Week 1 - Introduction, Understanding data using SPSS

Question 1

What are the two categories of data based on structure?

Answer 1

Structured data, unstructured data

Question 2

What is structured data?

Answer 2

organised, often numeric, easy to store in databases (e.g., transaction data).

Question 3

What is unstructured data?

Answer 3

not organised, harder to analyse (e.g., contracts, texts, voices, video).

Question 4

What are the four categories of data based on characteristics?

Answer 4

Descriptive data
Behaviour data
Interaction data
Attitudinal data

Question 5

What does descriptive data include?

Answer 5

attributes, characteristics, geo/demographics.

Question 6

What is does behaviour data include?

Answer 6

orders, transactions, payments, credit history.

Question 7

What does interaction data include?

Answer 7

emails, chat transcripts, click-streams.

Question 8

What does attitudinal data include?

Answer 8

opinions, preferences, needs, desires.

Question 9

What is Business Analytics (BA)?

Answer 9

The skills, technologies, and practices for continuous exploration of past business performance to gain insight and support fact-based decision making.

Question 10

What are the 4 main components of Business Analytics?

Answer 10

(Big) Data

Statistical & Quantitative Analysis

Explanatory or Predictive Models

Decision Making & Actions

Question 11

What types of questions does Business Analytics answer?

Answer 11

What happened? (Describe)

Why is this happening?

What if these trends continue?

What will happen next? (Predict)

What is the best that can happen? (Optimize)

Question 12

What are the three categories of analytics?

Answer 12

Descriptive Analytics
Predictive Analytics
Prescriptive Analytics

Question 13

What are descriptive analytics?

Answer 13

Using and visualising data (e.g., reports, scorecards, clustering) to understand performance.

Question 14

What are predictive analytics?

Answer 14

Using statistical/machine learning techniques (time-series, regression) to find relationships and predict outcomes.

Question 15

What are prescriptive analytics?

Answer 15

Using optimisation and simulation techniques to improve business performance, given objectives and constraints.

Question 16

Example of Descriptive Analytics.

Answer 16

Reports showing past sales trends, dashboards, and performance scorecards.

Question 17

Example of Predictive Analytics.

Answer 17

Using regression models to forecast future sales or customer churn.

Question 18

Example of Prescriptive Analytics.

Answer 18

Supply chain optimisation to minimise costs under constraints or scheduling staff to maximise efficiency.

Question 19

What are the two main types of structured data?

Answer 19

Categorical (qualitative)

Numerical (quantitative)

Question 20

What are the two subtypes of categorical (qualitative) data?

Answer 20

Nominal: Categories without order (e.g., gender, colors, countries).

Ordinal: Categories with implied order (e.g., rankings, education levels).

Question 21

What are the two subtypes of numerical (quantitative) data?

Answer 21

Interval: Numeric values where differences make sense, but no absolute zero (e.g., temperature in °C).

Ratio: Numeric values with a true zero, allowing all arithmetic operations, including ratios (e.g., weight, salary).

Question 22

Formula for the mean (x~?

Answer 22

= (∑x_i)/n

Question 23

Formula for standard deviation (s)?

Answer 23

s = √ (∑(x_i - x~)^2 /(n-1))

Question 24

What does skewness describe?

Answer 24

The shape of a distribution – whether it is symmetric, left-skewed (negative), or right-skewed (positive).

Question 25

What does right skewness (positive skew) mean?

Answer 25

A distribution with a long right tail (e.g., compensation levels, income distributions).

Question 26

Example of data with strong right-skewness?

Answer 26

Compensation levels – very varied, max = 16 × mean.

Question 27

How do compensation levels vary across industries?

Answer 27

Highest mean in Telecoms. Maximum overall in IT.

Question 28

What factors influence compensation besides industry?

Answer 28

Compensation is related to Net Sales, but other factors also play a role.

Question 29

Are there industry-specific compensation patterns?

Answer 29

Yes – fairly strong industry-specific relationships exist between Compensation and Net Sales.

Question 30

Pros and cons of a Dotplot

Answer 30

Pros: Simple, easy to construct. Shows individual data values. Good for small datasets. Cons: Not suitable for large datasets (too cluttered). Doesn’t show distribution shape well for big data.

Question 31

Pros and cons of a Histogram

Answer 31

Pros: Shows distribution (shape, spread, skewness). Useful for large datasets. Good for continuous variables like salary. Cons: Loses individual data details. Sensitive to bin width choice.

Question 32

Pros and cons of a Boxplot

Answer 32

Pros: Summarises data with median, quartiles, and outliers. Easy to compare multiple groups (e.g., industries). Good for skewed data like compensation. Cons: Doesn’t show the full distribution shape. Can be hard to interpret without practice.

Question 33

Pros and cons of a Pie Chart

Answer 33

Pros: Good for showing proportions/percentages. Easy for simple categorical data (e.g., % employees per industry). Cons: Hard to compare slices accurately. Not suitable for many categories. Misleading if values are close.

Question 34

Pros and cons of a Bar Chart

Answer 34

Pros: Good for categorical comparisons (e.g., mean pay by industry). Clear visual differences between groups. Handles many categories better than pie charts. Cons: Doesn’t show distribution within each category. Can oversimplify variation.

Question 35

What is the use of Boxplots by Industry in compensation analysis.

Answer 35

Compare pay distributions across industries. Show median pay, variation, and outliers. Highlights industries with high inequality or extreme salaries.

Question 36

What is the use of Mean Compensation by Industry (Bar Chart).

Answer 36

Shows average pay per industry. Easy to identify highest/lowest paying industries. Simpler than boxplots but hides variation within industries.

Question 37

What is the use of Scatterplot (Compensation vs. Sales).

Answer 37

Pros: Shows relationship between compensation and net sales. Identifies patterns or correlations (e.g., higher sales linked to higher pay). Can reveal industry-specific trends. Cons: Harder to interpret with overlapping points. Doesn’t show distribution within industries.