While education was inexorably moving online due to its many advantages, the benefits of online education could only be gained if the examinations too were conducted online. But conducting examinations online poses two major challenges – impersonation and excessive use of cheating aids - that thwart the integrity of the test outcomes.
DigiProctor’s path breaking technology ensures the
Highest Levels of Test Integrity.
Since students appear in online exams from distant locations, it is very difficult to verify who is actually appearing for the exam - someone else (a proxy) could appear for the exam on behalf of the legitimate student.
Students appearing in online exams from distant locations can seek help for answering questions in their exam from others or they may browse the Internet for answers, refer to books, cheat sheets etc. to answer questions.
Identity verification is carried out before students are allowed to take the test. It is aimed at stopping a proxy or impersonator from appearing for the test on behalf of the legitimate candidate.
Most online testing application use automated ID validation tools before the students enter their online exam. This activity is initiated before the exam begins. However, an impersonator can replace the student (whose ID has been validated) at any time during the test. Online Identity Verification needs to be a continuous activity – from when the test taker enters the exam till the time it is submitted to detect and check the impersonator appearing for the exam
How Online ID Verification Works?
Identity Verification is done with an AI-assisted Face Map. Test Takers’ identity can be verified using their headshot, ID card, or any other official identity card like Aadhaar, driving license, etc.
For online identity verification, headshots of students being verified and images of their ID cards are clicked using their computer’s camera.
Digital ID Verification
During the verification process, test taker’s headshot and images of their ID cards are sent to an AI face recognition engine, which automatically compares it with the student’s registered photo. After comparison, the AI- engine sends back a confidence level or percentage of match between the images, which is displayed on the proctor’s computer. AI engines are designed to auto-authenticate the students if their confidence levels are high and deter their test entry if the face map fails.
The process of AI-Assisted ID Verification is the same. The face-map confidence level is computed and displayed to aid and supplement remote proctors during the ID verification process. The images captured by the camera are sent to and displayed on the remote proctors' computer screens, where they are compared to verify the match. Only if the images match will the proctors (manually) approve the verification and allow students to enter their test.
Student Count Is Large
AI Assisted ID Verification mode is normally used when the number of students taking the test is very large, and number of proctors available to manually verify them is limited.
Examination Stake Is High
AI-Assisted Manual Verification mode is typically used in high stake exams where the number of students taking the test is less, and enough proctors are available to manually verify them.
Large Student Count For High Stake Exams
For high-stakes exams, the examination body of institutions mandates that a human proctor validates the test taker's identity. However, if the student count is high, this process may take some time. To avoid delays (in case of high volumes), DigiProctor's unique 'ID Verification Queue' feature allows students to auto-authenticate themselves to enter the test.
These test takers are visible to the proctor in the ID Verification Queue. Proctors can verify the identity of students manually during the examination. If the online ID verification fails the student will be disqualified and made to exit the test, they can no longer sit for the exam.
There is risk of students creating photoshopped versions of their identity cards with their headshot image on it replaced with that of an impersonator’s, to fool proctors during the identity verification.
To allay this risk, copies of the official photos of candidates held in records of the institution are uploaded into DigiProctor prior to the exam. While digitally verifying a candidate’s headshot/identity card image, the candidate’s official photo is also displayed on the proctors screen, which makes it easy for proctors to detect frauds.
Continuous Identity verification must be done to mitigate the risk of students getting their identity verified before the test and then getting someone else to appear for the test after the verification.
Most digital ID verification tools validate the user's identity only at the entry-level. To avoid such frauds, DigiProctor does continuous online identity verification during the test. For this, headshot images of students taking the exam are clicked in the background at periodic intervals and passed through the AI engine for face match to ensure that the person taking the test is legitimate and not an impersonator.
The students are permitted to start their test only after identity verification. If for any reason there is delay at the remote proctors’ end to manually verify the students (for example, due to high volumes), and if the test has already started, DigiProctor can be configured to allow the student to start the test even though not verified.
Such students are placed by DigiProctor in a verification queue and eventually when the proctor is available, the student would go through the verification process. If at that stage the student’s verification fails, the student will be exited out of the test automatically.
Select from Multiple Modes of ID Verification
Only AI-Verification: Automates students’ entry into the exams
AI-assisted Manual ID Verification: Students’ ID verified by Human Approver for High stake exams
Continuous Identity Validation: Prevents imposters from sitting for the exam on behalf of legitimate student
Are you conducting Online Exams and worried about the Test Results' Credibility?
It's time to add a layer of Digital Proctoring.