关于「买椟还珠,与科学广告」,《暗投授权与钩子归因》给出的定量结论是什么?

结论:示意数据里,等级相关为 0.629,置换 \(p\) 值 0.0169。十二条素材,就能看出方向。可这是因为,示意数据里,效应设得很大。真实效应若只有 0.4,要 38 条素材;若为 0.6,要 16 条。品牌一个季度,未必有这么多授权素材。

数据口径:书中参数均为示意,全部计算由书中程序生成;本书是非个人化经营分析,不构成经营或法律建议。 第六章用二叉树为跳单期权定价,合理预估 20.82 已拦住示意模型中全部跳单。

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: 0.0169, 0.4, 0.6, 0.629, 16, 38

hook score0.00 0.09 0.18 0.27 0.36 0.46 0.55 0.64 0.73 0.82 0.91 1.00
return rate0.01 0.07 0.07 0.12 0.04 0.11 0.16 0.14 0.09 0.10 0.14 0.13
Spearman rho 0.629; permutation p (one-sided) 0.0169
creatives needed for rho 0.438
creatives needed for rho 0.616
查看完整验算程序
# Chapter 7. p71: turning "bought the box, returned the pearl" into a test.
# Hypothesis from Han Feizi's parable: the louder the packaging (hook
# intensity score h), the larger the share of buyers who keep the box
# and send back the pearl (return rate). Synthetic panel of 12
# creatives (illustrative): return = 0.06 + 0.10 h + noise.
# Spearman rank correlation; one-sided permutation p-value from
# 20,000 shuffles with a fixed seed; then the number of creatives a
# test needs to detect rho = 0.4 and rho = 0.6 (Fisher z, alpha 0.05
# one-sided, power 0.80).
import math
import random
from statistics import NormalDist, correlation

rng = random.Random(1923)
Z = NormalDist().inv_cdf


def ranks(xs):
    order = sorted(range(len(xs)), key=lambda i: xs[i])
    r = [0.0] * len(xs)
    for pos, i in enumerate(order):
        r[i] = float(pos + 1)
    return r


H = [round(j / 11.0, 3) for j in range(12)]
RET = [round(0.06 + 0.10 * h + rng.gauss(0.0, 0.03), 4) for h in H]

if __name__ == "__main__":
    rh, rr = ranks(H), ranks(RET)
    rho = correlation(rh, rr)
    hits = 0
    shuffled = rr[:]
    for _ in range(20000):
        rng.shuffle(shuffled)
        if correlation(rh, shuffled) >= rho:
            hits += 1
    print("hook score:  " + " ".join("{:.2f}".format(h) for h in H))
    print("return rate: " + " ".join("{:.2f}".format(x) for x in RET))
    print("Spearman rho {:.3f}; permutation p (one-sided) {:.4f}".format(
        rho, hits / 20000))
    for target in (0.4, 0.6):
        n = math.ceil(((Z(0.95) + Z(0.80)) / math.atanh(target)) ** 2 + 3)
        print("creatives needed for rho {:.1f}: {}".format(target, n))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《暗投白名单授权经济学、三秒钩子生存分析与单层星探归因博弈母典》第 7 章「买椟还珠,与科学广告」(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-333 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026333,
  author    = {Fan, Yafu},
  title     = {Dark-Ads Whitelist Licensing Economics, Three-Second Hook Survival Analysis and Single-Tier Scout Attribution Games},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论