关于「实践与证伪方案:经营者清单、攻心文案检查器与预注册检验」,《恐惧文案的倒U与边界》给出的定量结论是什么?

结论:第八章给出经营者清单、攻心文案检查器、私讯跟进上限与预注册检验方案(主要比较每格 2,013 人)。书中以标注框区分已发表的证据、可运行的 Python 计算(19 段程序)、原典与作者的假说。

数据口径:书中参数均为示意;全部计算由书中程序生成;本书不是法律意见,也不构成个人化的商业或理财建议,对任何产品的功效不作判断。 第一章用「先接收、再顺从」的两步模型说明倒 U 型的顶点由注意与防御的速率共同决定,示意参数下顶点在剂量 3.901,并指出牙齿卫生实验的三个数据点恰在一条直线上。

English version

对照数据表

下表数字由本章 Python 程序原样打印,可重新运行复现。 程序只用 Python 标准库与结论吻合的打印值: 2013

comparisonpower 0.80power 0.90
high-high vs high-low2,0132,695
high-high vs low-high8,96612,002
low-high vs low-low30,39040,683
interaction contrast +0.012: n per cell 6,950
dose ladder 0.030, 0.036, 0.039, 0.037, 0.031
bend contrast +0.029: n per arm 4,264, total 21,320
查看完整验算程序
# Chapter 8. p81: sample sizes for the pre-registered copy test.
# Design: 2 x 2 (threat low / high, antidote low / high) plus a five-arm
# dose ladder. Planned conversion rates are illustrative guesses an
# owner must replace with her own baseline: low-low 0.030, high-low
# 0.026, low-high 0.034, high-high 0.042. Two-sided alpha 0.05.
# Two proportions: n = (z_a sqrt(2 pq) + z_b sqrt(p1q1 + p2q2))^2 / d^2.
# A contrast with weights c_i: n = (z_a + z_b)^2 sum c_i^2 p_i q_i / D^2.
import math
from statistics import NormalDist

Z = NormalDist()
ZA = Z.inv_cdf(0.975)
CELLS = {"low-low": 0.030, "high-low": 0.026, "low-high": 0.034,
         "high-high": 0.042}


def n_two(p1, p2, power):
    zb = Z.inv_cdf(power)
    pb = (p1 + p2) / 2
    num = (ZA * math.sqrt(2 * pb * (1 - pb)) +
           zb * math.sqrt(p1 * (1 - p1) + p2 * (1 - p2))) ** 2
    return math.ceil(num / (p1 - p2) ** 2)


def n_contrast(ps, cs, power):
    zb = Z.inv_cdf(power)
    d = math.fsum(c * p for c, p in zip(cs, ps))
    v = math.fsum(c * c * p * (1 - p) for c, p in zip(cs, ps))
    return math.ceil((ZA + zb) ** 2 * v / d ** 2), d


def main():
    print("comparison                      power 0.80  power 0.90")
    pairs = (("high-high vs high-low", "high-high", "high-low"),
             ("high-high vs low-high", "high-high", "low-high"),
             ("low-high vs low-low", "low-high", "low-low"))
    for name, a, b in pairs:
        print("{:<30} {:>11,} {:>11,}".format(
            name, n_two(CELLS[a], CELLS[b], 0.80),
            n_two(CELLS[a], CELLS[b], 0.90)))
    ps = [CELLS[k] for k in ("low-low", "high-low", "low-high",
                             "high-high")]
    n, d = n_contrast(ps, (1, -1, -1, 1), 0.80)
    print("interaction contrast {:+.3f}: n per cell {:,}".format(d, n))
    ladder = (0.030, 0.036, 0.039, 0.037, 0.031)
    n2, d2 = n_contrast(ladder, (-2, 1, 2, 1, -2), 0.80)
    print("dose ladder", ", ".join("{:.3f}".format(p) for p in ladder))
    print("bend contrast {:+.3f}: n per arm {:,}, total {:,}".format(
        d2, n2, 5 * n2))


if __name__ == "__main__":
    main()

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《恐惧唤醒倒U型曲线、具体化痛点成像与合规免责高转化文案边界》第 8 章「实践与证伪方案:经营者清单、攻心文案检查器与预注册检验」(版本 1.0.0). 5D Library. DOI 待发布

DOI 待发布 · 专著页正式发布后上线(本站暂无专著页) · 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}
}

本书其他章节结论