1|Page

Statistics-II

Contents
Statistics-II............................................................................................................................................................................... 1

1. Standard scores (z-score) and its properties....................................................................................................................3

a. Standard scores............................................................................................................................................................3

b. Z-score.........................................................................................................................................................................3

2. The nature and properties of the normal probability distribution....................................................................................5

a. Nature.......................................................................................................................................................................... 5

b. Properties.....................................................................................................................................................................6

c. Standard Normal Variate.............................................................................................................................................7

a. Why E(X) is called mean?.......................................................................................................................................7

b. Explanation of Standard Normal Variate.................................................................................................................7

d. Area Property...............................................................................................................................................................9

3. Applications of normal probability curve......................................................................................................................10

4. Skewedness and Kurtosis..............................................................................................................................................11

a. Skewness....................................................................................................................................................................11

b. Kurtosis......................................................................................................................................................................12

5. The Meaning of Correlation & the scatterplot of bivariate distributions......................................................................13

6. Correlation: A Matter of Direction, a matter of degree.................................................................................................14

7. The Coefficient of Correlation.......................................................................................................................................15

8. Calculating Pearson’s Correlation Coefficient from Deviation Scores.........................................................................17

9. Calculating Pearson’s Correlation Coefficient from Raw Scores.................................................................................18

10. Spearman’s Rank-Order Correlation Coefficient......................................................................................................19

11. Population and Sample..............................................................................................................................................21

12. Standard error of mean, SD, and r;............................................................................................................................22

13. Level of Significance & Type I and Type II errors....................................................................................................23

14. Degrees of Freedom; One-tailed and two-tailed tests; Null and alternate hypothesis...............................................24

15. Meaning and difference between parametric and non-parametric test......................................................................25

16. T-test and Chi-square test..........................................................................................................................................26

17. Glossary.....................................................................................................................................................................27
2|Page

a. Coefficient.................................................................................................................................................................27

b. Origin Vs. Scale.........................................................................................................................................................28

c. Covariance.................................................................................................................................................................30

d. Standard Deviation Vs. Variance...............................................................................................................................31

e. Distribution types......................................................................................................................................................32

i. Normal...................................................................................................................................................................32

ii. Theoretical.................................................................................................................................................................32

iii. Bivariate...................................................................................................................................................................32

iv. Multivariate...............................................................................................................................................................32

v. Poisson.......................................................................................................................................................................32

vi. Binomial...................................................................................................................................................................32
3|Page

1. Standard scores (z-score) and its properties

a. Standard scores
Numerical values that represent a person's performance relative to others in a particular area. They are commonly used in
psychology to assess and compare individuals' cognitive, academic, and social abilities. Standard scores have several
properties that make them useful for this purpose:

1. Mean: The mean (average) of standard scores is always 50, with 95% of scores falling between 35 and 65.

2. Variance: Standard scores have a variance of 10, meaning that scores can vary by as much as plus or minus 5 from the
mean.

3. Symmetry: The distribution of standard scores is symmetrical around the mean, with the same number of scores.

b. Z-score
Definition:

In statistics, a z-score is a measure of how a specific data point deviates from the mean or average of a given dataset. It is
measured in terms of standard deviations from the mean. This measure is extensively used across many fields, including
psychology, owing to its easy interpretability and normalization property.

Properties/Features of Z-score

1. A z-score can be positive or negative depending on whether the data point is above or below the mean of the dataset.

2. The z-score of the mean of the dataset is always 0.

3. A positive z-score signifies the data point is above the mean while a negative z-score indicates that it is below the
mean.

4. Z-score helps in normalizing the data, allowing for comparisons across different scales.

5. The standard deviation of a set of z-scores is always 1.

Merits of Z-score

1. Z-scores can help detect outliers in the data by highlighting data points that are significantly greater or smaller than the
mean.

2. It simplifies comparison of scores from different datasets or distributions by converting them into a common standard
scale.

3. In psychological studies, they help in easily comparing and interpreting individual scores with respect to the overall
group mean and variability.
4|Page

Demerits of Z-score

1. Z-scores do not offer an intuitive understanding and require understanding of the concept of standard deviations.

2. Z-scores cannot be accurately computed or assigned meaning for discrete and categorical data.

3. The original data values cannot be recovered from the z-score unless the mean and the standard deviation of the
distribution are known.

Applications of Z-score

1. In psychology, z-scores are used to interpret test scores and measures of central tendency and variability.

2. They have wide application in statistical analyses like regression, ANOVA and hypothesis testing.

3. In educational psychology, z-scores are used to compare student performance against the class average.

4. In clinical psychology, they are used to categorize a patient's response on a particular measure and find its deviation
from the norm.

Rationale

The rationale behind the use of z-scores lies in their ability to standardize variables which have different scales or units
into a single, unified scale. This enables direct comparison between different datasets, even with different
5|Page

2. The nature and properties of the normal probability distribution

a. Nature
The normal probability distribution, also known as the Gaussian distribution or bell curve, is a crucial concept in statistics
and probability theory. It is commonly used to model various phenomena in natural and social sciences due to its
characteristic properties. While it is challenging to provide a statistician's definition that precisely captures all aspects of
the normal distribution without being tautological or repetitive, I will strive to present a comprehensive explanation.

Rationale of the Normal Probability Distribution:

The rationale behind the normal probability distribution lies in the central limit theorem, which states that the sum or
average of a large number of independent and identically distributed random variables tends to follow a normal
distribution, regardless of the underlying distribution of the individual variables. This is a crucial property that makes the
normal distribution an essential tool in statistical analysis.

Phenomenology of the Normal Probability Distribution:

A normal distribution is symmetrically shaped like a bell, with a peak in the middle and tapering off gradually towards
both ends. The mean, median, and mode of a normal distribution all coincide at the center. The distribution is
characterized by two parameters: the mean (µ) and the standard deviation (σ). The mean determines the location of the
peak, while the standard deviation controls the spread or variability of the data.

Utility of the Normal Probability Distribution:

The normal distribution has numerous practical applications in various fields. Some of its utility includes:

1. Statistical Inference: Many statistical tests and estimation techniques assume that the data is normally distributed. This
assumption allows researchers to make valid inferences about the population based on sample data. Examples include
hypothesis testing, confidence intervals, and regression analysis.

2. Data Analysis: The normal distribution provides a useful framework for understanding and analysing data. It allows us
to calculate probabilities, determine outliers, assess the significance of deviations, and identify trends and patterns.

3. Quality Control and Process Monitoring: The normal distribution is commonly used in quality control processes to
monitor and control variability. It helps in setting control limits and identifying when a process is out of control.

4. Modelling and Simulation: The normal distribution serves as a basis for developing mathematical models and
simulations. It enables researchers to study and predict the behaviour of complex systems, such as financial markets,
weather patterns, and population dynamics.

Philosophical Perspective:

From a philosophical point of view, the normal distribution plays a pivotal role in statistical reasoning and scientific
inquiry. It reflects the nature of uncertainty and variability that exists in the real world. By embracing the normal
distribution, we acknowledge that random variation is inherent in many natural and social phenomena. It allows us to
6|Page

make sense of the world by quantifying uncertainty, drawing meaningful inferences, and making informed decisions
based on probabilistic reasoning.

In conclusion, the normal probability distribution is a fundamental concept in statistics with widespread applications
across various fields. Its rationale lies in the central limit theorem, and it exhibits characteristic phenomenology with a
bell-shaped curve. The normal distribution's utility encompasses statistical inference, data analysis, quality control,
modelling, and simulation. From a philosophical perspective, it provides a framework for understanding uncertainty and
making sense of complex phenomena.
7|Page

b. Properties
i. Bell-shaped curve; top of the bell is directly above the mean
ii. Curve is symmetrical about the line X=mean, (z=0)
iii. Since distribution is symmetrical, mean, median, and mode coincide.
iv. Since Mean=Median=mode, the ordinate at X= mean, (z=0) divides the whole area into two equal parts.
Since total area under normal probability curve is 1, the area to the right of the ordinate as well as to the left
of the ordinate at X=mean or z=0 is 0.5
v. By virtue of symmetry, the quartiles are equidistant from median or mean i.e.,
i. Q3 – Md = Md – Q1
ii. Q1 + Q3 = 2Md = 2(Mean)
vi. Since distribution is symmetrical, the moment coefficient of skewness is given by
8|Page

c. Standard Normal Variate

a. Why E(X) is called mean?
The logic behind calling the expected value the mean in a standard normal variable is based on the definition of the
expected value and the properties of the normal distribution. In probability theory, the expected value of a random
variable is a way to quantify the average value or the long-term average outcome of the variable.

For a continuous random variable, such as a standard normal variable, the expected value is calculated by integrating the
product of each possible outcome and its corresponding probability density function over the entire range of the variable.
In the case of a standard normal variable, the probability density function is given by the bell-shaped curve of the normal
distribution.

The mean, on the other hand, is a measure of the central tendency of a set of values. It represents the average value of the
data points. In the case of the standard normal variable, the mean is equal to 0, since the distribution is symmetric around
the mean.

In the context of the standard normal variable, the expected value (or average value) is equal to the mean because of the
symmetry and properties of the normal distribution. The normal distribution has a mean of 0 and a symmetric bell-shaped
curve, which means that the average outcome (expected value) is located at the center of the distribution (mean). This is
why the expected value of a standard normal variable is often referred to as the mean.

b. Explanation of Standard Normal Variate

The standard normal variate is a standardized form of a random variable that follows a normal distribution. The rationale
behind the standardization is to transform any normally distributed variable into a new variable with a mean of 0 and a
standard deviation of 1. This standardization has practical, philosophical, and mathematical significance, as outlined
below:

Practical Rationale:

1. Comparability: Standardizing variables enables easy comparison and interpretation, as they are transformed to a
common scale. This simplifies statistical analysis and allows researchers to make meaningful comparisons between
different variables.

2. Simplifying calculations: Standardizing variables eliminates the need for complex computations involving different
mean and standard deviation values. It simplifies various statistical calculations, such as finding probabilities, confidence
intervals, or conducting hypothesis tests

3. Z-scores interpretation: The standard normal variate, often represented as a z-score, provides insights into the relative
position of a data point within the distribution. Z-scores indicate the number of standard deviations an observation is from
the mean. Positive values indicate being above the mean, while negative values indicate being below the mean.

Philosophical Rationale:
9|Page

1. Phenomenological: The standard normal variate serves as a reference point that transcends specific distributions or
datasets. It enables the examination of the relative standing of a data point, irrespective of the original data's specific
characteristics. This allows for a more objective analysis and comparison of data.

2. Epistemological: Standardization promotes the comparability and interchangeability of data across different contexts. It
helps researchers gain a deeper understanding of the underlying nature of a specific phenomenon as they can compare
their observations with the expected behavior of a standard normal distribution.

3. Ontological: The standard normal variate reflects the theoretical model of a normal distribution with a mean of 0 and a
standard deviation of 1. By transforming variables to adhere to this model, researchers can unveil patterns, explore
statistical relationships, and make inferences based on theoretical assumptions.

In summary, the standard normal variate, with its mean of 0 and standard deviation of 1, has practical advantages such as
comparability and simplified calculations. From a philosophical standpoint, it serves as a reference point transcending
specific datasets and allows for objective analysis and comparison of data. The standardization process aligns with
phenomenological, epistemological, and ontological aspects, guiding researchers in their exploration, understanding, and
interpretation of data within the framework of a normal distribution.
10 | P a g e

d. Area Property
Certainly! The area property of the normal probability distribution refers to the relationship between the probability of
observing certain values and the area under the curve of a normal distribution. The normal distribution is a bell-shaped
curve that is commonly used to model many natural phenomena.

Here is an explanation of the area property in a simple and elaborate manner:

1. The normal distribution: The normal distribution is characterized by its mean and standard deviation. It is a symmetric
curve that is centered around its mean, with the majority of the values falling close to the mean and fewer values at the
tails.

2. The area under the curve: The area under the curve of a normal distribution represents the probability of observing
values within a certain range. The total area under the curve is always equal to 1, which means that the probability of
observing any value in the entire distribution is 1 or 100%.

3. Finding probabilities: The area property allows us to find probabilities associated with specific values or ranges within
the normal distribution. By calculating the area under the curve, we can determine the likelihood of observing values
within that range.

4. Using z-scores: To find the area under the curve, we often use z-scores. A z-score quantifies how many standard
deviations a particular value is away from the mean. With the help of z-scores, we can standardize and compare values
across different normal distributions.

5. Areas between values: We can calculate the area between two specific values in the normal distribution by subtracting
the area to the left of the lower value from the area to the left of the higher value. This provides us with the probability of
observing values within that range.

6. Areas to the left or right: We can also calculate the area to the left or right of a specific value in the normal distribution.
This gives us the probability of observing values less than or greater than that particular value.
11 | P a g e

7. Standard normal distribution: The standard normal distribution is a special case of the normal distribution with a mean
of 0 and a standard deviation of 1. It is widely used because it simplifies calculations and provides a common reference
point.

8. Using tables or software: To calculate areas under the curve or probabilities, we can use standard normal tables or
statistical software. These resources provide pre-calculated values based on z-scores, making it easier to find the desired
areas.

In summary, the area property of the normal probability distribution relates the probability of observing specific values or
ranges to the corresponding areas under the curve. By using z-scores, we can calculate these areas and determine the
likelihood of encountering certain outcomes in a normal distribution.

3. Applications of normal probability curve

In psychology and statistics, the normal probability distribution is frequently utilized for various purposes, including:

1. *Psychological Testing*: When designing psychological tests and assessments, the normal distribution is often used to
understand the distribution of scores and determine norms for interpretation. For example, IQ scores are often assumed to
follow a normal distribution.

2. *Measurement Error*: In psychological research, measurement errors are common. The normal distribution is used to
model these errors when assessing the reliability and validity of tests and measurements.

3. *Statistical Analysis*: Normal distribution is frequently assumed in statistical analyses. Many statistical tests, such as t-
tests and analysis of variance (ANOVA), assume that the data follows a normal distribution. Deviations from normality
can affect the validity of these tests.

4. *Personality Assessment*: Some personality traits are assumed to follow a normal distribution, and this assumption is
used in the development and interpretation of personality assessments.

5. *Clinical Psychology*: In clinical psychology, the normal distribution is used to understand the distribution of
symptoms, scores on clinical assessments, and other variables related to mental health.
12 | P a g e

6. *Psychometric Analysis*: When evaluating the psychometric properties of psychological tests, such as reliability and
validity, the normal distribution is used to assess the characteristics of test scores.

7. *Research Data*: In psychological research, the assumption of normality is often made when conducting inferential
statistics to test hypotheses. Researchers may check for normality in their data before applying specific statistical tests.

It's important to note that while the normal distribution is a common assumption in psychology and statistics, not all data
perfectly follows a normal distribution. In practice, researchers often assess the normality of their data and may use
alternative statistical methods when data deviate significantly from normality.
13 | P a g e

4. Skewedness and Kurtosis

a. Skewness
Skewness is a statistical measure that describes the asymmetry or lack of symmetry in the probability distribution of a
dataset. It quantifies the extent to which the data distribution deviates from perfect symmetry.

Key properties of skewness include:

1. *Direction of Skewness*:

- Positive Skewness (Right Skewed): In a positively skewed distribution, the tail on the right side is longer or fatter than
the left side. This means that the majority of data points are concentrated on the left side, and there are relatively few
larger values on the right.

- Negative Skewness (Left Skewed): In a negatively skewed distribution, the tail on the left side is longer or fatter than
the right side. This indicates that the majority of data points are concentrated on the right side, and there are relatively few
smaller values on the left.

2. *Magnitude of Skewness*:

- The greater the magnitude of skewness (either positive or negative), the more pronounced the asymmetry in the
distribution.

3. *Symmetrical Data*: If skewness is close to zero, it suggests that the data is approximately symmetrical, with a
balanced distribution of values on both sides of the mean.

4. *Influences on Mean and Median*: Skewness affects the relationship between the mean, median, and mode of a
dataset. In a positively skewed distribution, the mean is typically greater than the median, while in a negatively skewed
distribution, the mean is usually less than the median. The mode is the value where the distribution peaks.

5. *Applications*: Understanding skewness is important in various fields, including finance, economics, and data
analysis, as it can help interpret the shape and characteristics of data distributions. For instance, in finance, a positive
skewness might indicate an investment with the potential for large gains but also significant losses.
14 | P a g e

Skewness is a valuable tool for assessing the departure of data from a symmetric, bell-shaped (normal) distribution, and it
provides insights into the nature of the dataset's tail behavior and outliers. It's often used in data analysis and statistics to
better understand the underlying characteristics of the data.
15 | P a g e

b. Kurtosis
Kurtosis is a statistical measure that quantifies the "tailedness" or the relative concentration of data points in the tails of a
probability distribution compared to the center of the distribution. It describes the shape of a probability distribution and
how it deviates from a normal distribution.

Key features of kurtosis include:

1. *Leptokurtic and Platykurtic Distributions*:

- Leptokurtic: A distribution with positive kurtosis (excess kurtosis greater than 0) has relatively heavy tails, indicating
more extreme values and a higher peak near the mean. This means the data has more outliers and is more "peaked" or
"fat-tailed" compared to a normal distribution.

- Platykurtic: A distribution with negative kurtosis (excess kurtosis less than 0) has lighter tails, indicating fewer
extreme values and a flatter peak near the mean. This means the data has fewer outliers and is more spread out compared
to a normal distribution.

2. *Mesokurtic Distribution*:

- A distribution with kurtosis equal to 0 (excess kurtosis equal to 0) is referred to as mesokurtic. It has a shape similar to
a normal distribution with tails and central peak similar to the normal curve.

3. *Kurtosis and Normal Distribution*:

- The normal distribution has a kurtosis of 0 (mesokurtic). This means it has tails and a peak that are neither too heavy
nor too light, making it a reference point for comparing kurtosis in other distributions.

4. *Impact on Statistical Tests*:

- High kurtosis (positive or leptokurtic) indicates the presence of outliers and suggests that the data might not conform
to certain assumptions of statistical tests that assume normality, such as the t-test.

- Low kurtosis (negative or platykurtic) suggests that the data has less extreme values and is less prone to outliers.

5. *Applications*:

- Kurtosis is useful in various fields, including finance, where it can help identify financial instruments with extreme
risks (leptokurtic) or those with less extreme risks (platykurtic).
16 | P a g e

6. *Excess Kurtosis*:

- Excess kurtosis, also known as "kurtosis excess," is a common way to measure kurtosis. It subtracts 3 from the
kurtosis value, making a mesokurtic distribution have an excess kurtosis of 0.

Kurtosis is an important concept in statistics and data analysis, particularly for understanding the tails of data
distributions. It provides insights into the shape of the distribution and can guide decisions about data modeling and
statistical analysis.

5. The Meaning of Correlation & the scatterplot of bivariate distributions

Correlation

Correlation is a statistical measure that quantifies the degree and direction of a linear relationship between two variables.
It indicates how closely data points in a bivariate dataset cluster around a straight line, representing the relationship
between the two variables. The correlation coefficient, often denoted as "r," ranges from -1 to 1:

- A positive correlation (0 < r < 1) suggests that as one variable increases, the other tends to increase, indicating a
positive linear relationship.
- A negative correlation (-1 < r < 0) suggests that as one variable increases, the other tends to decrease, indicating a
negative linear relationship.
- A correlation of 0 (r = 0) implies no linear relationship between the variables.

Correlation does not imply causation; it only quantifies the strength and direction of the relationship between variables.

Scatterplot of Bivariate Distributions

A scatterplot is a graphical representation of a bivariate dataset that displays the relationship between two variables. Each
point on the plot represents a pair of values for the two variables. Here's how it works:

- The horizontal axis typically represents one variable, and the vertical axis represents the other variable.
- For each data point, you place a dot at the intersection of the values on the two axes.
- The scatterplot provides a visual overview of how the data points are distributed and whether there is a pattern or
relationship between the two variables.
17 | P a g e

In a scatterplot, different patterns can be observed:

- Positive Linear Relationship: Data points tend to cluster around an ascending straight line, indicating a positive
correlation.
- Negative Linear Relationship: Data points tend to cluster around a descending straight line, indicating a negative
correlation.
- No Linear Relationship: Data points appear randomly scattered with no discernible linear pattern, suggesting no
correlation.

Scatterplots are valuable for visualizing data, identifying outliers, and understanding the nature of relationships between
variables. They provide an initial impression of the data, which can help guide further statistical analysis and
interpretation.
18 | P a g e

6. Correlation: A Matter of Direction, a matter of degree

A high degree of correlation between two variables may occur for several reasons:

1. Causation: The variables are causally related, meaning changes in one variable directly cause changes in the other. In
such cases, a strong and positive correlation is expected. For example, there is a high degree of correlation between the
number of hours spent studying and academic performance because studying directly influences performance.

2. Common Cause: Both variables are influenced by a common cause or factor. This is known as a spurious or
confounding relationship. For instance, ice cream sales and the number of drownings are positively correlated, but the
common cause is hot weather; people buy more ice cream, and more people swim, leading to more drownings.

3. Linear Relationship: The relationship between the variables is linear, meaning that changes in one variable correspond
to proportional changes in the other. Linear relationships often result in high correlation coefficients. For example, if you
double all the values of one variable, the correlation remains the same in a linear relationship.

4. Data Accuracy: A high degree of correlation can also occur when the data is highly accurate and precise. Small
variations in one variable are accurately reflected in the other variable.

5. Lack of Outliers: High correlation can result from the absence of outliers or extreme values in the dataset. Outliers can
influence the correlation by pulling the relationship in different directions.

6. Homogeneous Data: The data is relatively homogeneous, meaning it follows a consistent pattern without much
variability. In such cases, the data points tend to cluster closely around the best-fit line, leading to a strong correlation.

7. Synchronicity: In time series data, two variables may exhibit a high degree of correlation if they move in sync,
responding to the same external events or trends simultaneously.

8. Measurement Accuracy: When both variables are accurately and precisely measured with minimal measurement error,
a strong correlation is more likely to be observed.
19 | P a g e

It's important to note that a high correlation does not necessarily imply causation. Careful analysis is required to
determine whether the correlation is a result of a causal relationship or other factors, such as common causes or statistical
artifacts.
20 | P a g e

7. The Coefficient of Correlation

The coefficient of correlation, often referred to as the "correlation coefficient," is a statistical measure that quantifies the
strength and direction of the linear relationship between two variables. It indicates how closely the data points in a
scatterplot cluster around a straight line, which represents the relationship between the two variables.

Key points about the coefficient of correlation:

1. *Range*: The correlation coefficient typically ranges from -1 to 1.

- A value of 1 indicates a perfect positive linear relationship, where an increase in one variable corresponds to an exact
increase in the other.

- A value of -1 indicates a perfect negative linear relationship, where an increase in one variable corresponds to a
decrease in the other.

- A value of 0 indicates no linear relationship between the variables.

2. *Direction*:

- If the correlation coefficient is positive, it implies a positive linear relationship, meaning that as one variable increases,
the other tends to increase.

- If the correlation coefficient is negative, it implies a negative linear relationship, indicating that as one variable
increases, the other tends to decrease.

3. *Strength*:

- The magnitude of the correlation coefficient represents the strength of the relationship. A coefficient closer to -1 or 1
indicates a stronger linear relationship, while a coefficient closer to 0 suggests a weaker relationship.

4. *Assumption*:

- The correlation coefficient assumes a linear relationship. It may not capture nonlinear relationships between variables.

5. *Pearson Correlation Coefficient*:

- The most commonly used correlation coefficient is the Pearson correlation coefficient (r), which is suitable for
measuring the strength and direction of a linear relationship between two continuous variables.
21 | P a g e

6. *Spearman and Kendall Correlation*:

- In cases where the relationship between variables is not necessarily linear or where the data is ordinal or ranked, other
correlation measures like the Spearman rank correlation and Kendall's tau are used.

The coefficient of correlation is widely used in various fields, including statistics, social sciences, economics, and natural
sciences, to assess the strength and nature of relationships between variables. It is a valuable tool for understanding how
changes in one variable relate to changes in another, which is essential for making informed decisions and predictions.
22 | P a g e

8. Calculating Pearson’s Correlation Coefficient from Deviation Scores

23 | P a g e

9. Calculating Pearson’s Correlation Coefficient from Raw Scores

24 | P a g e

10. Spearman’s Rank-Order Correlation Coefficient

Introduction

Spearman's rank correlation coefficient, denoted as ρ (rho), is a valuable statistical measure used to assess the strength
and direction of a monotonic relationship between two variables. It is particularly useful when dealing with non-
parametric data or data that do not meet the assumptions of parametric correlation measures.

Definition and Formula

Spearman's coefficient is calculated based on the ranks of the data points. The formula for Spearman's rank correlation
coefficient is:

ρ = 1 - {6Σd²}/{n(n²-1)}

Where:

- ρ represents Spearman's coefficient.

- d is the difference between the ranks of the corresponding data points.

- n is the number of data points.

Calculation Process

To compute Spearman's coefficient:

1. Rank the data points for both variables.

2. Calculate the differences (d) between the ranks of corresponding data points.

3. Square these differences (d²).

4. Sum all the squared differences (Σd²).

5. Plug Σd² into the formula to obtain the final coefficient ρ.

Interpretation

Spearman's coefficient can be interpreted as follows:

- If ρ is close to +1, it indicates a strong positive monotonic relationship, meaning that as one variable increases, the other
tends to increase.

- If ρ is close to -1, it suggests a strong negative monotonic relationship, indicating that as one variable increases, the
other tends to decrease.
25 | P a g e

- If ρ is close to 0, it implies no significant monotonic relationship, signifying independence between the variables.

Use Cases and Advantages

Spearman's coefficient is commonly used in various scenarios:

- When dealing with data that do not adhere to the assumptions of parametric correlation, such as normality.

- When working with ordinal or ranked data.

Its advantages include robustness to outliers and the ability to capture monotonic relationships, even if they are not
strictly linear.

Conclusion

In conclusion, Spearman's rank correlation coefficient is a valuable tool in statistics for assessing monotonic relationships
between variables. Its ability to handle non-parametric data and its robustness make it a crucial measure in data analysis
and research.

Additional Information

Spearman's coefficient was developed by Charles Spearman and is widely applied in fields such as social sciences,
psychology, and epidemiology, where ordinal or non-parametric data are prevalent. It complements other correlation
measures and provides valuable insights into data relationships.
26 | P a g e

11. Population and Sample

27 | P a g e

12. Standard error of mean, SD, and r;

28 | P a g e

13. Level of Significance & Type I and Type II errors

29 | P a g e

14. Degrees of Freedom; One-tailed and two-tailed tests; Null and

alternate hypothesis
30 | P a g e

15. Meaning and difference between parametric and non-parametric

test
31 | P a g e

16. T-test and Chi-square test

32 | P a g e

17. Glossary

a. Coefficient
In mathematics and statistics, a coefficient is a multiplicative factor or numerical value that is associated with a specific
term in an equation, expression, or mathematical relationship. Coefficients are used to represent the relative importance or
magnitude of individual components within a mathematical expression.

Here are a few common examples of coefficients in different contexts:

1. *Linear Equation*: In a linear equation like "y = mx + b," the coefficient "m" represents the slope of the line,
indicating how much "y" changes for a unit change in "x."

2. *Quadratic Equation*: In a quadratic equation like "ax^2 + bx + c = 0," the coefficients "a," "b," and "c" determine the
shape and characteristics of the parabolic curve.

3. *Polynomials*: In polynomial expressions like "ax^2 + bx + c," the coefficients "a," "b," and "c" are constants that
define the degree and shape of the polynomial.

4. *Regression Analysis*: In multiple regression analysis, coefficients represent the weights or contributions of different
independent variables to predict a dependent variable. For example, in the equation "y = a + bx + cz," the coefficients "a,"
"b," and "c" represent the impact of each independent variable on the dependent variable "y."

5. *Chemistry*: In chemical equations, coefficients represent the ratios of reactants and products in a balanced chemical
reaction. For example, in the equation "2H2 + O2 → 2H2O," the coefficients "2" indicate that two moles of hydrogen
react with one mole of oxygen to produce two moles of water.

Coefficients are essential in various mathematical and scientific contexts as they provide a way to quantify and express
relationships, proportions, and dependencies between different components within a system or equation. They help us
understand how changes in one variable affect other variables and are used for calculations, modeling, and analysis in
many fields.
33 | P a g e

b. Origin Vs. Scale

"Origin" and "scale" are terms commonly used in the context of data transformations, particularly when rescaling or
shifting data. Let's explain each term and how changes in both origin and scale affect data.

1. **Origin**:

- The origin refers to the point at which a scale or measurement system begins. It's often represented as the zero point on
the scale.

- In a one-dimensional context (e.g., on a number line), the origin is the point where the numerical value is zero.

- Changing the origin means shifting the entire scale, so the data values have a new reference point.

**Example of Changing the Origin**:

- Consider a temperature scale in degrees Celsius. The origin is set at 0°C, where water freezes. If you want to convert
these temperatures to Kelvin (where 0K is absolute zero), you would change the origin by adding 273.15. So, 0°C
becomes 273.15K, and 20°C becomes 293.15K.

2. **Scale**:

- The scale represents the range of values or the unit of measurement between points on a scale. It determines the size
and spacing of values.

- Changing the scale involves adjusting the intervals or units between data points without changing the origin.

**Example of Changing the Scale**:

- If you have a length scale in meters, you can change the scale by converting it to centimeters. For example, 1 meter is
equivalent to 100 centimeters. So, you change the scale by multiplying by 100, but the origin (0 meters) remains the
same.

How Changes in Origin and Scale Affect Data:

1. **Change in Origin**:

- Shifting the origin does not affect the relative differences between data points. It only changes the reference point.
34 | P a g e

- All data values are shifted by the same amount. If you add a constant "c" to every data point, the new data points will
be "data point + c."

2. **Change in Scale**:

- Changing the scale affects the spacing and relative proportions between data points.

- If you multiply every data point by a constant "k," the new data points will be "data point * k."

- This can either expand or compress the data values, making them larger or smaller, respectively, while preserving the
relative differences between them.

In data analysis, transformations involving changes in origin and scale are used to make data more interpretable,
standardize units, or prepare data for further analysis. Common examples include converting units (e.g., Celsius to
Fahrenheit) or rescaling data to a specific range (e.g., scaling data between 0 and 1 for machine learning algorithms).
These transformations help in data visualization, comparison, and modeling while preserving the underlying relationships
within the data.
35 | P a g e

c. Covariance
Covariance is a statistical measure that quantifies the degree to which two random variables change together. It indicates
whether there is a positive or negative relationship between the variables and the strength of that relationship. In other
words, covariance measures how two variables tend to move in relation to each other.

Key points about covariance:

1. **Formula**: The covariance between two variables X and Y is typically denoted as Cov(X, Y) and is calculated as:

`Cov(X, Y) = Σ [(Xi - μX) * (Y_i - μY)] / (n - 1)`

where:

- `X_i` and `Y_i` are individual data points.

- `μX` and `μY` are the means (average values) of X and Y, respectively.

- `n` is the number of data points.

2. **Interpretation**:

- A positive covariance (Cov(X, Y) > 0) indicates that as values of X increase, values of Y tend to increase, suggesting a
positive relationship or association between the variables.

- A negative covariance (Cov(X, Y) < 0) suggests that as values of X increase, values of Y tend to decrease, indicating a
negative relationship or association.

- A covariance of zero (Cov(X, Y) ≈ 0) implies that there is no strong linear relationship between the variables.
However, it does not necessarily mean there is no relationship; it just means that they do not tend to move together
linearly.

3. **Units**: The units of covariance are the product of the units of the two variables. This makes it difficult to compare
covariances between different variable pairs.

4. **Magnitude**: The magnitude of covariance does not have a standardized scale, so it can be challenging to assess the
strength of the relationship between variables based on covariance alone.
36 | P a g e

5. **Normalized Version**: To facilitate comparison, the correlation coefficient (r) is often used. It is derived from
covariance and has values between -1 and 1. A correlation of 1 or -1 indicates a perfect linear relationship, while a
correlation of 0 suggests no linear relationship.

Covariance is used in various fields, including statistics, finance, economics, and data analysis, to study the relationships
between variables. However, it has some limitations, such as its sensitivity to the units of measurement, which is why the
correlation coefficient is often preferred when measuring the strength and direction of linear relationships between
variables.
37 | P a g e

d. Standard Deviation Vs. Variance

Aspect Standard Deviation Variance

A measure of the spread or dispersion of data

Definition The square root of the variance.
points.

Unit of Measurement Same unit as the data. Square of the unit of the data.

Indicates the average distance of data points from Represents the average squared deviation from
Interpretation
the mean. the mean.

Usually larger in magnitude than standard

Magnitude Usually smaller in magnitude than variance.
deviation.

Less sensitive to outliers since data is not More sensitive to outliers due to squaring of
Sensitivity to Outliers
squared. differences.

Mathematical Formula σ = √(Σ(x - μ)² / N) σ² = Σ(x - μ)² / N

Preferred when considering the original data Preferred when analyzing the spread and
Use in Analysis
scale. variability of data.

May involve more complex calculations due to Calculations are relatively simpler due to
Calculation Ease
the square root. squared values.
38 | P a g e
39 | P a g e

e. Distribution types
i. Normal

ii. Theoretical
iii. Bivariate
iv. Multivariate
v. Poisson
vi. Binomial