Introduction
The article Understanding Statistical Tests by Neideen and Brasel provides insights into the statistical tool for a specific set of data. Moreover, the paper discusses the various methods of testing data. It analyzes the multiple parametric tests, including the Pearson product correlation, Anova, and student t-test. Additionally, the nonparametric tests included are chi-squared, Spearman rank coefficient, and the Mann Whitney U test. Generally, the paper critically analyzes whether the results are statistically valid.
Similarly, the article on Parametric and Nonparametric: Demystifying the Terms by Tanya Hoskins provides an evaluation of the differences in the procedures. Furthermore, it contains the definitions of the tests as methods of analyzing statistical data. It emphasizes that parametric tests rely on assumptions, whereas nonparametric procedures do not necessarily rely on assumptions. It, therefore, addresses the methods used to measure the validity of the premise.
Delegate your assignment to our experts and they will do the rest.
The article on Nonparametric Tests by Hui Bian states its various types. It discusses the one sample, independent, and related tests. It also includes several steps taken when analyzing data using SPSS. Additionally, it provides a correct definition of the terms used in the analysis. It, however, emphasizes that the tests are applied to data on an ordinal scale, but in some cases, a simple measure is used. Similarly, the chapter review is from the book titled Essentials of Biostatistics in Public Health by Lisa M. Sullivan. Chapter 10 compares and contrasts parametric and nonparametric procedures. Moreover, it assesses various applications where the latter is appropriate. Further, it performs and interprets the Mann Whitney U test.
Nonparametric Tests
Statistical tests are techniques that rely on the probability distribution when reaching for a conclusion on the validity of a hypothesis. They, therefore, make a generalization of a population from its sample. Generally, both parametric and nonparametric tests are classifications of statistical procedures (Hoskins, 2012). The hypothetical testing based on differences is divided into two. Parametric tests are techniques whose model defines specific conditions on the parameters of the population from which the sample is drawn. The set standards include observations that must be independent, from normally distributed people with the same variances, and the variables must be measured in at least an interval scale (Bian, 2003). Moreover, the data should have equal variance and the same standard deviation. In contrast, nonparametric procedures do not dependent on conditions about the parameters of the population from which the sample is drawn. They, therefore, do not require stable measurements. Moreover, they are applied on an ordinal or nominal scale.
Various parametric tests are used. First, the Pearson product correlation coefficient provides the value used in determining how much two variables from the same sample correlate to each other. Second, the student t-test establishes whether a difference occurs between the means of two similar data sets (Neideen & Brasel, 2007). Third, the analysis of variance evaluates the means and deviations to determine whether sets of data are different or similar. It mainly uses the F-ratio to analyze data.
Several nonparametric tests are carried out. One, the chi-squared, is typically used to compare numerous groups whose output and input variables are binary. Two, the Spearman rank coefficient determines how well two linear or non-linear data predict each other. Third, the Mann Whitney U test compares the changes between two independent groups when the dependent variable is either ordinal or continuous but not necessarily distributed (Neideen & Brasel, 2007). Additionally, the Kruskal-Wallis test utilizes ranks of ordinal data to analyze variance to measure whether the sets of data are similar to one another.
There are multiple situations in which nonparametric procedures are applied appropriately. First, when the underlying data do not meet the requirements of the population from which the sample is drawn. To begin with, skewedness ensures that parametric tests are less powerful since the mean is no longer the appropriate measure of the central tendency. Nonparametric tests, on the other hand, are perfect measures of skewed distributions represented medians. Second, they are used when the population from which the sample is drawn is too small. Its tools are appropriate when testing small sample sizes as it validates the spread of the data.
In contrast, parametric tests are reliable when measuring large samples. Third, when the data to be analyzed is nominal or ordinal. Parametric tests deal with only continuous sets of data. Nonparametric tests, however, deal with other types of variables such as technical and ordinal data. In such cases, these procedures are the most appropriate forms of testing data. According to Bian (2003), there are three types of nonparametric tests. Firstly, the one-sample criteria evaluate a particular field. A chi-squared test is used for categorical data; hence, it aids in testing the differences between the hypothetical and categorical variables. Moreover, the Kolmogorov-Smirnov procedure compares data from continuous fields. It, therefore, produces one sample test that evaluates whether the cumulative spread functions of an area is uniform or exponential. The binomial tests, however, deal with categorical data with two categories.
Secondly, the independent sample tests help in identifying the variations of two or more groups by using one or more nonparametric procedures. Thirdly, the related samples tests use the Wilcoxon signed-rank sum test to analyze the variances between two similar samples (Sullivan, 2017). Moreover, McNamara’s test aids in testing changes that occur in binary data.
Over-all, these articles and chapters have redefined my understanding of hypothesis testing. For instance, I also thought the Mann Whitney U test and the Wilcoxon signed-rank sum test are different. The papers have, however, provided adequate information on the examination by indicating that they are the same procedure. Moreover, the articles have aided in understanding the use of SPSS.
Conclusion
The articles and chapters provide insights relevant to healthcare settings. An example of how a nonparametric procedure in healthcare setup is when measuring individual pain using a visual analog scale. Zero represents no pain, while ten indicates extreme pain following a sample of ten characters. Ranks are, therefore, assigned to perform a nonparametric assessment. Initially, the data is arranged from the smallest to the largest. When conducting this examination, it is paramount to assess three sums of the ranks before analysis. The procedure for performing this evaluation is; one setting up a hypothesis and the level of significance. Two, selecting the most appropriate testing tool. Three set up conditions for decisions since; in some cases, when the data is too large or too small, the null hypothesis is rejected. Four, summarize the ranks created to enable testing of the statistic. When finalizing the evaluation, one should assess the data to suggest a conclusion.
In summary, the papers provide a detailed analysis of the similarities, differences, and types of statistical tests. The nonparametric is mainly used as an alternative technique to the parametric tests. In some cases, the latter is used to analyze relatively large data. I comprehended some limitations to the use of nonparametric tests. One, it is normally less potent than corresponding parametric procedures. It is, thus, less likely to reject the null hypothesis if the data is from a reasonable spread. Two, it becomes tedious to analyze data from large samples for data manipulation. The papers, however, provided the various methods of using SPSS when analyzing data using nonparametric tests.
References
Bian, H. (2003). Basic statistics.
Hoskin, T. (2012). Parametric and nonparametric: Demystifying the terms. In Mayo Clinic (pp. 1-5).
Neideen, T., & Brasel, K. (2007). Understanding statistical tests. Journal of surgical education, 64(2), 93-96.
Sullivan, L. M. (2017). Essentials of biostatistics in public health. Jones & Bartlett Learning.
