Business Decision Making: Analysis of Random Samples for Supermarket Chain

This presentation analyzes random samples of 25 customers each using simple random sampling and systematic random sampling to provide useful information for decision making for a supermarket chain. It covers descriptive statistics such as location, number of children, income, and debt, and highlights the potential reliability issues due to the differences between the two samples. It also discusses potential solutions such as determining the minimum sample size and using stratified sampling. Additionally, it covers inferential statistics and hypothesis testing to estimate population parameters and concludes that the results need to be more reliable for decision making.

BUSINESS DECISION
MAKING

 Population data comprising of 1000
customers
 Selection of random samples
◦ Simple Random Sampling (25 sample size)
◦ Systematic Random Sampling (25 sample size)
 Data focuses on key demographic features
including age, gender, debt, salary,
children, location, amount spent
Introduction

 The key objective is to provide useful
information to the supermarket chain in order
to provide assistance to them with regards to
decision making
 Key decision to made is whether to open a
store or not
 Key Tools
◦ Descriptive Statistics
◦ Inferential Statistics
Objective & Tools

 Aims at summarising the sample
characteristics without commenting on the
underlying population
 The key focus in this case is to summarise
the various demographic features for the
two samples and derive useful information
from the same
Descriptive Statistics

 Location of customers
◦ Both the samples indicated that majority of the
customers belonged to Zone C which might
indicate proximity to Zone C would lead to a
higher revenue generation
 Number of Children
◦ Higher count of children would typically lead to a
higher annual amount spent and hence locating
the store in an area with higher children would be
conducive for the supermarket
Useful Information – Descriptive
Statistics

 Income of customer
◦ It would be reasonable to expect that higher
income would culminate into higher average
spending and therefore the supermarket would
prefer being located in a zone with higher
average income
 Amount of debt and mortgage
◦ Consumers having higher debt and mortgage may
have lower disposable incomes and hence tends
to impact the yearly spending on retail
Useful Information
(Contd..)

 Descriptive statistics of the two samples are
not the same and in certain key attributes,
the difference between the samples is quite
staggering
 Hence, reliability issues tend to arise as the
underlying samples might not be
representative of the underlying population
 The conclusions drawn are incorrect
Reliability – The Key Issue

 Minimum sample size to be determined
based on the characteristics of the
population
 Two samples drawn of minimum sample size
which would provide more reliable results
 Stratified sampling may be useful for
ensuring even representation of key
attributes akin to popualtion
Potential Way Out

 The sample has also been used to estimate
the population parameters.
 Hypothesis test was deployed
◦ For both the samples, it was found that average
spending pattern by the consumers do not
support the claims made by the company
 Potentially indicative of the population not
being a suitable target segment
Inferential Statistics

 Random samples are derived using Excel
 Descriptive statistics provide vital information
about key attributes
 The results obtained by the two samples are not
same since neither is potentially representative
of population
 Results need to be more reliable for decision
making
Conclusion