أثر العلاقة بين حجم العينة والانحراف المعياري علي قوة الاختبارT لتوزيع بواسو

Abstract

Abstract:

    This paper aimed to recognize the power of statistical tests and that is to test the (T) of the sample per data that follow a Poisson distribution, which emanated from the Monte Carlo method to simulate the samples and the data generated from alternative distributions which follow a Poisson distribution. They are generated in 1000 various sizes from all distributions given and then determine the power of the test by comparing the test with the standard sample.As represented in the study, the problem of lack of recognition to the ways in which they can control the properties of the statistical test, and thus lead to no increase in the accuracy and strength of the test. The study also aims to develop a test of significance (importance) as an objective method by which to analyze the strength of the statistical test and its relationship to the level of statistical significance and magnitude of the impact through the analysis of the data generated using a program (MATLAB), and which has been reached on a number of the most important results that the greater the sample size increased the statistical power of the test (T) increases and also the value of a beta, and through the statements that follow a Poisson distribution, and therefore the higher the standard deviation less than the value of test and increase beta, according to data that follow a Poisson distribution value. Therefore this paper delivered (gave out) a number of recommendations, including: the researcher reaches at an appropriate sample, which gives tangible results interesting measure of statistical significance test that lies in increasing the power of the test and reduce the size of the beta value.