How to screen candidates with online tests

Screening candidates with online tests means sending applicants a short, standardized test early in the hiring process and using their scores to decide who moves forward, instead of relying only on a resume. It works because everyone answers the same questions under the same conditions, which makes the results comparable in a way resumes rarely are. This guide walks through where screening tests fit in your funnel, what to test for, and how to run the process without accidentally filtering out strong candidates. For the fuller picture of building an assessment program end to end, see our complete guide to online assessment for recruitment.

Why Screen Candidates With Online Tests (Instead of Resumes Alone)

A resume tells you what someone claims they've done. A screening test tells you what they can actually do, at least on the narrow set of things the test measures. Neither replaces the other - they answer different questions.

The practical case for adding a test is simple: resumes take a recruiter a few seconds each to skim, and that's not enough time to catch someone who's overstated a skill or undersold themselves. A short test applied consistently across every applicant surfaces a signal that a five-second scan can't. If you're weighing the two approaches directly, our breakdown of online assessment vs resume screening covers the trade-offs in more depth.

Where Screening Tests Fit in Your Hiring Funnel

Screening tests belong early, right after someone applies, before a recruiter spends time on a phone call. That ordering matters more than most teams realize.

A typical funnel looks like this:

• Application submitted
• Online screening test sent automatically
• Candidates above the pass mark move to a recruiter screen or phone call
• Shortlisted candidates go to hiring-manager interviews
• Final round and offer

Put a screening test any later than step 2, and you've already spent recruiter time on people the test would have filtered out anyway. That's the whole point of placing it where it sits - it does the sorting before a human has to.

What to Test For at the Screening Stage

Screening tests should be narrow on purpose. This isn't the stage for a deep technical exam or a personality battery - it's the stage for confirming baseline competence fast. For more on how this differs from a fuller evaluation later in the process, see what is a pre-employment assessment.

Most screening tests measure one or two of the following:

• Role-specific knowledge

  - terminology, tools, or concepts someone in the role would already know

• Basic aptitude

  - numerical reasoning, logical reasoning, or attention to detail, depending on the role

• Core hard skills

  - a short coding problem, a writing sample, or a spreadsheet task, kept intentionally brief

For technical roles specifically, the test often needs to go further than multiple choice - our guide on technical assessment for developers covers that case in detail. At the screening stage for most other roles, though, keep it short: 15-20 minutes is usually enough to get a usable signal without asking too much of someone who hasn't been interviewed yet.

Step-by-Step: How to Screen Candidates With Online Tests

Here's the process, in order:

• Define what "pass" looks like before you write a single question

  - Decide what skill or knowledge gap would actually disqualify someone for this role. If you can't name it, the test will end up testing the wrong thing.

• Pick a test format that matches the role

  - Multiple choice works for knowledge and aptitude checks. Written-answer or short technical tasks work better when you need to see reasoning, not just a final answer.

• Set a time limit and a pass mark

  - A tight timer keeps the test a genuine screen rather than something a candidate can research their way through. TunnelQuiz lets you set timers and pass marks per test, along with negative marking if guessing is a concern for a particular role.

• Send the test automatically after application

  - Manually emailing tests to each applicant defeats the purpose of screening at scale - automate the send so no candidate waits on a recruiter's schedule.

• Let scores do the first sort

  - Route everyone above the pass mark to the next stage and everyone below it to a polite rejection, without a human re-litigating each score.

• Spot-check the top scorers before the interview stage

  - A quick resume glance at your top 10-15% catches any obvious mismatch the test wouldn't know to flag, like a location or visa requirement.

• Review the test itself every few hiring cycles

  - If almost everyone passes or almost everyone fails, the test isn't discriminating between candidates and needs adjusting.

Reusable templates make step 7 far less painful - building the test once and reusing it across every opening for that role, rather than rebuilding it each time a requisition opens, is one of the more overlooked ways teams save time here.

Setting Pass Marks and Cutoffs Without Losing Good Candidates

A pass mark set too high filters out candidates who'd have been fine on the job but had a rough 15 minutes. Set too low, and the test stops doing its job.

A few things help get this right:

• Pilot the test on a small batch of applicants before rolling it out fully, and check whether your current strong performers would have passed it

• Avoid a single hard cutoff for borderline scores

  - a narrow band just below the pass mark can go to a quick manual review instead of an automatic rejection
• Revisit the pass mark whenever the role's requirements shift, rather than leaving it fixed indefinitely

• Watch for a pass rate that's near 100% or near 0%

  - both suggest the cutoff, not the candidates, is the problem

Keeping Screening Tests Fair and Cheat-Resistant

An online screening test is only useful if the score reflects the candidate's own ability. Unproctored, untimed tests taken from home are vulnerable to a second person helping, an open second monitor with answers pulled up, or simply unlimited time to search for solutions.

A few safeguards go a long way without turning a screening test into a heavily monitored exam:

• Shuffling question and option order so candidates can't easily compare answers with each other
• Tight, enforced time limits that make outside research impractical
• Basic integrity checks - tab/focus alerts and second-screen detection are two that TunnelQuiz includes as part of its browser-based test environment, without requiring any software install on the candidate's end

For a fuller look at this problem, including AI-assisted cheating specifically, see our guide on how to reduce cheating in remote hiring tests.

Common Mistakes When Screening With Online Tests

A few patterns show up repeatedly in screening programs that don't work well:

• Making the test too long

  - A screening test that takes an hour will get abandoned by strong candidates who have other offers in progress.

• Testing for things the job doesn't actually require

  - Every question should trace back to something the role genuinely needs.

• Never revisiting the pass mark

  - A cutoff set for one hiring market can quietly stop working as the candidate pool changes.

• Treating the score as the entire decision

  - A screening test narrows the pool - it shouldn't be the only input that decides who gets an offer.

• Skipping a trial run

  - Testing the test on internal team members first catches confusing wording before it costs you good candidates.

The Bottom Line

Screening candidates with online tests works when the test is short, the pass mark is deliberate, and the score is treated as one signal among several, not the whole decision. Get those three things right, and the rest of the process mostly takes care of itself. If you're setting up a screening test for the first time, TunnelQuiz's timers, pass marks, and reusable templates make it straightforward to build once and reuse across every opening for that role.