Week 8 Discussion Discuss the differences between non-parametric and parametric tests

Sample Answer

Parametric and non-parametric tests are two types of approaches that researchers use to test their studies. Parametric test has a standard dissemination that offers additional inferences and the design can be defined using statistics. Parametric tests offer generalizations when making statements about the mean of the core or parent population.

Parametric tests have standard dissemination that provide more inferences and can be statistically defined as a bell-shaped. Conversely, designs that do not fit this normal curve are non-parametric (Daraio et al., 2018). More specifically, parametric test data refers to the interval or ration and measures that men drawing more uncomplicatedness, hence the least influenced by the variables. On its part, non-parametric test is a hypothesis test that is not based on any underlying assumptions or generalizations.

This implies that it does not need population’s distribution to be referred by certain parameters. Non-parametric test is primarily based in variations in medians. Imperatively, it is also called the distribution-free test. The test assumes that all variables are measures on an ordinal or nominal level (Srividhya et al., 2021). The test is used when the independent variables are non-metric.

There are different examples to illustrate these two types of tests. For instance, the Pearson test and one-way analysis of variance (ANAOVA) are common parametric tests. The ANOVA tests can measure two groups that display a t-test (Orcan, 2020). Conversely, Spearman testing is a common example of non-parametric tests in statistical analysis, particularly correlation test. The Friedman’s ANOVA test is also an example of non-parametric test used in statistical analysis.
Investigators running the test must make certain assumptions.

Firstly, there must be asymmetric or normal distribution in situations where there is gathering of information with same differences. Secondly, investigators must also assume that data analysis on a project should show effective interactions with all possible variables. Again, the research should use the mean like the main measurement of tendency for the collected demographics.

References

• Daraio, C., Simar, L., & Wilson, P. W. (2018). Central limit theorems for conditional efficiency measures and tests of the “separability” condition in non‐parametric, two‐stage production models. The Econometrics Journal, 21(2), 170-191. https //doi.org/10.1111/ectj.12103
  Orcan, F. (2020). Parametric or non-parametric Skewness to test normality for mean comparison. International Journal of Assessment Tools in Education, 7(2), 255-265. https //doi.org/10.21449/ijate.656077
  
• Srividhya, K., Radhika, D. A., & M. Haripriya. (2021). Analysis of subsistence data by assessing the significance and subsistence time using non-parametric and semi-parametric accession. Psychology and Education Journal, 58(2), 5269-5277.
  https //doi.org/10.17762/pae.v58i2.2931