关于「实践与证伪方案:问诊的比特预算、话术红线与预注册检验」,《铁板神数的比特账》给出的定量结论是什么?
结论:第八章给出预注册的盲法检验,真实命中率为 0.10 时需 65 名求测者,并把结论蒸馏为一套客服问诊的话术骨架与红线。书中以标注框区分已发表的证据、可运行的 Python 计算(18 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;本书分析结构与概率,不对任何个人作任何占断,也不构成法律或理财建议。 第五章说明一句条文的期望信息量至多 0.5307 比特,按示意规模每分钟有 19.01 人出生。
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: 0.10, 65
| test | n | reject if hits >= | alpha | power |
|---|---|---|---|---|
| pair, chance vs 0.05 | 234 | 8 | 0.0059 | 0.9024 |
| pair, chance vs 0.10 | 65 | 4 | 0.0066 | 0.9004 |
| father, chance vs 0.20 | 109 | 17 | 0.0088 | 0.9009 |
| father, chance vs 0.30 | 37 | 8 | 0.0100 | 0.9053 |
查看完整验算程序
# Chapter 8. p81: Sample sizes for a pre-registered committed-cell test.
# The reader receives only the double-hour, commits in writing to one
# cell, and the cell's check verse is scored against facts sealed
# before the session. Chance hit rates come from p42: 0.01143 for the
# father-mother pair, 1/12 for the father's animal alone. We find the
# smallest n with a critical count c such that P(X >= c | chance)
# <= 0.01 and P(X >= c | claimed rate) >= 0.90, by exact binomials.
import math
def tail(n, c, p):
# P(X >= c) for X ~ Binomial(n, p)
return math.fsum(math.comb(n, k) * p ** k * (1 - p) ** (n - k)
for k in range(c, n + 1))
def design(p0, p1, alpha=0.01, power=0.90):
n = 5
while True:
c = 0
while tail(n, c, p0) > alpha:
c += 1
if tail(n, c, p1) >= power:
return n, c, tail(n, c, p0), tail(n, c, p1)
n += 1
cases = [("pair, chance vs 0.05", 0.01143, 0.05),
("pair, chance vs 0.10", 0.01143, 0.10),
("father, chance vs 0.20", 1 / 12, 0.20),
("father, chance vs 0.30", 1 / 12, 0.30)]
print(" test n reject if hits >= alpha power")
for name, p0, p1 in cases:
n, c, a, pw = design(p0, p1)
print(" {:<24}{:>4} {:>16} {:.4f} {:.4f}".format(
name, n, c, a, pw))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《铁板神数考刻分决策树、条文库仿射编码与信息熵套取下界》第 8 章「实践与证伪方案:问诊的比特预算、话术红线与预注册检验」(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-157 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026157,
author = {Fan, Yafu},
title = {Tieban Shenshu Ke-Fen Decision Trees, Affine Coding of the Verse Bank and Entropy Lower Bounds on Elicitation},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}