Some insights on pre-interview screening tests?

(Would appreciate personal experiences particularly from the employers’ side.)

  • I read somewhere that you generally need to score perfectly before being considered for an interview. Is this consistent with your experiences? Is this true even for entry level jobs?

  • I noticed some employers test you with practical tasks (e.g. write a function that set up event listeners, or debugging some code), and others test purely for your algorithmic skills (e.g. write an efficient function to solve a math problem). What could be inferred about an employer’s hiring intentions based on the kind of tests they use?

  • I read somewhere that someone passed some test cases before submitting their solution but upon submission, they failed the same test cases which passed the solution moments ago. In this person’s case obviously something is wrong, but… should we generally trust that the tests are reliable? If we suspect that a test question is bugged but don’t have solid proof, would it be a bad idea to try to bring it to the hirer’s attention before actually getting an interview?

  • What do programmers generally think of these screening tests? Welcome, hate, or necessary evil?

(Not an employer).

As with most of the questions, the answer is “it depends”.

If it’s something like Google, FB, Amazon etc. (usually abbreviated as FAANG), then I’d say you need to score perfectly, because the competition is fierce (go to reddit’s cscareerquestions and see for yourself).

My personal experience (in Europe, not applying to FAANG) those initial tests are to filter out people who can’t code at all, they’re not too hard (think FizzBuzz level hard) and they are not too common. More common are take home assignments where you’re given requirements and you have to implement them.

If it’s something like codility etc., then yes.

I personally not a fan of those 90+ minutes of algorithmic tests, but I’m ok with reasonable take home assignments or short tests to check my basic knowledge.


I was going to write a long answer going through each point, but then I realized that the answer to each is, “Could be yes, could be no, and everything in between in every possible combination. Every employer is different.”

For example, yes some will want a perfect test score. Some will write a test that is harder than they need to be because they want to see how you work.

As far as your test case question, yes sometimes they give you a few generalized test cases to help you make sure you have the basics. But then they might have other secret tests that hit every edge case they can think of - they want to see if you thought of and accounted for those edge cases. This is fairly common on algorithm testing sites, too.

Yeah, it sucks sometimes, but it’s just part of the dance.


Pre-interview tests and assignments vary wildly. Some places don’t use them at all. Some places rely on third party testing services. Some places expect you to spend an hour, other places expect you to spend 8 hours.

Programmers also differ in their opinions of them. Personally, I refuse to do them.


This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.