Question

Describe the condition in which a nonparemetric test would be a better selection than a paremetric test. illustrate your idea with a specific example of when you would use each type using similar variables for each examples.

Answer 1

Nonparametric checks don’t require that your records observe the normal distribution. They’re also recognized as distribution-free tests and can grant benefits in sure situations. Typically, people who perform statistical hypothesis tests are extra comfy with parametric exams than nonparametric tests.

Advantages of Nonparametric Tests

• Nonparametric tests assess the median which can be better for some study areas

nonparametric analyses furnish an advantage due to the fact they investigate the median alternatively than the mean. The mean is no longer always the higher measure of central tendency for a sample. Even even though you can perform a valid parametric analysis on skewed data, that doesn’t always equate to being the better method. Let me explain the usage of the distribution of salaries.

Salaries have a tendency to be a right-skewed distribution. The majority of wages cluster around the median, which is the point where 1/2 are above and half of are below. However, there is a long tail that stretches into the greater income ranges. This long tail pulls the imply some distance away from the central median value. The two distributions are ordinary for revenue distributions.

• Nonparametric tests are valid when our sample size is small and your data are potentially non normal

Use a nonparametric take a look at when your pattern measurement isn’t large enough to fulfill the necessities in the desk above and you’re now not sure that your facts follow the regular distribution. With small sample sizes, be conscious that normality checks can have inadequate energy to produce useful results.

This situation is difficult. Nonparametric analyses have a tendency to have decrease electricity at the outset, and a small pattern dimension only exacerbates that problem.

• Nonparametric tests can analyse ordinal data, ranked data and outliers

Parametric checks can analyze only continuous information and the findings can be overly affected by outliers. Conversely, nonparametric tests can additionally analyze ordinal and ranked data, and no longer be tripped up with the aid of outliers.

Sometimes you can legitimately remove outliers from your dataset if they characterize unusual conditions. However, on occasion outliers are a authentic phase of the distribution for a find out about area, and you ought to now not do away with them.

Parametric and Non-parametric tests for comparing two or more groups

It would appear prudent to use non-parametric assessments in all cases, which would retailer one the bother of checking out for Normality. Parametric checks are preferred, however, for the following reasons:

1. We are not often interested in a importance test alone; we would like to say something about the population from which the samples came, and this is high-quality completed with estimates of parameters and self assurance intervals.

2. It is hard to do bendy modelling with non-parametric tests, for instance allowing for confounding elements the use of more than one regression.

3. Parametric assessments normally have extra statistical strength than their non-parametric equivalents. In different words, one is more likely to discover full-size differences when they actually exist.