Skip to main content
↑

Proctoring Best Practices Guide

  • Updated
Hide
In this article

Overview

This guide has been developed to assist you and your organization with best practices for the various proctoring settings for your HackerRank assessments.

  1. Enable the copy/paste option to track and see the content copy-pasted from external sources. The Copy-Paste Frequency column in the excel report helps you to track the number of times a Candidate tries to copy paste a piece of code.

  2. Enable tab proctoring to track if the candidate has gone out of the tab and looked up some resources online.

  3. Enable Image Proctoring will force the candidate to keep the camera on for the entire test and the photos can help you identify suspicious activities including:
    • Any other person sitting with the candidate.
    • The candidate is looking through their phone or other gadgets.
    • A candidate is looking away from the laptop/desktop screen most of the time.
    • The candidate is not on screen (left the room) for a long time.

  4. Image Proctoring and Photo Identification options also help you verify if the right candidate has attempted the test.
  5. Avoid adding leaked questions to the test. Prior to inviting candidates to take a test, it is recommended to check over the test to make sure there are no leaked questions present. If there are leaked questions, it's highly recommended to remove and replace those questions. Want to learn more about the leaked questions indicator? Click Click here.
    • If questions have been leaked after the test has been attempted by the candidates:
    • Do not consider that for scoring or Look for plagiarism and copy-pasted content in the candidate’s code.

  6. One of the best ways to identify if the candidates have shared code snippets with each other is through our plagiarism detection mechanism. A detailed explanation can be found in the links below.
  7. Since MCQs or Sentence Completion question types have a single definite answer and it is much easier for candidates to share answers, we recommend the following;
    • Use sections while creating tests and add randomization by allowing the system to pick a set of different questions for each candidate. Since each candidate may get different questions, it is difficult for the candidates to split the test questions among themselves and share answers. Refer here - Section based testing (Creating test sections and how to use them) 
    • Enable the “shuffle questions” option, which would add another layer of randomization to the ordering of questions as well so that they cannot rely on question numbers for getting the answers. Refer here - Shuffle Questions and Question Points

    • Hide Question Title lets you disallow candidates from using a question title as a prompt on search engines to find answers to screening questions. While the actual question tiles are hidden from the candidates, they will be replaced with a generic title like Question 01, Question 02, etc. (Learn more in the ConfiguringTest Integrity Settings support article)

  8. If you still feel that the candidate might not have attempted the test in a genuine way, we would recommend that the candidate be asked to explain the approach and maybe even modify the code a bit during the live interview session. 

Related articles