Today, we’re beginning to roll out Insights for Tests/Questions on HackerRank for Work to help you understand how your tests and questions are performing.
Thousands of companies who’re looking for technical talent use HackerRank Tests to screen candidates effectively and make their recruitment process more streamlined and cost effective. Now for the first time you can gain insight into how your Tests and Questions are performing and how you can improve them over time.
Test Insights
Go to any HackerRank test and you’ll find a link to Test Insights in the left pane. Note that there’s also a link to Question Insights for each question added to the test.
The first thing you’ll notice is aggregate test statistics on the top right of the page. These include % invited candidates who end up attempting the test, median attempt time, median attempt time to full score and median score. Next, you’ll see a distribution of candidate scores. This gives a very good picture about the difficulty of the test.
Next, you go one step deeper and look at question-level statistics. This helps you understand
- How candidates are dividing their time across questions and whether the time limit for the test is sufficient or not
- Whether the score of the some questions is out of proportion to the time candidates take to solve them and you should adjust question scores
Last, you get to see the conversion funnel of candidates i.e. % of candidates who qualify the screening round.
Question Analytics
To check out the Insights for questions, you can open the question list of a test (as mentioned earlier) or go to Library.
The aggregate statistics, similar to the Test Insights, give a good idea about the difficulty of the question and the score distribution further gives a breakdown by candidates scores.
For coding questions, you see a very interesting breakdown by programming language. This helps you understand
- What languages are most commonly used by candidates to answer the question?
- Is the question relatively easier to solve in some languages than others? Some languages might give candidates an unfair advantage to solve a question much easily, which you might not want to allow going forward
Next, you see the usage of question over time. This tells you how long the question has been out there and whether it’s time to stop using it as it might already be popularly known to candidates.
We believe this is just the tip of the iceberg when it comes to empowering you to make your recruitment process more data-driven. Providing more insights around your HackerRank account usage, quantified returns of using HackerRank, industry-wide stats (including comparison to peers) and many more related things are among our top priorities today. We would love to hear your feedback on the Test/Question Insights and more about what insights will help your organisation.
It would be awesome!