Ensuring Fairness in Online English Language Assessment
- May 1
- 6 min read
How VTest Minimizes the Impact of Technology on Candidate Performance

Online English assessments offer many advantages over traditional paper-based formats. They provide broader access, greater flexibility in scheduling, and faster delivery of results.
However, online testing also introduces new challenges. The quality of the device used, the stability of the internet connection, the platform's user-friendliness, the remote monitoring methods, and even the proper functioning of the microphone can all influence the candidate's experience. These factors are not related to their actual level of English.
When technology interferes with performance, test results may no longer accurately reflect a candidate’s language ability. In assessment, this is known as technology-related irrelevant variance: it refers to score differences caused by the technological environment, rather than by the skill the test is intended to measure.
Ensuring fairness therefore requires more than well-designed test questions. It also involves securing testing conditions, preparing candidates for the digital environment, preventing technical issues, and implementing appropriate solutions when problems arise.
At Burlington VTest, this requirement is an integral part of the design of our online assessments.
Why Technology can Compromise the Fairness of an Online Test
In a paper-based test, the testing conditions are relatively uniform. Online, the situation is different. Candidates may use a variety of equipment, take the test in very different contexts, and encounter unforeseen technical difficulties.
These discrepancies can create inequalities between candidates, even when they have a comparable level of English. An unstable connection, an audio problem, a poor understanding of the interface, or an interruption during the test can artificially alter performance.
In this case, the score obtained no longer measures solely English proficiency. It also reflects, in part, the quality of the technical environment. This raises a direct issue of reliability, validity, and above all, fairness.
The VTest approach: Minimizing the Impact of Technological Factors
To mitigate these risks, VTest has implemented a coherent set of preventive and corrective measures. The objective is simple: to ensure that technology remains a tool for test administration, and not a factor that distorts the results.
1. Technical checks before the start of the exam
Before any test can be administered, VTest requires a complete technical check. This step ensures that the device used, the browser, the audio output, the microphone, and the internet connection meet the minimum requirements of the test.
The exam can only start when all the required elements are functioning correctly. If the candidate leaves the test and returns later, a new check is automatically initiated. This allows for the identification of any changes in the technical environment, such as a modification of settings, a change of device, or a degradation of the connection.
This procedure greatly reduces the risk of avoidable technical problems arising during the test.
2. Free resources to familiarize yourself with the test
One of the most frequent sources of difficulty in online testing is not language, but the interface. A candidate who is uncomfortable with digital navigation, on-screen reading, or the general functioning of an online test may be penalized, even if they possess the expected language skills.
To mitigate this effect, VTest provides free and unlimited familiarization resources. Candidates can view sample activities, learn how the interface works, understand response formats, watch explanatory videos, and access preparation guides.
This anticipation helps to limit performance gaps related to digital skills rather than English proficiency.
3. A policy for re-evaluation in case of actual disruption
Despite all precautions taken, some incidents may occur: internet outage, equipment failure, technical malfunction or significant external disruption.
When this type of situation prevents the candidate from correctly completing part of the test or clearly affects the quality of their response, VTest may allow a targeted retake of the relevant section, rather than the entire exam.
This policy applies only when the incident is real, verifiable, and documented using system data, records, or other available evidence. It protects candidates from unfair disadvantages while preserving the integrity of the assessment.
4. Automatic saving of responses
A technical problem should not erase the work already completed. That's why VTest automatically saves answers during administration.
For listening and reading comprehension activities, answers are saved at the item level. For speaking and writing activities, saving occurs at the task or relevant section level.
If an interruption occurs, the candidate can resume where they left off, without losing any answers already submitted. This feature reduces the risk of a score being penalized by a purely technical failure.
5. Secure loading of test content
Online language tests often rely on essential digital resources: audio files, images, instructions, or visual elements. If this content does not load correctly, the candidate may find themselves faced with an incomplete or unusable task.
To prevent this, VTest checks that resources are loaded correctly before each step. If a failure occurs, the system automatically attempts to reload the missing elements up to three times before displaying an error.
This mechanism improves the stability of the transfer experience, especially when the connection is fluctuating.
6. An interface designed to reduce unnecessary cognitive load
An assessment platform should not add unnecessary stress. Abrupt transitions, overly automated navigation, or unclear instructions can increase anxiety and distract the candidate.
The VTest interface was therefore designed to offer a clear and predictable pace. In most cases, the candidate progresses independently using a clearly marked button. For certain tasks, particularly oral or written expression, an optional preparation time is offered to allow for a better understanding of the instructions.
When certain automatic transitions are necessary to ensure standardization, they are clearly announced in advance.
7. Human supervision of AI-assisted processes
Automated systems offer speed and consistency, but they should never operate without oversight. Technical anomalies, poor-quality recordings, or atypical situations can skew a purely automated analysis.
This is why VTest relies on a human oversight approach. Professionals review cases where anomalies are detected or reported. This may include analyzing audio or video recordings, transcripts, or system data.
In addition, human support is available before, during and after the exam to help candidates in case of technical difficulties or questions related to the procedure.
8. Actively listening to candidate feedback
Not all technical problems are visible in the system logs. Therefore, the experience of the candidates is a valuable source of information.
VTest collects feedback through post-exam satisfaction surveys. These questionnaires help identify any recurring problems related to navigation, audio quality, understanding of instructions, or platform stability.
This data feeds into the continuous improvement of the system and helps to strengthen the overall fairness of procurement.
9. Multilingual instructions to avoid procedural misunderstandings
In an English assessment, the difficulties encountered by a candidate must not stem from a poor understanding of navigation or operating instructions.
VTest therefore provides, where relevant, procedural instructions and explanations in local languages. This includes, in particular, the test administration rules, the use of the interface, time management, and interactions with the system.
The content being assessed remains unchanged. This distinction is essential: the aim is not to simplify the test, but to reduce unnecessary obstacles related to the digital environment.
10. Accessibility across multiple device types
Not all candidates have access to the same equipment. Requiring specific equipment can exclude or disadvantage certain profiles, particularly young people, adults returning to studies, or candidates from less privileged backgrounds.
VTest therefore allows access to its exams on a wide range of compatible devices. However, depending on the type of test and the skills assessed, certain recommendations or limitations may apply in order to maintain the comparability of results.
This approach aims for a balance between accessibility and quality requirements.
11. The possibility of temporarily leaving the test under certain conditions
In paper-based exams, structured breaks or temporary exits are often possible. Online, candidates are sometimes required to remain connected without interruption, under penalty of losing their progress or their session.
This rigidity can increase fatigue, stress, or physical discomfort, especially during lengthy assessments. VTest therefore allows, according to the exam rules, for temporary departures from the session, enabling participants to resume later.
This feature relies on automatic backup, system checks upon return, and verification of the technical environment. It contributes to a more human and equitable handover.
A design conceived for a diverse pool of candidates
Candidates taking a VTest test do not constitute a homogeneous group. They may be young learners, adults in continuing education, people undergoing career changes, highly qualified candidates, or conversely, those with little familiarity with digital environments.
This diversity implies a particular responsibility. Designing a fair online test is not just about preventing failures. It is also about anticipating disparities in digital access, technological confidence, and experience with formal assessments.
At VTest, the principle is clear: technology should never become a hidden selection filter .
A requirement for fairness that is as much ethical as it is technical
The results of an English test can have significant consequences. They can play a role in a study project, a job application, career advancement, international mobility, or skills recognition.
In this context, even a slight technological bias can have disproportionate effects. Ensuring fairness is therefore not only a technical requirement: it is also an ethical responsibility.
This requires finding the right balance between security, standardization, accessibility and candidate experience.
Conclusion
Online assessment offers real opportunities, but it requires heightened vigilance. To be fair, a test must not only accurately measure language proficiency, but it must also neutralize, as much as possible, external factors that could artificially influence performance.
Through upstream technical checks, familiarization resources, automatic saving, structured retake procedures, human supervision and continuous improvement based on candidate feedback, VTest implements a comprehensive approach to reduce irrelevant variance related to technology.
The objective is constant: to ensure that the results reflect actual English skills, and not the quality of the material or the vagaries of the technical context.
