APA , citation ad reference. Please respond to this post
Oncology study involves the attempt to investigate how cancer can be diagnosed and treated through an evaluation of early symptoms in patients. The data obtained in such research is inferential on most occasions. Inferential statistical procedures can be categorized into two; parametric and non-parametric. When conducting researches, a majority of researchers can decide to use either parametric or non-parametric statistical analyses or test. However, the choice depends on the level of the data, such as nominal, ordinal, or continuous, that the researcher plans to examine. In the literal meaning, parametric tests involve making assumptions regarding the parameters of the population from which the researcher’s data is drawn. In contrary, non-parametric tests do not make such assumptions.
Parametric tests highly rely on the assumptions that the data that the researcher is testing resembles a particular distribution. For instance, a research on investigating the early signs and symptoms of cancer cannot be deemed to be nominal. This is because the symptoms vary from one patient to another. On the other hand, non-parametric tests are commonly referred to as distribution-free tests. This is because they do not involve limited assumptions to check in with regards to the distribution of the data. Additionally, parametric tests are highly selected when the dependent variable is being evaluated on a continuous scale. The non-parametric tests suit well when the dependent variable’s level of measurement is nominal or ordinal.
A good example of a parametric test is the t-tests and the analysis of variance. The investigator has to ensure that the underlying study population is normally distributed. Further, they must assume that the measures are deriving from an equal interval scale (Sullivan & Artino, 2013). For instance, the symptoms obtained during an oncology investigation can be analyzed using analysis of variance (ANOVA) to assess the difference between the early warnings or signs of cancer diseases. Through this approach, individuals and the health personnel will be able to mitigate the effects early enough before implication of the patient’s overall health. A Pearson correlation (r), which is a parametric test can be used to evaluate the relationship between unhealthy nutrition and cancer symptoms.
The non-parametric do not follow the assumptions made by an investigator while using parametric tests (Dergiades, Martinopoulos & Tsoulfidis, 2013). On some occasions, the non-parametric tests can be utilized as alternatives to parametric tests. For instance, t-test and analysis of variance (ANOVA) have non-parametric tests Mann-Whitney U test and the Kruskal-Wallis test respectively. The Spearman correlation (p) is an alternative to Pearson correlation and it is appropriate for application when at least one of the variables in a study is measured on an ordinal scale (Garson, 2014).
Finally, there are reasons behind a selection of either parametric or non-parametric test. The parametric tests have high performance when the spread of a sample data is different and when the investigator wishes to obtain high statistical power. Contrary, the selection of non-parametric tests can be as a result of a good representation of the investigator’s data by the median, a small sample size, and the data is ordinal or ranked. It is clear that making a decision on choosing between parametric and non-parametric tests is challenging. For instance, the oncology study might be involving both small sample size and non-normal data. According to various researchers, the representation of the center of distribution and sample size of the investigator’s data can dictate the choice of a statistical test.
Dergiades, T., Martinopoulos, G., & Tsoulfidis, L. (2013). Energy consumption and economic growth: Parametric and non-parametric causality testing for the case of Greece. Energy Economics, 36, 686-697.
Garson, G. D. (2014). Testing statistical assumptions. Asheboro, NC: Statistical Associates Publishing.
Sullivan, G. M., & Artino Jr, A. R. (2013). Analyzing and interpreting data from Likert-type scales. Journal of graduate medical education, 5(4), 541-542.