Controlled Versus “Random” Experiments: A Principle

  • Werner G. MüllerEmail author
  • Henry P. Wynn
Conference paper
Part of the Contributions to Statistics book series (CONTRIB.STAT.)


The contrast and tension between controlled experiment and passive observation is an old area of debate to which philosophers of science have made contributions. This paper is a discussion of the issue in the context of modern Bayesian optimal experimental design. It is shown with simple examples that a mixture of controlled and less controlled experiments can be optimal, and this is stated as a general principle. There is a short discussion of a wider theory in the last section.


Shannon Information Optimal Experimental Design Passive Observation Determinental Identity Optimum Allocation Problem 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    Bacon, F.: Novum organum scientiarum (1620). In: Jardine, L., Silverthorne, M. (eds.) Cambridge Philosophical Texts (2000)Google Scholar
  2. 2.
    Bernard, C.: Introduction à l’étude de la médecine expérimentale, Paris (1865). English edition, Dover, New York (1957)Google Scholar
  3. 3.
    Chaloner, K., Verdinelli, I.: Bayesian experimental design: a review. Stat. Sci. 10, 273–304 (1995)CrossRefzbMATHMathSciNetGoogle Scholar
  4. 4.
    El-Gamal, M.A., Palfrey, T.R.: Economical experiments: Bayesian efficient experimental design. Int. J. Game Theory 25, 495–517 (1996)CrossRefzbMATHMathSciNetGoogle Scholar
  5. 5.
    Hainy, M., Müller, W.G., Wynn, H.P.: Learning functions and approximate Bayesian computation design: ABCD. Entropy 16, 4353–4374 (2014)CrossRefMathSciNetGoogle Scholar
  6. 6.
    Kasy, M.: Why experimenters should not randomize, and what they should do instead. Technical report 36154, Harvard University OpenScholar Working Paper (2013)Google Scholar
  7. 7.
    List, J.A., Sadoff, S., Wagner, M.: So you want to run an experiment, now what? Some simple rules of thumb for optimal experimental design. Exp. Econ. 14, 439–457 (2011)Google Scholar
  8. 8.
    Mill, J.S.: A System of Logic. John W. Parker, West Strand, London (1843)Google Scholar
  9. 9.
    McLean, R.A., Sanders, W.L., Stroup, W.W.: A unified approach to mixed linear models. Am. Stat. 45, 54–64 (1991)Google Scholar
  10. 10.
    Pearl, J.: Causality: Models, Reasoning and Inference, 2nd edn. Cambridge University Press, Cambridge (2009)CrossRefzbMATHGoogle Scholar
  11. 11.
    Taguchi, G.: Taguchi on Robust Technology Development: Bringing Quality Engineering Upstream. ASME Press (American Society of Mechanical Engineers), New York (1992)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2016

Authors and Affiliations

  1. 1.Department of Applied StatisticsJohannes Kepler UniversityLinzAustria
  2. 2.Department of StatisticsLondon School of EconomicsLondonUK

Personalised recommendations