Current Approaches in Reporting Statistical Significance in Scientific Research: Errors and Truths
2022
Akbulut, Omer
Statistical thinking and data analysis constitute an important part of the scientific research process. The most common inferential statistical approach within the scope of data analysis is the “null hypotesis significance test, NHST” process. The final product of this process is the pvalue. In this respect, p is the most widely used and famous measure of inferential statistics. Researchers mostly interpret their findings on the p-value. However, important errors are made both in reporting and interpreting the p-value. In this article, firstly, the principles of accurate reporting of descriptive statistics are discussed. Then, the presentation errors of the p-value were shown with examples and correct spelling examples were given. It has been suggested that the P-value be given as three decimals and the exact value of the probability. Care should be taken to give the p=0.000 value taken from the computer printouts as p<0.001. The situation where researchers make more mistakes is that they interpret the articles on the p-value alone and make the interpretation wrong. At the end of the article, the mistakes made in terms of p-value and the results of the articles dealing with this subject in the literature are compiled.
显示更多 [+] 显示较少 [-]