关于「理性零食:环境可信度先验下的「不等」」,《棉花糖效应的收缩》给出的定量结论是什么?

结论:第二章重算 Kidd 等人的「理性零食」实验,两组平均等待之比为 3.98,并演示在等待毫无因果效应的世界里,理性的「不等」就能制造 0.208 的未控制斜率。

数据口径:书中参数为示意;全部计算由书中程序生成;本书不是心理评估、医疗或育儿意见,书中关于门店与社群的建议是非个人化经营建议。 书中以标注框区分已发表的证据、作者的假说、原典与可运行的 Python 计算(18 段程序)。

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: 3.98

mean wait reliable / unreliable: 722.43 s / 181.57 s
ratio of mean waits: 3.98
full 15 min, reliable : 9/14 = 0.643 CI [0.388, 0.837]
full 15 min, unreliable: 1/14 = 0.071CI [0.013, 0.315]
risk difference: 0.571
risk ratio: 9.0
odds ratio (+0.5 each cell): 15.5
Fisher exact p, two-sided: 0.0044
查看完整验算程序
# Chapter 2. p21: the size of the "rational snacking" effect
# 数据(出处):Kidd, Palmeri & Aslin (2013),每组 14 名儿童;可靠组平均
# 等待 722.43 秒,不可靠组 181.57 秒;等满 15 分钟者 9/14 对 1/14。
# 计算:均值比、风险差与风险比、Wilson 95% 区间、Fisher 精确检验(双侧,
# 超几何枚举)、加 0.5 校正的比值比。全部为对公开数字的精确再计算。
import math

N = 14
MEAN_REL, MEAN_UNREL = 722.43, 181.57
FULL_REL, FULL_UNREL = 9, 1
Z95 = 1.959964


def wilson(k, n):
    p = k / n
    den = 1 + Z95 ** 2 / n
    mid = (p + Z95 ** 2 / (2 * n)) / den
    half = Z95 * math.sqrt(p * (1 - p) / n + Z95 ** 2 / (4 * n * n)) / den
    return mid - half, mid + half


def fisher_two_sided(a, b, c, d):
    r1, c1, n = a + b, a + c, a + b + c + d

    def prob(x):
        return math.comb(c1, x) * math.comb(n - c1, r1 - x) / math.comb(
            n, r1)
    p0 = prob(a)
    lo, hi = max(0, r1 + c1 - n), min(r1, c1)
    ps = [prob(x) for x in range(lo, hi + 1)]
    return math.fsum(p for p in ps if p <= p0 * (1 + 1e-9))


def main():
    print("mean wait reliable / unreliable: {:.2f} s / {:.2f} s".format(
        MEAN_REL, MEAN_UNREL))
    print("ratio of mean waits            : {:.2f}".format(
        MEAN_REL / MEAN_UNREL))
    p1, p0 = FULL_REL / N, FULL_UNREL / N
    for lab, k in (("reliable", FULL_REL), ("unreliable", FULL_UNREL)):
        lo, hi = wilson(k, N)
        print("full 15 min, {:<10}: {}/{} = {:.3f}  CI [{:.3f}, {:.3f}]"
              .format(lab, k, N, k / N, lo, hi))
    print("risk difference                : {:.3f}".format(p1 - p0))
    print("risk ratio                     : {:.1f}".format(p1 / p0))
    a, b = FULL_REL, N - FULL_REL
    c, d = FULL_UNREL, N - FULL_UNREL
    orr = (a + 0.5) * (d + 0.5) / ((b + 0.5) * (c + 0.5))
    print("odds ratio (+0.5 each cell)    : {:.1f}".format(orr))
    print("Fisher exact p, two-sided      : {:.4f}".format(
        fisher_two_sided(a, b, c, d)))


if __name__ == "__main__":
    main()

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《延迟满足棉花糖实验复制收缩、家庭资源混杂与儿童自控干预边界母典》第 2 章「理性零食:环境可信度先验下的「不等」」(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-203 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026203,
  author    = {Fan, Yafu},
  title     = {The Shrinking Marshmallow Effect, Family-Resource Confounding and the Limits of Child Self-Control Interventions},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论