文本因果推断：用 LLM 与嵌入把文字变成可识别的因果变量

roles = {
    "披露是否提到裁员": "D 处理",
    "年报情绪得分": "Y 结果",
    "行业景气描述": "X 混淆",
}
for text_feature, role in roles.items():
    print(f"{text_feature:>16}  ->  {role}")

预期输出

      披露是否提到裁员  ->  D 处理
        年报情绪得分  ->  Y 结果
        行业景气描述  ->  X 混淆

import numpy as np
rng = np.random.default_rng(3)
n = 5000
D_star = rng.normal(size=n)          # true text construct
Y = 2.0 * D_star + rng.normal(size=n)
for sd in [0.0, 0.5, 1.0]:
    D_hat = D_star + rng.normal(scale=sd, size=n)  # measurement error
    beta = np.polyfit(D_hat, Y, 1)[0]
    print(f"noise sd={sd}: estimated beta = {beta:.2f}")

预期输出

noise sd=0.0: estimated beta = 2.00
noise sd=0.5: estimated beta = 1.60
noise sd=1.0: estimated beta = 1.00

# 正确做法：在 fold A 学表示/nuisance，在 fold B 估效应
# 错误做法：同一批数据既调表示又估效应 -> 乐观偏差
print("learn f_theta on fold A")
print("estimate tau on fold B")
print("never reuse outcome-supervised text features in-sample")

预期输出

learn f_theta on fold A
estimate tau on fold B
never reuse outcome-supervised text features in-sample

from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.linear_model import LinearRegression, LogisticRegression
from sklearn.model_selection import KFold
import numpy as np

X = TfidfVectorizer(max_features=30).fit_transform(docs).toarray()
y_hat, d_hat = np.zeros(n), np.zeros(n)
for tr, te in KFold(5, shuffle=True, random_state=1).split(X):
    y_hat[te] = LinearRegression().fit(X[tr], Y[tr]).predict(X[te])
    d_hat[te] = LogisticRegression(max_iter=500).fit(X[tr], D[tr]).predict_proba(X[te])[:, 1]

yr, dr = Y - y_hat, D - d_hat
theta = np.sum(dr * yr) / np.sum(dr ** 2)
print("naive:", round(np.polyfit(D, Y, 1)[0], 3))
print("text-adjusted:", round(theta, 3))

预期输出

naive: 1.70
text-adjusted: 1.05   # true effect = 1.0