What does Habit Curves and the 21-Day Myth conclude, in numbers, about “Scales, Noise and the Average Curve: SRHI, SRBAI and the Trap of Right Skew”?

Answer: Chapter 6 shows that the effect of one missed day is only 0.121 of a scale's minimal detectable change, and that the average curve moves more slowly than the median person.

Scope of the numbers: Its population parameters, forgetting rates and cue indices are illustrative; all calculations are generated by its programs; it is not medical or psychological advice, nor personalised business advice. Marked boxes separate published evidence, runnable Python calculations (19 programs), classical sources and the author's hypotheses.

中文版

Data table

The numbers below are printed verbatim by the chapter's Python program and can be reproduced by running it. standard-library Python onlyprinted values that match the answer: 0.121

itemsrrhoSEM (0-42)MDC95 (0-42)
40.50.8003.9610.98
70.50.8752.998.30
120.50.9232.296.34
40.60.8573.549.82
70.60.9132.687.42
120.60.9472.045.67
40.70.9033.078.50
70.70.9422.326.43
120.70.9661.774.91
one missed day as a share of the 7-item MDC95: 0.121
Show the full program
# Chapter 6. p61: How small a daily change can a habit scale see?
# Items are scored 0 to 6. With k items of standard deviation s_i and mean
# inter-item correlation r, the total has reliability (Spearman-Brown)
# rho = k r / (1 + (k - 1) r) and sd s_i sqrt(k + k (k - 1) r); the
# standard error of measurement is SEM = sd sqrt(1 - rho), and the
# smallest change detectable with 95 % confidence between two days is
# MDC = 1.96 sqrt(2) SEM. Scores are rescaled to the 0-42 range used by
# Lally et al. (2010) so the three forms can be compared. k = 4 is the
# SRBAI, 7 the Lally subset, 12 the full SRHI. s_i = 1.6 and r = 0.6 are
# illustrative, not published values.
import math

S_I = 1.6
print("  items   r     rho     SEM (0-42)   MDC95 (0-42)")
for r in (0.5, 0.6, 0.7):
    for k in (4, 7, 12):
        rho = k * r / (1 + (k - 1) * r)
        sd = S_I * math.sqrt(k + k * (k - 1) * r)
        sem = sd * math.sqrt(1 - rho) * 42.0 / (6.0 * k)
        mdc = 1.96 * math.sqrt(2.0) * sem
        print("  %5d  %4.1f   %5.3f   %10.2f   %12.2f"
              % (k, r, rho, sem, mdc))
ONE_MISS = 0.90          # largest gap after one missed day, from p41
k, r = 7, 0.6
rho = k * r / (1 + (k - 1) * r)
sd = S_I * math.sqrt(k + k * (k - 1) * r)
mdc = 1.96 * math.sqrt(2.0) * sd * math.sqrt(1 - rho) * 42.0 / (6.0 * k)
print("")
print("one missed day as a share of the 7-item MDC95: %.3f"
      % (ONE_MISS / mdc))

Single source and citation

Fan, Y. (2026). Asymptotic Habit-Formation Curves, Debunking the 21-Day Myth and a Cue-Based Automaticity Index (Version 1.0.0), Chapter 6: Scales, Noise and the Average Curve: SRHI, SRBAI and the Trap of Right Skew. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-197 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026197,
  author    = {Fan, Yafu},
  title     = {Asymptotic Habit-Formation Curves, Debunking the 21-Day Myth and a Cue-Based Automaticity Index},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book