🧠
Job title ≠ destiny
现代职业不一定存在于历史调查的分类里。别把“管理人员”的均值贴给 Product Manager。
CLASSIFICATION BIASAN EXTREMELY BIASED QUESTION · A FAIR TESTHeightly
Heightly
Suspect.
产品经理，真的比财务分析师矮吗？
把你的离谱假设带进实验室。Compare jobs, education and demographic groups — without confusing stereotypes with facts.
✳ BIAS IS A HYPOTHESIS, NOT A FINDING✳ CORRELATION ≠ DESTINY✳ SMALL SAMPLE, BIG ATTITUDE✳ DON'T DATE A P-VALUE✳ BIAS IS A HYPOTHESIS, NOT A FINDING
01 / The Bias Laboratory
Put your theory
on trial.
真实身高数据才能回答真实问题。默认展示合成模拟，可导入自己的匿名 CSV。
THE QUESTIONABLE HEIGHT EXPLORERSynthetic playground · 100% made-up participants
⚠ SIMULATION
Bring receipts, not vibes.
导入匿名 CSV，所有计算只在你自己的浏览器进行，不上传服务器。
ROWS MATCHING
—
MEAN HEIGHT
—
GROUPS VISIBLE
—
Average height by occupation
cm · 145—200 cm scale · n ≥ 10/group
⚠ Demo numbers are fictional. They are NOT height estimates for real professions or demographic groups.
THE GREAT OFFICE HEIGHT-OFF 🥊
Two roles. One extremely suspicious hypothesis.CHALLENGER A
VS
CHALLENGER B
⚠ Demo-only: randomly generated, fictional people.
02 / Ways to get fooled
The plot twists.
🎭
Simpson's paradox
不同职业的性别、年龄和地区构成可能不同。不做分层，平均值可能会骗人。
MIXING POPULATIONS📏
Sample ≠ population
谁会自愿提交身高？谁会夸大？样本偏差可能比两组平均身高差异还大。
SELECTION BIAS03 / Research, not rumours
Real science.
Absurd questions.
2002 · ERGONOMICS
Different jobs, different body measurements?
Hsiao and colleagues used US NHANES III data to study anthropometric differences among occupational groups. Not a PM-versus-FP&A scoreboard.
READ THE STUDY ↗2008 · NBER
Height, education & wages
Case, Paxson and Islam found an association between height and wages in the British Household Panel Survey; education and occupational selection helped explain the correlation.
READ THE PAPER ↗DATA / US NATIONAL SURVEYS
Where real data could come from
NHIS includes self-reported height and historical occupation categories. NHANES measures height and records education/background; records need suitable weights and definitions.
NHIS HEIGHT ↗ · NHANES ↗04 / Don't get sued by statistics
Our extremely
important fine print.
✅ What the site can do
- Visualize user-supplied, anonymized height observations.
- Group by occupation, education or self-reported ethnic background.
- Filter by age and gender; compare two occupations.
- Refuse group results when fewer than 10 observations remain.
🚫 What it cannot conclude
本网站不会声称某个种族、职业或学历的人“天生”更高，也不会拿群体平均数预测某个具体人的身高。用户上传的数据属于便利样本；没有概率抽样、权重及不确定性分析时，不能代表社会总体。演示数字均为合成数据，且不上传任何个人信息。