Non-Parametric and Parametric Tests

Non- parametric and parametric tests are statistical tests used in data analysis. Both are the main methodologies used to organize and draw inferences from sample data collected from a population. Non-parametric and parametric tests have a range of differences, and each has its assumptions. Choosing between non-parametric or parametric highly depends on the type of data available.

            A non-parametric test does not require any population distribution. It is a kind of hypothesis test which is not based on the underlying hypothesis. It is always based on the differences in media of the total population under study and thus is also known as a distribution-free test (Sprent & Smeeton, 2016). A parametric test draws its generalization from the mean of the population. It is based on the underlying hypothesis and includes the normal distribution of the variable.

            A good example of a non-parametric test is the Kruskal-Wallis test and the Mann-Whitney test. This can be applied when a researcher wants to determine the relationship between the number of hours of sleep and the frequency of falling ill since there is no normal distribution of illness among the people. In the case of parametric tests, z-test and t-test are the main examples (Kim, 2015). The parametric test assumes a normal distribution, and therefore parameters such as standard deviation can be applicable in this test.

            A non-parametric test is used when the underlying data does not meet the assumption of the population sample, or when the population sample is too small. On the contrary, a parametric test is used when the sample size is large because it is considered a powerful statistical test. While the non-parametric test does not make any assumptions, parametric tests assume that data have a normal distribution, homogeneity of variances, linearity, and independence.

            In conclusion, the two methods of statistical tests are widely used by researchers in the field of health and medicine. The choice of the test to use highly depends on the type of data available to the researcher.

References

Kim, T. K. (2015). T test as a parametric statistic. Korean journal of anesthesiology, 68(6), 540.

Sprent, P., & Smeeton, N. C. (2016). Applied non-parametric statistical methods. CRC press.