Home » Healthcare and Medicine » Nursing » Differences between Non-parametric and parametric Tests

# Differences between Non-parametric and parametric Tests

While the two statistical tools are necessary for deeper exploration of a data set, parametric tests such as f-test, t-test, and z-test are only used when information regarding the study subjects is already documented (Fagerland, 2012). Parametric tests also focus on distribution and utilize ratio and interval levels of measurement. A prime example of a parametric test is the distribution of students’ score in a classroom, a choice that is informed by the fact that students’ grade follows the long-standing normal distribution curve. Any set of data exploring a population-based issue conforms to the bell curve, where outlier scores are few, with the majority being average grades such as B, C, and D. In the event the distribution of a study subject is unknown, non-parametric tests are utilized (Gibbons & Chakraborti, 2011). The tests are based on arbitrary and make use of ordinal and nominal variables, with the focus being the median. While non-parametric tests are equally robust as parametric statistics, they are weak in documenting frequencies.

Assumptions to Run the Test

One key distinguishing feature in the choice of tests is distribution. While parametric tests are deemed more powerful in determining the significant effects, their use is limited by the normality of data (Dutta, 2014). On the other hand, non-parametric are distribution-free tests, where they are best applied to skewed data. Another consideration is the aim of the statistical engagements. For instance, in a skewed distribution, the preferred statistic in non-parametric tests because the focus is exploring the media rather than the mean. The choice of non-parametric is also considered in the event the sample size is not large enough to warrant conventional tests as well as the existence of outliers that cannot be eliminated.

Data Analysis Plan

With the study being based on concurrent nested approach, parametric tests will be the best statistic. The utility rests on the nature of the subject, where the incidences of bedsores follow a bell-shaped curve. The measures of central tendency to be reported for demographic variables will be mean and mode. Log-linear regression model will be used to address the study hypothesis, where it will be utilized in establishing the degree of association between the use of air mattress and incidences of bedsores.

Scroll to Top