28th mid-2(c)

Q-1: Karl Pearson’s Measure of Skewness: This is based on the relationship between the mean, mode, and standard
deviation. It is calculated as:

𝑀𝑒𝑎𝑛−𝑀𝑜𝑑𝑒
Skewness=𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑 𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛

If the mode is not known or cannot be reliably determined, the formula can be approximated as:

𝑀𝑒𝑎𝑛−𝑀𝑒𝑑𝑖𝑎𝑛
Skewness= 3 ∗ 𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑 𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛

Bowley’s Measure of Skewness: This is a quartile-based measure that focuses on the spread of the distribution in the
lower and upper halves. It is calculated as:

𝑄3 +𝑄1 −2𝑄2
Skewness= where Q1, Q2, and Q3 are the first quartile, median, and third quartile, respectively.
𝑄3 −𝑄1

Data Used:

Karl Pearson’s Method: Requires knowledge of the mean, mode, and standard deviation. It is based on the entire data
distribution.

Bowley’s Method: Relies only on quartiles (Q1, Q2, Q3) and is less influenced by extreme values or outliers.

1. Robustness:
o Karl Pearson’s Measure: Sensitive to extreme values and outliers, as it depends on the mean and mode.
o Bowley’s Measure: More robust in the presence of outliers, as it uses quartiles which focus on the central
portion of the data.
2. Applicability:
o Karl Pearson’s Method: Suitable for unimodal distributions (those with a clear mode) and when the
dataset is free from significant outliers.
o Bowley’s Method: Better suited for skewed distributions, data with outliers, or when the mode is difficult
to calculate or unreliable.

Preferred Situations for Each Method

1. Karl Pearson’s Method:

o When to Prefer:
▪ The dataset is normally distributed or close to normal.
▪ The mode can be accurately determined.
▪ There are no significant outliers in the dataset.
o Why:
▪ It gives a clear measure of skewness for symmetric and unimodal distributions.
▪ It is straightforward when the mean and mode are meaningful.
2. Bowley’s Method:
o When to Prefer:
▪ The dataset is non-normal or highly skewed.
▪ The data contains outliers or extreme values.
▪ Quartile measures (Q1, Q2, Q3) are more reliable than the mode.
o Why:
▪ It is less sensitive to extreme values, providing a more robust measure of skewness for skewed or
irregular data distributions.
▪ It requires less computation and is effective when the distribution lacks a clear mode.
In summary, Karl Pearson’s measure is preferred for clean, unimodal, and symmetrical data, while Bowley’s measure
is ideal for skewed distributions and datasets with outliers. The choice depends on the nature of the data and the specific
analysis requirements.

Q-2. Correlation is a statistical measure that quantifies the strength and direction of the relationship between two
variables. It indicates how changes in one variable are associated with changes in another.

• Positive Correlation: When one variable increases, the other also increases (e.g., height and weight).
• Negative Correlation: When one variable increases, the other decreases (e.g., speed and travel time).
• Zero Correlation: No linear relationship exists between the variables.

The correlation coefficient, denoted by r, ranges between -1 and 1:

• r=+1: Perfect positive correlation.

• r=-1: Perfect negative correlation.
• r=0: No linear correlation.

Significance of Correlation

1. Understanding Relationships: Helps identify whether two variables are related and the strength of their association.

2. Prediction: If two variables are correlated, the value of one variable can help predict the other.

3. Initial Analysis: Acts as a preliminary step to identify relationships before conducting deeper analyses like regression.

4. Data Insights: Correlation highlights patterns and trends in data, aiding decision-making in fields like economics,
finance, and science.

Does Correlation Always Imply Causation?

No, correlation does not always signify a cause-and-effect relationship. While correlation shows association, it does not
prove that one variable causes change in the other. Some reasons include:

1. Spurious Correlation: The relationship is coincidental or caused by chance (e.g., ice cream sales and shark attacks both
increase in summer due to higher temperatures).

2. Confounding Variables: A third variable influences both correlated variables, creating an illusion of causation (e.g.,
increased education levels and higher income may both be influenced by family background).

3. Reverse Causation: The causal relationship is opposite to what is assumed (e.g., poor health may reduce income rather
than low income causing poor health).

4. Non-Linear Relationships: Correlation captures only linear relationships. Variables may have a non-linear relationship
that is not revealed by correlation.

Correlation is a valuable tool for understanding relationships between variables, but it must be interpreted carefully. To
establish causation, additional analysis methods, such as controlled experiments, Granger causality tests, or regression
with robust controls, are needed. Misinterpreting correlation as causation can lead to incorrect conclusions and decisions.
Q-3. The correlation coefficient, often denoted by the symbol "r," is a statistical measure that quantifies the strength and
direction of the linear relationship between two variables. In other words, it tells us how closely two variables are related
to each other in a linear fashion.
What it measures:
The correlation coefficient measures the degree to which the relationship between two variables can be described by a
straight line. A perfect positive correlation (r = 1) means that as one variable increases, the other variable also increases
proportionally, and they lie perfectly on a straight line with a positive slope. A perfect negative correlation (r = -1) means
that as one variable increases, the other variable decreases proportionally, and they lie perfectly on a straight line with a
negative slope. A correlation coefficient of 0 indicates no linear relationship between the variables.
Interpreting the sign and magnitude of r:
Sign:

• Positive sign: Indicates a positive linear relationship, meaning that as one variable increases, the other variable
tends to increase as well.
• Negative sign: Indicates a negative linear relationship, meaning that as one variable increases, the other variable
tends to decrease.
Magnitude:

• The closer the magnitude of r is to 1 (either positive or negative), the stronger the linear relationship between the
variables.
• A magnitude of r close to 0 indicates a weak or no linear relationship.
4.

Given, 𝑥̅ =68,
𝑦̅ = 150

𝜎𝑥 =2.5, 𝜎𝑦 =20

r = 0.6
σx
here, X-x̅ = r * (Y- y̅)
σy

2.5
X=68+0.6* 20 *(200-150) [when weight is 200]

X=71.75”
5.
Rank by students of uni. A Rank by students of uni. B D=R1 -R2 D2
2 1 1 1
1 3 -2 4
5 2 3 9
3 4 -1 1
4 7 -3 9
7 5 2 4
6 6 0 0
N=7 7 ∑ 𝑫𝟐 = 𝟐𝟖

𝟔 ∑ 𝑫𝟐 𝟐𝟖
R= 1- =1- 𝟔 ∗ 𝟑𝟑𝟔 = 1- 0.5 = +0.5
𝒏𝟑 −𝒏
 28th final
Q -1a Statistics is the science of collecting, analyzing, interpreting, and presenting data to draw meaningful conclusions
and make informed decisions. It involves various methods and techniques to summarize, describe, and analyze data,
identify patterns and trends, and make predictions or inferences about a larger population based on a sample.
Importance of Statistics in Business and Industry:
Statistics plays a crucial role in various aspects of business and industry, helping organizations to:
i. Data-Driven Decision Making:

• Market Research: Understanding customer preferences, market trends, and competitor analysis.
• Product Development: Identifying new product opportunities, assessing product performance, and optimizing
product design.
• Financial Analysis: Forecasting financial performance, managing risk, and making investment decisions.
ii. Risk Management:

• Predictive Modeling: Forecasting future events and potential risks.

• Quality Control: Ensuring product quality and reliability.
• Financial Risk Assessment: Evaluating investment risks and managing portfolio diversification.
iii. Process Improvement:

• Data Analysis: Identifying areas for improvement in business processes.

• Six Sigma: A data-driven methodology for process improvement.
iv. Competitive Advantage:

• Data-Driven Insights: Gaining a competitive edge by understanding customer behavior and market dynamics.
• Innovation: Driving innovation through data-driven research and development.
• Effective Marketing: Targeting the right customers with the right messages.
In conclusion, statistics is an essential tool for businesses and industries to make informed decisions, manage risks,
improve processes, and gain a competitive advantage in today's data-driven world.
b. Modern statistical tools and techniques are essential for improving the quality of managerial decisions because they
provide a systematic framework for collecting, analyzing, interpreting, and presenting data. In today's dynamic business
environment, decision-making is increasingly data-driven. By using statistical tools, managers can minimize uncertainties,
identify patterns, forecast trends, and make informed decisions that enhance efficiency, profitability, and sustainability.
Statistics serve as a foundation for understanding market dynamics, customer behavior, operational efficiency, and
financial performance. For example, by analyzing customer purchasing trends, businesses can tailor their marketing
strategies to target specific customer segments, thereby optimizing resources.
Role of Statistics in Business Planning and Control:
1. Forecasting and Demand Estimation: A retail company uses time-series analysis to predict seasonal demand for its
products. For instance, an ice cream company may use historical sales data to anticipate higher sales during summer
months and plan inventory accordingly.
Impact: Forecasting ensures businesses have the right resources at the right time, reducing costs associated with
overstocking or understocking.
2. Quality Control: Manufacturing companies use statistical quality control (SQC) techniques like control charts to
monitor production processes. For instance, a car manufacturer checks defect rates in production to ensure products meet
quality standards.
Impact: Improved quality reduces waste, enhances customer satisfaction, and builds brand loyalty.
3. Market Research and Customer Insights: A company uses surveys and statistical analysis (e.g., regression analysis)
to understand customer preferences and identify factors influencing purchase decisions.
Impact: Insights gained help design better products, improve customer service, and enhance marketing campaigns.
4. Risk Analysis and Decision-Making: Financial institutions use probability distributions and Monte Carlo simulations
to assess risks in investment portfolios or loan approvals.
Impact: Risk management helps safeguard investments and ensures stability in financial planning.
5. Performance Monitoring and Control: A company uses Key Performance Indicators (KPIs) and dashboards with
statistical data to track sales, employee performance, or operational efficiency.
Impact: Monitoring metrics enables timely corrective actions and better resource allocation.
6. Cost-Benefit Analysis: Before launching a new product, a company conducts statistical cost-benefit analysis to assess
expected revenues versus costs.
Impact: This ensures resources are invested in projects with the highest return on investment.
7. Production and Inventory Management: Companies apply statistical models like Economic Order Quantity (EOQ) to
optimize inventory levels and reduce holding costs.
Impact: Effective inventory management minimizes operational disruptions and cost overruns.
Statistics plays a crucial role in the planning and control of business operations by reducing uncertainty and improving
decision-making processes. Tools like regression analysis, hypothesis testing, data visualization, and predictive modeling
empower managers to make evidence-based decisions that align with organizational goals. By integrating statistical
techniques, businesses can remain competitive, adapt to changing environments, and achieve sustainable growth.
c. Primary data refers to original data collected directly for a specific purpose, while secondary data is pre-existing data
gathered by others for different objectives. Primary data is more accurate but time-consuming and expensive, whereas
secondary data is cost-effective and readily available, though it may lack relevance or reliability.
Methods of Collecting Primary Data include:
1. Observation Method: Involves watching events or behaviors in real-time. It provides reliable data but is time-
consuming and prone to observer bias.
2. Interview Method: Data is collected through personal or telephonic interactions. It allows deeper insights but is costly
and may result in biased responses.
3. Questionnaire Method: Respondents answer pre-set questions, making it economical and easy to analyze. However, it
often suffers from low response rates.
4. Experiments: Conducted under controlled conditions to determine cause-and-effect relationships. This method is
precise but expensive and limited in generalizability.
5. Surveys: Useful for collecting data from large populations efficiently but depends on the honesty and clarity of
respondents. Each method has its merits and limitations. The choice of method depends on the research objective, budget,
and time constraints. Proper use of primary data collection methods ensures accurate and actionable insights for decision-
making.
d. A well-designed questionnaire ensures accurate, reliable, and useful data collection. Key essentials include:
1. Clarity and Simplicity: Questions should be clear, concise, and easy to understand, avoiding technical jargon or
ambiguity.
2. Relevance: Each question should align with the research objectives and provide meaningful information.
3. Logical Flow: Questions should be arranged in a logical sequence, starting with general topics and moving to specific
ones.
4. Brevity: The questionnaire should be brief to avoid respondent fatigue, but comprehensive enough to collect necessary
data.
5. Neutrality: Avoid leading questions or bias that might influence responses.
6. Variety of Question Types: Include a mix of closed-ended (e.g., multiple-choice, Likert scale) and open-ended
questions to gather quantitative and qualitative data.
7. Pilot Testing: Pre-test the questionnaire on a small sample to identify and rectify any issues.
8. Anonymity and Confidentiality: Assure respondents that their identities and responses will remain confidential to
encourage honesty.
Survey Planning for University Students in Dhaka
1. Objective: To understand the study habits of university students and assess the facilities available to them.
2. Target Population: University students in Dhaka across various disciplines and institutions.
3. Sampling Method: Use stratified random sampling to ensure representation from different universities, disciplines, and
genders.
4. Sample Size: Select a sufficiently large sample (e.g., 300–500 students) for statistical reliability.
5. Mode of Data Collection: Distribute questionnaires through online forms, email, or in-person surveys, depending on
accessibility.
6. Timeline: Allocate specific time for questionnaire preparation, data collection, analysis, and reporting (e.g., 4–6
weeks).
Draft Questionnaire
Section 1: Demographics 6. When do you prefer to study?
1. What is your age? • Morning
• Afternoon
• Below
• Night
• 20 20–25
• Above 25 7. Do you prefer individual study or group study? Why?
2. What is your gender? 8. What challenges do you face in maintaining a
consistent study routine?
• Male
• Female Section 3: Facilities
• Other
9. Are the library facilities at your university adequate?
3. Which university are you enrolled in?
• Yes
4. What is your field of study? • No

Section 2: Study Habits 10. How often do you use the library?

5. How many hours do you study daily (outside of • Daily

classes)? • Weekly
• Rarely
• Less than 2 hours
• 2–4 hours
• More than 4 hours
11. Do you have access to high-speed internet for
academic purposes? • 12. What improvements do you suggest for
better study facilities at your university?
• Yes
• No

Section 4: Feedback
13. Is there any additional information you would like to share about your study habits or the facilities available to you?
This structured questionnaire ensures relevant data is collected while maintaining respondent engagement and
confidentiality.
Q -2
a. A good average is a representative measure that summarizes a dataset effectively and provides meaningful insights. Its
essential properties include:
1. Simplicity: A good average should be easy to calculate and understand. It should be clear enough for interpretation by
both technical and non-technical audiences.
2. Representativeness: It should accurately reflect the characteristics of the entire dataset. The average should lie within
the range of the data.
3. Rigidity: The average should not be affected significantly by extreme values (outliers).
4. Stability: The average should remain stable across samples drawn from the same population, ensuring reliability.
5. Mathematical Properties: It should have mathematical properties that make it useful for further analysis, such as
algebraic manipulation.
6. Comprehensiveness: The average should consider all data points in the dataset.
7. Universality: I t should be applicable to various types of data and situations.
8. Capable of Further Analysis: The average should allow for further statistical calculations, such as standard deviation or
correlation.
Which Average Is the Best and Why?
The choice of the "best" average depends on the nature of the data and the purpose of the analysis. Common averages
include the mean, median, and mode, each with unique strengths and weaknesses:
1. Arithmetic Mean
Advantages:

• Simple to calculate and understand.

• Considers all data points.
• Has mathematical properties useful for further statistical analysis.
Disadvantages:

• Affected by outliers, making it less representative in skewed datasets.

2. Median
Advantages:

• Not affected by extreme values.

• Represents the central value in ordered data, making it useful for skewed distributions.
Disadvantages:

• Does not consider all data points.

• May not be suitable for further mathematical analysis.

3. Mode
Advantages:

• Useful for categorical data.

• Represents the most frequently occurring value in the dataset.
Disadvantages:

• May not exist or may not be unique.

• Does not consider all data points.
In most cases, the arithmetic mean is considered the best average due to its simplicity, comprehensiveness, and
mathematical applicability. However, when dealing with skewed data or outliers, the median is more appropriate because
it is not influenced by extreme values. For categorical data, the mode is ideal. Ultimately, the choice depends on the
dataset and the specific requirements of the analysis.
b.

Weekly wage fi xi fixi cf

1,200-1,400 8 1299.5 10396 8
1,400-1,600 12 1499.5 17994 20
1,600-1,800 20 1699.5 33990 40
1,800-2,000 30 1899.5 56985 70
2,000-2,200 40 2099.5 83980 110
2,200-2,400 32 2299.5 73584 142
2,400-2,600 18 2499.5 44991 160
2,600-2,800 7 2699.5 18896.5 167
2,800-3,000 6 2899.5 17397 173
3,000-3,200 4 3099.5 12398 177
N=177 ∑ 𝑓𝑖𝑥𝑖 = 370611.5
∑ 𝒇𝒊𝒙𝒊
i. 𝒙
̅= =370611.5/177 =2093.85
𝑵

Median =177/2 =88.5th term. which lies in the class of 2000-2200

𝑛
−𝑝𝑐𝑓 88.5−70
Median = L + 2 ∗ 𝑤 =2000+ ∗ 200 =2092.5
𝑓 40

𝑓1 40−30
Mode = L + 𝑓1+𝑓2 ∗ 𝑤 =2000+(40−30)+(40−32) ∗ 200 =2000+111.11=2111.11

ii. Q1= 177/4=44.25th term which is in the class 1800-2000

𝑛
−𝑝𝑐𝑓 44.25−40
Q1 = L + 4
∗ 𝑤 = 1800+ *200 =1823.33
𝑓 30

Q3=3(177/4) =132.75th term which is in the class 2200-2400

3𝑛
−𝑝𝑐𝑓 132.75−110
Q3 = L + 4
∗ 𝑤 = 2200+ *200 =2342.2
𝑓 32

Required limits for the middle 50% =1823.3 to 2342.2

iii.

Target classes Calculation Target frequencies

1750-1800 1 5
20*4
1800-2000 30 30
2000-2200 40 40
2200-2250 32* 4
1 8
N=83
Required percentage of the workers who earned between Tk. 1750 and Tk.2250 is (83/177) ×100=46.89%
Iv

Target classes Calculation Target frequencies

2500-2600 18*2
1 9
2600-2800 7 7
2800-3000 6 6
3000-3200 4 4
N=26
Required percentage of the workers who earned more than Tk. 2500 is (26/177) ×100 =14.69%
v.

Target classes Calculation Target frequencies

1200-1400 8 8
1400-1600 12 12
1600-1800 20 20
1800-2000 30 30
N=70
Required percentage of the workers who earned less than Tk. 2000 is (70/177) ×100 =39.55%
𝑠.𝑑
c. coefficient of variation, C.V =𝑚𝑒𝑎𝑛× 100

C.V for Mahmudullah = (15/50) *100 =30

C.V of mustahfiz = (12/48) *100 =25
C.V of Shanto = (2/12) *100=16.67
Less C.V indicates more consistent player
C.V for Shanto is the least. Hence the batsman III is more consistent.
Based on consistency Shanto should be selected. if selector consider performance, for highest mean (50) Mahmudullah
might be considered.
𝑋1 𝑁1 +𝑋2 𝑁2 +𝑋3 𝑁3 (305∗20)+(300∗25)+(340∗40) 27200
d. 𝑋123= = = = 320
𝑁1 +𝑁2 +𝑁3 20+25+40 85

here, d1=|𝑋1 − 𝑋123 | =|305 − 320| =15

d2=|𝑋2 − 𝑋123 |=|300 − 320|=20

d3=|𝑋3 − 𝑋123 |=|340 − 320|=20 /

𝑵𝟏 𝝈𝟐𝟏 +𝑵𝟐 𝝈𝟐𝟐 +𝑵𝟑 𝝈𝟐𝟑 +𝑵𝟏 𝒅𝟐𝟏 +𝑵𝟐 𝒅𝟐𝟐 +𝑵𝟑 𝒅𝟐𝟑 𝟐𝟎∗𝟓𝟎𝟐 +𝟐𝟓∗𝟒𝟎𝟐 +𝟒𝟎∗𝟒𝟓𝟐 +𝟐𝟎∗𝟏𝟓𝟐 +𝟐𝟓∗𝟐𝟎𝟐 +𝟒𝟎∗𝟐𝟎𝟐 𝟐𝟎𝟏𝟓𝟎𝟎
𝜎123 = √ =√ =√ =48.69
𝑵𝟏 +𝑵𝟐 +𝑵𝟑 𝟐𝟎+𝟐𝟓+𝟒𝟓 𝟖𝟓

Thus, the combined average wage is 320 and the combined standard deviation is 48.69
 Q-3
a. Conditional probability is the probability of an event occurring given that another event has already occurred. It is
denoted as and is calculated as:
p(A∩B)
P(A|B) = , provided p(B) ≠0
p(B)

Two events are said to be independent of each other if the occurrence or non-occurrence of one event in any trial does not
affect the occurrence of the other event in any trial. Ex: If we tossed a coin three times, the result of the 1st, 2nd and 3rd
would be independent each other.
Given, P (A+B) =P(A)+P(B),
where P(A+B) indicates unions of A and B, P(A∪B).
The formula for the probability of the union of two events is:

P(A∪B) =P(A)+P(B)−P(A∩B), substitute P (A+B) =P(A)+P(B)into the formula,

P(A)+P(B)= P(A)+P(B)−P(A∩B)
P(A∩B) = 0
When P(A∩B) =0, it means the events A and B cannot occur simultaneously, which is the definition of mutually exclusive
events.
So, P (A+B) =P(A)+P(B) is mutually exclusive
However, for two events to be statistically independent, the condition is:

P(A∩B) = P(A)⋅P(B)
In this case, P(A∩B) =0,
0= P(A)⋅P(B)
So, unless P(A)=0 or P(B)=0 the condition for independence is not satisfied. Thus, A and B are not independent but
mutually exclusive.
b.
Dhaka Sylhet Chittagong Khulna Total
Yes 45 55 60 50 210
No 35 45 35 45 160
No option 5 5 5 5 20
Total 85 105 100 100 390
210
i. The total number of consumers who prefer brand A is, P(A) = =0.5385
390
60
ii. Probability of prefers brand A and from Chittagong, P (A∩C) = 390 = 0.1538

iii. This is a conditional probability,

P(A∩C) 60/390
P (A |𝐶 ) = = 100/390 = 0.6
P(C)

P(K∩A) 50/390
iv. P(K|𝐴) = = =0.238
P(A) 210/390

c. Given, P(X) =0.7 so, P (𝑋̅) =1-0.7 =0.3

P (𝑌̅ ) =0.5 so, P(Y) =1-0.5=0.5

 ̅∪𝒀
P (𝑿 ̅)= 0.6

̅ ∩𝒀
Not selected in any firms, P (𝑿 ̅)= P (𝑿
̅ ) + P (𝒀
̅)-P (𝑿
̅∪𝒀
̅)= 0.3+0.5 - 0.6 = 0.2

̅∩𝒀
The probability that the persons will be selected in one of the two firms P (X∪ 𝒀)=1- P (𝑿 ̅)= 1-0.2 =0.8

̅∪𝒀
OR, P (𝑿 ̅)=0.6

≫1-P (X∩ 𝒀)=0.6

≫ P (X∩ 𝒀)= 0.4

SO, required probability, P (X∪ 𝒀)= P(X)+P(Y)- P (X∩ 𝑌)= 0.7+0.5-0.4=0.8

d.

E1 E2 E3 Total
P(Ei) 0.5 0.375 0.125 1
P (E|Ei) 0.85 0.65 0.6
P (E∩Ei) = P(Ei)* P(E|Ei) 0.425 0.24375 0.075 ∑ P(Ei) ∗ P(E|Ei) =0.74375
P(E1)∗ P(E1)
(I) P (E1|E) =∑ 𝑃(𝐸𝑖)∗ 𝑃(𝐸|𝐸𝑖) =0.425/0.74375 =0.571

P(E2)∗ P(E2)
(ii) P (E2|E) =∑ 𝑃(𝐸𝑖)∗ 𝑃(𝐸|𝐸𝑖) =0.24375/0.74375 =0.3277

P(E3)∗ P(E3)
(iii) P (E3|E) =∑ 𝑃(𝐸𝑖)∗ 𝑃(𝐸|𝐸𝑖) =0.075/0.74375 =0.1008

Q-4
a. Regression analysis and correlation analysis are statistical techniques used to study relationships between variables, but
they differ in purpose and approach.
Regression analysis examines the relationship between a dependent variable and one or more independent variables, often
to establish causation or make predictions. It provides a mathematical model, such as y= β0+β1x+∊, that quantifies how
changes in the independent variable(s) influence the dependent variable. Regression can be univariate (one predictor) or
multivariate (multiple predictors) and is widely used for forecasting, hypothesis testing, and decision-making.
Correlation analysis, on the other hand, measures the strength and direction of a linear association between two variables.
The correlation coefficient (r) ranges from -1 to 1, where values close to -1 or 1 indicate strong negative or positive
associations, and values near 0 indicate no relationship. However, correlation does not imply causation; it only quantifies
the degree of association.
The key differences are:
Objective: Regression predicts or explains one variable based on others, while correlation quantifies the degree of
association.
Causality: Regression can suggest causal links (with proper assumptions), but correlation cannot.
Output: Regression provides an equation; correlation provides a single statistic (r).
Variables: Regression handles one dependent variable and multiple predictors, whereas correlation is limited to two
variables.
In summary, correlation is about association, while regression is about modeling and prediction. Both are essential in data
analysis, serving complementary roles.
 b. The regression line represents the relationship between the independent variable(s) and the dependent variable,
providing a predicted value of the dependent variable for given values of the independent variable(s). However, this
prediction is not exact—it is only an estimate based on the observed data. There is often a difference between the actual
observed values and the values predicted by the regression line, called residuals.
The standard error of estimate is a statistical measure used to assess the degree of uncertainty or variability in the
predictions made by the regression line. It quantifies the average distance that the observed data points deviate from the
predicted values.
Formula for Standard Error of Estimate (SEE)

̂)𝟐
∑(𝒚𝒊−𝒚𝒊
The standard error of estimate is calculated as: SEE = √ 𝒏−𝒌−𝟏

Where:
yi = observed values of the dependent variable

̂ = predicted values of the dependent variable from the regression line

𝑦𝑖
n = total number of observations
k= number of independent variables in the regression model
Interpretation
1. Small SEE: A smaller standard error of estimate indicates that the data points are closer to the regression line, meaning
the model's predictions are more accurate and reliable.
2. Large SEE: A larger standard error of estimate suggests that the predictions deviate significantly from the actual
observed values, implying less reliability.
Why is SEE important?

• Uncertainty Measurement: It provides a quantitative measure of the uncertainty or variability in the predictions.
• Model Comparison: SEE can be used to compare the accuracy of different regression models. A model with a
smaller SEE is generally preferred.
• Confidence Intervals: SEE is used to construct confidence intervals around the predicted values, giving a range
within which the actual value is likely to fall.
In essence, while the regression line gives a best-fit prediction, the standard error of estimate reveals the level of
confidence we can have in those predictions.
c. calculation of correlation coefficient

Age (X-49) dx2 Blood pressure Y-145 dy2 dxdy

X dx Y dy
56 +7 49 147 +2 4 14
42 -7 49 125 -20 400 140
36 -13 169 118 -27 729 351
47 -2 4 128 -17 289 34
49(A) 0 0 145(A) 0 0 0
42 -7 49 140 -5 25 35
60 +11 121 155 +10 100 110
72 +23 529 160 +15 225 345
63 +14 196 149 +4 16 56
55 +6 36 150 +5 25 30
∑ 𝑋 = 522 ∑ 𝑑𝑥 = 32 ∑ 𝑑𝑥2 = 1202 ∑ 𝑌 = 1417 ∑ 𝑑𝑦 =-33 ∑ 𝑑𝑦2 = 1813 ∑ 𝑑𝑥𝑑𝑦 = 1115
*Assumed mean is =49
𝐍 ∑ 𝐝𝐱𝐝𝐲−∑ 𝐝𝐱 ∑ 𝐝𝐲 (10∗1115)−[32∗(−33)] 12206
(i) co efficient of correlation is given by = = =13689 =0.892
√10∗1202−322√10∗1813−(−33)2
√𝐍 ∑ 𝐝𝟐𝐱 −(∑ 𝐝 𝐱 )𝟐 ∗√𝐍 ∑ 𝐝𝟐𝐲 −(∑ 𝐝 𝐲)𝟐

There is a high degree of positive correlation between age and blood pressure.
(ii) the least square regression equation of Y on X is given by,
̅= byx (X-𝑋̅); 𝑋̅ = 522/10 = 52.2, 𝑌̅ =1417/10 = 141.7
Y-Y
𝑁 ∑ 𝑑𝑥𝑑𝑦−∑ 𝑑𝑥 ∑ 𝑑𝑦 (10∗1115)−[32∗(−33)]
And byx = = =1.11
𝑁 ∑ 𝑑𝑥2 −(∑ 𝑑𝑥 )2 10∗1202−322

Substituting these values in the above equation we have,

Y-141.7= 1.11(X-52.2)
Y=83.758+1.11X
This is the required least square regression equation of Y on X.
(iii) when X=45
Y=83.758+1.11*45=133.708
Hence the most likely blood pressure of a woman of 45 years is 134.
Q-05
a. 28th mid -2(c) Q1
b. Skewness and kurtosis are statistical measures used to describe the shape of a frequency distribution, which provides
insights into the nature of data beyond measures of central tendency (like mean) and dispersion (like variance).
1. Skewness: Skewness measures the degree and direction of asymmetry in a frequency distribution:

• Symmetrical distribution: Skewness = 0. Examples include normal distributions.

• Positively skewed distribution: Skewness > 0. The tail on the right side (higher values) is longer or fatter.
• Negatively skewed distribution: Skewness < 0. The tail on the left side (lower values) is longer or fatter.

Role:

• Understand data symmetry: Skewness helps identify whether the data is evenly distributed around the mean.
• Highlight extreme values: A high positive or negative skew may indicate outliers or uneven spread in the data.
• Impact on statistical methods: Many statistical techniques assume normality (zero skewness), so understanding
skewness is essential for model selection and adjustments.
2. Kurtosis: Kurtosis measures the "tailedness" or sharpness of a frequency distribution:

• Mesokurtic distribution: Kurtosis = 3. It matches the kurtosis of a normal distribution.

• Leptokurtic distribution: Kurtosis > 3. The distribution has heavier tails and a sharper peak, indicating more
extreme values.
• Platykurtic distribution: Kurtosis < 3. The distribution has lighter tails and a flatter peak.
Role:

• Assess data outliers: Higher kurtosis (leptokurtic) suggests more extreme values in the tails.
• Shape of distribution: Helps in determining whether the distribution is flat or sharply peaked.
 • Model suitability: Some models are sensitive to kurtosis, requiring transformations or robust approaches if
kurtosis deviates significantly from 3.
In Analyzing Frequency Distributions:

• Skewness and kurtosis together help assess deviations from normality, which is critical in many statistical
analyses.
• Decision-making: By analyzing these measures, analysts can better understand data patterns, choose appropriate
statistical tests, and identify potential anomalies in the data.
• Practical implications: For example, in economics or finance, skewness and kurtosis provide insights into market
returns, risk assessment, or resource distribution.
c. Sampling is the process of selecting a subset of individuals, items, or observations (called a sample) from a larger
population to analyze and draw conclusions about the entire population. It is often used when studying the entire
population is impractical or costly.
Types of Sampling
Sampling methods are broadly classified into probability sampling and non-probability sampling:
1. Probability Sampling: In probability sampling, every member of the population has a known and equal chance of being
selected. This ensures unbiased representation of the population.
Types of Probability Sampling:
i. Simple Random Sampling:

• Every individual has an equal chance of being selected.

• Example: Drawing names from a hat to select participants for a survey.
ii. Stratified Sampling:

• The population is divided into subgroups (strata) based on shared characteristics, and samples are taken
proportionally from each stratum.
• Example: Dividing a university’s students by department (e.g., Science, Arts, Commerce) and selecting
proportional samples from each group.

iii. Systematic Sampling:

• A starting point is chosen randomly, and then every individual is selected.

• Example: Selecting every 10th customer entering a store for a survey.
iv. Cluster Sampling:

• The population is divided into clusters (often geographically), and some clusters are randomly selected. Then, all
members of selected clusters are sampled.
• Example: Studying the literacy rate by selecting random villages and surveying all residents within those villages.
2. Non-Probability Sampling: In non-probability sampling, not all members have an equal chance of being selected. It is
often quicker and less expensive but may introduce bias.
Types of Non-Probability Sampling:
i. Convenience Sampling:

• Samples are chosen based on their availability or ease of access.

• Example: Interviewing people at a shopping mall for a quick opinion survey.
ii. Judgmental (Purposive) Sampling:

• Samples are selected based on the researcher’s judgment and expertise.

• Example: Selecting specific experts to participate in a panel discussion.
iii. Quota Sampling:

• The population is divided into subgroups, and samples are selected to meet predetermined quotas for each
subgroup.
• Example: Ensuring a survey includes 40% males and 60% females to match population proportions.
iv. Snowball Sampling:

• Existing participants recruit additional participants, often used in hard-to-reach populations.

• Example: Researching drug addiction by asking current participants to refer others they know.
Why Sampling Is Important

• Cost and Time Efficiency: Studying a sample saves resources compared to surveying an entire population.
• Feasibility: Some populations (e.g., all ocean fish) are too large to study entirely.
• Accuracy: When conducted properly, sampling can provide reliable insights about a population.
d.

Age of fi xi fixi xi-µ fi (xi-µ) fi (xi-µ)2 fi (xi-µ)3 fi (xi-µ)4

workers

25-30 30 27.5 825 -11.6 -348 4037 -46827 543192

30-35 22 32.5 715 -6.6 -145.2 958 -6325 41744
35-40 27 37.5 1012.5 -1.6 -43.2 69 -111 177
40-45 35 42.5 1487.5 3.4 119 405 1376 4678
45-50 21 47.5 997.5 8.4 176.4 1482 12447 104553
50-55 18 52.5 945 13.4 241.2 3232 43310 580352
Total N=153 5982.5 0* 10183 3870 1274696
*Rounded by 0.2 due to fraction calculation
∑ 𝑓𝑖𝑥𝑖
Mean (µ) = = 5982.5/153= 39.10
𝑛

∑ 𝑓𝑖(𝐱𝐢−µ) 0
First central moment 𝜇1 = =153 =0
𝑛

∑ 𝑓𝑖(𝐱𝐢−µ)2 10183
second central moment 𝜇2 = = 153 =66.56
𝑛

∑ 𝑓𝑖(𝐱𝐢−µ)3 3870
third central moment 𝜇3 = = 153 =25.294
𝑛

∑ 𝑓𝑖(𝐱𝐢−µ)4 1274696
fourth central moment 𝜇4 = = =8331.346
𝑛 153

µ2 25.2942
Coefficient of skewness 𝛽1 = µ33 = 66.563 = 0.00217
2

Skewness is close to zero indicating the distribution is fairly symmetrical.

 Q-6
(a) The survey has some strengths, such as gender balance and potential representativeness within Dhaka University.
However, its adequacy is limited by the small sample size, potential lack of diversity in disciplines, and restricted
generalizability beyond Dhaka University. To improve the study, the researcher could consider expanding the sample size,
including students from multiple universities, and ensuring a more diverse representation of academic disciplines and
career stages.
(b) (i) Negative Correlation: A negative correlation occurs when two variables move in opposite directions. If one
increases, the other decreases.
Example: The price of a product and its demand—when the price of a product rises, its demand generally falls.
ii) Perfectly Positive Correlation: A perfectly positive correlation means that two variables move in exactly the same
direction with a correlation coefficient of +1.
Example: The amount of raw material used and the total cost of production—if raw material usage doubles, the total cost
also doubles.
iii) Coefficient of Determination (R²): The coefficient of determination (R²) measures how well independent variables
explain variations in a dependent variable. It ranges from 0 to 1, where 1 means perfect predictability.
Example: If R² = 0.8 in a regression of income on education level, it means 80% of income variation is explained by
education level.
iv) Zero Correlation: Zero correlation means that there is no relationship between two variables; changes in one do not
affect the other.
Example: Shoe size and intelligence—knowing a person's shoe size does not help predict their intelligence level.
v) Linear Relation: A linear relation means that the relationship between two variables can be represented by a straight
line, either increasing or decreasing at a constant rate.
Example: A worker’s salary and years of experience—if a company gives a fixed annual raise, salary increases linearly
over time.
(c) let,
X be the percentage of men in the group
Y be the percentage of women in the group
So, X+Y =100
(32∗𝑋)+25∗𝑌)
Combined Mean age = 100

{32(100−𝑌)}+25∗𝑌) 3200−32𝑌+25𝑌
> 30 = =
100 100

>3000=3200-7Y
>Y=200/7 =28.54%
So, X= 100-28.54=71.46%
(d) student-1: (85*20%) + (89*10%) + (94*10%) +(87*25%) + (90*35%) =88.55
Student-2: (78*20%) + (84*10%) + (88*10%) +(91*25%) +(92*35%) =87.75
Student-3: (94*20%) + (88*10%) + (93*10%) + (86*25%) +(89*35%) =89.55
Student-4: (82*20%) + (79*10%) + (88*10%) + (84*25%) +(93*35%) =86.50
 27th final
Q-1
(a) The statement "Statistics are numerical statements of facts, but all numerically stated facts are not statistics"
emphasizes the distinction between general numerical facts and statistical data.
Statistics involve the systematic collection, classification, analysis, and interpretation of numerical information related to a
specific phenomenon. For example, stating that the average income of workers in a factory is $50,000 is statistical
because it is derived from analyzing multiple data points.
However, not all numerical facts qualify as statistics. A simple number, such as "John is 6 feet tall," is a numerical fact but
not a statistic since it is an isolated piece of information. Statistical data must be aggregated, comparable, analyzable, and
related to a specific study.
For data to be considered statistical, it must be collected in a systematic manner and used for analysis to derive
meaningful conclusions. In contrast, standalone numerical facts lack analytical context and do not contribute to statistical
insights.
In summary, while statistics always involve numerical facts, not all numerical facts can be classified as statistics. Only
when data is systematically collected and analyzed does it hold statistical significance, aiding in decision-making and
problem-solving.
(b) Statistics is broadly classified into descriptive statistics and inferential statistics, each serving different purposes in
data analysis.
Descriptive Statistics: Descriptive statistics focus on summarizing and organizing data to make it easily interpretable. It
does not involve drawing conclusions beyond the given data set. Common measures include:

• Measures of central tendency: Mean, median, and mode.

• Measures of dispersion: Range, variance, and standard deviation.
• Data visualization: Charts, graphs, and tables.
For example, if a teacher calculates the average score of students in a class, it is descriptive statistics because it simply
summarizes the data without making any predictions.
Inferential Statistics: Inferential statistics go beyond describing data; they help in making predictions and generalizations
about a larger population based on a sample. Key techniques include:

• Hypothesis testing (e.g., t-tests, chi-square tests).

• Confidence intervals (estimating population parameters).
• Regression analysis (predicting relationships between variables).
For instance, if a researcher surveys 500 people to estimate the average income of a country’s population, they use
inferential statistics because they infer conclusions beyond the sample.
Key Differences

• Descriptive statistics summarize and present data without generalizing.

• Inferential statistics analyze sample data to make predictions about a larger population.
Both types are essential in data analysis, with descriptive statistics providing a foundation for inferential techniques.
(c) (i) highest value =88, lowest value = 50
Taking class interval 5 frequency distribution is given below,

Class interval Tally Frequency Mid-point

50-55 ||| 3 52
55-60 |||| 4 57
60-65 |||| 5 62
65-70 |||| 4 67
70-75 |||| 4 72
75-80 ||| 3 77
80-85 ||| 3 82
85-90 |||| 4
N=30
All companies have profit more than or equal 50
Numbers of companies with profit more than 50 are 28 (excluding two 50)
The percentage of companies having income more than 50 = (28/30) *100 =93.33%
(ii)

A histogram is drawn with class intervals on the x-axis and frequency on the y-axis. A frequency polygon is plotted using
the midpoints of each class (e.g., midpoint of 50–54 is 52) and connected by straight lines.
(60-65) class have the highest number of companies. after that there is a fluctuation until (65-85). Then there is an
increase.
Q-2
(a) Classification and tabulation are two methods used to organize and present data, each serving a distinct purpose.
Classification involves sorting data into distinct categories or groups based on shared characteristics. It helps to simplify
and interpret data by organizing it into manageable units. This method can be qualitative, like categorizing individuals by
occupation or gender, or quantitative, such as grouping data into ranges like income levels or age brackets. Classification
aids in understanding patterns and trends within data but doesn’t necessarily provide detailed comparisons.
On the other hand, tabulation involves arranging data in a table format with rows and columns, often making it easier to
analyze, compare, and summarize. This method is primarily used for numerical data and involves organizing it
systematically to highlight relationships or distributions. A table may include totals, averages, percentages, or other
measures to provide a comprehensive view of the data. It allows for clearer comparisons between different data points and
is particularly useful when presenting complex information.
In short, classification groups data based on common characteristics, whereas tabulation organizes data into a structured
table, enhancing clarity and comparison. Both methods are essential tools for data analysis, depending on the complexity
and type of data being dealt with.
(b) There are various graphs and charts commonly used to present business data, each serving a specific purpose to
enhance understanding and analysis:
1. Bar diagram: A bar diagram (also known as a bar chart) is a graphical representation of data where individual bars
(rectangles) represent the values of different categories. The length or height of each bar is proportional to the value or
frequency it represents. Bar diagrams are simple yet effective in comparing different groups or categories of data.
2. Line Graphs: Ideal for showing trends over time, line graphs plot data points along a continuous line. They help track
changes in variables and visualize patterns, such as sales performance over several months or stock price fluctuations.
3. Pie diagram: A pie chart is a circular graph divided into slices to illustrate numerical proportions. Each slice represents
a category or part of the whole, with the size of each slice being proportional to the quantity it represents. Pie charts are
often used to show percentage or proportional data, making it easy to visualize how individual parts contribute to the total.
4. Histograms: A histogram is a type of bar chart used to represent the distribution of numerical data by grouping data
points into ranges (called bins or intervals). Unlike regular bar charts that compare discrete categories, histograms show
the frequency of data points within each range, helping to visualize the distribution and spread of data.
[Link]: An ogive is a line graph that represents the cumulative frequency or cumulative relative frequency of a dataset. It
helps in visualizing how data accumulates across different ranges or intervals. By plotting the cumulative frequencies, the
ogive provides insights into the overall distribution and helps determine percentiles such as the median or quartiles.
6. frequency polygon: A frequency polygon is a graphical representation of a frequency distribution, similar to a
histogram, but with a line graph instead of bars. It is used to visualize the shape of a dataset’s distribution by plotting the
midpoints of the class intervals (or bins) and connecting the points with straight lines. The frequency polygon is
particularly useful for comparing multiple datasets or observing trends in data distributions over time.
(c) ogive by less than method

Profits No of companies
Less than 20 8
Less than 30 20
Less than 40 40
Less than 50 64
Less than 60 79
Less than 70 89
Less than 80 96
Less than 90 99
Less than 100 100

It is clear from the graph that the number of companies getting profits is less than Tk. 75 crores is 92 and the number of
companies getting profits less than Tk. 45 crores is 51. Hence the number of companies getting profits between
Tk. 45 crores and Tk. 75 crores is = 92-51=41
 Q-3
(a) Central tendency refers to the statistical concept that identifies the center or typical value of a dataset. It is used to
describe a single value that represents the entire distribution of data, providing an overview of where most data points are
concentrated. Central tendency helps summarize the data by locating its center, making it easier to compare and interpret.
The central value is significant because it helps simplify and summarize large sets of data. Instead of analyzing every data
point, a single central value provides a general sense of where most observations lie. This is particularly useful in
decision-making, forecasting, and comparisons. For example, the central value can serve as a reference point when
comparing different groups or time periods, and it plays a key role in understanding distributions and making predictions.
Properties of a good measure of central tendency:
Based on all observations: The measure should consider all data points within the set, not just a subset.
Easy to calculate and understand: The calculation should be straightforward and easily interpretable.
Not unduly affected by outliers: Extreme values in the data should not significantly influence the central value.
Suitable for further statistical analysis: The chosen measure should allow for further mathematical operations and
comparisons.
The mean is typically the most appropriate measure of central tendency when the data is symmetrical and free from
significant outliers. It takes into account all the data points and provides a precise average. However, when the data is
skewed or has outliers, the median is often the better choice because it is more resistant to extreme values and provides a
better representation of the "typical" value. In cases of categorical data, the mode might be the most useful measure of
central tendency.
(b)(i)
Daily fi c.f. xi fixi
Mean (µ) =
∑ 𝑓𝑖𝑥𝑖
=
127200
= 424 wages
𝑁 300 250-300 25 25 274.5 6862.5
median = size of N/2 or 150th observation 300-350 30 55 324.5 9735
350-400 55 110 374.5 20597.5
hence median lies in the class (400-450) 400-450 70 180 424.5 29715
𝑁 450-500 67 247 474.5 31791.5
−𝑝.𝑐.𝑓 150−110
median = L + 2
*i = 400+ *50 500-550 39 286 524.5 20455.5
𝑓𝑖 70
550-600 14 300 574.5 8043
= 400 + 28.5714 =428.5714 n=300 ∑ 𝒇𝒊𝒙𝒊 = 𝟏𝟐𝟕𝟐𝟎𝟎

highest number of observations 70 lies in the class (400-450)

∆1 (70−55)
mode = L + ∆1+∆2 *i = 400 + (70−55)+(70−67) * 50 =400+41.6667 = 441.6667

(ii) Q3 = (300/4) *3 =225th observation

Q3 lies in the class (450-500)
3𝑁
−𝑝.𝑐.𝑓 225−180
So, Q3 = L + 4
*i = 450 + *50 = 450+33.5821 = 483.5821
𝑓𝑖 67

Q1 = (300/4) =75th observation. Q1 lies in the class (350-400)

𝑁
−𝑝.𝑐.𝑓 75−55
So, Q1 = L + 4
*i = 350 + *50 = 350 + 18.182 =368.182
𝑓𝑖 55

Wages limit of central 50% employees is (368.182-483.5821)

(iii) since Mode > Median > Mean, the distribution appears to be left-skewed (negatively skewed). This suggests that
most of employees earn towards the higher and of the wage distribution, but some lower wages pull the mean down.
 Q-4
(a) Variation" in statistics refers to the degree of difference or spread among data points within a set, essentially indicating
how much the values deviate from each other and from the central tendency (mean) of the data; the more spread out the
data, the higher the variation, and vice versa.
Common methods to measure variation
Range: The difference between the highest and lowest values in a data set.
Merits: Simple to calculate, provides a quick overview of the spread.
Demerits: Highly sensitive to outliers, ignores information about the distribution within the range.
Interquartile Range (IQR): The difference between the third quartile and the first quartile, representing the spread of the
middle 50% of the data.
Merits: Less affected by outliers than the range, provides information about the middle portion of the distribution.
Demerits: Ignores information outside the interquartile range.
Mean Deviation: The average absolute difference between each data point and the mean.
Merits: Relatively easy to understand, considers all data points.
Demerits: Can be sensitive to outliers, not as mathematically robust as other measures.
Variance: The average of the squared deviations from the mean.
Merits: Gives more weight to larger deviations from the mean, suitable for further statistical analysis.
Demerits: Can be heavily influenced by outliers due to the squaring process, not easily interpretable as it is not in the
same units as the original data.
Standard Deviation: The square root of the variance, representing the average deviation from the mean.
Merits: Most commonly used measure of variation, easily interpretable as it is in the same units as the data, less affected
by outliers compared to variance.
Demerits: Can still be affected by outliers to some extent.
(b) Measuring variation is crucial for understanding the spread and consistency of data. It helps assess consistency,
revealing how reliable or predictable data is, and is essential for managing risk in fields like finance, where higher
variation signals more uncertainty. Variation also improves prediction accuracy, as it indicates how spread-out data points
are from the mean. It aids in identifying outliers or anomalies, which could signify errors or unusual events. Additionally,
it supports comparisons between datasets, showing which one is more consistent. In industries, measuring variation is key
to quality control, highlighting areas for improvement. It provides insights into data distribution, helping choose the right
statistical tools. Overall, variation is vital for decision-making, identifying trends, and ensuring informed, data-driven
choices across various sectors, from business to public health. Understanding variation enhances the accuracy and
reliability of conclusions drawn from data.
A good measure of variation should possess, as far as possible, the following properties:
(i) It should be simple to understand. (ii) It should be easy to compute. (iii) It should be rigidly defined.
(iv) It should be based on each and every observation of the distribution.
(v) It should be amenable to further algebraic treatment.
(vi) It should have sampling stability.
(vii) It should not be unduly affected by extreme observations.
(c)

Profits xi fi fixi cf xi-µ fi(xi-µ)2

10-20 15 8 120 8 -17.2 2366.72
20-30 25 12 300 20 -7.2 622.08
30-40 35 20 700 40 2.80 156.8
40-50 45 6 270 46 12.80 983.04
50-60 55 4 220 50 22.80 2079.36
N=50 ∑ 𝑓𝑖𝑥𝑖 = 1610 ∑ fi(xi − µ)2 =6208
Mean (µ) =1610/50 = 32.20

∑ fi(xi−µ)2 6208
Standard deviation 𝝈 = √ =√ =√124.16 = 11.1427
𝑁 50

𝜎 11.1427
C.V = µ ∗ 100 = *100 = 34.60%
32.20

Third quartile = 3N/4 =150/ 4 = 37.5th observation

37.5th observation lies in class (30-40)
3𝑁
−𝑝.𝑐.𝑓 37.5−20
Q3 = L + 4
*i = 30 + *10 =30+8.75=38.75
𝑓𝑖 20

Q-5
(a) Regression is a statistical method used to analyze the relationship between two or more variables. It helps in predicting
the value of one variable based on the value of another variable. Essentially, it measures the average probable change in
one variable (the dependent variable) when there is a certain change in another variable (the independent variable).
Example:
Let's consider an example where we want to understand how the number of hours studied (independent variable) affects a
student's exam score (dependent variable).

• Independent variable (X): Hours studied

• Dependent variable (Y): Exam score
Suppose we have the following data:
Hours studied(X) Exam score(Y)
Now, we run a linear regression to find the equation of the line that best 1 55
fits the data. The regression model might result in an equation like this: 2 60
3 65
Y = 50 + 5X
4 70
This means that for every additional hour studied (a change of 1 in X), 5 75
the exam score (Y) increases by an average of 5 points.
Interpretation:
The slope of the line (5 in this case) tells us the average probable change in the exam score for each additional hour of
studying.
If a student increases their study time by 1 hour, we predict that their exam score will increase by 5 points, on average.
Thus, regression measures how much the dependent variable (exam score) is likely to change in response to changes in
the independent variable (hours studied).
(b) The correlation coefficient and regression coefficients have several important properties.
Correlation coefficient
Range: The correlation coefficient ranges from -1 to 1.
Symmetry: The correlation coefficient is symmetric, meaning rxy = ryx.
Independence: The correlation coefficient is independent of the change of origin, but not of scale.
Perfect correlation: A correlation of -1 indicates a perfect negative correlation, while a correlation of 1 indicates a perfect
positive correlation.
No linear relationship: A correlation of 0 indicates that the two variables have no linear relationship.
Regression coefficients
Change of origin: Regression coefficients are independent of the change of origin.
Change of scale: Regression coefficients are not independent of the change of scale.
Relationship to correlation coefficient: The correlation coefficient is the geometric mean of the two regression
coefficients.
Relationship to each other: If one regression coefficient is greater than 1, then the other must be less than 1.
Relationship to sign: If one regression coefficient is positive, then the other must also be positive.
(c) (i)

X R1 Y R2 (R1-R2)2
5 1.5 20 5 12.25
10 5 16 3 4
18 8 23 7.5 0.25
16 7 20 5 4
5 1.5 20 5 12.25
10 5 23 7.5 6.25
9 3 8 2 1
10 5 4 1 16
N= 8 ∑ 𝐷2 = 56

𝟏 𝟏 𝟏 𝟏 𝟏 𝟏 𝟏 𝟏
𝟔{∑ 𝑫𝟐 +𝟏𝟐(𝒎𝟑𝟏 −𝒎𝟏 )+𝟏𝟐(𝒎𝟑𝟐 −𝒎𝟐 )+𝟏𝟐(𝒎𝟑𝟑−𝒎𝟑 )+𝟏𝟐(𝒎𝟑𝟒 −𝒎𝟒 ) 𝟔{𝟓𝟔+𝟏𝟐(𝟐𝟑 −𝟐)+𝟏𝟐(𝟑𝟑 −𝟑)+𝟏𝟐(𝟑𝟑 −𝟑)+𝟏𝟐(𝟐𝟑 −𝟐) 𝟔[𝟓𝟔+𝟎.𝟓+𝟐+𝟐+𝟎.𝟓)
R= = =
𝑵𝟑 −𝑵 𝟖𝟑 −𝟖 𝟓𝟎𝟒

= 0.726
(ii)

x y xy x2 y2
5 20 100 25 400
10 16 160 100 256
18 23 414 324 529
16 20 320 256 400
5 20 100 25 400
10 23 230 100 529
9 8 72 81 64
10 4 40 100 16
∑x=83 ∑y=134 ∑xy=1436 ∑x2=1011 ∑y2=2594
 𝑁∑xy−∑x∑y (8∗1436)−(83∗134) 11488−11122
r= = = 34.6266∗52.8772 = 0.199≈ 0.2
√𝑁∑𝑥 2−(∑ 𝑥)2 √𝑁∑𝑦2 −(∑ 𝑦)2 √(8∗1011)−(83)2 √(8∗2594)−(134)2

⸫Coefficient of determination r2 =0.22 =0.4

Only 4% of the variation in y is explained by x meaning the relationship between two variables is weak.
(iii)
 Q-6
(a) i. Probability: Probability is a measure of the likelihood that a particular event will occur, expressed as a number
between 0 and 1, where 0 indicates impossibility and 1 indicates certainty. Values between 0 and 1 represent the degree of
uncertainty about the events occurrence. It is calculated as the ratio of favorable outcomes to the total number of possible
outcomes in a sample space.
ii. Experiment: An experiment is a process or procedure that generates a set of outcomes. In probability theory, an
experiment refers to any process that produces observable results, such as rolling a die, flipping a coin, or drawing a card
from a deck.
iii. Exclusive Events (Mutually Exclusive Events): Two or more events are said to be mutually exclusive (or disjoint) if
they cannot occur at the same time. In other words, the occurrence of one event prevents the occurrence of the other(s).
For example:
- When flipping a coin, the events "heads" and "tails" are mutually exclusive because both cannot happen simultaneously.
Mathematically, for two events A and B: 𝑷(𝑨 ∩ 𝑩) =0
iv. Conditional Probability: Conditional probability is the probability of an event occurring given that another event has
already occurred. It is denoted as P (A | B) and is calculated using the formula:
𝑷(𝑨∩𝑩)
P (A | B) = 𝑷(𝑩)

v. Bayes' Theorem: Bayes' Theorem is a mathematical formula used to determine the probability of an event based on
prior knowledge of related conditions. It is expressed as:
P(B | A)∗ P(A)
P (A | B) = P(B)

• P (A | B) is the probability of event A given event B,

• P (B | A) is the probability of event B given event A,
• P(A) is the prior probability of event A,
• P(B) is the prior probability of event B.
This theorem is fundamental in updating probabilities when new evidence is introduced, often used in statistics, machine
learning, and decision-making processes.
(b) A probability function describes the likelihood of different outcomes in a random experiment. It assigns probabilities
to each possible outcome while ensuring that:
1. The probability of any outcome is between 0 and 1.
2. The sum of all probabilities equals 1.
There are two main types of probability functions:
Probability Mass Function (PMF) – Used for discrete random variables. It gives the probability of each possible outcome.
Probability Density Function (PDF) – Used for continuous random variables. It represents probabilities over an interval,
where the total area under the curve equals 1.
Different Probability Distributions: Probability distributions describe how probabilities are assigned to different values
of a random variable. Some common distributions include:
Discrete Distributions:
1. Bernoulli Distribution – For a single trial with two possible outcomes (success/failure).
2. Binomial Distribution – For multiple independent Bernoulli trials.
3. Poisson Distribution – For the number of events occurring in a fixed interval (e.g., number of calls at a call center).
4. Geometric Distribution – Models the number of trials until the first success.
Continuous Distributions:
1. Normal Distribution (Gaussian) – Symmetric, bell-shaped distribution used in many real-world phenomena.
2. Exponential Distribution – Describes waiting times between events in a Poisson process.
3. Uniform Distribution – All outcomes within a range are equally likely.
4. Gamma Distribution – Generalization of the exponential distribution.
Properties of Normal Distribution
The Normal Distribution is a continuous probability distribution characterized by:
1. Bell-Shaped Curve – Symmetric around the mean (μ).
2. Mean = Median = Mode – The highest point of the curve is at the mean.
3. Defined by Two Parameters – Mean (μ) and standard deviation (σ).
4. 68.95-99.7 Rule (Empirical Rule) –

• 68% of data falls within 1σ of the mean.

• 95% within 2σ.
• 99.7% within 3σ.
5. Total Area Under the Curve = 1 – Ensures valid probability distribution.
6. Symmetry – The left and right halves are mirror images.
7. Asymptotic to the X-Axis – The curve extends infinitely in both directions without touching the x-axis.
4000−5000
(c) z1 = = -1 [left of the mean]
1000

6500−5000
z2 = = 1.5 [right of the mean]
1000

So, 34.13%workers fall between 4000 and 5000 and 43.32% fall between 5000 and 6500.
⸫ 34.13+43.32= 77.45% or (15000*77.45%) =11618 workers have a salary between 4000 and 6500

Mid -1
Q-1
(a) Statistics is the branch of mathematics that deals with the collection, analysis, interpretation, and presentation of
numerical data. It helps in making informed decisions by identifying patterns, trends, and relationships in data.

Functions of Statistics

1. Collection of Data – It involves gathering data from various sources using surveys, experiments, and observations.

2. Organization & Classification – Data is arranged systematically into tables, graphs, and charts for better understanding.

3. Presentation of Data – The data is represented visually through diagrams, histograms, and frequency distributions.

4. Analysis & Interpretation – Using statistical methods like mean, median, standard deviation, and regression analysis to
derive meaningful insights.
5. Forecasting & Decision Making – Based on past trends, statistics help in predicting future outcomes and guiding
decision-making.

6. Formulation of Policies – Governments and businesses use statistics to design economic policies, marketing strategies,
and production plans.

(b) Statistics plays a crucial role in business decision-making and management by providing a data-driven approach to
solving problems.

1. Business Decision-Making

• Helps managers analyze consumer behavior, market trends, and demand patterns.
• Assists in pricing strategies and product development.

2. Quality Control: Businesses use statistical tools like Six Sigma and control charts to maintain product quality.

3. Market Research & Consumer Insights: Surveys and statistical analysis help understand customer preferences and
improve marketing strategies.

4. Financial Analysis & Risk Management:

• Businesses use statistics for budgeting, forecasting, and investment analysis.

• Helps in identifying and managing business risks.

5. Human Resource Management: Employee performance evaluation, salary distribution, and workforce planning are
done using statistical models.

6. Economic & Production Planning: Helps in optimizing production processes, inventory management, and resource
allocation.

statistics is an essential tool in business and management as it aids in making strategic, evidence-based decisions to
maximize efficiency and profitability

Q-2
(a)

Aspect Data Information

Definition Raw facts and figures without any Processed or organized data that is
meaning meaningful.
Nature Unprocessed, scattered Processed, structured
Format Numbers, symbols, text etc. Reports, graphs, summaries
example List of customer purchase Sales report showing trends
(b)Types of Data
1. Qualitative Data – Descriptive data (e.g., customer feedback, colours, opinions).
2. Quantitative Data – Numerical data (e.g., sales figures, temperature).
o Discrete Data – Countable values (e.g., number of employees).
o Continuous Data – Measurable values (e.g., weight, height).
Primary Data and Its Collection Methods: Primary data is firsthand data collected directly from sources for a specific
purpose. It is original, reliable, and up-to-date.
Methods of Collecting Primary Data

• Surveys & Questionnaires

o Structured set of questions asked to respondents.
o Used in market research and customer feedback.

• Interviews
o Face-to-face, telephone, or online conversations to gather detailed insights.
o Useful in qualitative research.

• Observation
o Directly watching and recording behaviours or events.
o Used in traffic studies, consumer behaviour analysis.

• Experiments
o Conducting controlled studies to analyse cause-and-effect relationships.
o Used in product testing and medical research.

• Focus Groups
o Small group discussions to collect opinions and feedback.
o Common in advertising and product development.

• Field Studies
o Collecting data by visiting real-life settings.
o Used in environmental studies, anthropology.
Primary data is useful because it is specific and reliable, but it can be time-consuming and expensive to collect.
Q-3

(a) Data presentation helps in understanding, analysing, and interpreting data effectively. The main ways of presenting
data are:

1. Textual Presentation
• Data is presented in written or descriptive form.
• Example: "The company's sales increased by 20% in the last quarter."
• Suitable for small datasets but not for large numerical data.
2. Tabular Presentation
• Data is arranged in rows and columns for easy comparison.
• Useful for summarizing numerical data effectively.
• Example:

Year Sales Profit

2022 50,000 10,000
2023 60,000 15,000
(b)

Daily profits No of enterprise

0-50 12
50-100 18
100-150 27
150-200 20
200-250 17
250-300 6

Q-04
(a) Central tendency refers to the statistical measure that identifies the center or typical value of a data set. It is a way to
describe the central or average position of a distribution of data. Measures of central tendency help summarize a set of
data by determining the central point, around which other data points cluster.
Measures of Central Tendency
There are three main measures of central tendency:
1. Arithmetic Mean (Average)
o The sum of all the values in the data set divided by the number of values.
o Formula: Mean = ∑X/N
2. Median
o The middle value in a data set when it is arranged in ascending or descending order.
o If the number of values is odd, the median is the middle value; if even, it's the average of the two middle
values.
3. Mode
o The value that appears most frequently in a data set.
o A data set may have one mode, more than one (bimodal, multimodal), or none if no value repeats.
The Arithmetic Mean is the most commonly used measure of central tendency due to several reasons:
1. Simplicity
o It is easy to compute and understand, making it accessible for most types of data.
2. Mathematical Properties
o The mean is algebraically tractable, making it useful for further analysis. It can be used in mathematical
and statistical operations, such as variance and standard deviation.
3. Representativeness
o The mean takes every data point into account, making it a comprehensive measure of the overall data.
4. Applicability
o The mean is ideal for interval and ratio data (where both order and the exact difference between values
matter).
 5. Efficiency
o It provides a good summary of the data when the distribution is symmetric, and there are no extreme
outliers.
However, the mean can be sensitive to extreme values (outliers), which is why in some skewed distributions, the median
may be a better measure of central tendency.
(b) 27th final Q-3b
Q-5
(a) The Lorenz Curve is a graphical representation of income or wealth distribution within a population. It is used to
show the degree of inequality in a distribution, such as income or wealth, by comparing the cumulative share of the
population to the cumulative share of income.

How it Works:
• The x-axis represents the cumulative percentage of the population (from the poorest to the wealthiest), typically
divided into equal percentiles.
• The y-axis represents the cumulative percentage of total income or wealth held by that percentage of the
population.
• The line of perfect equality is a 45-degree diagonal line, where everyone has the same income or wealth.
• The Lorenz curve typically lies below this line, and the further the Lorenz curve is from the line of perfect
equality, the greater the inequality.
The Gini coefficient, a measure of inequality, can be derived from the Lorenz curve by calculating the area between the
Lorenz curve and the line of perfect equality.
Why is the Range of Data Required?
The range of a data set is a measure of the spread or dispersion of the values. It is calculated by subtracting the smallest
value from the largest value in the data set.
Importance of Range:
1. Understand the Spread of Data
o The range provides a simple measure of how spread out the values are in a data set, giving a quick idea of
the variability in the data.
2. Identify Outliers
o A large range may indicate the presence of outliers or extreme values, which can significantly affect the
mean or other measures of central tendency.
3. Comparison Across Data Sets
o Range helps compare the dispersion of different data sets. A large range implies more variability, while a
smaller range indicates less spread.
4. Initial Data Assessment
o It is often the first step in understanding the data, as it gives a broad overview of the data's extent. It can
be useful in identifying trends or areas that may need further analysis.
Limitations of Range: The range only considers the smallest and largest values, and therefore, it can be influenced by
outliers, not reflecting the overall distribution well.
(b)

RUN (A) X-𝐗̅ (X-𝐗̅) 2 RUN (B) Y- 𝐘̅ ̅)2

(Y- 𝐘
114 56 3136 15 -5 25
45 -13 163 25 5 25
0 -58 3364 18 -2 4
31 -27 729 30 10 100
75 17 289 11 -9 81
102 44 1936 4 -16 256
198 140 19600 23 3 9
8 -50 2500 21 1 1
0 -58 3364 31 11 121
7 -51 2601 22 2 4
∑X=580 ̅)2 =35688
∑(X-𝐗 ∑𝒀 = 20 ̅)2 =626
∑(Y- 𝐘

∑X 580 ∑Y 200
̅
X= 𝑁 = 10 = 58 ̅
Y= 𝑁 = 10 = 20

∑(X−X̅ )2 35688 ∑(Y−Y̅ )2 626

S.D (A), 𝜎 = √ =√ = 59.76 S.D (B), 𝜎 = √ = √ 10 = 7.91
𝑁 10 𝑁

𝟓𝟗.𝟕𝟔 𝟕.𝟗𝟏
Coefficient of variance(A) = *100 = 103.03% Coefficient of variance(B) = 𝟐𝟎 *100 = 39.55%
𝟓𝟖

Even though A has a higher average run, its performance is inconsistent compared to B.
2nd mid
1(a) Skewness and Kurtosis are statistical measures that describe the shape and characteristics of a frequency
distribution.
Skewness: Skewness measures the asymmetry or the lopsidedness of a probability distribution. If the distribution is
skewed to the right (positively skewed), the tail on the right side is longer or fatter, indicating that the majority of the data
points are concentrated on the left side. If the distribution is skewed to the left (negatively skewed), the tail on the left side
is longer or fatter, indicating that the majority of the data points are concentrated on the right side.
Skewness = 0 indicates a perfectly symmetric distribution.
Role in frequency distribution analysis: Skewness helps in understanding the direction of the data spread. If the data is
highly skewed, measures of central tendency like the mean might not reflect the typical value well. It's important to adjust
statistical methods when dealing with skewed data, as normal distribution-based methods may not be valid.
Kurtosis: Kurtosis measures the "tailedness" of the probability distribution, or the extent to which data points are outliers.

• Leptokurtic distributions (kurtosis > 3) have heavy tails and more outliers, meaning they are more prone to
extreme values.
• Platykurtic distributions (kurtosis < 3) have light tails and fewer outliers, meaning data points are more
concentrated around the mean.
• Mesokurtic distributions (kurtosis = 3) resemble the normal distribution with a moderate level of outliers.
Role in frequency distribution analysis: Kurtosis helps to understand the risk of extreme values or outliers in the data.
In financial analysis, for example, a leptokurtic distribution indicates a higher risk of extreme losses or gains. It can affect
the precision and accuracy of statistical inferences, especially in hypothesis testing and regression analysis.
S0, Skewness informs you about the asymmetry of the distribution and how the data is shifted. Kurtosis informs you about
the presence of outliers and the shape of the distribution's tails. Both are crucial for understanding the full picture of a data
set, beyond just measures like the mean and standard deviation, and they guide decisions on which statistical techniques to
use.
(b)

Consumption (Kw. Hours) Mid- fi fixi xi- µ fi(xi-µ)2

point(xi)
0-10 5 6 30 -20.9 2620.86
10-20 15 25 375 -10.90 2970.25
20-30 25 36 900 -0.9 29.16
30-40 35 20 700 9.10 1656.2
40-50 45 13 585 19.10 4742.53
N=100 ∑fixi = 2590 ∑ fi(xi-µ)2 = 12019
2590 ∆1 11
Mean = = 25.90 Mode = L + ∆1+∆2 *i = 20 +11+16 ∗ 10 = 24.07
100

fi(xi−µ)2 12019
Standard deviation 𝝈 = √ =√ =√120.19 = 10.963
𝑁 100

𝑚𝑒𝑎𝑛−𝑚𝑜𝑑𝑒 25.90−24.07
Coefficient of skewness = = = 0.167
𝜎 10.963

Q-2
(a) Correlation is a statistical measure that describes the strength and direction of a relationship between two variables. It
quantifies how changes in one variable are associated with changes in another. Correlation values range from -1 to +1:

• Positive Correlation (+1): As one variable increases, the other also increases.
• Negative Correlation (-1): As one variable increases, the other decreases.
• Zero Correlation (0): No relationship between the variables.
Does Correlation Signify Cause and Effect?
No, correlation does not imply causation. Just because two variables are correlated does not mean one causes the other to
change.
For example: Ice cream sales and drowning incidents may be positively correlated, but eating ice cream does not cause
drowning. Instead, both increase due to hot weather (a third factor).
Causation requires further statistical tests or controlled experiments to establish a cause-and-effect relationship.
Rank correlation (Spearman’s) is used for ordinal (ranked) data and measures a monotonic relationship, while Karl
Pearson’s correlation is used for continuous (interval/ratio) data and measures a linear relationship. Rank correlation is
less affected by outliers, whereas Pearson’s is more sensitive to extreme values. Rank correlation is applied in non-
parametric cases, while Pearson’s is used when data is normally distributed and requires a parametric approach.
(b)

X (X-Mean) x2 Y (Y-mean) y2 xy
x y
48 14 196 45 10 100 140
35 1 1 20 -15 225 -15
17 -17 289 40 5 25 85
23 -11 121 25 -10 100 110
47 -13 169 45 10 100 130
∑X= 170 0 776 ∑Y= 170 0 550 280
Mean (X)= 170/5 = 34 Mean (Y) =175/5 =35
∑𝒙𝒚 𝟐𝟖𝟎
r= = = +0.429
√∑ 𝒙𝟐 ∑ 𝒚𝟐 √𝟕𝟕𝟔∗𝟓𝟓𝟎

it is a moderate case of positive correlation between marks in accountancy and statistics.

 Q-3
(a) Linear regression is a statistical technique used to model the relationship between a dependent variable and one or
more independent variables. It helps in predicting outcomes based on given inputs. The equation for simple linear
regression is Y= a+bx+∊, where Y is the dependent variable, X is the independent variable, a is the intercept, b is the
slope, and ∊ is the error term. For multiple variables, the model extends to multiple linear regression.
Uses in Business and Industry:
1. Sales Forecasting – Predicting future sales based on advertising spend and market trends.
2. Price Optimization – Understanding how pricing changes affect demand.
3. Risk Management – Identifying financial risks and fraudulent activities.
4. Customer Retention – Predicting customer churn to improve retention strategies.
5. Manufacturing Quality Control – Analyzing how factors like temperature and pressure affect product quality.
6. HR Analytics – Forecasting employee turnover and assessing training effectiveness.
7. Marketing Analytics – Measuring the impact of campaigns on sales and engagement.
8. Supply Chain Optimization – Forecasting demand to maintain optimal inventory levels.
Linear regression is widely used in business decision-making, helping organizations optimize performance, reduce risks,
and improve efficiency.
(b)
regression equation of X on Y regression equation of Y on X
𝜎 𝜎
X-𝑋̅ = r 𝜎𝑥 (Y-𝑌̅ ) Y-𝑌̅ = r 𝜎𝑦 (X-𝑋̅ )
𝑦 𝑥

5 25
X-20=0.8∗ 25 (Y-120) Y-120=0.8∗ (X-20)
5

X-20= 0.16Y-19.2 Y-120= 4X-80

X= 0.16Y-0.8 Y= 4X+40

Q-4
(a) Probability is a measure of the likelihood that a particular event will occur. It is expressed as a number between 0 and
1, where 0 means the event is impossible and 1 means it is certain. The probability of an event A is given by:
P(A) = Number of favorable outcomes / Total number of outcomes
Example:
Suppose you toss a fair six-sided die. The probability of rolling a 4 is:
P (4) = 1/6 since there is only one favorable outcome (rolling a 4) out of six possible outcomes (1, 2, 3, 4, 5, 6).
Mutually Exclusive Events: Two events are mutually exclusive if they cannot happen at the same time. In other words,
the occurrence of one event prevents the occurrence of the other.
Example:
When rolling a six-sided die:

• Event A: Rolling an even number 2,4,6

• Event B: Rolling an odd number 1,3,5
Since a number cannot be both even and odd at the same time, Events A and B are mutually exclusive. Mathematically:

P (A ∩ B) = 0
For mutually exclusive events, the probability of either occurring is:

P (A ∪B) = P(A) + P(B)

Bayes' Theorem: Bayes' Theorem describes how to update the probability of an event based on new evidence. It is
expressed as:
𝑃(𝐵|𝐴) 𝑃(𝐴)
P(A|B) = P(B)

where:

• P(A|B) is the probability of event given that B has occurred.

• 𝑃(𝐵|𝐴) is the probability of event A given that has occurred.
• P(A) and P(B) are the independent probabilities of events A and B.
Example:
A factory produces 60% of its products from Machine X and 40% from Machine Y. Defective rates are 2% for Machine X
and 5% for Machine Y. If a product is randomly selected and found to be defective, Bayes' Theorem helps find the
probability that it came from Machine X.
Using Bayes’ Theorem:
P (X | D) = [P(D|X) P(X)] / P(D) = [(0.02*0.6) / (0.02*6) (0.05*0.4)] = 0.375=37.5%
So, if a product is defective, there is a 37.5% chance it came from Machine X.

(b) let event A and B denote that the computer A and B are sold respectively,
Then, P(A)=0.60, P(B)=0.40

And Probability of selling both computers simultaneously P (A∩ 𝐵)= P(A)*P(B)=0.60*0.40 =0.24 [independent events]
Probability of selling at least one computer is given by,
P(A∪B) = P(A)+P(B)- P (A∩ 𝐵)=0.6+0.4-0.24 =0.76
𝑷(𝑨) 𝟎.𝟔𝟎
Probability of sold computer is to be A=P (A|A∪ 𝐁)= 𝑷(𝑨∪𝑩) =𝟎.𝟕𝟔 = 𝟎. 𝟕𝟖𝟗𝟓

26th final
Q-1(a) The data type collected by the accountant on historical costs is quantitative (numerical) data, specifically
continuous data if the costs include precise values (e.g., $10,523.75) or discrete data if recorded in whole numbers (e.g.,
$10,524).
Justification:
1. Numerical Nature – Historical costs are expressed in monetary terms, making them numerical rather than categorical.
2. Objective Measurement – These costs are based on actual transactions and accounting records, ensuring they are
measurable and not subjective.
3. Used for Analysis – Accountants use historical cost data for financial reporting, auditing, and decision-making, which
requires numerical analysis.
(b) Both arithmetic mean (AM) and geometric mean (GM) are capable of further algebraic treatment because they adhere
to key mathematical properties that allow for manipulation, transformation, and integration into more complex analyses.
Arithmetic Mean (AM) and Its Algebraic Properties
𝑥1+𝑥2 +𝑥3………+𝑥𝑛
The arithmetic mean, defined as: AM = 𝑛

Linear Property: If all values in a dataset are multiplied by a constant c, the arithmetic mean also gets multiplied by c, i.e.,
AM (cx1, cx2, ..., cxn) = c* AM (x1, x2, ..., xn).
Summation Representation: Since AM is based on summation, it integrates well into algebraic expressions and statistical
formulas.
Geometric Mean (GM) and Its Algebraic Properties
The geometric mean, given by: GM = (x1*x2……...xn)1/n
Multiplicative Property: If each data point is raised to a power p, the geometric mean is also raised to p:
GM (x1p, x2p, ..., xnp) = (GM (x1, x2, ..., xn)) p
1
log GM =𝑛 ∑ 𝑙𝑜𝑔𝑥𝑖

Why Algebraic Treatment Matters?

The ability to manipulate AM and GM algebraically makes them powerful tools in economics, finance, and statistics.

• AM is widely used for trend analysis, average returns, and cost estimations.
• GM is crucial in compound interest calculations, economic growth rates, and investment performance evaluation.
These properties ensure that AM and GM are not just abstract mathematical concepts but practical tools for solving real-
world problems.
(c) The 3.8 million search queries made on Google represent big data, specifically unstructured and semi-structured data,
with key characteristics that define its complexity and usefulness.
Type of Data
1. Unstructured Data – Most search queries are text-based and do not follow a predefined format (e.g., questions,
keywords, sentences).
2. Semi-structured Data – Some search queries may have metadata, such as timestamps, location, and user preferences,
which add structure but do not fit into traditional databases like relational tables.
Characteristics of This Large-Volume Data (Big Data Properties - 5Vs)
1. Volume – The sheer number of search queries (millions per minute) makes this data massive and requires scalable
storage and processing systems.
2. Velocity – Data is generated at an extremely high speed, as new queries are constantly entered in real time.
3. Variety – Queries come in different formats, including text, voice searches, images, and multi-language inputs.
4. Veracity – The accuracy of the data varies, as users may enter misspelled words, misleading queries, or false
information.
5. Value – Search queries are valuable for insights into user behavior, trends, and marketing strategies, making them
crucial for businesses and data analysts.
Google’s search data is a prime example of big data, requiring advanced analytics, machine learning, and artificial
intelligence for processing. It is used for improving search algorithms, targeted advertising, and predicting trends globally.
 𝟒𝟒+𝟐𝟓+𝟏𝟎+𝟏𝟎+𝟏𝟎
(d) AM= = 19.8%
𝟓
Year Diminishing value (for Log x
∑logX
G.M.=Antilog ( ) =78.92 the value of tk.100)
𝑁
X
The diminishing average percentage of depreciation 1 100-44=56 1.7482
2 100-25=75 1.8751
= 100-78.92= 21.08% 3 100-10=90 1.9542
AM gives 19.8%, while GM gives 21.08%. 4 100-10=90 1.9542
5 100-10=90 1.9542
GM is more appropriate because depreciation follows a ∑logX=9.4859
multiplicative process rather than an additive one. Since
each year's depreciation is applied to the diminished value, the geometric mean better represents the average rate of
depreciation over time.
The arithmetic mean tends to overestimate or underestimate in cases of percentage changes applied successively.
Thus, the Geometric Mean (21.08%) is the correct approach for this problem.
Q-2
(a) The four scales of measurement are:
1. Nominal Scale Categorizes data without any inherent order. It only identifies or labels items.
Example: Types of fruits (Apple, Banana, Orange).
2. Ordinal Scale: Categorizes data with a meaningful order but without a fixed difference between values.
Example: Customer satisfaction levels (Satisfied, Neutral, Dissatisfied).
3. Interval Scale: Measures data with equal intervals between values but no true zero point.
Example: Temperature in Celsius (30°C is 10°C warmer than 20°C, but 0°C is not an absence of temperature).
4. Ratio Scale: Measures data with equal intervals and a true zero point, allowing for meaningful ratios.
Example: Height in centimeters (0 cm means no height, and 160 cm is twice as tall as 80 cm).
(b)

CGPA fi xi pcf fixi fi(xi-µ)2

1-2 5 1.5 5 7.5 11.25
2-3 25 2.5 30 62.5 6.25
3-4 40 3.5 70 140 10
N=70 ∑fixi=210 ∑ fi(xi-µ)2 =27.5
Mean µ = 210/70=3
𝑁
−𝑃𝐶𝐹 35−30
Median M = L + 2
∗ 𝑖 = 3+ ∗ 1 = 3+0.125 = 3.125
𝑓𝑖 40

∆1 15
Mode = L+ ∆1+∆2 * i = 3+15+40 ∗ 1 = 3+0.2727=3.2727

Range = maximum value – minimum value = 4-1=3

Variance 𝝈𝟐 = 27.5/70=0.39286 Standard deviation , 𝝈 = √0.39286 = 0.6268

• The mean (3.0) is a good summary of the dataset since it considers all values.
• The median (≈ 3) is useful, especially if outliers were present.
• Mode (3 – 4 range) shows the most common category.
• Since there’s low variability (σ ≈ 0.63), mean is the best measure of central tendency.
 • For dispersion, standard deviation (0.63) is better than range (3), as it considers all data points.
• Interpretation
• Most students have a CGPA around 3.0 – 3.5.
• The data is not highly dispersed, meaning students' performance is relatively consistent.
• Since the mean ≈ median ≈ mode, the data is symmetrically distributed (no major skewness).
Q-3
(a) We use relative measures of dispersion to compare variability between different datasets, especially when they have
different units or scales. Unlike absolute measures (like range or standard deviation), which depend on the units of data,
relative measures express dispersion as a proportion or percentage, making them useful for comparisons.
Reasons for Using Relative Measures of Dispersion
1. Comparison Across Different Datasets: If two datasets have different units (e.g., income in dollars vs. height in
centimeters), absolute dispersion measures cannot be directly compared.
Example: A standard deviation of 500 for income may seem large, but for population size, it might be small.
2. Scalability: Relative measures adjust for differences in magnitude, making comparisons meaningful regardless of
dataset size.
Example: Comparing stock market volatility across different companies with varying stock prices.
3. Unit-Free Interpretation: Relative measures (like coefficient of variation) do not depend on measurement units,
allowing for universal interpretation.
Example: A coefficient of variation (CV) of 10% for salaries and 15% for house prices means house prices are more
variable, regardless of currency units.
4. Better Insight into Variability: Helps in understanding risk vs. reward in investments, economics, and business decision-
making.
Example: In finance, a stock with a higher relative dispersion (CV) is considered more volatile.
Common Relative Measures of Dispersion
𝜎
Coefficient of Variation (CV) = 𝑥 ∗ 100
𝑄 𝑄
Quartile Coefficient of Dispersion = 𝑄3−𝑄1
3+ 1

Relative Range = range / mean

Relative measures of dispersion help in comparing datasets with different scales, sizes, and units, making them essential in
economics, finance, and social sciences.
(b)

Income fi xi fixi fi(xi-µ)2

0-1999 25 999.5 24987.5 136110722.2
2000-3999 25 2999.5 74987.5 2777722.223
4000-5999 40 4999.5 199980 111111555.6
N=90 299955 ∑ fi(xi-µ)2 =250000001
Mean µ= 299955/90=3332.83
∆1 15
Mode = L+ ∆1+∆2 * i = 4000 + 15+40 ∗ 2000 = 4000+545.45= 4545.45

𝐟𝐢(𝐱𝐢−µ)𝟐 250000001
Standard deviation = √ =√ = 1666.67
𝑁 90
 𝑚𝑒𝑎𝑛−𝑚𝑜𝑑𝑒 3332.83−4545.45
Skewness 𝛽1 = = = -0.7276
𝜎 1666.67

A skewness of -0.7276 indicates that the distribution is negatively skewed, meaning it has a longer or fatter tail on the left
side. In other words, the majority of the data points are concentrated on the right, with some smaller values pulling the
mean to the left. Generally, a skewness value between -1 and -0.5 suggests moderate negative skewness. So, with a
skewness of -0.7276, the distribution is slightly negatively skewed.
90𝑁
−𝑃𝐶𝐹 81−50
P90 =L+ 100 𝑓𝑖 ∗ 𝑖 = 4000 + ∗ 2000 = 5550
40

10𝑁
−𝑃𝐶𝐹 9−0
P10 =L+ 100
∗𝑖=0+ ∗ 2000 =720
𝑓𝑖 25

𝑁
−𝑃𝐶𝐹 22.5−0
Q1= L+ 4 ∗𝑖=0+ ∗ 2000 = 1800
𝑓𝑖 25

3𝑁
−𝑃𝐶𝐹 67.5−50
Q3 = L+ 4
∗ 𝑖 = 4000 + ∗ 2000 = 4875
𝑓𝑖 40

𝑄.𝐷 (4875−1800)/2 1537.5

Kurtosis 𝛽2 = P90−P10 = = 4830 = 0.3183
5550−720

Since this value is less than 0.263 (normal distribution kurtosis using this method), the data is platykurtic, meaning the
distribution has lighter tails and a flatter peak compared to a normal distribution. This suggests that student incomes are
more evenly spread out with fewer extreme values.
Q-04
(a) Correlation analysis is a statistical method used to determine the strength and direction of the relationship between two
or more variables. It helps in understanding whether and how strongly pairs of variables are related. The result of
correlation analysis is usually expressed through a correlation coefficient, such as Pearson's correlation coefficient (r),
which ranges from -1 to +1:

• +1 indicates a perfect positive correlation (as one variable increases, the other also increases).
• -1 indicates a perfect negative correlation (as one variable increases, the other decreases).
• 0 indicates no correlation.
Significance of Studying Correlation:
1. Understanding Relationships: Correlation analysis helps identify whether there is any association between variables,
such as whether an increase in one variable is linked to an increase or decrease in another.
2. Predictive Power: Knowing the correlation between variables allows for predictions. For example, if you know the
correlation between advertising expenditure and sales, you might predict future sales based on changes in the advertising
budget.
3. Simplifying Complex Data: In cases with multiple variables, correlation analysis helps simplify and highlight the most
significant relationships, making data easier to interpret.
4. Decision Making: In business, economics, and social sciences, correlation analysis is crucial for informed decision-
making. For instance, understanding the correlation between interest rates and inflation can help policymakers set
effective monetary policies.
5. Identifying Causal Relationships (with caution): While correlation does not imply causation, studying correlations can
help generate hypotheses about causal relationships that may be tested with further research.
correlation analysis is important because it offers insight into the connections between variables, aids in predictions, and
supports more informed decision-making.
(b) Properties of Correlation Coefficient
1) The range of correlation coefficient is -1 to +1.
2) The correlation coefficient is unit free measure.
3) The correlation coefficient between two variables is symmetry.
4) If r = +1 then the correlation is perfect positive correlation.
5) If r=1 then the correlation is perfect negative correlation.
6) If r = 0 then no/zero correlation.
7) If r > 0 then correlation is positive correlation.
8) If r < o then correlation is negative correlation.
(c)

Month [Link] (X-𝑿 ̅) ̅ )2

(X-𝑿 Sales (Y-𝒀̅) ̅)2
(Y-𝒀 xy
X x Y y
Jan. 50 -70 4900 1600 -1000 1000000 70000
Feb 60 -60 3600 2000 -600 360000 36000
March 70 -50 2500 2200 -400 160000 20000
April 90 -30 900 2500 -100 10000 3000
May 120 0 0 2400 -200 40000 0
June 150 30 900 2600 0 0 0
July 140 20 400 2800 200 40000 4000
August 160 40 1600 2900 300 90000 12000
September 170 50 2500 3100 500 250000 25000
October 190 70 4900 3900 1300 1690000 91000
Total 1200 0 22200 26,000 0 3640000 261000
𝑋̅ = 1200/10 = 120 𝑌̅ = 26,000/10 = 2600
∑ 𝑥𝑦 261000 261000
r= = = 284267.4797 = 0.918
√∑ 𝑥 2∗∑ 𝑦2 √22200∗3640000

There is a very high degree of positive correlation between advertising expenditure and sales.
Q-05
(a) Regression analysis estimates the relationship between a dependent variable and one or more independent variables. It
measures the average change in the dependent variable for a unit change in an independent variable, holding other factors
constant. This means that regression does not capture individual variations but provides an overall trend based on
observed data.
For example, in a simple linear regression model:

Y = 𝛽 0 + 𝛽 1X + 𝜀
Y is the dependent variable,
X is the independent variable,

𝛽 0 is the intercept (the value of Y when X=0),

𝛽 1 is the regression coefficient, which measures the average change in Y for a one-unit increase in X,
𝜀 is the error term.
Regression Coefficients & Their Interpretation

Regression coefficients (𝛽 0 𝛽 1, 𝛽 2………,) quantify the relationship between independent variables and the dependent
variable.
Simple Linear Regression (One Independent Variable):

• 𝛽 1 indicates how much Y changes on average for a one-unit increase in X.

• If 𝛽 1>0, there is a positive relationship (as X increases, Y increases).
• If 𝛽 1<0, there is a negative relationship (as X increases, Y decreases).
• If 𝛽 1= 0, X has no effect on Y.
Multiple Regression (More Than One Independent Variable):
Each coefficient (𝛽 1, 𝛽 2………,)) measures the effect of its corresponding independent variable while holding all other
independent variables constant.
b) Important Characteristics of Regression
1. Direction of Relationship: Determines whether the relationship between variables is positive or negative.
2. Magnitude of Relationship: The size of the regression coefficients tells us how strong the effect of an independent
variable is.
3. Predictive Power: Regression allows for predictions, but accuracy depends on model fit (e.g., R-squared value).
4. Linearity Assumption: In linear regression, the relationship between variables is assumed to be linear.
5. Impact of Outliers: Extreme values can distort the estimated regression coefficients.
6. Dependence on Data Quality: Results depend on the accuracy and completeness of the data used.
7. Multicollinearity: When independent variables are highly correlated, it can distort coefficient estimates.
8. Homoscedasticity: The variance of the residuals (errors) should be constant across all values of the independent
variable(s).
9. Error Term Assumptions: The error terms should be normally distributed and uncorrelated for valid inference.
(c) (i)

Number of ads Cans purchased XY X2

X Y
3 11 33 9
7 18 126 49
4 9 36 16
2 4 8 4
0 7 0 0
4 6 24 16
1 3 3 1
2 8 16 4
Total = 23 66 246 99
𝑋̅ = 23/8 = 2.875 𝑌̅ = 66/8 =8.25
∑ 𝑋𝑌−𝑛 𝑋̅ 𝑌̅ 246−8∗2.875∗8.25
b= = =1.7110
∑𝑋 2−𝑛𝑋̅ 2 99−8∗𝟐.𝟖𝟕𝟓𝟐

a = 𝑌̅ -b𝑋̅ = 8.25 – 1.7110*2.875 = 3.331

thus, 𝑌̂ = a+ bX = 3.331+1.7110X
(ii)

X Y ̂
𝒀 ̂)2
(Y-𝐘
3 11 8.464 6.43
7 18 15.308 7.247
4 9 10.175 1.381
2 4 6.753 7.579
0 7 3.331 13.462
4 6 10.175 17.430
1 3 5.042 4.17
2 8 6.753 1.555
59.254
̂2
∑(𝑌−𝑌) 59.254
standard error of estimate, se= √ =√ = 3.1426
𝑛−2 6

(iii)

• The equation suggests that when no ads are seen (X=0), approximately 3.33 cans are purchased.
• For each additional advertisement viewed, the number of cans purchased increases by about 1.71.
• The standard error of 3.14 indicates that the actual number of cans purchased typically deviates by about 3.14
cans from the predicted values.

Mid -1(c)
(1) Understanding these concepts is essential for analyzing and interpreting data effectively. Here's why each one is
important:
a. Statistics: Statistics helps in collecting, analyzing, interpreting, and presenting data. It allows for making informed
decisions based on data patterns, trends, and probabilities.
b. Histogram: A histogram visually represents the distribution of a dataset, helping to understand frequency, spread, and
the shape of data distribution. It is crucial for detecting patterns like normality, skewness, or outliers.
c. Dispersion: Dispersion measures how spread-out data points are in a dataset. Common measures include variance,
standard deviation, and range. It helps in understanding the variability and consistency of data.
d. Central Tendency: Measures of central tendency (mean, median, and mode) summarize data by providing a central
value. They help in understanding the general trend of a dataset.
e. Bar Chart: A bar chart is used for comparing categorical data. It visually represents differences in quantities, making it
easier to analyze trends and comparisons across categories.
f. Skewness and Kurtosis: Skewness measures the asymmetry of a data distribution. A skewed dataset may indicate biases
or trends that require attention. Kurtosis measures the "tailedness" of a distribution, indicating whether data have extreme
values (outliers) that could affect statistical analysis.
These concepts are fundamental in economics, finance, social sciences, and business analytics, helping to make data-
driven decisions and predictions
(2)(a) There is no single "best" measure of central tendency; the choice depends on the nature of the data and its
distribution:
1. Mean (Arithmetic Average) – Best when data is symmetrical and has no extreme outliers.
Example: Income levels in a controlled study with no extreme values.
Limitation: Affected by outliers and skewed data.
2. Median (Middle Value) – Best when data has skewness or outliers, as it is not affected by extreme values.
Example: Income distribution in a country where a few billionaires significantly increase the mean.
Limitation: Ignores the actual values of most data points.
3. Mode (Most Frequent Value) – Best for categorical data or when analyzing the most common occurrence.
Example: Finding the most common shirt size sold in a store.
Limitation: May not represent the dataset well if multiple modes exist or if data is continuous.
When might the others be used?

• Use Mean when data is normally distributed (e.g., heights of students in a classroom).
• Use median when data is skewed (e.g., house prices in a city).
• Use mode when dealing with categorical data (e.g., favorite ice cream flavors).
b. Relative measures of dispersion (like coefficient of variation, quartile deviation, or relative standard deviation) are
needed when comparing variability between datasets with different units or scales.
Reasons for Using Relative Measures of Dispersion:
1. Comparing different datasets:
Example: Comparing income variability between two countries (where one country’s currency has larger numerical
values).
2. Understanding risk in financial data:
Example: In investment, Coefficient of Variation (CV) = Standard Deviation / Mean helps compare risk across stocks with
different average returns.
3. Standardizing variability:
Example: Comparing student test scores across different exams where the total marks differ.
Absolute measures (variance, standard deviation) are useful within the same dataset, but relative measures help compare
across datasets.

25TH FINAL
Q-01(a) Statistics is a branch of mathematics that deals with the collection, organization, analysis, interpretation, and
presentation of data. It helps in making informed decisions based on numerical evidence.
Scope of Statistics in Business and Management:
Statistics plays a crucial role in various aspects of business and management, including:
1. Business Decision Making: Helps in analyzing market trends, customer preferences, and financial performance.
2. Marketing and Market Research: Used in demand forecasting, consumer behavior analysis, and advertising
effectiveness.
3. Finance and Investment: Helps in risk analysis, portfolio management, and financial forecasting.
4. Production and Operations Management: Assists in quality control, inventory management, and process optimization.
5. Human Resource Management: Used for workforce planning, performance evaluation, and salary structuring.
6. Economics and Policy Making: Helps in economic forecasting, inflation analysis, and impact assessment of policies.
(b) Despite its advantages, statistics has some limitations:
1. Does Not Provide Complete Information: Statistics only provide quantitative analysis and may ignore qualitative
factors.
2. Can Be Misleading: Data manipulation or improper interpretation can lead to incorrect conclusions.
3. Requires a Large Sample Size: Small or biased samples may not represent the whole population accurately.
4. Cannot Establish Causality: Correlation does not imply causation; external factors may influence results.
5. Dependence on Assumptions: Statistical methods rely on certain assumptions, which, if incorrect, may lead to wrong
conclusions.
6. Time-Consuming and Costly: Collecting and analyzing data can be expensive and time-intensive.
(c) Primary data refers to firsthand information collected directly from the source for a specific purpose. It is obtained
through methods such as surveys, interviews, experiments, and observations. Since it is gathered specifically for a
research objective, it is often more accurate and relevant. However, collecting primary data can be time-consuming and
costly.
Secondary data, on the other hand, is information that has already been collected and compiled by someone else for a
different purpose. It is found in sources such as government reports, books, research papers, and company records.
Secondary data is easier and cheaper to obtain but may not always be fully relevant or up-to-date for the current study.
The key differences between the two are their sources, purpose, originality, cost, and reliability. While primary data is
more specific and accurate, secondary data is useful for gaining background knowledge and saving time.
For example, a company conducting a customer satisfaction survey collects primary data, whereas using government
census data to study market trends involves secondary data. Both types of data are essential in business and research,
depending on the needs and constraints of the study.
Q-02
(a) The weighted mean is an important statistical measure that gives different weights to values based on their relative
importance. Unlike the simple arithmetic mean, where all values are treated equally, the weighted mean accounts for
variations in significance, making it more accurate in many real-world applications.
Key Significance of Weighted Mean:
1. More Accurate Representation: It reflects the true impact of values by assigning higher importance to more significant
data points.
2. Useful in Business & Economics: Used in financial analysis, price indices, and economic indicators, such as calculating
GDP or inflation rates.
3. Application in Academics: In student grading, different assignments and exams may carry different weights, ensuring
fair evaluation.
4. Better Decision Making: Helps businesses analyze sales, profits, and customer preferences when some factors
contribute more than others.
5. Financial Portfolio Analysis: Used in stock market investments, where different assets have different levels of influence
on the overall portfolio performance.
Overall, the weighted mean is crucial in situations where all data points do not have equal significance, ensuring a more
meaningful and fair analysis.
∑ 𝑥𝑤 14(16.5)+10(19.00)+2(25.00)
𝑋̅= ∑ 𝑤 = = 17.038
14+10+2
(b)

(i)The histogram shows that the lower tail of the distribution is fatter than the upper tail
(ii)q-If each ski lift chair holds two people but is limited in total safe weight capacity to 400 pounds, what can the operator
do to maximize the people capacity of the ski lift without exceeding the safe weight capacity of a chair? Do the data
support your proposal?
Ans- There are very few peoples are weighing 180 pounds or above. So that the operator affords to pair each person who
appears to be heavy person with low weigh person. This can be done without exceeding the safe weight capacity of a
chair.
Q-3
(a) Yes! Measures of central tendency, also called statistics of location, describe where the center of a dataset lies. These
measures summarize a distribution by identifying a typical or central value around which the data points are distributed.
The three main measures are:
1. Mean (Arithmetic Average): The mean is calculated by summing all values and dividing by the total number of
observations.
Formula: Mean = ∑X / N
2. Median (Middle Value): The median is the middle value when data is arranged in ascending order. If there is an even
number of observations, the median is the average of the two middle values. It is not affected by outliers, making it a
better measure for skewed distributions.
3. Mode (Most Frequent Value): The mode is the most frequently occurring value or class in the dataset. A dataset can be
unimodal (one mode), bimodal (two modes), or multimodal (more than two modes). It is useful for categorical data and
for identifying the most common weight class in your dataset.
Why Are These Important?

• They summarize large datasets into a single representative value.

• They help compare distributions (e.g., comparing ski lift users' weights across different days).
• They provide insight into the shape of the distribution (e.g., a large gap between the mean and median suggests
skewness).
(b) (i)

Ages Tally Frequency (fi) xi fixi c.f.

40-49 |||| 4 44.5 178 4
50-59 |||| 4 54.5 218 8
60-69 ||| 3 64.5 193.5 11
70-79 || 2 74.5 149 13
80-89 |||| || 7 84.5 591.5 20
∑fi=20 ∑fixi = 1330
(ii) sample mean 𝑋̅= ∑fixi / n = 1330/20 = 66.5
The average age of COVID-19 patients admitted is 66.5 years.
 𝑁
−𝑝𝑐𝑓 10−8
(iii) median = L + 2 ∗ 𝑖 = 60 + *10 = 60 + 6.67 =66.67
𝑓𝑖 3

∆1 5
Mode = L + ∆∆2 ∗ 𝑖 = 80+ 5+7 ∗ 10 = 80 + 4.17= 84.17

(iv) The mean (66.5) is slightly lower than the median (66.67).
This suggests that the distribution is likely left-skewed (or
negatively skewed). In a negatively skewed distribution, the longer
tail is on the left side of the peak, and the mean is typically pulled
in the direction of the skew, which makes it smaller than the
median.
The mode (84.17) is the highest value, indicating that the most
frequent observations are clustered around the higher values.
The median (66.67) is slightly higher than the mean and represents the middle value. The mean (66.5) is lower because it
is pulled toward the lower values due to the left skew.
24th final
1.(b) The variable is quantitative and discrete, as it represents the number of patents granted, which are counted in whole
numbers.
Measurement Level: The variable uses the Ratio measurement scale because:

• It involves numerical data.

• There is a true zero point (a company can have zero patents).
• Differences and ratios between values are meaningful.
Characteristics of the Variable:

• Quantitative: Measured numerically.

• Discrete: Countable in whole numbers.
• Ratio scale: True zero, and meaningful comparisons are possible.
Characteristics of the Measurement Scale:

• Ordered: The values can be ranked.

• Meaningful Zero: Zero represents no patents.
• Equal Intervals: Differences between values are consistent.
• Absolute Zero: Represents complete absence of patents.
2.a -- Dispersion refers to the spread or variability of data points in a dataset. It helps in understanding how much the data
deviates from the central value (mean).
Measures of Dispersion:
1. Range: The difference between the maximum and minimum values.
2. Variance: The average of the squared differences from the mean.
3. Standard Deviation: The square root of the variance, providing spread in the original units.
4. Interquartile Range (IQR): The range between the first quartile (Q1) and third quartile (Q3), representing the middle
50% of the data.
Absolute Measures of Dispersion: These are measures of spread expressed in the same units as the data. They give the
actual dispersion in the original units.
Examples: Range, Variance, Standard Deviation.
Relative Measures of Dispersion: These measures allow for comparisons between datasets with different units or scales.
They are dimensionless, meaning they are ratios or percentages that standardize the dispersion.
Examples: Coefficient of Variation (CV), which is the ratio of the standard deviation to the mean, expressed as a
percentage:
𝜎
CV= 𝜇 ∗ 100

Five Number Summary:

A summary of a dataset's distribution, consisting of:
1. Minimum: The smallest value.
2. Q1: First quartile (25th percentile).
3. Median: The middle value (50th percentile).
4. Q3: Third quartile (75th percentile).
5. Maximum: The largest value.
The Five Number Summary provides a quick overview of the dataset’s spread, central tendency, and skewness.
(b)(i)

Life in Model-X Mid-point- fixi fi(xi-𝝁)2 Model Y Mid-point- fixi fi(xi-𝝁)2

years fi xi fi xi
2-4 1 3 3 22 2 3 6 40.141
4-6 7 5 35 50.65 7 5 35 43.052
6-8 12 7 84 5.71 9 7 63 2.074
8-10 10 9 90 17.16 12 9 108 27.725
10-12 4 11 44 43.82 2 11 22 24.781
12-14 1 13 13 28.2 1 13 13 30.47
Total N=35 269 167.54 N=33 247 168.243
Mean for model X =269/35 = 7.69
Mean for model Y = 247/33 = 7.48
So, model X has a slightly higher average life than model Y.

∑ 𝐟𝐢(𝐱𝐢−𝝁)𝟐
(ii) standard deviation of model X 𝜎X= √ = √167.54/35 = 2.19
𝑁

∑ 𝐟𝐢(𝐱𝐢−𝝁)𝟐
standard deviation of model Y 𝜎y = √ = √168.243/33 = 2.26
𝑁

∆1 5
Mode of model X = L + ∆∆2 ∗ 𝑖 = 6 + 5+2 ∗ 2 = 7.43
∆1 3
Mode of model Y = L + ∗𝑖 =8+ ∗ 2 = 8.46
∆∆2 3+10

𝑚𝑒𝑎𝑛−𝑚𝑜𝑑𝑒 7.69−7.43
Coefficient of skewness of model X = = = 0.12
𝜎𝑋 2.19

𝑚𝑒𝑎𝑛−𝑚𝑜𝑑𝑒 7.48−8.46
Coefficient of skewness of model Y = = = -0.43
𝜎𝑦 2.26
Model X: Slightly positively skewed (0.12) → Data is almost symmetric, with a small tail on the right.
Model Y: Negatively skewed (-0.43) → Data is left-skewed, meaning more values are concentrated on the higher side.
(iii) As model Y has slightly high standard deviation, it has more variable than model X.

Model-X Mid-point-xi fi(xi-𝝁)4 Model Y Mid-point-xi fi(xi-μ)4

fi fi
1 3 483.83 2 3 805.65
7 5 366.53 7 5 264.79
12 7 2.72 9 7 0.48
10 9 29.45 12 9 64.06
4 11 480.14 2 11 307.04
1 13 795.02 1 13 928.45
N=35 2157 N=33 2352.47
2157
𝜇
Kurtosis of model X = 𝜎44 = 35
= 2.67
2.194
2352.47
𝜇4
Kurtosis of model Y = 𝜎4 = 33
= 2.73
2.264

Model X has a lower kurtosis (2.67), meaning its data is more spread out and has fewer outliers. Model Y has a slightly
higher kurtosis (2.73), meaning its distribution is closer to normal but still slightly platykurtic. Neither model has a
leptokurtic distribution (kurtosis > 3), so neither has extreme outliers.
Q-3
number of male in “mildly support” class
(a)(i) P (Male mildly supports) = = 25/80 = 0.3125
total number of male polled

𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑓𝑒𝑚𝑎𝑙𝑒 𝑖𝑛 “𝑢𝑛𝑑𝑒𝑐𝑖𝑑𝑒𝑑” 𝑐𝑙𝑎𝑠𝑠

(ii) P (female undecided) = =12/80 = 0.15
total number of female polled

(iii) P(Strongly or mildly support) = (44+44)/160= 0.55

Opinion Frequency
(combined)
SS 44
(iv) Relative frequency. MS 44
U 25
MO 26
SO 21
TOTAL 160

24th mid

1(a)i-mode = 3median -2mean

Median =(mode+2mean)/3 = [42+2*46.2]/3 = 44.8

(ii) mean = 2665/18 = 148.056

(iii) mode = 3median -2mean

Mean = (3median-mode)/2 = [3*28.37-23.92]/2 = 30.595

(b)

Sales fi xi fixi cf
20-25 3 22 66 3
25-30 5 27 135 8
30-35 17 32 544 25
35-40 9 37 333 34
40-45 5 42 210 39
N=39 1288
(i) 𝒙
̅ = 1288/39 = 33.026
∆1 12
(ii) mode = L + ∆∆2 ∗ 𝑖 = 30 + 12+8 ∗ 5 =33
𝑁
−𝑝𝑐𝑓 19.5 − 8
(iii)median = L + 2
∗ 𝑖 = 30 + ∗ 5 = 30 + 3.38 = 33.38
𝑓𝑖 17

(iv) Since the three values are very close, the distribution appears approximately symmetric. There is no significant
skewness in the data.
Q-2

(a) (i) To present the data visually, a bar chart would be a

suitable choice. In the bar chart, each product type
(Pepsodent, Close Up Renue, Colgate, Fresh Gel, close
up) will be plotted on the x-axis, and the number of
customers who prefer each product will be plotted on
the y-axis.

(II) The data is categorical, as it classifies customers' preferences by the type of toothpaste. The distribution of customer
preferences shows that Colgate is the most preferred brand, with 49 customers choosing it. On the other hand, Fresh Gel
is the least preferred, with only 16 customers. This kind of data can help the company understand which brands are more
popular and guide their decisions for stock and marketing strategies in Dhaka City.

(b)(i)
ROI% fi xi filogxi 𝒇𝒊
𝒙𝒊
10-15 2 12 2.1584 0.1667
15-20 3 17 3.6913 0.1765
20-25 10 22 13.4242 0.4545
25-30 4 27 5.7255 0.1481
30-35 1 32 1.5051 0.0313
N=20 26.5045 0.9771
∑ 𝑓𝑖𝑙𝑜𝑔𝑥𝑖
G.M.= antilog ( ) = antilog (26.5045/20) =antilog (1.325225) =21.146
𝑁
𝑁
H.M.= 𝑓𝑖 = 20/0.9771=20.4687
∑𝑥𝑖

(ii) The Geometric Mean (GM) = 21.146 is slightly higher than the Harmonic Mean (HM) = 20.4687, which is expected
since HM is always less than or equal to GM. The small difference suggests that the data points are relatively close in
value, indicating minimal variation.
Q-3(a) Central tendency refers to the statistical measure that identifies a single value as the most representative of an
entire dataset. It aims to describe a dataset by finding a central or typical value around which the data is distributed. The
three main measures of central tendency are the Mean, Median, and Mode.
1. Mean (Arithmetic Mean): The mean is the sum of all values in a dataset divided by the number of values.
Formula: Mean =∑ X / N

• X represents individual values

• N is the total number of values
Example: Consider the dataset: 5, 10, 15, 20, 25
Mean = (5 + 10 + 15 + 20 + 25)/5 = 15
Use Cases:

• Used in economics for average income, GDP per capita, and stock market returns.
• In business, it helps calculate average sales, costs, or profits.
2. Median: The median is the middle value when the data is arranged in ascending or descending order. If the dataset has
an even number of observations, the median is the average of the two middle values.
Example: Odd number of observations: Dataset: 3, 7, 9, 12, 14 → Median = 9 (middle value)
Even number of observations: Dataset: 4, 8, 10, 12 = Median = (8 + 10) / 2 = 9
Use Cases:

• Preferred over the mean when data has outliers (e.g., median income is often used instead of mean income to
avoid distortion by extremely high earnings).
• Used in real estate to find the median house price.
3. Mode: The mode is the most frequently occurring value in a dataset. A dataset can have:

• One mode (Unimodal)

• Two modes (Bimodal)
• Multiple modes (Multimodal)
Example: Dataset: 2, 3, 4, 4, 5, 5, 5, 6, 7
Mode = 5 (since it appears the most)
Use Cases:

• Used in marketing to determine the most popular product size, color, or brand preference.
• In education, mode helps identify the most common exam score.
24th mid-2
1(a) (i) Classical probability is based on the assumption that all possible outcomes of an experiment are equally likely.
The probability of an event occurring is calculated as:
Number of favorable outcomes
P(A) =
𝑇𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑝𝑜𝑠𝑠𝑖𝑏𝑙𝑒 𝑜𝑢𝑡𝑐𝑜𝑚𝑒𝑠

Example: Consider rolling a fair six-sided die. The probability of rolling a 3 is: P(3)=1/6
Since each face of the die has an equal chance of appearing, classical probability applies here.

(ii) Mutually Exclusive Events: Two events are mutually exclusive if they cannot occur at the same time. This means
their intersection is empty: P(A∩B) = 0
Example:
• When rolling a die, the events A: rolling an even number {2,4,6} and B : rolling an odd number {1,3,5} are
mutually exclusive because a single roll cannot be both even and odd.
• If we draw a card from a deck, the events A: drawing a red card and B: drawing a black card are mutually
exclusive.
For mutually exclusive events:

P(A∪B) =P(A)+P(B)
(iii) The total probability theorem states that if we have a set of mutually exclusive and exhaustive events B1, B2..., Bn
that cover the entire sample space, then the probability of any event A can be found as:
P(A)=P(A∣B1) P(B1) +P(A∣B2) P(B2) +...+P(A∣Bn) P(Bn)
or in general

P(A)= ∑P(A∣Bi) P(Bi)

Example: Suppose a factory has three machines producing items:
• Machine M1 produces 30% of items,
• Machine M2 produces 50% of items,
• Machine M3 produces 20% of items.
The probability of a defective item from each machine is:
• P(D∣M1) =0.01

• P(D∣M2) =0.02

• P(D∣M3) =0.05
The probability of getting a defective item is:
P(D)= (0.01×0.3) +(0.02×0.5) +(0.05×0.2) =0.023

(iv) The multiplication rule states that the probability of two events A and B happening together (joint
probability) is given by: P(A∩B) = P(A)P(B∣A)
or if A and B are independent: P(A∩B) = P(A)P(B)

Example 1 (Dependent Events):

A box contains 3 red and 5 blue balls. If we draw two balls without replacement,
• P(A) = probability of drawing a red ball first = 3/8
• P(B∣A) = probability of drawing a red ball again = 2/7
• P(A∩B) =3/8×2/7 = 3/28

Example 2 (Independent Events): If we toss a coin and roll a die, the probability of getting heads and rolling
a 4 is:

P(H∩4) = P(H)P(4)=1/2×1/6=1/12
(v) Bayes' Theorem: Bayes' Theorem is a mathematical formula used to determine the probability of an event
based on prior knowledge of related conditions. It is expressed as:
P(B | A)∗ P(A)
P (A | B) = P(B)

• P (A | B) is the probability of event A given event B,

• P (B | A) is the probability of event B given event A,
• P(A) is the prior probability of event A,
• P(B) is the prior probability of event B.
This theorem is fundamental in updating probabilities when new evidence is introduced, often used in statistics,
machine learning, and decision-making processes.
(b)(i)

(ii) Total paths = 3*3*2=18

Number of paths equal to 12=1
probability of completing the work within target time = 1/18 = 0.056
(ii) Number of paths equal to 11 =0
Number of paths equal to 13 = 3
probability of completing the work one month later = 3/18= 0.167
(iii) Number of paths equal to 10 =0
Number of paths equal to 14 = 5
probability of completing the work two month later = 5/18=0.278
 Q-2
(a) Correlation is a statistical measure that expresses the extent to which two variables are linearly related. It shows
whether and how strongly pairs of variables are related to each other, meaning how one variable changes in relation to the
other. The correlation coefficient ranges from -1 to +1:

• +1 indicates a perfect positive linear relationship.

• -1 indicates a perfect negative linear relationship.
• 0 means no linear relationship.
Types of Correlation:
1. Positive Correlation: When one variable increases, the other variable also increases. The correlation coefficient will be
between 0 and +1.
Example: Height and weight of individuals. As height increases, weight tends to increase as well.
2. Negative Correlation: When one variable increases, the other decreases. The correlation coefficient will be between 0
and -1.
Example: The number of hours spent watching TV and academic performance. As TV watching increases, academic
performance often decreases.
3. Zero or No Correlation: When the variables show no predictable relationship. The correlation coefficient will be close
to 0.
Example: Shoe size and intelligence. There is no relationship between the two.
4. Perfect Correlation: A perfect linear relationship where one variable changes in exact proportion to the other. The
correlation is either +1 or -1.
Example: The relationship between the temperature in Celsius and Fahrenheit.
Distinction between Correlation and Regression:
Correlation measures the strength and direction of the relationship between two variables but does not imply a cause-and-
effect relationship. If we have data on hours studied and exam scores, we can use correlation to see if there is a
relationship between the two, and if so, how strong it is
Regression is a statistical method used to model the relationship between a dependent variable and one or more
independent variables. It predicts the value of the dependent variable based on the values of the independent variable(s). If
we want to predict a student's exam score based on the number of hours studied, we would use regression analysis.
(b)
X Y XY X2 Y2
9 36 324 81 1296
7 25 175 49 625
8 33 264 64 1089
4 15 60 16 225
7 28 196 49 784
5 19 95 25 361
5 20 100 25 400
6 22 132 36 484
∑X=51 ∑Y=198 ∑XY=346 ∑ X2 =345 ∑ Y2 =5264

𝑋̅=51/8=6.375
̅ = 198/8 = 24.75
Y
 ∑ 𝑋𝑌−∑ 𝑛𝑋̅ 𝑌̅ 1346−8(6.375)(24.75)
b= ∑ 𝑋 2−𝑛𝑋̅ 2
= = 4.2138
345−8∗6.3752

a=̅ ̅ = -2.1130
Y -bX

𝑌̂ = -2.1130+4.2138X

(ii) when X=10

𝑌̂ =-2.1130+4.2138*10=40.025

when X=12

𝑌̂ =-2.1130+4.2138*12=48.4526

∑ 𝒀𝟐 −𝒂∑𝒀−𝒃∑𝑿𝒀
(iii)𝒔𝒆 = √ = 1.3291
𝒏−𝟐

Q-3
(a)(i) marginal probability is the probability of an event occurring without considering other variables:

P(M) = 60/200 = 0.3

P(F)= 140/200 =0.7

P(E)= 135/200 =0.675

P(I) = 65/200 =0.325

(ii) joint probability is the probability of two events happening together:

P (M∩ 𝐸) = 38/200 =0.19

P (M∩ I) = 22/200=0.11

P (F∩ 𝐸) = 97/200=0.485
P (F∩ I) = 43/220=0.215

(iii) conditional probability is the probability of an event given that another event has already occurred :
P (E∩𝑀)
Probability of being Efficient given Male: P(E|M) = = 0.19/0.3 =0.633
𝑃(𝑀)

P (I∩M)
Probability of being Inefficient given Male: P(I|M) = = 0.11/0.3=0.367
𝑃(𝑀)

P (E∩𝐹)
Probability of being Efficient given Female: P(E|F) = =0.485/ 0.7=0.6929
𝑃(𝐹)

P (I∩F)
Probability of being Inefficient given Female: P(I|F) = = 0.215/0.7=0.307
𝑃(𝐹)

P (M∩𝐸)
Probability of being Male given Efficient: P(M|E) = = 0.19/0.675=0.2815
𝑃(𝐸)

P (F∩𝐸)
Probability of being Female given Efficient: P(F|E) = = 0.485/0.675=0.7185
𝑃(𝐸)

P (M∩I)
Probability of being Male given Inefficient: P(M|I) = =0.11/0.325=0.3385
𝑃(𝐼)

P (F∩I)
Probability of being Female given Inefficient: P(F|I) = = 0.215/0.325=0.6615
𝑃(𝐼)
(iv) Union probability is the probability that at least one of two events occurs, calculated using the formula:

1. Probability of being Male or Efficient: P (M or E) = P(M) + P(E) – P (M ∩ E) = 0.30 + 0.675 - 0.19 = 0.785
2. Probability of being Male or Inefficient: P (M or I) = P(M) + P(I) – P (M ∩I) = 0.30 + 0.325 - 0.11 = 0.515

3. Probability of being Female or Efficient = P (F or E) = P(F) + P(E) – P (F ∩E) = 0.70 + 0.675 - 0.485 = 0.89

4. Probability of being Female or Inefficient: P (F or I) = P(F) + P(I) – P (F ∩I) = 0.70 + 0.325 - 0.215 = 0.81

23rd final

1.(a) Statistics plays a vital role in both business and economics by providing data-driven insights for decision-
making, forecasting, and policy development. Its applications help improve efficiency, minimize risks, and optimize
resource allocation.

Scope of Statistics in Business

1. Market Research & Consumer Behavior – Analyzes consumer preferences, purchasing patterns, and demand
trends to guide marketing strategies.

2. Quality Control & Production – Ensures product consistency, reduces defects, and improves manufacturing
processes through statistical tools like Six Sigma.

3. Financial Analysis & Investment Decisions – Assists in stock market predictions, risk assessment, and portfolio
management using statistical models.
4. Business Forecasting – Helps estimate future sales, demand, and revenue using past trends and predictive
analytics.

5. Human Resource Management – Evaluates employee performance, determines salary structures, and predicts
workforce trends.

6. Supply Chain & Logistics – Optimizes inventory management, transportation, and demand forecasting to enhance
operational efficiency.
Scope of Statistics in Economics

1. Economic Planning & Policy Making – Governments use statistical data for economic development strategies,
resource allocation, and policymaking.
2. Demand & Supply Analysis – Helps in understanding market equilibrium, pricing strategies, and consumption
patterns.

3. Inflation & Price Index Measurement – Used to calculate price indices like CPI and WPI to monitor inflation and
cost-of-living changes.

4. Income & Wealth Distribution Studies – Analyzes disparities in income distribution and supports policies for
economic equality.
5. National Income Estimation (GDP, GNP) – Measures economic growth and overall national productivity.

Overall, statistics is a powerful tool that enables businesses and governments to make informed decisions, predict
future trends, and enhance economic stability.
(b)(i) A frequency distribution simplifies data analysis by organizing data into intervals, making patterns and
trends more apparent. The steps to create a frequency distribution include determining the number of classes,
calculating the class width, setting class intervals, tallying data, and creating the table.

Maximum value 968

Minimum value 27
Range = 968-27= 941

Number of classes = 1+3.322log (48)= 6.58 ≈ 7

Class width= 941/7≈ 135

Class interval Tally Frequency

25-160 |||||||||||||| 17
160-295 ||||||| 8
295-430 |||||| 7
430-565 |||||| 7
565-700 ||| 3
700-835 |||| 4
835-970 || 2
N=48
(ii) Cumulative frequency polygon by more than method

More than 25 48
“ “ 160 31
“ “ 295 23
“ “ 430 16
More than 565 9
“ “ 700 6
“ “ 835 2

(iii) the lowest 10% ending balances can be considered as the 10 th percentiles of the data.
10
The 10th percentile = (100 ∗ 48)th observation

≈5th observation

= 57 [5th lowest observation]

Students with balances below 57 will have to pay the service charge. Students with balances 57 or above will not be
charged.

Q-2
(a) The arithmetic mean is a common measure of central tendency, but it can be heavily influenced by extreme
values or outliers. For example, in a dataset with most values clustered around a certain range but a few extreme
values, the mean can shift toward these extremes, giving a distorted representation of the data's central tendency.
In such cases, alternative measures may provide a better reflection of the central value.
The mode is the most frequent value in a dataset and is particularly useful for categorical data or when identifying
the most common value. It is not affected by extreme values, making it ideal for datasets with repeated categories.
The median represents the middle value when the data is ordered and is less sensitive to outliers, making it
appropriate for skewed distributions or datasets with extreme values. It provides a better central measure when
the data is not symmetrically distributed.
The geometric mean is used for datasets involving products, percentages, or growth rates. It reduces the impact of
extreme values by averaging on a multiplicative scale, making it suitable for data spanning several orders of
magnitude.
The harmonic mean is useful for rates, such as speed or efficiency, and is weighted toward smaller values. It’s
appropriate when the data involves ratios or rates, as it gives more importance to smaller values, mitigating the
impact of large extremes.
In summary, choosing the appropriate measure of central tendency depends on the nature of the data and whether
there are extreme values or outliers that could distort the analysis.

(b) (i)

Class interval fi cf
10-20 12 12
20-30 30 42
30-40 X 42+x
40-50 65 107+x
50-60 Y 107+x+y
60-70 25 132+x+y
70-80 18 150+x+y
N=229
Given median 46
Median class = 40-50
114.5−42−𝑋
Median = 40 + ∗ 10
65

72.5−𝑋
46 = 40+ ∗ 10
65

390=725-10x

X=335/10=33.5
Again, 229=150+x+y
229=150+33.5+y

Y=45.5

Class interval fi xi fixi

10-20 12 14.5 174
20-30 30 24.5 735
30-40 33.5 34.5 1155.75
40-50 65 44.5 2892.5
50-60 45.5 54.5 2479.5
60-70 25 64.5 1612.5
70-80 18 74.5 1341
N=229 ∑fixi=10390.5
Mean µ = 10390.5/229=45.37
 ∆1 31.5
Mode = L + ∆1+∆2 ∗ 10 = 40+31.5+19.5 *10 = 40 + 6.18= 46.18

(ii) The small differences between the measures (Mean-Median =

-0.63, Median-Mode = -0.18) suggest mild negative skewness.
The distribution is mostly symmetric but slightly elongated on the
left.

(b)

Bursting fi(A) xi fixi(A) fi(B) fixi(B)

pressure
5-10 2 7 14 9 63
10-15 9 12 108 11 132
15-20 29 17 493 18 306
20-25 54 22 1188 32 704
25-30 11 27 297 27 729
30-35 5 32 160 13 416
N=110 2260 N = 110 2350
Mean (A)=2260/110=20.5454
Mean(B) = 2350/110=21.3636

So, B set has the highest average bursting pressure

(ii)

xi fi(A) fi(xi-µ)2 fi(B) fi(xi-µ)2

7 2 367.95 9 1856.81
12 9 657.2 11 964.45
17 29 364.53 18 342.74
22 54 114.26 32 12.96
27 11 458.28 27 857.76
32 5 656.04 13 1470.73
2618.26 5505.45
2618.26
𝜎𝐴 = √ =4.878 C.V(A) = (4.878/20.5454) *100 = 23.74
110

5505.45
𝜎𝐴 = √ =7.075 C.V(B) = (7.075/21.3636) *100 = 33.12
110

A has more uniform pressure

(iii) Manufacturer A’s bags would be preferred if uniformity is important (e.g., for packaging fragile goods). If
bursting strength is the priority, Manufacturer B’s bags are better. Since variance is significantly different, A’s bags
seem like the safer choice for reliability.

Q-3
(a) Gupta – page 184 ill.12

(b)Gupta-page 146 -ill15[similar]

(c)(i)

stem leaf
0 8,8,9
1 8
2 3,4,5,5,6,6,9
3 0,5,5,6,9,9
4 0,0,0,1,1,2,2,3,4,5,5,5,8
(ii)Q1=30/4=7.5TH term = (25+25)/2 =25

25% of the banks have a dividend payout ratio of 25.

𝟏𝟓𝒕𝒉+𝟏𝟔𝒕𝒉 𝒕𝒆𝒓𝒎
Q2 = = (36+39) /2 = 37.5
𝟐

50% of the banks have a dividend payout ratio of 37.5.

Q3= (30/4) *3= 22.5th term = (41+42)/2 = 41.5

75% of the banks have a dividend payout ratio of 41.5.

(iii) minimum value 8, maximum value 48

Q1=25 Q2=37.5 Q3=41.5

8----------------------------------------------------------48

The left whisker (8 to 25) is longer than the right whisker (41.5 to 48), and the median (37.5) is closer to Q3
(41.5). This indicates left-skewed (negative skew) data. Most banks cluster in the 25–42 range, but a few have
very low payouts (e.g., 8, 9).

Q-4

(a) Regression: Regression is a statistical method used to examine the relationship between one dependent
variable and one or more independent variables. It helps in predicting the value of the dependent variable based on
changes in the independent variable(s).

Example: Suppose a company wants to analyze how advertising expenditure (independent variable) affects sales
revenue (dependent variable). Using linear regression, we can estimate the equation:
Sales = a + b * Advertising Expenditure

• a is the intercept (value of sales when expenditure is zero).

• b is the slope (how much sales increase for a unit increase in expenditure).

2. Correlation: Correlation measures the strength and direction of the linear relationship between two variables. It
is represented by the correlation coefficient (r), which ranges from -1 to +1:

• r=+1→ Perfect positive correlation (as one variable increases, the other increases).
• r=-1 → Perfect negative correlation (as one increases, the other decreases).
• r=0 → No correlation (no linear relationship).

Example: If we analyze the relationship between hours studied and exam scores, and find, it means that studying
more hours is strongly associated with higher exam scores.
Properties of Regression with Explanation

1. The coefficient of correlation is the geometric mean of the two regression coefficients. Symbolically:

r = √𝑏𝑥𝑦 𝑏𝑦𝑥

2. If one of the regression coefficients is greater than unity, the other must be less than unity, since the value of the
coefficient of correlation cannot exceed unity. For example, if b xy = 1.2 and byx = 1.4, r would be √1.2 ∗ 1.4 =1.29
which is not possible.

(3) Both the regression coefficients will have the same sign, i.e., they will be either positive or negative. In other
words, it is not possible that one of the regression coefficients is having minus sign and the other plus sign.

(4) The coefficient of correlation will have the same sign as that of regression coefficients. i.e., if regression
coefficients have a negative sign, r will also have negative sign and if the regression coefficients have a positive sign,
r would also be positive. For example: if bxy=-0.2 and byx = - 0.8, r = - √0.2 ∗ 0.8 = −0.4

5. The average value of the two regression coefficients would be greater than the value of coefficient of correlation.
In symbols (bxy+byx)/2>r. For example, if bxy=0.8 and byx=0.4, the average of the two values would be
(0.8+0.4)/2=0.6 and the value of r would be √0.8 ∗ 0.4 = 0.566which is less than 0.6.
(b) (i)
 23rd mid-2(A+C)

1. Regression and Multiple Regression in Business Decision-Making

1. Simple Regression: Simple regression models the relationship between a dependent variable (Y) and one
independent variable (X). It helps businesses predict outcomes and quantify the impact of a single factor.

Example: A coffee shop wants to understand how daily temperature (X) affects iced coffee sales (Y). By running a
simple regression, they find that sales increase by 10 units for every 1°C rise in temperature. This insight helps
them adjust inventory and staffing during hotter days.

2. Multiple Regression: Multiple regression incorporates multiple independent variables (X₁, X₂, ..., Xₙ) to predict a
dependent variable. It isolates the effect of each predictor, enabling more nuanced decision-making.

Example: An e-commerce company predicts monthly revenue (Y) using variable like ad spend (X₁), website traffic
(X₂), and discount rate (X₃). The model reveals that a 10% increase in ad spend boosts revenue by $5k, while a 5%
discount reduces revenue by $2k. This guides budget allocation and pricing strategies.

Assumptions for Regression Analysis

1. Linearity: The relationship between variables is linear (e.g., sales vs. ad spend).
2. Independence of Errors: Residuals (errors) are uncorrelated (no hidden patterns).

3. Homoscedasticity: Residuals have constant variance across predicted values.

4. Normality of Errors: Residuals are approximately normally distributed.

5. No Multicollinearity (for multiple regression): Independent variables are not highly correlated (e.g., avoid using
both "price" and "discount rate" if they are inversely related).

6. No Autocorrelation: Residuals are not correlated over time (critical for time-series data).
Assumptions for Correlation Analysis

1. Linearity: Relationship between variables is linear (Pearson’s r).

2. Continuous Data: Variables are interval/ratio (Pearson) or ordinal (Spearman’s rank).

3. Bivariate Normality (for Pearson): Both variables are normally distributed.

4. No Outliers: Extreme values can distort correlation coefficients.

5. Homoscedasticity: Data points are evenly spread around the regression line.
Regression and correlation provide actionable insights for businesses, such as optimizing ad budgets or pricing
strategies. However, violating assumptions (e.g., ignoring multicollinearity) can lead to flawed decisions. Always
validate assumptions before interpreting results.

2. R² and Adjusted R² are two key metrics used to evaluate the performance of a multiple regression model. Here's
a simple explanation of both:
i. R² (Coefficient of Determination): R² measures the proportion of the variance in the dependent variable (Y) that
is explained by the independent variables (X’s) in the regression model.
𝑆𝑆𝑇
Formula: R2 = 1 - 𝑆𝑆𝑅 :SST = Total Sum of Squares (total variation in Y): SSR= Sum of Squared residuals

Interpretation: R² ranges from 0 to 1. If R² = 0.85, it means 85% of the variability in the dependent variable is
explained by the model. A higher R² means a better fit of the model.
Limitation: In multiple regression, R² always increases when you add more independent variables—even if they are
not meaningful.
ii. Adjusted R²: Corrected for Number of Predictors

Adjusted R² modifies R² to account for the number of independent variables in the model. It penalizes the addition
of variables that don’t improve the model much.
(1−𝑅 2)(𝑛−1)
Formula: R2 =1- ( )
𝑛−𝑘−1

k = number of independent variables

Interpretation: Adjusted R² increases only if the new variable improves the model more than by chance. It can
decrease if unnecessary variables are added. It's more reliable than R² when comparing models with different
numbers of predictors.
Example: Let’s say you build two regression models to predict sales:

Model 1: Uses Advertising and Price

R² = 0.88
Adjusted R² = 0.86

Model 2: Adds Weather and Competitor Activity

R² = 0.90
Adjusted R² = 0.85

Even though R² increased in Model 2, Adjusted R² dropped. This suggests that the new variables didn’t add real
value, and Model 1 might be a better choice.

4. (i) Here’s the scatter plot of the data. You can clearly see a
downward trend — as the number of problems increases, the
number of tickets purchased generally decreases, which visually
confirms the negative correlation.

(ii)

X Y X-𝑿̅ ̅
Y-𝒀 x2 y2 xy
x y
3 5 0.125 -1.25
7 0 4.125 -6.25
4 5 1.125 -1.25
2 4 -0.875 -2.25
0 13 -2.875 6.75
4 3 1.125 -3.25
1 9 -1.875 2.75
2 11 -0.875 4.75
TOTAL =23 50
̅
𝑿 = 23/8 = 2.875
̅= 50/8 = 6.25
𝒀