关于「实证闭环:实战工具与检查器规格」,《声香签名与气味记忆》给出的定量结论是什么?

结论:第八章给出蒸馏小工具「声香签名对齐与跨感官一致度试算器」、按四件套分组的检查器规格、台账与预注册检验,主检验每组 276 人。空间投放香氛与音乐须事先公示、可选择离开、顾及气味敏感人群,不针对未成年人。书中以标注框区分已发表的证据、可运行的 Python 计算(22 段程序)、原典与作者的假说。

数据口径:书中参数为示意,全部计算由书中程序生成。 本书不是医疗、心理咨询或法规意见,不宣称任何身心功效。

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: 276

A. people per arm, liking, both contrasts
d0.25: 305 per arm, 915 in all
d0.30: 212 per arm, 636 in all
d0.40: 119 per arm, 357 in all
B. people per arm, one-week recognition
0.45 vs 0.55474 per arm
0.45 vs 0.60210 per arm
C. simulated, d0.30, 2,000 trials per plan
per-contrast power 0.80: n212 per arm, both pass 0.680
per-contrast power 0.90: n276 per arm, both pass 0.827
查看完整验算程序
# Chapter 8. p82: sample size for the pre-registered three-arm test.
# Arms: music only, scent only, congruent music + scent. Primary
# outcome: liking of the space (1-7, sd 1.2). Two confirmatory
# contrasts (combination versus each single cue), Bonferroni alpha
# 0.025 each, two-sided, power 0.80, normal approximation. Secondary:
# one-week cue recognition, a proportion. A fixed-seed simulation of
# 2,000 trials checks the joint power of the primary plan, since the
# combination must beat both single cues. Effect sizes are
# illustrative guesses, to be replaced by the pilot.
import math
import random
from statistics import NormalDist

ND = NormalDist()
ZA = ND.inv_cdf(1 - 0.025 / 2)
ZB = ND.inv_cdf(0.80)


def n_means(d, power=0.80):
    zb = ND.inv_cdf(power)
    return math.ceil(2.0 * ((ZA + zb) / d) ** 2)


def n_props(p1, p2):
    pb = (p1 + p2) / 2.0
    a = ZA * math.sqrt(2 * pb * (1 - pb))
    b = ZB * math.sqrt(p1 * (1 - p1) + p2 * (1 - p2))
    return math.ceil(((a + b) / (p2 - p1)) ** 2)


if __name__ == "__main__":
    print("A. people per arm, liking, both contrasts")
    for d in (0.25, 0.30, 0.40):
        n = n_means(d)
        print("   d = {:.2f}: {:>4} per arm, {:>5,} in all".format(
            d, n, 3 * n))
    print("B. people per arm, one-week recognition")
    for p1, p2 in ((0.45, 0.55), (0.45, 0.60)):
        print("   {:.2f} vs {:.2f}: {:>4} per arm".format(
            p1, p2, n_props(p1, p2)))
    rng = random.Random(39282)
    sd = 1.2
    print("C. simulated, d = 0.30, 2,000 trials per plan")
    for power in (0.80, 0.90):
        n = n_means(0.30, power)
        crit = ZA * sd * math.sqrt(2.0 / n)
        both = 0
        for _ in range(2000):
            m = rng.gauss(0.0, sd / math.sqrt(n))
            s = rng.gauss(0.0, sd / math.sqrt(n))
            c = rng.gauss(0.30 * sd, sd / math.sqrt(n))
            if c - m > crit and c - s > crit:
                both += 1
        print("   per-contrast power {:.2f}: n = {} per arm, both pass "
              "{:.3f}".format(power, n, both / 2000))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《声学节律与嗅觉分子跨模态耦合、普鲁斯特记忆唤醒与瞬时状态锚定母典》第 8 章「实证闭环:实战工具与检查器规格」(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-392 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026392,
  author    = {Fan, Yafu},
  title     = {Cross-Modal Coupling of Acoustic Rhythms and Olfactory Cues, Proustian Memory Retrieval and Rapid State-Conditioning Anchors},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论