Nutrinaut learning library
Learn With Nutrinaut
Concrete examples, visible arithmetic and the limits of consumer data. Read a field note, then open the matching tool when you need to apply the method.
Start with your question
Two wearable screens disagree
Check the source → check dates → compare coverage.
Use the Data Forensics Lab for a browser-local timestamp example.
A food number looks wrong
Check the label basis → check recipe yield → check the arithmetic.
Before interpreting a weekly average, check diary completeness.
An AI insight suggests a change
Check its evidence → check context and missing entries.
Separate an observation from a recommendation before acting.
Wearables & sleep
Start with dates, duration and data sources before comparing a trend or reading a summary score.
Time Asleep, Time in Bed and the Night Window
Compare sleep categories, gaps and a separate in-bed total with an interactive coverage visual.
Read field note → Interactive lab · Data provenanceNutrinaut Data Forensics Lab
Inspect timestamps, duration definitions and possible duplicate imports in a browser-local sample.
Open interactive lab → Field note 13 · Calendar checksOne Night, Two Dates: Reading Wearable Sleep Around Clock Changes
Midnight, travel and clock changes can give one observation several date labels. Here is how to check the interval before changing the diary.
Read field note → Field note 12 · Training inputsBefore You Trust a Workout Calorie Estimate, Check the Minutes
An estimate can look precise while using the wrong minutes. Check the duration, its source and its definition before deciding what the calorie number means.
Read field note → Field note 07 · Wearable comparisonApple Watch and Garmin Readiness Disagree: What to Check First
Check scales, timestamps and missing records when two wearables disagree. Includes a downloadable comparison worksheet.
Read field note → Field note 01 · Wearable dataBefore You Trust a Wearable Trend, Check Its Source
Check sources and dates, separate daily summaries from workouts, and work through three missing-night scenarios.
Read field note → Field note 06 · RecoveryHow to Read Recovery Signals Without a Readiness Verdict
Separate a wearable measurement, a baseline, its context and the decision that follows.
Read field note →Food & calculations
Check diary completeness, portion sizes and arithmetic before interpreting a food or energy estimate.
One Pot, Different Portions: Audit a Recipe Before Logging It
Match ingredient basis, weigh the finished yield and separate whole-pack shopping cost from the amount used.
Read field note → Field note 15 · Calorie categoriesCan You Add Workout Calories to Your Watch’s Daily Total?
Use a calorie ledger to check categories, time windows and whether a workout is already included.
Read field note → Field note 10 · Diary completenessAn Empty Food Log Is Not a Calorie Deficit
Check what was recorded before interpreting a low average. Missing, pending and zero are not interchangeable.
Read field note → Field note 09 · Food arithmeticWhy Calories and Macros Do Not Always Match
Four worked examples and a repeatable check for the gap between a calorie total and its protein, carbohydrate and fat.
Read field note → Field note 03 · Label arithmeticOne Label, Three Portions: A Serving-Size Audit
Use the FDA sample label to see exactly how a one-, one-and-a-half- or whole-container portion changes the numbers.
Read field note → Field note 04 · Model sensitivityWhy Two Energy Estimates Can Both Look Plausible
A calculation audit showing how activity assumptions, not a hidden change in metabolism, can widen an energy estimate.
Read field note → Field note 05 · Food loggingWhat AI Meal Estimates Can—and Cannot—Tell You
Follow an estimate through portion and ingredient changes to the final saved meal.
Read field note →AI & context
Separate an observation from its interpretation, and check what evidence actually supports a recommendation.
When Should an AI Insight Stop?
Change one evidence condition at a time and see when a narrow statement still follows.
Open interactive lab → Field note 11 · AI evidenceBefore You Follow an AI Insight: Check the Evidence
Six questions, three worked examples and a downloadable card for checking the reasoning behind an app recommendation.
Read field note → Field note 08 · Reading your dataHow to Read Health-App Data Without Jumping to Conclusions
Six checks, three everyday examples and a reusable summary for making sense of food logs, wearable trends and AI insights.
Read field note → Field note 02 · Lifestyle tagsUse Lifestyle Tags as Context, Not Proof of Cause
Keep yes, no and unknown separate with a downloadable 14-day journal.
Read field note →