Research Method and Statistics

Ques. 1. Discuss the nature of nonparametric tests.

Ans. 1. Nonparametric tests are statistical tests that do not assume a specific
distribution of the data. They are used when the assumptions of parametric tests cannot
be met, such as when the data is not normally distributed or when the sample size is too
small.
• Nonparametric tests are also known as distribution-free tests because they do
not rely on any assumptions about the underlying distribution of the data.
Nonparametric tests are typically less powerful than parametric tests, but they
can still provide useful information about the data.
• Nonparametric tests are often used in fields such as psychology, sociology, and
medicine, where the data may not follow a normal distribution.
• Nonparametric tests are typically less powerful than parametric tests, but they
are more robust and can be used in a wider variety of situations.
• Nonparametric tests are often used to analyze data from small samples, or when
the data is not normally distributed. Examples of nonparametric tests include
Mann-Whitney U test, Wilcoxon signed-rank test, Kruskal-Wallis test, and
Spearman's rank correlation coefficient. These tests can be used to compare two
or more groups, measure relationships between variables, and assess
differences in distributions.

1. "Wilcoxon rank-sum test"

The Wilcoxon rank-sum test is a nonparametric statistical test used to compare two
independent samples. It is used to determine if there is a statistically significant
difference between the two samples. The test works by ranking the values from both
samples and then comparing the ranks of each sample. If the ranks of one sample are
significantly higher than those of the other, then it can be concluded that there is a
statistically significant difference between the two samples.

2. "Mann-Whitney U test"
The Mann-Whitney U test can be used to compare two independent samples, such as
comparing the performance of two different treatments or groups.
The test statistic is the difference between the means of the two samples. The U test is
usually computed as:
U=β+ε
where β is the Background Error Rate, ε is the standard error of the mean, and n is the
number of samples.
For both continuous and dichotomous data, the U test statistic is a measure of the
association between two groups of variables. It is used to determine whether there is a
significant difference in the rates of occurrence of different types of events between the
two groups. The U test is specific to tests of association, and does not consider any
potential confounding factors. It is used when testing hypotheses about whether two
groups have different rates of occurrence of events. The U test statistic can be
interpreted as indicating whether there is a significant difference in the rates of
occurrence between the two groups in terms of proportions (or degrees of separation).
When this happens, it indicates that there is a difference in the frequencies of
occurrence among the two groups, which could be indicative of different risk factors or
different types of events.

3. Kruskal-Wallis
The Kruskal-Wallis test is used to measure the distribution of data. It is composed of a
test statistic, k, and a weight, W. The purpose of the Kruskal-Wallis test is to determine
whether the data are distributed normally.
k = -(1 + w)

4. "Spearman's rank correlation coefficient"

The Spearman's rank correlation coefficient is a statistic that measures the correlation
between two variables. It is used to measure the strength of the association between
the two variables. The Spearman's rank correlation coefficient is typically used to
measure the relationship between two pairs of variables, such as a pair of variables in a
model. The Spearman's rank correlation coefficient is computed by calculating the
correlation between the two pairs of variables and then averaging the results. The
Spearman's rank correlation coefficient is usually represented in terms of a value, such
as r. The Spearman's rank correlation coefficient can be used to measure the strength
of the relationship between two pairs of variables. The value of the Spearman's rank
correlation coefficient can be used to determine whether there is a significant
relationship between the two pairs of variables.
Different types of nonparametric tests are: - p-value - t-value - k-value - alpha - beta -
gamma - Chi - Cramer's Rule p-value is a statistic used to Different types of
nonparametric tests are:

- p-value
- t-value
- k-value
- alpha
- beta
- gamma
- Chi
- Cramer's Rule

p-value is a statistic used to determine whether the sample is significantly different

from the null hypothesis of the test. It is a value that tells us how likely it is that the data
would be different if there were no difference in the variables. The test can be based on
a chi-squared statistic or a Fisher's exact test.

t-value is an statistic that tells us how often the data are different from the null
hypothesis of the test. It is also a value that tells us how many times there are data that
are different from the null hypothesis. The test can be based on a t-test or an analysis of
variance.

k-value is a statistic that tells us how often the data are different from the null
hypothesis of the test. It also gives us information about how many times there are data
that are different from the null hypothesis. The test can be based on a k-test or an
analysis of variance.

alpha is a statistic that tells us how effective (or powerful) the testing method was in
finding differences between groups of samples. alpha can be expressed as a
percentage or as an absolute value.

beta is a statistic that tells us how often (or often not) there are differences between
groups of samples. beta can be expressed as a percentage or as an absolute value.

gamma is a statistic that tells us about how much variation there is in tests within
groups of samples (ie, within each group there is more variation than between groups).
Gamma can be expressed as a percentage or as an absolute value. determine whether
the sample is significantly different from the null hypothesis of the test. It is a value that
tells us how likely it is that the data would be different if there were no difference in the
variables. The test can be based on a chi-squared statistic or a Fisher's exact test. t-
value is an statistic that tells us how often the data are different from the null hypothesis
of the test. It is also a value that tells us how many times there are data that are
different from the null hypothesis. The test can be based on a t-test or an analysis of
variance. k-value is a statistic that tells us how often the data are different from the null
hypothesis of the test. It also gives us information about how many times there are data
that are different from the null hypothesis. The test can be based on a k-test or an
analysis of variance. alpha is a statistic that tells us how effective (or powerful) the
testing method was in finding differences between groups of samples. alpha can be
expressed as a percentage or as an absolute value. beta is a statistic that tells us how
often (or often not) there are differences between groups of samples. beta can be
expressed as a percentage or as an absolute value. gamma is a statistic that tells us
about how much variation there is in tests within groups of samples (ie, within each
group there is more variation than between groups). Gamma can be expressed as a
percentage or as an absolute value.

Ques. 2. Describe anyone nonparametric test in detail with the help of a

demonstration or algorithm.

A nonparametric test is a test that does not allow for order of data. For example, if you
want to test whether a person is taller or shorter than another person, you would use a
nonparametric test. The test would compare the heights of the two people and report
the results in degrees.

One common nonparametric test is the Mann-Whitney U Test (also known as the
Wilcoxon Rank Sum Test). This test is used to compare two independent samples of
data and determine if there is a statistically significant difference between them. It can
be used to determine if two groups have different means, medians, or distributions.

The Mann-Whitney U Test works by ranking all of the data points from both samples
and then summing the ranks. If the two samples have different means, it will result in a
large sum for one of the samples and a small sum for the other. The test then uses this
information to calculate the probability that the two samples come from different
populations.

To illustrate, let's take an example of two samples of height measurements - Sample A

and Sample B.

Sample A contains 5 people with heights: 74, 76, 77, 78, 81 inches
Sample B contains 4 people with heights: 72, 75, 77, 79 inches

The first step is to rank all of the data points from both samples. For Sample A this
would be 1st = 74 inches (lowest), 2nd = 76 inches, 3rd = 77 inches, 4th = 78 inches
and 5th = 81 inches (highest). For Sample B this would be 1st = 72 inches (lowest), 2nd
= 75 inches, 3rd = 77 inches (tied for 3rd place in Sample A) and 4th = 79 inches
(highest).

Once all of the data points have been ranked from both samples we can then sum up
their ranks. For sample A this would be 1+2+3+4+5=15. For sample B this would be
1+2+3+(3)+(4)=13. Since sample A has a higher rank sum than sample B we can
conclude that it has a higher mean height than sample B with 95% confidence level.

The Mann-Whitney U Test is a powerful nonparametric test that can be used to compare
two independent samples. It is easy to understand and simple to implement, making it
one of the most popular nonparametric tests.