方法对比

	普通最小二乘法 (OLS) 回归	面板数据固定效应模型	泊松回归与负二项回归
领域	计量经济学	计量经济学	计量经济学
方法族	Regression model	Regression model	Regression model
起源年份	2019	2014	1998
提出者	Wooldridge (textbook treatment); classical least squares	Hsiao (textbook treatment); within transformation of panel data	Cameron & Trivedi (textbook treatment); Hilbe (negative binomial)
类型	Linear regression	Panel data regression	Generalized linear model for count data
开创性文献	Wooldridge, J. M. (2019). Introductory Econometrics: A Modern Approach (7th ed.). Cengage Learning. ISBN: 978-1337558860	Hsiao, C. (2014). Analysis of Panel Data (3rd ed.). Cambridge University Press.	Cameron, A. C. & Trivedi, P. K. (1998). Regression Analysis of Count Data. Cambridge University Press.
别名	ordinary least squares, classical linear regression, linear regression, en küçük kareler regresyonu	fixed effects model, within estimator, panel fixed-effects regression, Panel Veri — Sabit Etkiler Modeli	count regression, log-linear count model, negative binomial regression, Poisson / Negatif Binom Regresyon
相关	5	5	4
摘要	Ordinary Least Squares is the classical linear regression method that explains a continuous outcome as a linear combination of predictors. It estimates the coefficients by minimising the sum of squared residuals, and under the Gauss-Markov assumptions these estimates are the best linear unbiased estimator (BLUE).	The Panel Data Fixed Effects model estimates relationships from panel data (the same units observed over several time periods) while controlling for unit- and/or time-specific effects, supporting causal inference. It is developed as the within estimator in standard treatments such as Hsiao's Analysis of Panel Data (2014).	Poisson regression is a generalized linear model for count outcomes — events tallied as non-negative integers such as hospital admissions, accidents, or article counts. It models the log of the expected count as a linear function of the predictors, and is developed in the standard count-data treatment of Cameron and Trivedi (1998); when the counts are over-dispersed, the closely related negative binomial model (Hilbe, 2011) is preferred.