关于「滑屏深度生存曲线:威布尔风险率、删失与断崖段」,《叙事传输与长文滑屏节律》给出的定量结论是什么?
结论:第五章把滑屏深度当作生存时间:负老化时前两屏流失 0.347,而平均滑屏深度把 8.55 屏的真实意愿记成 5.43 屏。
数据口径:书中参数均为示意;全部计算由书中程序生成;本书不是法律、广告合规或理财方面的个人化建议,也不教任何欺骗手法。 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: 5.43, 8.55
| depth | true S | naive share | Kaplan-Meier |
|---|---|---|---|
| 2 | 0.653 | 0.653 | 0.653 |
| 4 | 0.500 | 0.497 | 0.497 |
| 6 | 0.398 | 0.391 | 0.391 |
| 10 | 0.268 | 0.202 | 0.260 |
| 16 | 0.161 | 0.079 | 0.155 |
| sessions censored by the page end: 0.218 | |||
| naive mean depth 5.43; true mean depth 8.55 screens |
查看完整验算程序
# Chapter 5. p52: "Average scroll depth" versus a survival estimate.
# Readers' willingness to scroll is Weibull (K = 0.7, median 4 screens).
# Pages differ in length (LENGTHS screens); a reader who reaches the end
# of a page is censored, not lost: we never learn how far they would have
# gone. We compare the naive share "reached 10 screens" among all
# sessions with the Kaplan-Meier estimate and with the truth.
# Fixed-seed Monte Carlo; parameters are illustrative.
import math
import random
import statistics as st
rng = random.Random(20261112)
K, MED = 0.7, 4.0
LAM = MED / math.log(2) ** (1 / K)
LENGTHS = (6, 10, 16, 24)
N = 4000
obs = []
for _ in range(N):
d = LAM * (-math.log(1 - rng.random())) ** (1 / K)
length = rng.choice(LENGTHS)
obs.append((min(d, length), d < length))
obs.sort()
def km(at):
s, risk = 1.0, len(obs)
for t, event in obs:
if t > at:
break
if event:
s *= 1 - 1 / risk
risk -= 1
return s
print("depth true S naive share Kaplan-Meier")
for x in (2, 4, 6, 10, 16):
true = math.exp(-(x / LAM) ** K)
naive = len([1 for t, _ in obs if t >= x]) / N
print("{:5d} {:6.3f} {:11.3f} {:12.3f}".format(x, true, naive, km(x)))
cens = len([1 for _, e in obs if not e]) / N
print("sessions censored by the page end: {:.3f}".format(cens))
print("naive mean depth {:.2f}; true mean depth {:.2f} screens".format(
st.fmean([t for t, _ in obs]), LAM * math.gamma(1 + 1 / K)))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《叙事沉浸传输、认知失调破防与高转化长文滑屏多巴胺节律》第 5 章「滑屏深度生存曲线:威布尔风险率、删失与断崖段」(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-141 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026141,
author = {Fan, Yafu},
title = {Narrative Transportation, Cognitive-Dissonance Disarming and Dopaminergic Scroll Rhythms in High-Conversion Long-Form Copy},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}