SESSION AUG/SEP 2023

PROGRAMME MASTER OF BUSINESS ADMINISTRATION (MBA)

SEMESTER I
COURSE CODE & NAME DMBA103-STATISTICS FOR MANAGEMENT

SET 1

1 Define statistics. Explain various functions of statistics. Also discuss the key limitations of
statistics. 2+4+4
Ans 1.
Statistics for Management: Definition, Functions, and Limitations
1. Definition of Statistics: Statistics can be understood in two main contexts:

a) Singular Sense: In this context, 'statistics' refers to the science and art of collecting, presenting,
analyzing, and interpreting data in order to make informed decisions. It provides methods and
tools for dealing with variability, uncertainty, and making inferences about populations from
samples.

b) Plural Sense: When used in the plural, 'statistics' refer to numerical facts or data points
collected systematically. For instance, the GDP growth rate of a country, the average age of
employees in an organization, etc., are statistics.

2. Functions of Statistics:

a) Simplification: Large volumes of raw data can be overwhelming. Statistics helps in

condensing that data into understandable and interpretable forms, making it easier for individuals
and organizations to comprehend and make use of it.
b) Establishing Correlations: Statistics helps in establishing a relationship between two or more
variables. For example, the relationship between advertising expenditure and sales can be
determined using statistical methods.

c) Forecasting: Through regression and time series analysis, statistics can predict future trends or
occurrences based on past data. Companies often use these forecasts for budgeting, planning, and
decision-making.

d) Testing Hypotheses: In research, statistics allow scientists and researchers to test hypotheses
and make inferences about populations. Statistical tests can help determine if an observed effect
is genuine or if it could have occurred by random chance.

e) Formulating Policies: Governments and organizations utilize statistical analyses to formulate

policies. For example, census data (which is a collection of statistics about a population) can help
a government decide where to allocate resources.

f) Comparative Study: Statistics can help in comparing data sets to determine their relative
position. For instance, companies can compare their performance against industry averages or
benchmarks.

3. Limitations of Statistics:

a) Qualitative Aspects Ignored: Statistics mainly deals with quantitative data. Qualitative aspects,
like intelligence, attitude, and honesty, which don't have numerical values, are generally
overlooked.

b) Doesn't Represent Individual Items: Statistics gives an overview or average and doesn't
represent individual data points. For instance, if the average income in a city is $50,000, it
doesn’t mean every individual earns that amount.

c) Can Be Misleading: If not properly used or interpreted, statistics can be misleading. A

common adage is, "There are three types of lies: lies, damned lies, and statistics." This speaks to
the potential for data to be manipulated or misrepresented to support a specific narrative.
d) Only Suitable for Homogeneous Data: Statistics is applicable to aggregates of similar data
points. It is not suitable for heterogeneous or dissimilar data.

e) Results Can Be Affected by Extreme Values: Outliers or extreme values can skew statistical
results. For example, if in a sample of ten people, nine have an income of $30,000 and one has
an income of $1,000,000, the average income will be heavily influenced by the outlier.

f) Requires Expertise: Proper collection, analysis, and interpretation of data require expertise.
Without proper understanding, there's a risk of drawing incorrect conclusions.

In conclusion, while statistics provide valuable tools for understanding and interpreting data, it's
imperative to approach them with caution and understanding. One should not blindly accept
statistical results but rather critically evaluate and understand the methods and assumptions
behind them.

2. Define Measurement Scales. Discuss Qualitative and Quantitative data in detail with
examples. 2+8

Ans 2.

Measurement Scales

Measurement scales refer to various types of categories into which data can be classified. These
scales are foundational in the field of statistics and research, as they dictate the type of statistical
technique that can be used to analyze the data. There are primarily four types of measurement
scales: nominal, ordinal, interval, and ratio.

1. Qualitative Data

Qualitative data, often termed as categorical data, deals with characteristics and descriptors that
can be observed but not measured. In other words, it's about understanding an attribute rather
than counting it or measuring it with exactitude. Qualitative data can be divided based on the
measurement scale into:
Nominal Scale: It classifies data into distinct categories in which no order or hierarchy is
implied. The data in this scale is purely labeled and has no specific order. For instance, the color
of a car (red, blue, green) or gender (male, female, other) can be categorized under the nominal
scale.

Ordinal Scale: This scale classifies data into categories that can be ranked or ordered. Though
there is a sequence, the exact difference between the ranks is unknown. An example would be
the rating of a movie on a scale of 1-5. While we know that a rating of 4 is better than 3, we don't
know by how much.

Examples of Qualitative Data:

 A person's religion or ethnicity.

 The type of car someone drives.
 The preference of ice cream flavors among a group (like vanilla, chocolate, strawberry).

2. Quantitative Data

Quantitative data refers to data that can be measured and assigned a specific numerical value. It's
about quantities and is expressed in terms of numbers. Based on the measurement scale,
quantitative data can be divided into:

Interval Scale: This type of scale offers consistent, logical intervals between numbers. The data
has order, and there is a definite meaningful difference between data points. However, it lacks an
absolute zero. For example, temperature on the Celsius scale is an interval scale. While we can
say 30°C is hotter than 20°C, we cannot say it's twice as hot.

Ratio Scale: This is the most powerful scale among all. It has all the properties of the interval
scale with an added benefit of an absolute zero point, which means it has a clear definition of
zero. Examples include age, income, or weight. For instance, if a person weighs 60kg, and
another weighs 120kg, we can definitively say that the second person is twice as heavy as the
first.

Examples of Quantitative Data:

  The height of students in a class.
 The number of books sold by a store in a month.
 The temperature of a city throughout the week.

To summarize, qualitative data is more about understanding characteristics, feelings, or

perceptions, and it uses nominal and ordinal scales. Quantitative data, on the other hand, is
numeric and measurable and utilizes interval and ratio scales. Both types of data are essential in
research and statistics, with each offering unique insights and analysis possibilities. Recognizing
the difference between them and understanding their nature is crucial for proper data analysis
and interpretation.

3. Discuss the basic laws of Sampling theory. Define following Sampling techniques with
help of examples:
Stratified Sampling
Cluster Sampling
Multi-stage Sampling 4+6

Ans 3.

Laws of Sampling Theory

Sampling is a fundamental concept in statistics, which deals with the selection of a subset of
individuals from within a statistical population to estimate characteristics of the whole
population. There are basic laws of sampling theory that guide how sampling should be done to
ensure that the sample is representative of the entire population.

Law of Statistical Regularity: Even if the population is heterogeneous, if the sample is random
and sufficiently large, it will still be representative of the population. This law suggests that
random samples exhibit the same properties and characteristics as the population from which
they are drawn.
Law of Inertia of Large Numbers: This law states that large samples tend to be stable and less
affected by sampling fluctuations than small samples. In other words, as the sample size
increases, the variability of sample statistics decreases, approaching the true population
parameter.

Sampling Techniques

1. Stratified Sampling: Definition: Stratified sampling is a method in which the population is

divided into mutually exclusive and exhaustive subgroups, called strata, and then a random
sample is taken from each stratum. The idea is to represent all sections or strata of the population
in the sample.

Example: Imagine a university with a student population made up of 60% undergraduates, 30%
postgraduates, and 10% doctoral students. If a researcher wants to sample 100 students, he would
select 60 undergraduates, 30 postgraduates, and 10 doctoral students to ensure that each group is
proportionately represented.

2. Cluster Sampling: Definition: In cluster sampling, the entire population is divided into
clusters or groups, and then a random sample of these clusters is selected. All or a random
sample of the elements within the chosen clusters are then surveyed.

Example: Let's say a researcher wants to understand the teaching quality in public high schools
of a state. Instead of surveying all schools, the state is divided into districts (clusters). From
these, a random sample of districts is chosen, and every high school within the selected districts
is surveyed.

3. Multi-stage Sampling: Definition: Multi-stage sampling is a more complex form of cluster

sampling in which two or more levels of units are embedded one in the other. The process
involves dividing the population into clusters and then randomly selecting clusters for further
sampling.

Example: Consider a nation-wide survey on the reading habits of high school students. In the
first stage, several states within the country are randomly selected. In the second stage, within
those selected states, certain cities or districts are randomly chosen. In the third stage, a few
schools within those cities or districts are chosen. Finally, within those schools, students are
randomly selected to be surveyed about their reading habits.

In conclusion, sampling is a crucial methodology in statistics and research that allows for
effective data collection and insights without examining an entire population. The choice of
sampling technique largely depends on the research objectives and the nature of the population.

SET 2

1 Define Business Forecasting. Explain various methods of Business Forecasting. 10

Ans 1.

Business Forecasting Definition and Methods

Business Forecasting refers to the process of estimating or predicting future trends, demands, and
developments in a particular business or industry based on historical data, market analysis, and
relevant business insights. It aids businesses in making informed decisions, planning for the
future, managing resources effectively, and anticipating market changes.

Methods of Business Forecasting:

1. Judgmental or Qualitative Forecasting:

Expert Opinion: This method involves seeking the opinion of experts in the respective field to
predict future outcomes. Their insights, based on experience and expertise, can guide business
decisions.

Market Research: By conducting surveys or interviews with potential customers or clients,

businesses can gauge future demand for their product or service.
Delphi Technique: A panel of experts is asked to provide their forecasts independently. Their
responses are then aggregated, and the process is repeated until a consensus is reached.

2. Time Series Analysis:

Moving Averages: This method involves averaging data over a specific number of periods to
smooth out short-term fluctuations and highlight long-term trends.

Exponential Smoothing: It assigns decreasing weights to older data, giving more emphasis to
recent observations.

Decomposition: This involves breaking down time series data into its individual components
like seasonality, trend, and cyclic patterns.

3. Causal or Regression Models:

Linear Regression: This method establishes a relationship between a dependent variable and
one or more independent variables. The strength and nature of the relationship can then be used
to forecast future values.

Multiple Regression: Extends the simple linear regression model by considering multiple
independent variables.

4. Econometric Models:

This combines economic theory with statistical methods. Econometric models consider multiple
equations and relationships between different variables to forecast outcomes.

5. Simulation Models:

These are computer-based models that recreate real-world scenarios. By altering specific
variables or conditions within the simulation, the potential impact on the business can be
observed and analyzed.

6. Extrapolation:
This involves extending historical data into the future based on identified patterns or trends.
While it's simple and easy to use, its accuracy may diminish over longer forecasting horizons.

7. Technology and Artificial Intelligence:

Neural Networks: Inspired by the human brain's structure, neural networks can detect and adapt
to non-linear patterns in large datasets, making them suitable for complex forecasting tasks.

Machine Learning: Advanced algorithms can analyze vast amounts of data, learn from it, and
make predictions based on identified patterns.

In conclusion, the choice of a forecasting method largely depends on the nature of the business,
the available data, the forecasting horizon, and the specific objectives of the forecast. No method
is universally best for all situations; therefore, businesses often combine multiple techniques to
increase accuracy and reliability. Proper business forecasting is essential for strategic planning,
budgeting, and anticipating future challenges and opportunities.

2. What is Index number? Discuss the utility of Index numbers 5+5

Ans 2.

Index Number

An index number is a statistical tool that measures the relative change in a particular variable or
a group of related variables over time or different geographical locations. Typically, it quantifies
the changes in terms of percentages. Index numbers are often used to track economic indicators,
such as inflation, production, and the cost of living. The base year or location is given a value of
100, and subsequent values are calculated to represent the relative change from this base value.

Utility of Index Numbers:

1. Economic Analysis: Index numbers are extensively used to study economic trends and
formulate policies. For instance, the Consumer Price Index (CPI) provides insights into
the inflationary trends in an economy and can influence monetary policies.
2. Cost of Living Analysis: The Cost of Living Index (COLI) is essential to understand the
changing living standards of people. It helps in wage negotiations and salary adjustments
based on the increase or decrease in the cost of living.

3. Comparison Over Time and Space: Index numbers allow for comparisons across
different time periods and geographical areas. For instance, a company might want to
compare its production levels over different years or compare its sales in different
regions.

4. Business Forecasting: Organizations can use index numbers to predict future trends. For
instance, if the sales index of a product has been steadily rising, it may indicate increased
demand in the future, allowing companies to adjust their production or marketing
strategies accordingly.

5. Deflating: Index numbers help in converting current values into constant values. For
example, when comparing the GDP of two different years, it's essential to account for
inflation. Using GDP deflators, which are a type of index number, we can get a real GDP
figure that's free from the influence of inflation.

6. Policy Formulation: Governments use index numbers like the Wholesale Price Index
(WPI) and CPI to formulate various policies. For instance, a rising CPI may signal the
need for tighter monetary policies to control inflation.

7. Performance Measurement: Index numbers can be used to measure the performance of

various sectors in the economy. A rising industrial production index might indicate a
booming industrial sector, whereas a declining index can signal a recession.

8. Adjustment in Financial Instruments: Index numbers are also used to adjust values in
various financial instruments like treasury bonds. For instance, inflation-indexed bonds
use inflation index numbers to adjust the principal and interest payments.

9. International Comparisons: Index numbers allow for the comparison of economic

indicators across different countries. This can be particularly useful for international
businesses, policymakers, and researchers.
 10. Basis for Constructing Other Statistical Measures: Often, index numbers serve as a
foundational tool in the creation of other statistical measures. For example, the Human
Development Index (HDI) uses education, income, and life expectancy indices to derive a
comprehensive measure of human development.

In essence, index numbers are powerful statistical tools that provide a snapshot of relative
changes in various variables. They play a crucial role in the world of economics, business, policy
formulation, and more, making them indispensable for analysts, policymakers, businesses, and
researchers. The ability to track changes and derive meaningful insights from data over time and
across locations adds significant value to decision-making processes in various fields.

3. Discuss various types of Estimators. Also explain the criteria of a good estimator. 5+5
Ans 3.
Various Types of Estimators

Estimators are statistical measures that give an approximation of a particular population

parameter based on the information available from a sample. There are several types of
estimators:

1. Point Estimator: This gives a single value as an estimate of the population parameter.
For instance, the sample mean (X̄ ) is a point estimator of the population mean (μ).

2. Interval Estimator: Instead of giving a single value, an interval estimator provides a

range of values as the estimate for the population parameter. A confidence interval is an
example where we can say we are "95% confident that the population parameter lies
between a and b."

3. Unbiased Estimator: An estimator is unbiased if the expected value of the estimator

equals the parameter it is estimating. In other words, on average, the estimator hits the
target. For example, the sample mean is an unbiased estimator of the population mean.
 4. Consistent Estimator: Consistency implies that as the sample size (n) increases, the
estimator gets closer and closer to the parameter it is estimating.

5. Minimum Variance Unbiased Estimator (MVUE): Among the class of unbiased

estimators, an MVUE has the lowest variance. This ensures that there's minimal
dispersion of the estimator around the parameter being estimated.

6. Maximum Likelihood Estimator (MLE): The MLE method finds the parameter value
that maximizes the likelihood function, which represents the probability of observing the
sample data given the parameter.

7. Bayesian Estimator: This type of estimator incorporates prior knowledge or beliefs (in
the form of a prior distribution) about the parameter. The data from the sample and this
prior are combined to provide a posterior distribution, which is used for estimation.

Criteria for a Good Estimator

1. Unbiasedness: A good estimator should be unbiased, meaning its expected value should
equal the true parameter value. This ensures that there's no systematic error.

2. Consistency: As the sample size grows, a good estimator should converge to the true
parameter value.

3. Efficiency: Among various estimators, the one with the lowest variance is preferred. An
efficient estimator provides estimates closer to the true value with less variability.

4. Sufficiency: An estimator is sufficient if it captures all the information in the data about
the parameter. No other estimator based on the same data should provide additional
information about the parameter.

5. Robustness: A robust estimator performs well across various situations, especially when
there are deviations from assumptions, such as non-normality.
 6. Minimum Mean Square Error (MMSE): This criterion combines both bias and
variance. An estimator with a smaller mean square error is preferred since it's closer on
average to the true parameter value.

7. Ease of Computation: In practical situations, the ease with which an estimator can be
computed is crucial. Complex estimators that are hard to compute might be less useful,
even if they are theoretically optimal.

In conclusion, the choice of estimator depends on the situation and the specific properties one
desires. However, unbiasedness, consistency, and efficiency are often the most sought-after
properties. Practical considerations, like ease of computation and robustness, also play a
significant role in the selection of estimators.