Verbally, significance means something important for the research, while in statistics, significance means something true, not necessarily important. There are many opinions among scientists on the scientific value of a significance test. The vast majority of scientists criticize the use of significance test advising to return to more scientific methods of data analysis. The first point is that the test does not give us exactly what we want. The significance test results show us the probability of our data, given the assumption that there are no differences in the sample. Meanwhile, we need to know the probability that there are differences in the sample, given our data. (Significance in Statistics & Surveys)
The test of significance shows the probability that the changes in sample data occurred randomly, not due to the program. The reliability coefficient is chosen to be 0.5. Many scientists believe that it is somewhat vacuous to declare that the reliability coefficient is a nonzero number. The test of significance shows the maximum probability of type I error, or the error of accidental rejection of true hypothesis Ho. The greater the test coefficient the higher the probability of type I error. On the opposite side the smaller the test coefficient - the higher the probability of type II error. Choosing large size of data sample may result in high coefficient of significance even if the difference between hypotheses is very small (Significance test). The effect size test is used to calculate the widespread of significance difference.
There are many scientists criticizing the significance test as well as there are many researchers who do not fully understand the test.
Bibliography
Significance in Statistics & Surveys. (n.d.). Retrieved from Creative Research Systems: http://www.surveysystem.com/signif.htm
Significance test. (n.d.). Retrieved from http://psychology.wikia.com/wiki/Significance_test
PAGE \* MERGEFORMAT 1
Type I and Type II …