Guide

Section-wise vs whole-test timers: which to use

One deadline for the whole paper, or a deadline per section. It looks like a settings detail and is actually a decision about what the exam measures.

Whole-test timing

One clock for the attempt. Candidates spend their time wherever they choose — skip the hard section, bank easy marks, come back if there is time.

This measures total performance under time pressure *including the candidate’s own time allocation*. That is a real skill and often the one you want. It is also why a strong candidate can score well while never opening a section they find hard, which is fine if you only care about the total and a problem if you care about the parts.

Section-wise timing

Each section gets its own deadline, and the total time is their sum. Time left over in one section does not carry into the next, and an expired section auto-submits.

This measures each competency separately, and it is what you need when a section has a minimum standard of its own, when your paper mirrors a real exam that works this way, or when a batch keeps ignoring one section entirely.

Choosing

Use whole-test when: The total is the outcome, triage is part of the skill, or the paper is short enough that section boundaries are artificial.
Use section-wise when: You will report or gate on per-section performance, you are mirroring an exam that is section-timed, or you want to force engagement with every part.
Watch the navigation policy: Section timing interacts with whether candidates can jump between sections, return to previous questions, or read a section after it expires. Section-wise timing with free jumping is a confusing combination — pick one discipline.

Decide before the first candidate starts

On this platform, the timing model and section durations lock the first time a candidate starts or resumes an attempt. Everything else stays editable; changing the timing model means duplicating the test into a fresh draft.

That is deliberate. Changing the clock mid-series would mean two cohorts sat measurably different exams under one test record, and every comparison across them would be quietly wrong. Run a small pilot with your own team before a batch sits it.

Whichever you pick, the clock is the server’s

Both models anchor to the server-recorded start of the attempt. The countdown a candidate sees is a display; the deadline is computed server-side, so a refresh, a dropped connection or a changed system clock does not extend anything. On expiry, whatever has been answered is submitted.

Common questions

Can unused time in one section carry over to the next?

No. Under section-wise timing each section has its own deadline and the total is their sum; time left over does not roll forward. If you want candidates to allocate time freely, use whole-test timing.

Can I switch timing model after candidates have started?

Not on the same test. The timing configuration locks on the first attempt so in-flight and historical results stay comparable. Duplicate the test into a fresh draft to change it.