关于「干预的天花板:消退、乐观上界与两个杠杆」,《棉花糖效应的收缩》给出的定量结论是什么?
结论:第六章给出训练等待的乐观上界:即使把系数当作因果、效果永不消退,十年后的学业收益也只有 1.77 分(均值 100、标准差 15 的量表),用控制后的系数只有 0.61 分。
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)。
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: 0.61, 1.77, 15
| gain at 15 in SD, delta = 0.5 SD of waiting | ||||
| beta | none | h=5y | h=2y | h=1y |
| raw 0.236 | 0.1180 | 0.0275 | 0.0031 | 0.0001 |
| early 0.081 | 0.0405 | 0.0094 | 0.0011 | 0.0000 |
| full 0.050 | 0.0250 | 0.0058 | 0.0007 | 0.0000 |
| same in points on a mean-100, SD-15 scale: | ||||
| raw 0.236 | 1.77 | 0.41 | 0.05 | 0.00 |
| early 0.081 | 0.61 | 0.14 | 0.02 | 0.00 |
| full 0.050 | 0.38 | 0.09 | 0.01 | 0.00 |
| share kept after 10.5 y with h = 2: 0.0263 |
查看完整验算程序
# Chapter 6. p61: the upper bound on what a waiting drill can buy
# 模型:若某项训练让 4 岁半儿童的等待能力提高 delta 个标准差,15 岁成就
# 的期望增益 = beta * delta * 2^(-10.5 / h),h 为训练效应的半衰期(年),
# 10.5 年为 4 岁半到 15 岁的间隔;beta 取 Watts 等 (2018) 三种模型的系数
# (出处),并假定它们是因果的(这是对干预最乐观的读法)。
# delta = 0.5 与半衰期 1、2、5 年及“不消退”均为示意。
import math
BETAS = [("raw 0.236", 0.236), ("early 0.081", 0.081),
("full 0.050", 0.050)]
YEARS, DELTA = 10.5, 0.5
def gain(beta, delta, half):
keep = 1.0 if half is None else 2 ** (-YEARS / half)
return beta * delta * keep
def main():
halves = (None, 5.0, 2.0, 1.0)
head = "".join("{:>10}".format("none" if h is None else
"h={:.0f}y".format(h)) for h in halves)
print("gain at 15 in SD, delta = {:.1f} SD of waiting".format(DELTA))
print("{:<13}{}".format("beta", head))
for lab, b in BETAS:
row = "".join("{:>10.4f}".format(gain(b, DELTA, h)) for h in halves)
print("{:<13}{}".format(lab, row))
print("same in points on a mean-100, SD-15 scale:")
for lab, b in BETAS:
row = "".join("{:>10.2f}".format(15 * gain(b, DELTA, h))
for h in halves)
print("{:<13}{}".format(lab, row))
print("share kept after 10.5 y with h = 2: {:.4f}".format(
2 ** (-YEARS / 2.0)))
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 6 章「干预的天花板:消退、乐观上界与两个杠杆」(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}
}