In the wake of increasing cases of Alzheimer’s disease worldwide, many scholars have opted to conduct various researches to establish its underlying causes. Most statistical tests are chosen based on formulated hypotheses. This essay elaborates on both parametric and non-parametric tests by providing various examples, applications, and underlying assumptions that the researcher must meet to accomplish them.
Parametric and Non-Parametric Tests
Parametric statistical tests are based on assumptions that specific properties of a population are normally distributed. Various forms of parametric tests include the t-test, z-test, F-test, and Analysis of Variance (ANOVA) (Geisser & Johnson, 2006).
On the other hand, non-parametric statistical tests fall under the null hypotheses. Various examples of non-parametric tests include the Chi-square tests, Fisher Exact Probability, Mann-Whitney, Kruskal-Wallis, Wilcoxon Signed-Rank, and Friedman tests (Burns & Groves, 2009).
A difference between the two tests is depicted in the assumption. A parametric test takes on a hypothesis that works in a uniformly distributed population, while a non-parametric test is used without an assumption of a particular distribution (Geisser & Johnson, 2006). Parametric tests are also assertive since they produce a more specific outcome to a well-designed question than non-parametric methods (Burns & Groves, 2009).
Usage of both tests
Most researchers use non-parametric tests in situations where the assumptions that are meant to warrant the use of parametric tests are not elaborated or clear. Conditions such as ranking, the existence of outliers, generalized null hypotheses, and/or usage of ordinal scales can also warrant the usage of non-parametric tests (Geisser & Johnson, 2006). Parametric tests are often used to study populations that have normal distributions and fixed parameters.
Assumptions that must be met by a researcher to run the tests
One of the assumptions that a researcher must meet during the study is distribution, which can be a theory of normality in cases where parametric tests are run. Parametric tests are more powerful than non-parametric techniques if a normal distribution assumption is achieved. Non-parametric tests favor infinite distribution tests (Bridge & Sawilowsky, 1999).
Data Analysis Plan
The analysis procedure that will be applied in the study of Alzheimer’s disease entails examining, coding, categorizing, tabulating, recombining, and sorting data based on demographic information and gender among others (Geisser & Johnson, 2006). Coding will be based on open, axial, and selective processes. This situation will enable reduction, reconstruction, and cleaning of data.
A Chi-square test will be used to address the hypothesis since the sample is chosen randomly. Independent scores and null hypotheses are also taken into account. Measures of central tendency including the mean, mean, and median will be used to determine demographic variables (Bridge & Sawilowsky, 1999).
Data Analysis for Demographic Variables
The demographic variables that will be considered include age, education, and experience among others. These variables will be explored using univariate analysis or tables. This method will be easily accomplished using the SPSS software version 21 or advanced functions of the Ms-Excel. The univariate analysis will show details concerning age distribution, central tendencies, and measures of dispersions such as range, standard deviation, and variance (Sugiyono, 2008).
Data Analysis Plan for Study Variables
The research will be based on the following hypothesis.
- H1: Public awareness of lifestyle issues such as lack of exercise and poor diet among others reduces the progression of Alzheimer’s disease.
Variables that will be analyzed include public awareness of lifestyle (independent) and progression of Alzheimer’s disease (dependent) (Sugiyono, 2008). In this case, inferential statistics will be used to elaborate on the general condition of the study.
Bridge, P., & Sawilowsky, S. (1999). Increasing physicians’ awareness of the impact of statistics on research outcomes: comparative power of the t-test and Wilcoxon rank-sum test in small samples applied research. Journal of clinical epidemiology, 52(3), 229-35.
Burns, N., & Groves, S. (2009). The Practice of Nursing Research: Appraisal, Synthesis, Generation of Evidence. Philadelphia, US: Saunders.
Geisser, S., & Johnson, W. (2006). Modes of parametric statistical inference. Hoboken, NJ: John Wiley & Sons.
Sugiyono, J. (2008). Statistical for Research. Bandung, Indonesia: Alfabeta Press.