MANCOSA

Bachelor of Commerce in Information and Technology Management

ANALYTICAL TECHNIQUES
UNIT 1: INTRODUCTION TO STATISTICS
Study Notes | Prepared from Dr H. Matsongoni's Lecture Slides & Module Guide

Unit Learning Outcome

• Understand the scope of analytical techniques and the role of statistics in business
decision-making.

1. What is Statistics?
A statistic is an algebraic expression that combines scores into a single number. But modern-
day statistics goes beyond just numbers — it deals primarily with statistical inference, meaning
the process of drawing conclusions about a population based on sample data.

Statistics has been described as:

• Turning data into information
• Data-based decision making
• The technology of the Scientific Method

The scientific approach to decision-making follows this flow:

Hypothesis → Data → Conclusion
Statistics consists of principles and methods for:
1. Designing studies
2. Collecting data
3. Presenting and analysing data
4. Interpreting the results

📌 Key Concept
Without context, data is meaningless. For example, knowing that 4,500 Toyota Corollas were sold in
August means nothing unless you know how many were sold in July, or how competitors performed.
Data becomes information only when it has a basis of comparison, a known source, and a clear
purpose.

2. Why Study Statistics?

There are three main reasons every business professional needs statistical literacy:
 5. To evaluate published numerical facts — knowing when to trust them, question them,
or reject them based on how they were collected.
6. Professional necessity — most business roles require interpreting sample results or
using statistical methods to make informed workplace decisions.
7. Legal relevance — probability and statistical inference are increasingly used in court
proceedings (e.g. regression analysis as evidence).

📌 Real-World Relevance
Fields such as economics, business, psychology, law, computer science, education, and biology all
require at least one course in statistics.
Any manager who lacks numerate proficiency will be at a disadvantage in today's business
environment.

3. What Do Statisticians Do?

• Design surveys and experiments to minimise costs while obtaining sufficient
information.
• Seek the best method for analysing data and making inferences for a given sampling
situation.
• Summarise and analyse data.
• Provide a measure of the goodness of an inference (error of prediction) — i.e., they
quantify the probability of reaching a correct conclusion.

4. Key Definitions

Term Definition
Population The complete set of all elements of interest in a study.
Sample A subset of the population selected to represent it.
Statistical Inference The process of drawing conclusions about a population based on sample
data.
Parameter A numerical characteristic of a population (e.g. the population mean).
Statistic A numerical characteristic calculated from a sample.
Census A study that includes every element in the population — no sampling.
Sampling Error The difference between a sample statistic and the true population
parameter; occurs when the sample is not fully representative.
Population Frame A list, map, directory, or other source used to represent the population for
sampling purposes.
5. Sampling vs. Census

5.1 Reasons for Sampling

• Saves money — studying the whole population is often too costly.
• Saves time — smaller samples can be analysed more quickly.
• Broader scope — with given resources, sampling allows a wider range of data to be
gathered.
• Protects the product — when the research process is destructive, sampling
preserves the remaining product.
• Only feasible option — when the population is inaccessible, sampling is the only way
to gather data.

5.2 Reasons for a Census

• To eliminate the risk that a random sample is not representative of the population.
• When decision-makers are uncomfortable relying solely on sample information.

6. Population Frame
A population frame is a list, map, or directory used to represent the population from which the
sample is drawn. Two problems can arise:

Over-registration
The frame contains all target population members plus additional elements that don't belong.
Example: Using a chamber of commerce directory as the frame for a target population of
women-owned businesses — the directory includes all businesses, not only women-owned
ones.

Under-registration
The frame does not contain all members of the target population.
Example: Using a chamber of commerce directory as the frame for all businesses in a city —
not all businesses may be registered members.

7. Sampling Techniques

7.1 Random (Probability) Sampling

In random sampling, every unit of the population has an equal probability of being selected. A
chance mechanism is used, which eliminates selection bias.

Technique Description
 Simple Random Every member of the population has an equal chance of selection. Selections
are independent of each other.
Stratified Random The population is divided into subgroups (strata) and random samples are
drawn from each stratum.
Systematic Random Every nth member of the population is selected after a random starting point.
Cluster (Area) The population is divided into clusters (often geographic), and whole clusters
are randomly selected for study.

7.2 Non-Random (Non-Probability) Sampling

In non-random sampling, not every member has an equal chance of selection. This introduces
selection bias and makes inferential statistical analysis inappropriate.

Technique Description
Convenience Elements are chosen for the convenience of the researcher (e.g. surveying
people in a nearby mall).
Judgment Elements are selected based on the researcher's personal judgment about
who is most representative.
Quota Elements are selected until a predetermined quota for certain categories is
met (e.g. 50 men, 50 women).
Snowball Survey subjects are recruited through referrals from existing participants —
common in hard-to-reach populations.

8. Sampling Errors and Non-Sampling Errors

8.1 Sampling Error

Occurs when the sample is not perfectly representative of the population. This is expected in
any sampling exercise — the key is to minimise it through proper random sampling techniques.

8.2 Non-Sampling Errors

These are errors that arise from problems other than the sampling process itself:
• Missing data, recording errors, data entry errors, and analysis errors.
• Poorly conceived study concepts, unclear definitions, and defective questionnaires.
• Response errors — when respondents do not know the answer, are unwilling to
disclose it, or exaggerate their responses.

📌 Important Rule
Data from non-random samples is NOT appropriate for analysis using inferential statistical methods.
Only probability (random) samples produce results that can be statistically generalised to the
population.
9. Data Collection Methods

9.1 Types of Data

• Primary data — collected directly from elements of the population or a selected
sample (e.g. salary surveys, market research surveys).
• Secondary data — data collected from records that hold primary data (e.g. Statistics
South Africa collects primary data that is then used by others as a secondary
source).

9.2 Ways of Collecting Primary Data

Observation and Experimentation
Data is collected by measuring, counting, or observing subjects. Raw data is generated from
trials conducted under controlled conditions (e.g. clinical trials for new medications). This
method produces consistent results when experiments are replicated under the same
conditions.

Other Methods
• Personal (personnel) interviews — face-to-face discussions with respondents.
• Postal or mail surveys — questionnaires sent to respondents via mail.
• Telephone interviews — structured questions asked over the phone.
• Email and internet surveys — digital questionnaires distributed online.
• Self-administered questionnaires — respondents complete forms independently.

10. Unit Summary

Topic Core Idea

What is Statistics? The science of collecting, analysing, and interpreting data to support
decisions.
Population vs. Sample Population = all elements; Sample = selected subset used to
represent the population.
Random Sampling Four types: Simple, Stratified, Systematic, and Cluster. All give every
element a known probability of selection.
Non-Random Sampling Four types: Convenience, Judgment, Quota, Snowball. Not suitable
for inferential statistical analysis.
Errors Sampling errors: sample not representative. Non-sampling errors:
data entry, poor design, or response issues.
 Data Types Primary = collected first-hand. Secondary = obtained from existing
records or published sources.

11. Self-Test Questions

Use these questions to test your understanding of Unit 1:

8. Define statistics and explain the difference between descriptive and inferential
statistics.
9. Distinguish between a population and a sample. Give one example of each in a
business context.
10. List and briefly describe the four types of random sampling techniques.
11. Why is data from a non-random sample not suitable for inferential statistical
analysis?
12. What is a population frame? Explain what over-registration and under-registration
mean.
13. Explain the difference between sampling error and non-sampling error. Give an
example of each.
14. Differentiate between primary data and secondary data. Give one real-world example
of each.
15. A retail company wants to survey 200 customers about their shopping experience.
Which sampling method would you recommend, and why?

Prescribed Reading
• Anderson, D., et al., (2020), Statistics For Business And Economics, 6th Edition,
Cengage Learning, South Africa.

Recommended Reading
• Keller, G., and Gaciu, N., (2020), Statistics For Management And Economics, 2nd
Edition, Cengage Learning, South Africa.

End of Unit 1 Study Notes