What does Fear Copy: the Inverted U and Its Limits conclude, in numbers, about “Does the Inverted U Exist? The Monotone Verdict of Meta-Analysis, Stacked Individual Peaks and the Power of Three-Arm Studies”?

Answer: Chapter 5 faces the counterintuitive meta-analytic verdict that within the range of existing studies stronger fear generally works better and evidence for an inverted U is weak; it explains how both can hold through stacked individual peaks and statistical power (with 200 people per level the chance of detecting a real bend is only 0.719), and states the conditions for refutation.

Scope of the numbers: All parameters are illustrative; all calculations are generated by the book's programs; the book is not legal advice nor personalised business or financial advice, and makes no judgement about the efficacy of any product. 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: 0.719, 200

true adoption: 0.1554, 0.2728, 0.2023
true curvature -0.1878, true linear 0.0469
n per armfinds Ufinds risefinds fall
500.2260.1020.005
1000.4240.1330.002
2000.7190.2280.001
4000.9540.4210.000
8001.0000.6890.000
Show the full program
# Chapter 5. p52: how often a three-arm study detects the bend.
# True adoption follows the p21 model (mixed audience Beta(3, 3), no
# antidote), at three doses 0.5, 1.5 and 2.5: a real inverted U. Each
# arm has n readers; each adopts with the true probability. A study
# "finds the U" if the curvature contrast y1 - 2 y2 + y3 is negative at
# z < -1.96; it "finds a rise" if the linear contrast y3 - y1 is
# positive at z > 1.96. 2,000 simulated studies per n, fixed seed.
# For scale: Tannenbaum et al. (2015) pooled a mean effect d = 0.29.
import math
import random

K = 0.35
DOSES = (0.5, 1.5, 2.5)
SIMS = 2000


def surv(x, a=3, b=3):
    n = a + b - 1
    return math.fsum(math.comb(n, j) * x ** j * (1 - x) ** (n - j)
                     for j in range(a))


def true_p(f):
    t = 1 - math.exp(-K * f)
    return t * surv(t)


def main():
    ps = [true_p(f) for f in DOSES]
    print("true adoption:", ", ".join("{:.4f}".format(p) for p in ps))
    print("true curvature {:.4f}, true linear {:.4f}".format(
        ps[0] - 2 * ps[1] + ps[2], ps[2] - ps[0]))
    rng = random.Random(20261112)
    print("  n per arm   finds U   finds rise   finds fall")
    for n in (50, 100, 200, 400, 800):
        u = r = fall = 0
        for _ in range(SIMS):
            y = []
            for p in ps:
                k = 0
                for _ in range(n):
                    if rng.random() < p:
                        k += 1
                y.append(k / n)
            v = [max(q * (1 - q), 0.25 / n) / n for q in y]
            cz = (y[0] - 2 * y[1] + y[2]) / math.sqrt(
                v[0] + 4 * v[1] + v[2])
            lz = (y[2] - y[0]) / math.sqrt(v[0] + v[2])
            u += cz < -1.96
            r += lz > 1.96
            fall += lz < -1.96
        print("  {:>9}   {:>7.3f}   {:>10.3f}   {:>10.3f}".format(
            n, u / SIMS, r / SIMS, fall / SIMS))


if __name__ == "__main__":
    main()

Single source and citation

Fan, Y. (2026). The Inverted-U of Fear Arousal, Concrete Pain-Point Imagery and the Compliance Boundary of High-Conversion Copy (Version 1.0.0), Chapter 5: Does the Inverted U Exist? The Monotone Verdict of Meta-Analysis, Stacked Individual Peaks and the Power of Three-Arm Studies. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-143 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026143,
  author    = {Fan, Yafu},
  title     = {The Inverted-U of Fear Arousal, Concrete Pain-Point Imagery and the Compliance Boundary of High-Conversion Copy},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book