# 1 Parametric is when the researcher has information

1 Parametric is when the researcher has information

Discussion question #1

Parametric is when the researcher has information on the population and specific assumption are made about the population. A non-parametric test, also known as the Hypothesis test, is not based on any underlying assumptions and the researcher does not have information on the population. T test (Student’s T-Test) is a parametric test where for example, lets say one day you take a natural remedy to help you sleep and record how long it keeps you asleep. Then another day you take a prescribed medication and record how long it keeps you asleep (Statistics How To, n.d.). You ask your friends what their experience was and determine which one was more effective. Kruskal Wallis Test is a non-parametric test, which determines is the median between 2 or more groups are different (Stephanie, 2017).
Reference
Statistics How To. (n.d.). T Test (Student’s T-Test): Definition and Examples. Retrieved from http://www.statisticshowto.com/probability-and-sta…
Stephanie. (2017). Kruskal Wallis H Test: Definition, Examples & Assumptions. Retrieved from http://www.statisticshowto.com/kruskal-wallis/
Surbhi, S. (2016). Difference Between Parametric and Nonparametric test. Retrieved from https://keydifferences.com/difference-between-para…

Discussion question 2
Nonparametric Testing
Nonparametric testing may be used when criteria for parametric testing is not completely met or the study is assumed to have an abnormal distribution of data (Gray, Grove, & Sutherland, 2017). This type of testing is considered less powerful as compared to parametric testing (Kitchen, 2009). Instances where nonparametric testing would be used from abnormal distribution are when outcomes are ranked or ordinal variables are present, definite outliers are present, and the outcome has clear limits of detection (Sullivan, 2017). These tests have less ability to detect differences and have an increased risk of a type II error secondary to ranking of original data (Gray, Grove, & Sutherland, 2017). An example may include ranking of various types of trauma injuries over the course of a certain amount of time.
Parametric
Parametric testing has specific assumptions that must be adhered to in order to produce valid results and are more likely to find a significant effect (Gray, Grove, & Sutherland, 2017). If the specific criteria or “assumptions” are not met, the testing most likely may be conducted in the nonparametric manner. Parametric tests include data that follows a normal distribution pattern and are considered more powerful from a statistical analytic standpoint as compared to nonparametric tests (Gray, Grove, & Sutherland, 2017) (Kitchen, 2009). Assumptions specific to parametric testing are that the variances must be similar and able to be calculated with normal/close to normal distribution, variables are continuous and independent with measurement at a minimum of interval or ordinal data, and sampling is random (Gray, Grove, & Sutherland, 2017). The t-test is the most used statistical test for parametric data (Gray, Grove, & Sutherland, 2017) (Kitchen, 2009). Examples of parametric statistics would be the average hospital length of stay (LOS) for a certain age group with femur fractures where a mean would be calculated with an acceptable standard deviation.
Assumptions
Assumptions specific to parametric testing are that the variances must be similar and able to be calculated with normal/close to normal distribution, variables are continuous and independent with measurement at a minimum of interval or ordinal data, and sampling is random (Gray, Grove, & Sutherland, 2017). This type of test is appropriate if normalcy of data is known, independence of variables is not influential on one another, and variance in groups are relatively similar (Gray, Grove, & Sutherland, 2017).
References
Gray, J., Grove, S., & Sutherland, S. (2017). Burns and Groves The Practice of Nursing Research: Appraisal, Synthesis, and Generation of Evidence; Edition 8. Retrieved from South University Online Library: https://digitalbookshelf.southuniversity.edu/#/boo…
Kitchen, C. M. (2009, April 01). Nonparametric versus parametric tests of location in biomedical research. American Journal of Opthamology, 147(4), 571-572. doi:doi: 10.1016/j.ajo.2008.06.031
Sullivan, L. (2017, May 04). Bumc.bu.ed. Retrieved from Boston University School of Public Health: http://sphweb.bumc.bu.edu/otlt/mph-modules/bs/bs70…