arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2601.18072stat.MEstat.APstat.CO
收藏

共线性与样本量对线性回归结果的影响:一项模拟研究

The effect of collinearity and sample size on linear regression results: a simulation study

National University of Singapore(新加坡国立大学);National University Polyclinics, National University Health System(新加坡国立大学诊所,新加坡国立健康系统);Department of Family Medicine, National University Health System(新加坡国立健康系统家庭医学系);Centre for Research in Health Systems Performance (CRiHSP), National University of Singapore(新加坡国立大学卫生系统绩效研究中心 (CRiHSP));Division of Informatics, Imaging and Data Sciences, Faculty of Biology, Medicine and Health, The University of Manchester(曼彻斯特大学生物、医学与健康学院信息学、成像与数据科学部)

Stephanie CC van der Lubbe, Jose M Valderas, Evangelos Kontopantelis

AI总结:

该研究通过模拟分析了共线性与样本量对线性回归结果的影响,发现共线性在小样本中会降低精度,但在大样本中影响较小,同时指出VIF阈值应结合样本量和潜在偏倚进行合理应用。

摘要:

背景:共线性会放大OLS系数的方差,扩大置信区间并降低推断的可靠性。然而,固定方差膨胀因子(VIF)的阈值往往被统一应用于样本量差异极大的研究中,尽管共线性是一个有限样本问题。我们量化了共线性与样本量对线性回归性能的联合影响,并提供了解释VIFs的实用指导。\n\n方法:我们模拟了样本量N=100-100,000和共线性水平VIF=1-50的数据。对于每个场景,我们生成1,000个数据集,拟合OLS模型,并评估覆盖率、平均绝对误差(MAE)、偏差、传统功率(置信区间排除0)和精度保证(概率95%置信区间落在预设的围绕真实效应的范围内)。我们还评估了偏倚、不恰当设定的设置,通过排除相关预测因子来研究偏倚放大。\n\n结果:在正确设定下,共线性未对名义覆盖率产生实质性影响,也未引入系统性偏差,但会减少小样本的精度:在N=100时,即使轻微共线性(VIF<2)会放大MAE并显著降低两种功率指标,而在N≥50,000时,即使VIF=50,估计结果仍稳健。在不恰当设定下,共线性会强烈放大偏倚,增加误差,降低覆盖率,并显著降低精度保证和传统功率,即使在低VIF下。\n\n结论:VIF阈值不应机械应用。共线性必须与样本量和潜在的偏倚来源相关联;仅通过移除预测因子来降低VIF可能会通过遗漏变量偏倚恶化推断。附带的热图提供了在研究规模和建模假设下的实用参考。

英文摘要:

Background: Multicollinearity inflates the variance of OLS coefficients, widening confidence intervals and reducing inferential reliability. Yet fixed variance inflation factor (VIF) cut-offs are often applied uniformly across studies with very different sample sizes, even though collinearity is a finite-sample problem. We quantify how collinearity and sample size jointly affect linear regression performance and provide practical guidance for interpreting VIFs. Methods: We simulated data across sample sizes N=100-100,000 and collinearity levels VIF=1-50. For each scenario we generated 1,000 datasets, fitted OLS models, and assessed coverage, mean absolute error (MAE), bias, traditional power (CI excludes 0), and precision assurance (probability the 95% CI lies within a prespecified margin around the true effect). We also evaluated a biased, misspecified setting by omitting a relevant predictor to study bias amplification. Results: Under correct specification, collinearity did not materially affect nominal coverage and did not introduce systematic bias, but it reduced precision in small samples: at N=100, even mild collinearity (VIF<2) inflated MAE and markedly reduced both power metrics, whereas at N>=50,000 estimates were robust even at VIF=50. Under misspecification, collinearity strongly amplified bias, increasing errors, reducing coverage, and sharply degrading both precision assurance and traditional power even at low VIF. Conclusion: VIF thresholds should not be applied mechanically. Collinearity must be interpreted in relation to sample size and potential sources of bias; removing predictors solely to reduce VIF can worsen inference via omitted-variable bias. The accompanying heatmaps provide a practical reference across study sizes and modelling assumptions.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑