Does Interview Time of Day Actually Affect Your Score? What a 104,000-Exam Study Found

The Study: 104,552 Oral Exams Found a Clear Midday Peak

Published in Frontiers in Psychology in July 2025, a research team led by Carmelo Vicario at the University of Messina pulled archival records from their university's academic platform: 104,552 in-person oral exam assessments, given by 680 different examiners across 1,243 courses and 78 degree programs, covering 19,116 students between October 2018 and February 2020 — a window chosen specifically to end before COVID pushed exams online.

Every exam had a recorded start time between 8 a.m. and 4 p.m. and a binary outcome: pass or fail. The researchers weighted results by course credit value to roughly account for differences in exam difficulty, then ran the numbers by hour.

The overall pass rate across the full dataset was 57.4%. But it wasn't flat across the day. Pass rates peaked at noon, at 72.4%, with 11 a.m. and 1 p.m. close behind at 67.4% each. They bottomed out at 4 p.m., at 51.3% — barely above a coin flip, and over twenty percentage points below the noon peak. Early morning wasn't great either: 8 a.m. sessions passed at 54.3%, just above the 4 p.m. low.

• Noon: 72.4% pass rate — the peak
• 11 a.m. and 1 p.m.: 67.4% each — not statistically different from noon
• 9 a.m.: 56.4%
• 8 a.m.: 54.3%
• 4 p.m.: 51.3% — the low point

The pattern held up statistically (the researchers' analysis of variance was significant at p < 0.001), and the shape of the curve — low, rising to a midday peak, falling again — matched a bell curve closely enough that the authors' own fit tests couldn't distinguish it from one.

“Students examined at noon passed 72% of the time. The same pool of students, examined at 4 p.m., passed 51% of the time.”

The Catch: This Is Exam Data, Not Job-Interview Data

Here's where most coverage of this study — and certainly most "best time to interview" articles that will eventually cite it secondhand — gets ahead of what it actually shows. The paper studied university oral exams. It does not mention job interviews, hiring, or recruiters anywhere in its results. The word "workplace" appears exactly once, in a single sentence suggesting the pattern could be worth studying in other high-stakes evaluation settings — a hypothesis for future research, not a finding about this one.

The authors are also candid about what their own data can't tell them. Because they only recorded exam date, start time, course, and pass/fail outcome — no individual scores, no demographic data — they can't separate two very different explanations: it's possible the time-of-day pattern reflects students performing differently depending on when they're assessed, or examiners evaluating differently depending on when they're doing the assessing, or some mix of both. The paper lists chronotype, sleep, fatigue, assessor breaks, and mood as plausible contributing factors, and explicitly calls the underlying mechanism "speculative" pending further research.

One more limitation worth knowing: all of this comes from a single Italian university over a roughly 16-month window. It's a genuinely large sample, but it's one institution's scheduling norms, one country's academic culture, and one specific evaluation format — a live oral exam, not a video call, panel interview, or one-way recorded assessment.

What the Hiring Industry Itself Claims — And Why to Treat It Cautiously

Separately from the Messina study, a claim specifically about job interviews has been circulating in HR and recruiting trade coverage: organizational psychologist Shiran Danoch, of a recruitment-analytics firm called Informed Decisions, has reportedly analyzed "thousands of job interviews" and found that interviewers tend to score more harshly later in the day, with candidates interviewed in the morning coming out ahead on average.

This blog treats that claim the way it treats any vendor-sourced statistic: real, possibly directionally useful, and not something to repeat as settled fact. There's no public paper, no named journal, no disclosed sample size or methodology, and no independent replication — just secondhand trade-press coverage of an analysis run by a company whose business is selling interview-related analytics. That's a meaningfully lower evidence bar than the peer-reviewed, methodologically transparent Messina study above, even though both point in a similar rough direction.

Two different claims, from two different kinds of sources, both suggesting a midday-ish advantage — that's worth noting. It is not the same as proof that your 4 p.m. interview is working against you.

It's worth being this careful about sourcing precisely because hiring-psychology claims travel fast and get flattened along the way — the same thing has happened to what the HR round is actually evaluating, where the popular explanation and the research-backed one turned out to be different stories.

If This Effect Is Real, It's Probably About the Evaluator — Not You

Assuming some version of this pattern does hold for job interviews, the more useful question isn't "is 4 p.m. bad" — it's "bad for what reason," because that changes what, if anything, is worth doing about it.

Nothing about either source suggests candidates themselves perform worse late in the day. The far more plausible mechanism, and the one the Messina researchers themselves point to, sits on the evaluator's side: an interviewer who has already sat through three other conversations since lunch has less attention, more decision fatigue, and a harder time staying generous than the same interviewer at 11 a.m., fresh and still calibrating for the day.

That fits a broader pattern this blog has covered before: a hiring decision isn't a single clean judgment, it's an accumulation of small, imperfect signals the interviewer is processing in real time — and the research on how and when that judgment actually forms is itself full of surprises.

If that's the mechanism, it means a time-of-day effect, if it exists, is a property of the interviewer's state that day — not a verdict on your performance, your preparation, or your fit for the role. There's nothing you can rehearse or fix about it, because it was never about you in the first place.

“If timing matters at all, it's the interviewer's attention running low — not your performance.”

What's Actually Worth Doing With This (and What Isn't)

Given the gap between "a large, honest, but tangential study" and "an unpublished vendor claim," the sensible response is modest, not a scheduling strategy:

• If a recruiter genuinely offers you a free choice of slots with no other constraint, picking something in the late-morning-to-early-afternoon range is a reasonable, low-cost default — treat it as a mild tiebreaker, not a meaningful edge.
• Don't turn down a slot you'd otherwise prefer — earlier in the week, more prep time, a day you're less exhausted — just to chase a theoretically better hour. The preparation trade-off almost certainly outweighs whatever marginal timing effect exists.
• Don't ask to reschedule an already-confirmed interview over this. Flexibility and reliability are themselves signals recruiters weigh, and there's no published evidence strong enough to justify spending that goodwill on a time slot.
• Don't treat a 4 p.m. slot as a loss condition going in. Research on decision fatigue in evaluators is a reason for an interviewer to self-check, not a reason for a candidate to walk in already defeated — that mindset affects your actual performance far more reliably than the clock does.

Interviewer attention and fatigue show up in other places worth knowing about too — including why interviewers keep asking questions they already know the answer to, a pattern that has more to do with testing your composure than catching you out.

None of this changes what still does the heavy lifting in an interview: specific, structured, evidence-backed answers delivered with real preparation behind them. If timing is genuinely a coin flip you can't fully control, the move is to put your energy into the parts you can — which is less about luck and more about reps.

Structure matters here too: a well-rehearsed story still falls apart under a tired evaluator's follow-up questions if it was only ever built to survive a first pass, so the same prep that protects you from a bad time slot also protects you from a sharp one.

Key Takeaways

If you remember only a few things, remember these:

• A 2025 Frontiers in Psychology study of 104,552 university oral exams found pass rates peaked at noon (72.4%) and bottomed out at 4 p.m. (51.3%) — a real, large, peer-reviewed dataset.
• That study is about exams, not job interviews — the authors never studied hiring, and they call their own proposed explanations (chronotype, fatigue, mood) "speculative."
• A separate claim specifically about job interviews exists from a recruitment-analytics consultancy, but it has no published methodology or peer review — treat it as directional industry chatter, not established fact.
• If a time-of-day effect on evaluation is real, the likelier mechanism is interviewer attention and fatigue, not candidate performance — it isn't a verdict on you.
• If given a real choice, a late-morning slot is a reasonable low-cost pick — but it's not worth rescheduling for, and it matters far less than how prepared you actually are.