In statistical research, the reliability of the results obtained is of utmost importance because only confidence in their significance and correctness can be a predictor of a competent interpretation of the entire study. In other words, confidence in the significance of the results obtained must be high so that the researcher can create truly valuable conclusions and implications for his or her work. The primary tool for determining the statistical significance of the results is the p-value. The p-value is a mathematical model, a number that shows the viability of the null hypothesis; it is this criterion that ultimately makes the decision to reject or retain the null hypothesis depending on the critical threshold of significance (Resnick). Thus, the p-value is an extremely common tool in statistical studies of any type, so it is reasonable to investigate whether the value and validity of the results can be conveyed without using this parameter.
In fact, demonstrating the significance and value of results is impossible without the use of a p-value since the statistical paradigm of existing studies is deeply based on this model. However, it is pertinent to note that the focus vector of this parameter is rapidly changing. Specifically, a p-value is used to compare this criterion to a critical threshold, typically 0.05 (Wasserstein et al. 1). Resnick reports that the level of this threshold may be too high to allow invalid, unreliable results into publication (Resnick). Parameter boundary conditions, whether 0.049 or 0.051, also create some confusion in the results because they involve mathematical rounding: consequently, such results may not be meaningful in reality. In other words, this creates a situation in which the p-value can be misleading and suggest deliberately false results; therefore, a lower critical threshold needs to be standardized. In other words, we cannot discard the p-value within the current statistics, but we must advocate a more critical approach to using this parameter and not take any result for which the p-value<0.05 as automatically reliable.
Works Cited
Resnick, Brian. “What a Nerdy Debate About P-Values Shows About Science — And How to Fix It.”Vox, Web.
Wasserstein, Ronald L., Allen L. Schirm, and Nicole A. Lazar. “Moving to a World Beyond “p< 0.05”.” The American Statistician, vol. 73, no. 1, 2019, pp. 1-19.