Time of day.

Does the time of day you taste change your blind results, and does tasting at the same time make them steadier?

Status: Open
Protocol: Version 1.0, 2026-10-18
Consent: S.time-of-day.v2
Investigator: John Schuster, Drunken Unicorn Productions
Preregistration: Being posted to the Open Science Framework; the link will appear here
Ethics review: Not yet reviewed; we intend independent review before any results are published
Target: At least 392 daytime and 392 evening triangle rounds
Runs until: The target is reached, or 2029-10-18

Why this study

Two controlled studies find that smell sensitivity is lowest in the morning and highest in the evening: 56 adults tested four times in one day had their best odor thresholds between 8 and 10 p.m. and their worst between 8 and 10 a.m. (Mai, Rosbach & Hummel, Rhinology, 2023), and in 37 adolescents on a forced-desynchrony schedule, sensitivity peaked around 9 p.m., with no one peaking between about 2 a.m. and 10 a.m. (Herz et al., Chemical Senses, 2017). Sweet taste runs the other way: in 91 adults, sweet thresholds were lowest at 8 a.m. and highest at 10 p.m. (Nakamura et al., Diabetes, 2008). Hunger sharpens smell, so meals matter as well as the clock. The wine trade’s rule that tasting is best in late morning rests on practice, not on a published study. All of this measures detection thresholds, the faintest smell or taste a person can notice. Whether any of it changes how well a taster tells two whiskeys apart or describes one has not been tested. A 2023 review of the taste literature found that many studies don’t even record the time of day they test (Costanzo, Nutrition Reviews, 2023).

What taking part means

Nothing extra. Taste whenever you normally do. The app notes your local time with each session and triangle round. When your tasting sessions and triangle rounds happen, in your local time; your triangle results; your repeat checks; how long since you last ate, as you already enter it; and your sweet-to-dry ratings. The app has recorded local time only since this study opened, so only data from then on is used.

To join: in the app, turn on research contribution in Settings → Research, then open Time of day under Research studies, read it, and agree. You can leave with one tap.

Hypotheses

H1 (primary). Triangle accuracy differs between evening rounds (5 p.m. to midnight, local time) and daytime rounds (6 a.m. to 5 p.m.). The test is two-sided, because the published evidence points different ways for smell and for taste.
H2 (secondary). In repeat checks, two pours of the same whiskey tasted within three clock hours of each other (whatever the date) get more consistent descriptions than two pours tasted further apart in clock time.
H3 (secondary). Tasters who keep to a regular tasting time have steadier blind results: the more a taster’s tasting times spread across the day, the more their triangle accuracy swings from night to night.

Design

Observational, within-taster, repeated measures. Local time is the UTC timestamp plus the offset the device reported when the data was filed.

Who can take part. Adults 21 or older in the United States who use Savorant, have research contribution turned on, and have joined this study in the app.

Outcomes

Primary. Difference in proportion of triangle rounds correct, evening minus daytime, from a mixed-effects model.

Secondary and exploratory.

Repeat-check consistency, same clock window versus different clock window (linear mixed model, random intercept for taster).
For tasters with at least six nights of three or more triangle rounds: the correlation between how widely their tasting times spread (circular standard deviation of local hour) and how much their accuracy swings from night to night (beta-binomial overdispersion across nights). One-sided Spearman correlation, α = 0.05, with a 95% CI; repeated with each taster’s overall accuracy and number of nights as covariates.
Triangle accuracy across the day as a smooth curve (cosinor fit of local hour), exploratory.
Effect of minutes since eating on triangle accuracy, and whether it explains the time-of-day effect, exploratory.
Sweet-to-dry ratings by time of day, exploratory, given the morning peak in sweet sensitivity.

Sample size

At least 392 daytime and 392 evening triangle rounds (to detect 55% versus 45% correct, α = 0.05 two-sided, power 0.80: n = 391.3 per window, rounded up). For H2, at least 99 same-window and 99 different-window repeat checks (d = 0.4, α = 0.05 two-sided, power 0.80: n = 98.1 per group, rounded up). For H3, at least 68 tasters with six or more qualifying nights (correlation 0.3, α = 0.05 one-sided, power 0.80: n = 67.5 by the Fisher z method, rounded up).

Analysis

Primary: mixed-effects logistic regression, correct ~ window (evening vs daytime) + rounds already run that night, random intercepts for taster and for pair, two-sided α = 0.05. The difference is reported in percentage points with a 95% CI. Rounds between midnight and 6 a.m. are excluded from the primary analysis and reported separately.

Only data filed after a taster joins is used.

When a study reports more than one secondary test, p-values are adjusted with the Holm method.

Limitations we know about

Tasters choose when to taste, so time of day travels with other things: evenings are after work and dinner, weekends differ from weekdays. Rounds already run that night and minutes since eating are adjusted for, but this remains an observational comparison, not an experiment. A taster’s body clock (chronotype) isn’t measured. For H3, tasters who keep a regular time may differ in other ways (more experienced, more deliberate), so a correlation would not show that the regular time itself causes steadier results.

Standards every Savorant study follows

Preregistration. This page is the protocol of record, versioned and dated, and is posted to the Open Science Framework (OSF) before any study data is analyzed. Hypotheses, outcomes, sample size and analysis are fixed here in advance.
Primary and secondary outcomes. Each study names one primary outcome, tested at the stated alpha. Everything else is labeled secondary or exploratory and reported as such, with effect sizes and 95% confidence intervals rather than bare p-values.
No peeking. No interim analysis of the outcomes. The only checks before the sample size is reached are for data quality (duplicates, impossible values), and they never look at results.
Repeated measures. One taster contributes many rounds, so analyses use mixed-effects models with the taster as a random effect, or per-taster summaries, never pooled rounds treated as independent.
Exact methods for small counts. Triangle results are analyzed against the exact binomial with chance at one in three, the way ISO 4120 sets out the triangle test, adapted here for repeated measures.
Exclusions. Rounds are excluded only for the reasons listed in advance: test accounts, the founder’s accounts, rounds filed in a preview build, and duplicates. Every exclusion is counted and reported.
Missing data. Tasters who stop early are kept for every analysis their data supports. The number who stop, and when, is reported.
Deviations. Any change after data collection starts is logged on this page as a dated amendment with its reason, and reported in the paper.
Reporting. Results are reported following the STROBE guideline for observational studies, with the analysis code published alongside.
Conflict of interest. The study is run by the company that sells the app. Preregistration, publishing regardless of outcome, and public analysis code are how that interest is kept out of the result.

The consent you agree to

This is the exact text shown in the app before you join.

Amendments

None. This is the first version of the protocol.

All open studies · The evidence review · Privacy policy · palate@savorant.com