关于「系统架构与运行规格:申请流程、公开登记册、周度台账与预注册检验」,《达人认证的信号与分级》给出的定量结论是什么?
结论:第八章给出申请流程、数据表、只追加的公开登记册、周度台账、认证录取率与分级切点试算器及预注册检验方案。书中以标注框区分已发表的证据、可运行的 Python 计算(20 段程序)、原典与作者的假说。
数据口径:书中的达人人数、品牌数、打分误差、成本、溢价与审计参数均为示意值,全部计算由书中程序生成;本书是非个人化经营建议,不构成法律、税务或投资建议;书中的认证标准公开可测,费用与是否通过无关,不暗示任何政府或法定资格,可以经审计撤销,不制造虚假稀缺,推荐奖励只限单层、只给直接推荐人,不涉及任何多层分润。 书中以标注框区分已发表的证据、可运行的 Python 计算(20 段程序)、原典与作者的假说。
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: 20
| power 0.8 | |
| H1 creators per group (effect 0.10) | 142 |
| H1 if the effect is only 0.05 | 566 |
| H2 blind re-audits (0.10 vs 0.20) | 69 |
| H3 eligible creators per arm | 963 |
| power 0.9 | |
| H1 creators per group (effect 0.10) | 190 |
| H1 if the effect is only 0.05 | 757 |
| H2 blind re-audits (0.10 vs 0.20) | 102 |
| H3 eligible creators per arm | 1,289 |
| at 20 re-audits a week H2 needs 4 weeks |
查看完整验算程序
# Chapter 8. p82: sample sizes for the three pre-registered tests.
# Two-sided alpha 0.05, power 0.8 and 0.9. Normal approximations.
# H1 (holder premium): newly certified vs eligible not-yet-certified
# creators, outcome = post-minus-pre mean log inquiries per creator
# (collapsed as in p42); SD of that difference 0.30, effect 0.10.
# H2 (credibility): blind re-audits; null false-pass rate 0.10 against
# the alternative 0.20 that would trigger a standard review (one-sided
# alpha 0.05 for this monitoring test).
# H3 (coarse grades attract applicants): eligible creators randomised by
# cohort to a pass/fail or a three-level display; outcome = application
# rate, 0.30 against 0.36.
# All effect sizes and SDs are illustrative planning values.
import math
from statistics import NormalDist
ND = NormalDist()
def z(p):
return ND.inv_cdf(p)
def n_means(sd, delta, power, alpha=0.05):
return math.ceil(2 * ((z(1 - alpha / 2) + z(power)) * sd / delta) ** 2)
def n_one_prop(p0, p1, power, alpha=0.05):
a = z(1 - alpha) * math.sqrt(p0 * (1 - p0))
b = z(power) * math.sqrt(p1 * (1 - p1))
return math.ceil(((a + b) / (p1 - p0)) ** 2)
def n_two_prop(p0, p1, power, alpha=0.05):
pbar = (p0 + p1) / 2
a = z(1 - alpha / 2) * math.sqrt(2 * pbar * (1 - pbar))
b = z(power) * math.sqrt(p0 * (1 - p0) + p1 * (1 - p1))
return math.ceil(((a + b) / (p1 - p0)) ** 2)
for power in (0.8, 0.9):
print("power %.1f" % power)
print(" H1 creators per group (effect 0.10): %d" % n_means(
0.30, 0.10, power))
print(" H1 if the effect is only 0.05: %d" % n_means(
0.30, 0.05, power))
print(" H2 blind re-audits (0.10 vs 0.20): %d" % n_one_prop(
0.10, 0.20, power))
print(" H3 eligible creators per arm: %s" % format(n_two_prop(
0.30, 0.36, power), ","))
weeks = math.ceil(n_one_prop(0.10, 0.20, 0.8) / 20)
print("at 20 re-audits a week H2 needs %d weeks" % weeks)唯一数据源与引用
范雅福(Fan, Y.). (2026). 《行业认证标准自举、斯宾塞信号分级与达人主动申领动力学母典》第 8 章「系统架构与运行规格:申请流程、公开登记册、周度台账与预注册检验」(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-358 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026358,
author = {Fan, Yafu},
title = {Bootstrapping Industry Accreditation Standards, Spence Signalling Tiers and Inbound Creator Certification Magnetics},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}