Marginal effect

from Wikipedia, the free encyclopedia

As Marginal effect , even marginal effect or border effect , in the multivariate data analysis of the effect referred to an independent has on the dependent variable when it is changed by one unit and the other independent variables are held constant ( ceteris paribus ).

In simple KQ regression , the marginal effects correspond to the values ​​of the regression coefficients (beta values). In the case of nonlinear regression models, the marginal effects are not constant. Therefore, the following average effect indicators are used:

  • Marginal effects on mean briefly MEMs for English Marginal Effects at the Means
  • Average marginal effects, short AMEs for English Average Marginal Effects and
  • Marginal effects for representative units of investigation, MERs for short for English Marginal Effects at Representative values

See also

Literature on special applications

  • MJ Hanmer, K. Ozan-Kalkan: Behind the curve: Clarifying the best approach to calculating predicted probabilities and marginal effects from limited dependent variable models. In: American Journal of Political Science. Volume 57, No. 1, 2013, pp. 263-277.
  • Carina Mood: Logistic regression: Why we cannot do what we think we can do, and what we can do about it. In: European Sociological Review. Volume 26, No. 1, 2010, pp. 67-82.
  • I. Fernández-Val: Fixed effects estimation of structural parameters and marginal effects in panel probit models. In: Journal of Econometrics. Volume 150, No. 1, 2009, pp. 71-85.
  • S. Hoderlein, E. Mammen: Identification of marginal effects in nonseparable models without monotonicity. In: Econometrica. Volume 75, No. 5, 2007, pp. 1513-1518.
  • W. Greene: Marginal effects in the censored regression model. In: Economics Letters. Volume 64, No. 1, 1999, pp. 43-49.
  • LN Christofides, T. Stengos, R. Swidinsky: On the calculation of marginal effects in the bivariate probit model. In: Economics Letters. Vol. 54, No. 3, 1997, pp. 203-208.
  • WH Greene: Marginal effects in the bivariate probit model. 1996.

Individual evidence

  1. http://oldprof.typepad.com/a_dash_of_insight/2007/07/thinking-about-.html
  2. http://nd.edu/~rwilliam/stats3/Margins02.pdf
  3. http://www.nd.edu/~rwilliam/stats/Margins01.pdf
  4. http://www.stata-journal.com/sjpdf.html?articlenum=st0086