TESTINGTHE RESEARCH HYPOTHESIS
Mostresearchers in testing their hypothesis commonly apply the use ofalpha level of 0.05. The reason for preferring the alpha level isthat, the significance level of 0.05 is the standard condition,particularly in the sector of business and science. However, not allfields may prefer to use the standard level of testing the hypothesisif the result of the using the null hypothesis is exceptionallyserious. For instance, if the application of the null hypothesis mayresult to injuries, a deadly disease occurring or may even lead todeaths then it is important to avoid the type one error. In suchexperiments, it is important to reject the null hypothesis even inthe circumstances that the null hypothesis is true[ CITATION Bar15 l 1033 ].The reason for taking such a critical step is to avoid making type 1error irrespective the presence of the significant level. Hence, forexperiments mentioned above with serious results, there is need tomake the alpha level smaller than the standard level commonly used.For such experiments that cannot endure the standard level chance ofbeing wrong, one can apply a lower significant level of 1% or 0.1%.For example, medical doctors may use the lower significant level of0.001 if there is a likelihood of death, injury or a serious disease.
Alternatively,researchers may choose to accept a higher significant level that thestandard level, if the penalty of being wrong is particularlynegligible. For instant, the field of insurance may opt to use asignificant level of 0.1, which is higher than 0.05 a standardsignificant level however in rare situations[ CITATION Bar15 l 1033 ].To note, it is common to use a lower significant level than thestandard level but rare to use a higher significant level. Thepractice ensure to reduce the type 1 error as much as possible
Barsalou,M. A., & American Society for Quality. (2015). TheASQ pocket guide to statistics for six sigma black belts.