关于「身份标签的自我实现:贝叶斯自我知觉与名词效应的缩水」,《身份标签与自我说服的可证伪修辞学》给出的定量结论是什么?
结论:第一章用贝叶斯自我知觉模型说明身份标签只移动先验处于中段的读者,名词标签的最大增益只有 0.0468;小样本研究的量级错误可达 18.64 倍,由此提出核心反常识命题:名词身份标签在大样本现场条件下的提升接近于零。
数据口径:书中参数为示意;全部计算由书中程序生成;本书不是法律意见,也不是个人化的财务或经营建议。 书中以标注框区分已发表的证据、可运行的 Python 计算(19 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: 0.0468
| prior | noun post | verb post | uplift noun | uplift verb |
|---|---|---|---|---|
| 0.02 | 0.0316 | 0.0249 | 0.0046 | 0.0020 |
| 0.10 | 0.1509 | 0.1220 | 0.0204 | 0.0088 |
| 0.25 | 0.3478 | 0.2941 | 0.0391 | 0.0176 |
| 0.44 | 0.5570 | 0.4955 | 0.0468 | 0.0222 |
| 0.60 | 0.7059 | 0.6522 | 0.0424 | 0.0209 |
| 0.80 | 0.8649 | 0.8333 | 0.0259 | 0.0133 |
| 0.95 | 0.9682 | 0.9596 | 0.0073 | 0.0038 |
| noun: shift peaks at prior 0.4415, max uplift 0.0468 | ||||
| verb: shift peaks at prior 0.4721, max uplift 0.0223 | ||||
| mean noun uplift over uniform priors: 0.0314 | ||||
| mean noun uplift for priors <= 0.10: 0.0118 |
查看完整验算程序
# Chapter 1. p11: Who moves when a copy line hands the reader a label.
# Model: Bayesian self-perception. A reader holds a prior b that "I am
# the kind of person who invests in X". A noun label ("you are an
# investor") is treated as evidence with likelihood ratio LR; a verb
# wording ("you invest") as weaker evidence. Action probability is a
# linear link a0 + a1 * belief. All parameters are illustrative.
# The shift b' - b peaks exactly at b* = 1 / (1 + sqrt(LR)).
import math
A0, A1 = 0.05, 0.40 # illustrative action link
LR_NOUN, LR_VERB = 1.60, 1.25 # illustrative evidence strengths
def post(b, lr):
return b * lr / (b * lr + 1.0 - b)
def act(b):
return A0 + A1 * b
print("prior noun post verb post uplift noun uplift verb")
for b in (0.02, 0.10, 0.25, 0.44, 0.60, 0.80, 0.95):
pn, pv = post(b, LR_NOUN), post(b, LR_VERB)
print("{:5.2f} {:9.4f} {:9.4f} {:11.4f} {:11.4f}".format(
b, pn, pv, act(pn) - act(b), act(pv) - act(b)))
for name, lr in (("noun", LR_NOUN), ("verb", LR_VERB)):
bs = 1.0 / (1.0 + math.sqrt(lr))
print("{}: shift peaks at prior {:.4f}, max uplift {:.4f}".format(
name, bs, act(post(bs, lr)) - act(bs)))
# average uplift over a reader pool: priors on a fixed grid of 99 points
grid = [i / 100 for i in range(1, 100)]
mean_n = math.fsum(act(post(b, LR_NOUN)) - act(b) for b in grid) / 99
low = [b for b in grid if b <= 0.10]
mean_low = math.fsum(act(post(b, LR_NOUN)) - act(b) for b in low)
print("mean noun uplift over uniform priors: {:.4f}".format(mean_n))
print("mean noun uplift for priors <= 0.10: {:.4f}".format(
mean_low / len(low)))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《身份标签自我实现、群体归属召唤与高客单零抗拒自我说服修辞学》第 1 章「身份标签的自我实现:贝叶斯自我知觉与名词效应的缩水」(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-142 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026142,
author = {Fan, Yafu},
title = {Self-Fulfilling Identity Labels, In-Group Belonging Calls and the Rhetoric of Resistance-Free High-Ticket Self-Persuasion},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}