关于「性别相似性:重叠系数与按性别分流的误分上界」,《加工广度与后悔安全垫》给出的定量结论是什么?
结论:文献给出的量级很清楚:多数心理变量的性别差异落在接近零或小效应区间,组内差异远大于组间差异。d 为 0.30 时,按性别分流只比抛硬币高出六个百分点,组内方差是组间方差的 44.4 倍;规则 G 要追平一张分层页面,d 得大到 1.79。尖端组成的悬殊,是正态尾部的几何效应,不能拿来给多数人贴标签。
数据口径:梅耶斯-列维的选择性加工假说讲的不是感性对理性,而是加工广度:启发式加工者看少数显著线索,综合加工者把全部线索都算进去,线索彼此一致或过弱过强时差异消失。 海德的性别相似性假说与后续元合成显示,多数心理变量的性别差异很小,组内差异远大于组间差异;d 为 0.30 时,按性别猜加工风格的最好命中率只有 0.560,按性别分流要追平一张分层页面,d 须达到 1.79。
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: 0.30, 44.4
| d OVL best acc misclass var share within/between | |
| 0.05 | 0.980 0.510 0.490 0.0006 1,600.0 |
| 0.10 | 0.960 0.520 0.480 0.0025 400.0 |
| 0.20 | 0.920 0.540 0.460 0.0099 100.0 |
| 0.30 | 0.881 0.560 0.440 0.0220 44.4 |
| 0.35 | 0.861 0.569 0.431 0.0297 32.7 |
| 0.50 | 0.803 0.599 0.401 0.0588 16.0 |
| 0.65 | 0.745 0.627 0.373 0.0955 9.5 |
| 0.80 | 0.689 0.655 0.345 0.1379 6.2 |
| 1.00 | 0.617 0.691 0.309 0.2000 4.0 |
| Cohen's labels for d: small 0.2, medium 0.5, large 0.8 | |
| A random member of the higher group outscores a random member | |
| of the lower group with probability Phi(d / sqrt 2): | |
| d | 0.10: 0.528 |
| d | 0.30: 0.584 |
| d | 0.50: 0.638 |
查看完整验算程序
# Chapter 2. p21: overlap, misclassification and variance share for d.
# Two normal groups with equal sd and standardised mean gap d.
# OVL = 2 Phi(-|d|/2) (Inman and Bradley 1989); the best a gender-only rule
# can do with equal group sizes is accuracy Phi(|d|/2); the share of total
# variance carried by group membership is d^2 / (d^2 + 4). The d values
# span Hyde's bands: close to zero (<= 0.10), small (0.11 to 0.35),
# moderate (0.36 to 0.65), large (0.66 to 1.00).
import statistics
PHI = statistics.NormalDist().cdf
DS = (0.05, 0.10, 0.20, 0.30, 0.35, 0.50, 0.65, 0.80, 1.00)
if __name__ == "__main__":
print(" d OVL best acc misclass var share within/between")
for d in DS:
ovl = 2 * PHI(-d / 2)
acc = PHI(d / 2)
share = d * d / (d * d + 4)
ratio = 1.0 / (d * d / 4)
print(" {:.2f} {:.3f} {:.3f} {:.3f} {:.4f} {:>9,.1f}"
.format(d, ovl, acc, 1 - acc, share, ratio))
print("Cohen's labels for d: small 0.2, medium 0.5, large 0.8")
print("A random member of the higher group outscores a random member")
print("of the lower group with probability Phi(d / sqrt 2):")
for d in (0.10, 0.30, 0.50):
print(" d = {:.2f}: {:.3f}".format(d, PHI(d / 2 ** 0.5)))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《性别消费决策异质性、综合信息加工假说与女性高客单转化工程母典》第 2 章「性别相似性:重叠系数与按性别分流的误分上界」(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-373 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026373,
author = {Fan, Yafu},
title = {Gendered Consumption Decision Heterogeneity, the Selectivity Hypothesis and High-Conversion Architecture for Female Markets},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}