In practice, parametric and non-parametric tests cannot be utilized concurrently. The assumptions that have traditionally guided their use are an underlying factor in their exclusive use. While parametric tests are thought to be more powerful in proving actual significant effects, their application is constrained by data normality (Ghasemi & Zahediasl, 2012). Non-parametric tests, on the other hand, are distribution-free tests that work best with skewed data. Similarly, their application is not constrained by prior information. However, in the case of parametric testing, they must be utilized in datasets from verified population-based challenges (Ghasemi & Zahediasl, 2012). However, the difference in the application of the tests does not mean that nothing is known about the subject being explored. For instance, both ANOVA and Kruskal-Wallis test are used in examining differences among means of categorized data. However, there use is determined by the set is conforming to bell curve. Besides, the choice of non-parametric is also considered in the event the sample size is not large enough to warrant conventional tests as well as existence of outliers that cannot be eliminated (Harrar, 2009). Another consideration is the aim of the statistical engagements. For instance, in a skewed distribution, the preferred statistic in non-parametric tests because the focus is exploring the media rather than the mean.
Despite the exclusivity when undertaking statistical explorations that emanate from real-life incidences, non-parametric and parametric tests can be used together when conducting research with dummy data. While it violates assumptions that guide the applicability of each test, the application is based on Monte Carlo simulations, where the focus is platitudinal when experts are examining the truism of the tools in concluding what is already known (Scott et al., 2016).
References
Ghasemi, A. & Zahediasl, S. (2012). Normality Tests for Statistical Analysis: A Guide for Non-Statisticians. International Journal Of Endocrinology And Metabolism, 10(2), 486-489. doi:10.5812/ijem.3505
Harrar, S. (2009). Asymptotics for tests on mean profiles, additional information and dimensionality under non-normality. Journal Of Statistical Planning And Inference, 139(8), 2685-2705. doi:10.1016/j.jspi.2008.12.008
Scott, S., Blocker, A., Bonassi, F., Chipman, H., George, E., & McCulloch, R. (2016). Bayes and big data: the consensus Monte Carlo algorithm. International Journal Of Management Science And Engineering Management, 11(2), 78-88. doi:10.1080/17509653.2016.1142191
