The Assumptions of Parametric statistical testing versus the Assumptions of Nonp

The Assumptions of Parametric statistical testing versus the Assumptions of Nonparametric tests
COLLAPSE
A statistical test presents a quantitative decision-making mechanism for a process or processes. The objective is to determine whether sufficient evidence exists to “reject” a process-related conjecture or hypothesis. Additionally, hypothesis tests consider the uncertainty of the sample estimate. When they presume that the population follows a specific distribution, such as the normal distribution, with a set of parameters, the hypothesis tests are parametric. In contrast, nonparametric tests are utilized when certain population assumptions cannot be made. Typically, rank or ordinal data necessitate nonparametric analysis. Nonparametric tests are also known as distribution-free procedures. Since nonparametric tests make fewer assumptions than their parametric counterparts, they are more robust. Non-parametric models differ from parametric models in that the structure of the model is not specified a priori but is instead determined by the data. The term non-parametric does not infer that these models have no parameters, but rather that the number and character of the parameters are variable and not predetermined(Kaur & Kumar, 2015).
In contrast, nonparametric approaches are preferable to parametric approaches for analyzing large datasets due to their low computational cost and lack of modeling assumption requirements(Alhusain & Hafez, 2018). The assumptions for the nonparametric examination are lower than those for a parametric test, and it has been stated that it is preferable to use the nonparametric test if the assumptions are not met. When the underlying distributions have hefty tails or extreme skewness, however, these nonparametric tests are much more potent than their parametric counterparts(Kitchen, 2009).
Therefore, different authors of textbooks on statistical modeling have expressed varying degrees of concern regarding the use of variable selection methods in practice, dependent on their personal experience and the research fields where weak or strong theories may predominate. We presume that the statistician is responsible for the design and analysis of statistical studies, but is working in an interdisciplinary environment with the opportunity to discuss the meaning of variables with applied life scientists, who are frequently the principal investigators(Heinze et al., 2018). Consequently, any variable selection applied to a linear predictor model with correlated IVs will always alter the interpretation of the effects. This is of great importance in explanatory or descriptive models where the interpretability of regression coefficients is of interest.
References:
Alhusain, L., & Hafez, A. M. (2018). Nonparametric approaches for population structure analysis. Human Genomics, 12(1), 25. https://doi.org/10.1186/s40246-018-0156-4
Heinze, G., Wallisch, C., & Dunkler, D. (2018). Variable selection – A review and recommendations for the practicing statistician. Biometrical Journal. Biometrische Zeitschrift, 60(3), 431–449. https://doi.org/10.1002/bimj.201700067
Kaur, A., & Kumar, R. (2015). Comparative Analysis of Parametric and Non-Parametric Tests.
Kitchen, C. M. R. (2009). Nonparametric versus parametric tests of location in biomedical research. American Journal of Ophthalmology, 147(4), 571–572. https://doi.org/10.1016/j.ajo.2008.06.031
this is was qoustion :
Discuss the assumptions of parametric statistical testing versus the assumptions of nonparametric tests. Discuss why a researcher would select a nonparametric approach based on the data and when they would select parametric tests for their data set. Does it matter what type of variables have been collected in the dataset?