This paper considers the problem of hypotheses testing of the mean of a Poisson distribution. Accordingly we consider the following test statistics: Wald, WCC, Score (S), FT, VS, RVS, Exact and Bayes test statistics. A simulation study based on both one and two sided alternatives has been conducted to compare the performances of the test statistics. The study suggests that for a large sample size, all proposed test statistics except VCC and FT perform well in the sense of correct type I error rate of the test and power. However, for a small sample size, Score and VS have better type I error rate and power properties than the other test statistics.
Golam Kibria, B. M. and George, Florence
Empirical Comparison of Some Test Statistics for Testing the Mean of a Poisson Distribution,
Applications and Applied Mathematics: An International Journal (AAM), Vol. 6,
1, Article 6.
Available at: https://digitalcommons.pvamu.edu/aam/vol6/iss1/6