All Collections
Proctoring Tests
Preventing cheating with ChatGPT during Tests
Preventing cheating with ChatGPT during Tests
Priyanka Khandagale avatar
Written by Priyanka Khandagale
Updated over a week ago

WeCP is committed to ensuring a fair and unbiased testing environment for candidates. To maintain the integrity of our assessments, we have implemented several measures to prevent candidates from seeking assistance from external sources, including ChatGPT. This article aims to provide an overview of the steps we take to ensure the authenticity of test results and maintain a level playing field for all participants.

  1. Secure Browser Environment: WeCP utilizes a secure browser environment for conducting tests. This environment restricts access to external websites and applications, including ChatGPT. Candidates are unable to open additional tabs or windows during the assessment, ensuring that they cannot seek assistance or engage in any form of cheating.

  2. Automatic Clipboard Deletion: To prevent candidates from copying questions or answers or relevant information from external sources, including ChatGPT, WeCP has implemented a feature that automatically deletes the clipboard contents during the test. This ensures that candidates cannot copy answers or cheat by pasting information obtained from unauthorized sources. By eliminating the possibility of using the clipboard to gain an unfair advantage, we further reinforce the integrity of our assessments and maintain a level playing field for all candidates.

  3. Keystroke Analysis and Plagiarism Detection: WeCP employs advanced technology to analyze candidates' keystrokes and detect patterns that may indicate potential cheating. This analysis helps identify instances where a candidate may be copying or pasting information from external sources, including ChatGPT. Our plagiarism detection algorithms compare candidate responses against a vast database of known sources to ensure the authenticity of their answers.

  4. Auto-stopping of the test: In case the candidates goes out of the test window more than allowed number of times, the system automatically stops the test for the candidate.

  5. Real-Time Proctoring: WeCP employs real-time proctoring to monitor candidates throughout the duration of the test. Trained proctors use video and audio feeds to observe candidates' activities and detect any suspicious behavior. This constant monitoring acts as a deterrent and reduces the likelihood of candidates attempting to use ChatGPT or any other external resources.

  6. Randomized Question Ordering: To further prevent candidates from seeking help from ChatGPT, WeCP randomizes the order in which questions appear on the test. This approach ensures that candidates cannot collaborate or receive assistance from external sources by sharing specific question details with others before or during the assessment.

  7. Time Limit and Question Difficulty: WeCP sets time limits for each test, which creates a sense of urgency for candidates to complete the assessment within the allotted timeframe. Additionally, our assessments are carefully designed to include questions of varying difficulty levels, ensuring that the test accurately assesses candidates' skills and knowledge. This approach minimizes the reliance on external sources and discourages the use of ChatGPT for seeking help.

Conclusion: Though our system cannot stop a candidate use ChatGPT from external means such as mobile phone etc, we're constantly discovering ways to capture instances of cheating to improve our system. WeCP places emphasis on maintaining the integrity of our assessments and preventing candidates from seeking external assistance, including ChatGPT, during tests. Through a combination of secure browser environments, real-time proctoring, keystroke analysis, plagiarism detection, randomized question ordering, time limits, and question difficulty levels, we create a fair and unbiased testing environment. These measures work together to ensure that candidates' skills and abilities are accurately evaluated, providing organizations with reliable insights into the candidates' capabilities.

Did this answer your question?