关于「明说是糖丸:开放标签安慰剂的三次检验」,《开放标签与价格标签》给出的定量结论是什么?
结论:第一章复算开放标签安慰剂的三次检验:2010 年试点第 21 天的效应量为 0.788,2021 年三臂试验中开放标签相对不服药为 0.434、与双盲安慰剂相差无几,2025 年元分析中自评效应为 0.39、客观效应只有 0.09。
数据口径:书中参数为示意;全部计算由书中程序生成;本书不是医疗建议,不宣称任何产品或做法对身体或健康有任何作用,书中流程与工具为非个人化经营建议。 书中以标注框区分已发表证据、可运行的 Python 计算(21 段程序)、原典与作者假说。
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: 0.09, 0.39, 2021, 2025
| estimate SMD SE P(sup) NNT | |
| 2021, 11 trials | 0.72 0.168 0.695 2.6 |
| 2025 overall | 0.35 0.046 0.598 5.1 |
| 2025 clinical | 0.47 0.054 0.630 3.8 |
| 2025 non-clinical | 0.29 0.064 0.581 6.2 |
| 2025 self-report | 0.39 0.048 0.609 4.6 |
| 2025 objective | 0.09 0.105 0.525 19.7 |
| 2025 trim-and-fill | 0.28 0.048 0.578 6.4 |
| self-report minus objective | 0.30, SE 0.115, z 2.60, p 0.0093 |
| objective share of self-report effect | 0.231 |
| 2025 clinical over 2021 estimate | 0.653 |
| implied spread of true effects (tau, approx) | 0.265 |
| share of settings with true SMD below 0 | 0.094 |
查看完整验算程序
# Chapter 1. p13: Two meta-analyses, four years apart, and the gap
# between self-reported and objective outcomes.
# Published inputs: von Wernsdorff et al. (2021), Sci Rep: 11 trials,
# SMD 0.72 (95% CI 0.39 to 1.05). Fendel et al. (2025), Sci Rep: 63
# comparisons, n 4554, SMD 0.35 (0.26 to 0.44), prediction interval
# -0.17 to 0.87; clinical 0.47 (0.36 to 0.57); non-clinical 0.29 (0.17
# to 0.42); self-report 0.39 (0.30 to 0.49); objective 0.09 (-0.12 to
# 0.29); trim-and-fill 0.28 (0.18 to 0.37). SEs from the CIs, the
# z test of the self-report vs objective gap and the conversions are ours.
# 两次元分析:效应量的缩水与「自评 vs 客观」的落差。
import math
from statistics import NormalDist
Z = NormalDist()
Z975 = Z.inv_cdf(0.975)
ROWS = [("2021, 11 trials", 0.72, 0.39, 1.05),
("2025 overall", 0.35, 0.26, 0.44),
("2025 clinical", 0.47, 0.36, 0.57),
("2025 non-clinical", 0.29, 0.17, 0.42),
("2025 self-report", 0.39, 0.30, 0.49),
("2025 objective", 0.09, -0.12, 0.29),
("2025 trim-and-fill", 0.28, 0.18, 0.37)]
def nnt(d):
# Kraemer and Kupfer: 1 / (2 * P(superiority) - 1)
return 1 / (2 * Z.cdf(d / math.sqrt(2)) - 1)
print("estimate SMD SE P(sup) NNT")
se = {}
for name, d, lo, hi in ROWS:
se[name] = (hi - lo) / (2 * Z975)
print("{:<19} {:5.2f} {:.3f} {:.3f} {:5.1f}".format(
name, d, se[name], Z.cdf(d / math.sqrt(2)), nnt(d)))
gap = 0.39 - 0.09
s = math.hypot(se["2025 self-report"], se["2025 objective"])
z = gap / s
print("self-report minus objective: {:.2f}, SE {:.3f}, z {:.2f}, p {:.4f}"
.format(gap, s, z, 2 * (1 - Z.cdf(z))))
print("objective share of self-report effect: {:.3f}".format(0.09 / 0.39))
print("2025 clinical over 2021 estimate: {:.3f}".format(0.47 / 0.72))
tau = (0.87 - 0.35) / Z975
print("implied spread of true effects (tau, approx): {:.3f}".format(tau))
print("share of settings with true SMD below 0: {:.3f}"
.format(Z.cdf(-0.35 / tau)))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《开放标签安慰剂预测编码、价格标签期望增益与杯弓蛇影反安慰剂母典》第 1 章「明说是糖丸:开放标签安慰剂的三次检验」(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-201 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026201,
author = {Fan, Yafu},
title = {Open-Label Placebos and Predictive Coding, Price-Tag Expectancy Gains and the Nocebo of the Bow in the Cup},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}