LBL Chronotype Profile — Methodology & Validation

What this tool measures (and what it does not)

The LBL Chronotype Profile measures biological/trait chronotype — the circadian preference your body would have if life imposed no schedule. It is a self-report instrument structured to capture the trait dimension of morningness-eveningness, calibrated to the population (age, sex, life stage) the respondent actually belongs to rather than to 1970s university students.

The Profile does not measure:

• Your current behavioral sleep patterns — when you actually sleep on workdays vs. free days. That is measured by the companion LBL Sleep-Cognition Optimizer.
• Your dim-light melatonin onset (DLMO) directly — laboratory saliva sampling is required for that. The Profile produces an estimated DLMO window based on the score, with explicit uncertainty bounds.
• Any sleep disorder. Persistent inability to sleep at chronotype-appropriate times warrants clinical evaluation by a sleep medicine specialist.

The output is a continuous 0–100 morningness score paired with a descriptive 5-band classification (Strongly Evening / Evening-Leaning / Intermediate / Morning-Leaning / Strongly Morning), an age-sex calibrated percentile, an estimated DLMO window, and a trait-stability confidence indicator.

The Profile measures who you are biologically. The Sleep-Cognition Optimizer measures what you are doing. Together they compute the biology-behavior gap — the most clinically meaningful output of the pair.

Why LBL-original items, not MEQ or MCTQ paraphrase

The Horne & Östberg (1976) Morningness-Eveningness Questionnaire (MEQ) and the Munich ChronoType Questionnaire (MCTQ; Roenneberg 2003) are foundational research instruments. Both are copyrighted, neither has a permissive commercial-use license, and both require per-use permission from their authors or distributors for commercial deployment.

Earlier consumer chronotype tools — including our previous v0 Chronotype Test — used the MEQ-19 framework with "items written in our own voice" as a paraphrase strategy. We concluded that this strategy carries unresolvable legal ambiguity. The contemporary chronotype literature has also moved on substantially since 1976; modern work emphasizes the continuum (not discrete types), age-sex-stratified norms (not student-derived universal cutoffs), and genetic/biological grounding (not behavior alone).

This Profile v1.0 (May 2026) is rebuilt from the contemporary 2019–2025 literature as the scientific foundation, with LBL-original items written to measure the relevant constructs. The MEQ and MCTQ are cited as references in a broader literature, alongside Jones 2019, Taillard 2004, Burgess 2015, Carskadon 2011, Roenneberg 2007, and Allebrandt 2014 — they are not administered.

Instrument structure

The Profile consists of 14 items in 5 sections. Section weights reflect their role in the scoring algorithm.

Items Q1–Q7 produce a 7-item raw morningness score with theoretical range 7 (strongest evening on all dimensions) to 35 (strongest morning on all dimensions). This is normalized to a 0–100 scale. Items Q11–Q14 then apply calibration adjustments to produce the final calibrated score. Items Q8–Q10 modify the result interpretation, not the score itself.

Scoring algorithm — full pseudocode

The exact algorithm running in your browser. The code is open in the page source.

function computeRawMorningness(answers):
    items = [q1, q2, q3, q4, q5, q6, q7]
    sum = sum_of(answers[items])     # range 7..35
    rawScore = ((sum - 7) / 28) * 100
    return clip(rawScore, 0, 100)

function applyCalibration(rawScore, age, sex, lifeStage, latitude):
    calibrated = rawScore
    adjustments = []

    # Age-based adjustment
    if age < 18:
        calibrated += 5
        adjustments.append("adolescent phase delay calibration")
    elif age >= 60:
        calibrated -= 5
        adjustments.append("older adult phase advance")
    elif 45 <= age < 60:
        calibrated -= 2
        adjustments.append("midlife Taillard 2004 calibration")

    # Sex adjustment (only meaningful under ~age 45)
    if sex == "F" and age < 45:
        calibrated -= 3
        adjustments.append("female under 45 (~3 pts more morning)")
    elif sex == "M" and age < 45:
        calibrated += 1

    # Life-stage modifiers (interpretation only, no score shift)
    if lifeStage in ("perimeno", "pregnant"):
        adjustments.append("hormonal life stage noted")

    # Latitude modifier (noted, minimal effect)
    if latitude in ("far-north", "far-south"):
        adjustments.append("extreme photoperiod noted")

    return clip(calibrated, 0, 100), adjustments

function assignBand(score):
    if score < 20:  return "Strongly Evening"
    if score < 40:  return "Evening-Leaning"
    if score <= 60: return "Intermediate"
    if score <= 80: return "Morning-Leaning"
    return "Strongly Morning"

function estimateDLMO(score):
    # Linear mapping: score 0 → ~12:30 AM, score 50 → ~10:00 PM, score 100 → ~7:30 PM
    baseMinutes = 24*60 + 30        # 24:30
    offsetPerPoint = 3              # minutes earlier per morningness point
    center = baseMinutes - (score * offsetPerPoint)
    return (center - 22 minutes, center + 22 minutes)   # ~45-minute window

function checkTraitStability(q8_teen, q9_shift, calibratedScore):
    if q8_teen == 0:                 # "don't remember"
        return "medium", "no developmental anchor"

    teenExpected = 20 + (q8_teen - 1) * 15   # map 1..5 to expected 20..80
    diff = abs(calibratedScore - teenExpected)

    if diff < 20:  return "high",   "consistent with lifelong pattern"
    if diff < 35:  return "medium", "moderate shift from teen pattern"
    return "low", "substantial environmental phase shift suggested"
    

All computation runs locally in the user's browser. No inputs leave the device.

Age-sex calibration — rationale and references

Chronotype is not constant across the lifespan or between sexes. The original 1976 Horne & Östberg cutoffs were derived from 150 university students aged 18–32 and do not generalize well. Calibration adjustments are based on established literature:

The magnitudes are conservative — each is well within the range supported by multiple studies. We deliberately under-adjust rather than over-adjust to avoid manufacturing differences that aren't there. Latitude is collected for future research but applies minimal score shift in v1.0.

Score-band derivation

The 5 bands are soft descriptive points on a continuum, not fixed types. Band boundaries are placed at 20-point intervals on the 0–100 calibrated morningness scale:

The proportions are approximations based on aggregated chronotype distribution data from large epidemiologic samples (e.g., Fischer et al. 2017, n=53,689 US adults; Roenneberg 2007 review). The bands are intentionally symmetric around 50 (Intermediate) to convey the bell-shaped continuum nature of chronotype.

Both the continuous score and the band are displayed in the result. The score serves the data-curious user; the band serves the navigation/identity/sharing use case ("I'm Evening-Leaning" reads more naturally than "I'm 32 on the morningness scale").

Estimated dim-light melatonin onset (DLMO)

DLMO is the most direct biological marker of circadian phase. In a sleep laboratory, it is measured by serial saliva samples in dim light, typically 1–3 hours before usual sleep onset.

Self-report chronotype scores correlate with DLMO at r ≈ 0.5–0.7 across published studies (Burgess et al. 2015 meta-analysis). The correlation is meaningful but not precise — chronotype is a useful proxy, not a substitute, for laboratory DLMO measurement.

We map the calibrated 0–100 morningness score to a DLMO window using a linear relationship:

The estimate is presented as a ~45-minute window, NOT a precise time. It is explicitly framed as an approximation. The mapping is a linear approximation; the true relationship is non-linear at the extremes, but the linear form is adequate for the educational use case.

Trait stability check

Chronotype is a partly heritable trait (12–42% heritability per twin studies; 351 GWAS loci per Jones et al. 2019) but it can also be modulated by environment. The trait-stability check distinguishes lifelong biological pattern from recently-acquired environmental shift.

Q8 (teenage pattern) and Q9 (recent 5-year shift) are compared against the current preference score (Q1–Q7). The teen response is mapped to an expected score range (Q8=1 ↔ expected score ~20, Q8=5 ↔ expected score ~80), and the difference from the user's current calibrated score is computed.

Confidence levels:

• High confidence — current preference matches teenage pattern (difference < 20 points). Result reflects stable biological trait.
• Medium confidence — moderate shift (20–34 points). Common during normal lifespan transitions; might also indicate environmental phase shift.
• Low confidence — substantial shift (35+ points). Suggests environmentally-induced phase shift; the user's biology may be more extreme than their current preferences suggest. Sleep-Cognition Optimizer recommended to measure the gap.

The trait-stability check does not change the score itself. It changes how the result is interpreted and helps the user understand when their result might be reflecting trait biology vs. recent environmental influence.

Validation strategy

Validation is staged across four levels. Each level addresses a distinct type of validity claim and has explicit completion status.

Users should treat the Profile as a structured self-assessment of trait chronotype, not as a clinically validated diagnostic instrument. The Level 1 and Level 2 evidence supports its use as an educational tool for self-assessment, longitudinal personal tracking, and as input to the companion Sleep-Cognition Optimizer.

Edge cases and handling

• Adolescents (under 18): Result is calibrated for adolescent norms (+5 point morning shift). The result page includes a note that under-18 users should consult a parent or guardian about sleep changes.
• Shift workers: The Profile measures trait preference, not behavioral adaptation to shift schedules. Shift workers can take the Profile to identify their underlying preference, then use the Optimizer to measure the gap between preference and current shift schedule.
• Recent jet lag: Users should answer based on typical (pre-jet-lag) preference, not current sleep-deprived state.
• Pregnancy / postpartum: Q13 captures life stage. Result is noted as potentially transiently shifted; users may want to retake post-pregnancy for a stable trait reading.
• Perimenopause / menopause: Q13 captures life stage. Hormonal transitions can cause temporary chronotype shifts; result is noted with caution.
• "Don't remember" on Q8: If the user cannot recall their teenage pattern, trait-stability check downgrades to "medium" confidence with explanation.
• Mixed cognitive/physical peak times: Q3 and Q4 are scored independently. A user with morning physical peak but evening cognitive peak will get a composite score that reflects the average — and may sit closer to Intermediate than to either extreme. This is correct behavior given the construct.

Limitations

The Profile inherits limitations common to all self-report chronotype questionnaires, plus several specific to this v1.0 implementation.

Self-report limitations apply to any chronotype questionnaire:

• Recall bias — asking about preferences requires accessing semantic memory of typical patterns, which is noisier than measuring actual behavior.
• Social-desirability bias — cultural framing of morningness as virtuous ("the early bird gets the worm") may inflate scores in some users.
• Domain inconsistency — users whose preferences differ across domains (e.g., morning physical peak, evening cognitive peak) may produce composite scores that don't fit any single band cleanly.
• Cross-cultural validity — reference norms used for age-sex calibration are derived primarily from European/North American populations. Cross-cultural validity is untested in v1.0.

Implementation-specific limitations of v1.0:

• Level 3 (convergent validity vs. MEQ/MCTQ) and Level 4 (DLMO criterion validity) studies are not complete.
• Test-retest reliability over 4-week and 12-week intervals has not been measured.
• The DLMO mapping is a linear approximation; the true relationship is non-linear at extremes.
• Latitude calibration is collected but applies minimal score shift; effect sizes await empirical refinement.
• The Profile is designed for ages 13+; under-13 users should not use this tool.

What this Profile is NOT:

• Not a clinical diagnostic instrument for circadian rhythm sleep-wake disorders.
• Not a genetic test (despite providing genetic context in the result).
• Not a substitute for actigraphy or laboratory DLMO measurement.
• Not appropriate for users with suspected delayed sleep phase disorder, advanced sleep phase disorder, Non-24, or shift work disorder — clinical evaluation is needed.

Independent review

The Profile v1.0 instrument design, scoring algorithm, age-sex calibration, and validation framework were reviewed by:

• Abiot Y. Derbie, PhD — Postdoctoral research fellow specializing in neurodevelopmental neuroimaging, multimodal connectivity, and machine-learning prediction models. Reviewed instrument construct definition and scoring algorithm.
• Eskezeia Y. Dessie, PhD — Researcher in behavioral and brain sciences. Reviewed item wording, edge-case handling, and clinical safety framing.

The Profile is an educational self-assessment tool, not a clinical instrument. The reviewers are not clinical sleep medicine specialists. Users with suspected sleep disorders should consult a sleep medicine specialist (typically a physician board-certified in sleep medicine).

Version log

The next scheduled review is November 13, 2026. Triggers for unscheduled review include: publication of major new chronotype validation studies; identification of methodology errors via corrections; user-reported edge cases not covered above; updates to age-sex norm distributions in the literature.

Methodology FAQ

Last reviewed

This methodology page was last reviewed on May 13, 2026. The next scheduled review is November 13, 2026.

How to cite this methodology

Related pages