On the basis of the comparison principle, we have developed a universal algorithm for testing hypotheses, making this procedure simple and transparent. The algorithm is based on the formula for calculating the probability of filling the bucket of the experimental histogram by the probability distribution of the random process source. It has been established that the probability of filling the bucket is described by the binomial distribution. The relation of this formula to the order measure formula has been shown.
Similar content being viewed by others
References
E. V. Chernukho, Comparison principle as an alternative of the substitution principle in metrology and statistics, in: Heat and Mass Transfer-2008, A. V. Luikov Institute of Heat and Mass Transfer of the National Academy of Sciences of Belarus, Minsk (2009), pp. 418–423.
E. V. Chernukho, Solution of the problem of estimating arbitrary distribution parameters by the data of a repetitive experiment, Inzh.-Fiz. Zh., 83, No. 2, 403–409 (2010).
Author information
Authors and Affiliations
Additional information
Translated from Inzhenerno-Fizicheskii Zhurnal, Vol. 83, No. 3, pp. 566–573, May–June, 2010.
Rights and permissions
About this article
Cite this article
Chernukho, E.V. Universal algorithm for testing statistical hypotheses with respect to the distribution. J Eng Phys Thermophy 83, 602–610 (2010). https://doi.org/10.1007/s10891-010-0382-7
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10891-010-0382-7