Annals of Applied Statistics

On regression adjustments in experiments with several treatments

David A. Freedman

Full-text: Open access


Regression adjustments are often made to experimental data. Since randomization does not justify the models, bias is likely; nor are the usual variance calculations to be trusted. Here, we evaluate regression adjustments using Neyman’s nonparametric model. Previous results are generalized, and more intuitive proofs are given. A bias term is isolated, and conditions are given for unbiased estimation in finite samples.

Article information

Ann. Appl. Stat., Volume 2, Number 1 (2008), 176-196.

First available in Project Euclid: 24 March 2008

Permanent link to this document

Digital Object Identifier

Mathematical Reviews number (MathSciNet)

Zentralblatt MATH identifier

Models randomization experiments multiple regression estimation bias balanced designs intention-to-treat


Freedman, David A. On regression adjustments in experiments with several treatments. Ann. Appl. Stat. 2 (2008), no. 1, 176--196. doi:10.1214/07-AOAS143.

Export citation


  • Altman, D. G., Schulz, K. F., Moher, D. et al. (2001). The revised CONSORT statement for reporting randomized trials: Explanation and elaboration., Ann. Internal Medicine 134 663–694.
  • Chattopadhyay, R. and Duflo, E. (2004). Women as policy makers: Evidence from a randomized policy experiment in India., Econometrica 72 1409–1443.
  • Cook, N. R., Cutler, J. A., Obarzanek, E. et al. (2007). Long term effects of dietary sodium reduction on cardiovascular disease outcomes: Observational followup of the trials of hypertension prevention., British Medical J. 334 885–892.
  • Cox, D. R. (1956). A note on weighted randomization., Ann. Math. Statist. 27 1144–1151.
  • Freedman, D. A. (2008). Randomization does not justify logistic regression. Available at,
  • Freedman, D. A. (2007). On regression adjustments to experimental data., Adv. in Appl. Math. Available at
  • Freedman, D. A. (2006). Statistical models for causation: What inferential leverage do they provide?, Evaluation Review 30 691–713.
  • Freedman, D. A. (2005)., Statistical Models: Theory and Practice. Cambridge Univ. Press, New York.
  • Gertler, P. (2004). Do conditional cash transfers improve child health? Evidence from PROGRESA’s control randomized experiment., American Economic Review 94 336–341.
  • Hill, A. B. (1961)., Principles of Medical Statistics, 7th ed. The Lancet, London.
  • Hoeffding, W. (1951). A combinatorial central limit theorem., Ann. Math. Statist. 22 558–566.
  • Hoeffding, H. (1963). Probability inequalities for sums of bounded random variables., J. Amer. Statist. Assoc. 58 13–30.
  • Höglund, T. (1978). Sampling from a finite population: A remainder term estimate., Scand. J. Statist. 5 69–71.
  • Kempthorne, O. (1952)., The Design and Analysis of Experiments. Wiley, New York.
  • Neyman, J. (1923). Sur les applications de la théorie des probabilités aux experiences agricoles: Essai des principes., Roczniki Nauk Rolniczych 10 1–51. [In Polish. English translation by D. M. Dabrowska and T. P. Speed (1990). Statist. Sci. 5 465–480 (with discussion).]
  • Rossouw, J. E., Anderson, G. L., Prentice, R. L. et al. (2002). Risks and benefits of estrogen plus progestin in healthy postmenopausal women: Principal results from the Women’s Health Initiative randomized controlled trial., J. American Medical Association 288 321–333.
  • Victora, C. G., Habicht, J. P. and Bryce, J. (2004). Evidence-based public health: Moving beyond randomized trials., American J. Public Health 94 400–405.