement of institutional strategic imperatives that may include student wellbeing, student success, student retention, financial viability, staff enrichment, and student employability.

Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

J o u r n a l P r e -p r o o f 

Proposed features of an online examination interface design and its optimal values

Proposed bio-authentication system for question bank in learning management systems

Development of continuous authentication system on android-based online exam application

Developing a new e-exam platform to enhance the university academic examinations: The case of Lebanese French University

The effect of using electronic exams on students' achievement and test takers' motivation in an English 101 course

The law and the outlaw: is legal prohibition a viable solution to the contract cheating problem

Online course evaluations: Review of literature and a pilot study

Online exams for blended assessment. Study of different application methodologies

Postgraduate students' perceptions toward online assessment: The case of the faculty of education, Umm Al-Qura university

Education for a Knowledge Society in Arabian Gulf Countries

Implementing Web 2.0 technologies in higher education: A collective case study

Enhancing teaching through constructive alignment

E-exams in engineering education-online testing of engineering competencies: Experiences and lessons learned

A comparison of online vs. proctored final exams in online classes

Financial aid and attainment among students in a state with changing demographics

Exploring the influence of facial verification software on student academic performance in online learning environments

Changing students, faculty, and institutions in the twenty-first century

Impeding students' efforts to cheat in online classes

Immersive learning technologies: Realism and online authentic learning

The mixed methods appraisal tool (MMAT) version 2018 for information professionals and researchers. Education for Information

Utilizing webcam-based proctoring to deter misconduct in online exams

Three types of integrated course designs for using mobile technologies to support creativity in higher education

History of New York State Regents Exams

College student cheating: The role of motivation, perceived norms, attitudes, and knowledge of institutional policy

Exams in algebra in Russia: Toward a history of high stakes testing

A decade of adoption: How the internet has woven itself into American life

University reform in the post-massification era in Japan: analysis of government education policy for the 21st century

Situational and personal causes of student cheating

Online proctored versus unproctored low-stakes internet test administration: Is there differential test-taking behavior and performance?

Student Verification System for Online Assessments: Bolstering Quality and Integrity of Distance Learning

Utilizing online exams: A case study

Have college students' attitudes toward older people changed

Students' perceptions of BYOD open-book examinations in a large class: a pilot study

An informal history of how law schools evaluate students, with a predictable emphasis on law school final exams

E-exams and exam process improvement

Use of e-exams in high stakes law school examinations: student and staff reactions

Effects of online testing on student exam performance and test anxiety

An integrated approach to preempt cheating on asynchronous, objective, online assessments in graduate business classes

Objective structured clinical exams: a critical review

Medical subject headings

Online and print inclusive design and legibility considerations

The paradoxical future of digital learning

A pilot study of an electronic exam system at an Australian University

The efficacy of final examinations: A comparative study of closed-book, invigilated exams and open-book, open-web exams

Distinguished Scholar Invited Essay: Reflections on the role of character in business education and student leadership development

The research and design of online examination system