Chapter 1: Introduction to

Data Analysis
Data Analysis Course
Agenda
1. Data Overview
• What is data?
• Types of Data: Qualitative and Quantitative
• Measurement Scales
• Data Sources: Primary and secondary
• Post-Data Collection Procedures
2. Descriptive Statistics
• Types of data: Cross-sectional, Time series and Panel
• Tables for appropriate data representation by type
• Descriptive statistics for each type of data
3. Applications
4. Review on statistics in business
What is Data?
• Data:set of values such as numbers, text, or symbols collected for
storage, summary, analysis and decision-making.
• Dataset: collection of data collected in a particular study.
• Elements: entities on which data are collected.
• Variable: characteristic of interest for the elements.
• Observation: the set of measurements obtained for a particular
element.
• A dataset with n elements contains n observations.
• The total number of data values in a complete dataset is the number
of elements multiplied by the number of variables.
What is Data?
Element Observation
Variables
names

Student ID Age Major GPA

S001 20 Business 3.5
S002 22 Computer Science 3.8
S003 19 Psychology 3.2
S004 21 Engineering 3.6
S005 20 Economics 3.4

Dataset
Types of Data
• Qualitative:
• Labels or names used to identify an attribute of each element
• Can also be called categorical data
• Use either the nominal or ordinal scale of measurement
• Can be either numerical or non-numerical
• Quantitative:
• Indicate how many or how much
• Is always numerical
• Ordinary arithmetic operations (mean, median, etc.) are meaningful for this
type of data
Quantitative Data Types
• Discrete: Countable values
• Example: number of students in this classroom, the set of all positive integers
• Continuous: Infinite values within a range, cannot be counted
• Example: temperature, time, weight, height
Example: Qualitative Data
Student ID Favorite Color Preferred Learning Style Country of Origin
001 Blue Visual Canada
002 Green Auditory Brazil
003 Red Kinesthetic India
004 Yellow Visual Australia
Example: Quantitative Data
Student ID Age Test Score (%) Hours Studied
001 20 88 10
002 22 75 7
003 19 92 12
004 21 81 9
Measurement Scales
• How variables are defined and categorized
• Describes the nature of data in the dataset
• Determines the type of analysis can be used for the data of interest
• There are 4 types:
• Nominal scale
• Ordinal scale
• Interval scale
• Ratio scale
Nominal Scale
• Classifies data into categories
• No intrinsic ranking or order within categories
• Can only be compared for equality (only A = B or A ≠ B)
• Example: Blood type (A, B, O, or AB), gender, nationality
Ordinal Scale
• Classifies data into categories
• There is ranking and order within categories
• The interval between ranking is not equal
• Example: Education level, rankings in a competition
Interval Scale
• Equal intervals between values
• No true zero point
• Addition and subtraction can be applied but not multiplication and
division
• Example: Temperature in Celsius or Fahrenheit, IQ scores
Ratio Scale
• Equal intervals between values
• True zero point (point where 0 means “none”)
• Supports addition, subtraction, multiplication, and division
• Ratio between values has meaning
• Example: Height, age, distance
 Data

Qualitative Quantitative

Non-
Numerical Numerical
numerical

Nominal Ordinal Nominal Ordinal Interval Ratio

Operation Nominal Scale Ordinal Scale Interval Scale Ratio Scale
Equality Comparison
Order Comparison
Addition /
Subtraction
Multiplication /
Division
Mode
Median
Mean
Other Statistical
Operations
Data Sources
• Data can be primary or secondary based on collection method.
• Primary data:
• Data collected directly from original sources through methods such as
surveys, interviews, experiments, or observations, for the specific purpose of
a current research project.
• Original and first-hand information
• Tailored to the researcher's needs
• Time-consuming and costly to gather
• High accuracy and relevance
Data Sources
• Secondary data:
• Data that has been previously gathered and made available by other sources,
such as government agencies, organizations, or researchers, and is reused for
a different analysis or purpose
• Quick and cost-effective to obtain
• May not perfectly match current research needs
• Data quality varies depending on the original source
• Example: Government data, academic journals, online databases
Data Sources
Aspect Primary Data Secondary Data
Data collected firsthand for a Data previously collected by others
Definition
specific research purpose for a different purpose
Original sources such as surveys, Existing sources such as books,
Source
interviews, observations reports, websites, and databases
Direct methods: surveys, Indirect methods: accessing
Collection Method
experiments, observations published data or reports
Time Required Time-consuming Time-saving
Expensive (requires resources and
Cost Usually low-cost or free
effort)
Specifically tailored to current May not fully match current
Relevance to Research
research needs research objectives
Depends on the credibility of the
Data Accuracy High (if properly collected)
original source
Full control over how data is No control over how data was
Control Over Data
collected originally collected
Field surveys, experiments, direct Government reports, academic
Examples
interviews articles, statistical databases
Post-Data Collection Procedures
• Data preparation
• Data cleaning: identify and correct errors, missing values, and inconsistencies
• Data coding: Convert qualitative data into numerical or symbolic codes
• Data validation: Confirm the consistency, completeness, and accuracy of data
• Data analysis and interpretation
• Data transformation: format data for analysis
• Data analysis: apply statistical or thematics methods to the data
• Data interpretation: extract findings and insights from the results of the analysis
• Documentation and storage
• Report writing: present your findings in a structured report/presentation
• Data storage and backup
Types of data
• Based on time dimension and number of units, data can be
categorized into the following:
• Cross-sectional data: Data collected at a single point in time from multiple
units.
• Time series data: Data collected over time from a single unit.
• Panel data: Data collected over time from multiple units.
Cross-sectional data
• Data collected at a single point in time (or over a very short period)
across multiple subjects.
• Captures a snapshot of a population or phenomenon.
• Allows for comparison between different subjects at one moment.
• Commonly used in surveys, market research, and census studies.
Cross-sectional data
• Data on 4 households with the number of motorbikes of each
households and the monthly electricity bill for the month of June
2025.
Monthly Electricity Bill
Household ID Number of Motorbikes
(VND)
H001 2 1,050,000
H002 1 870,000
H003 3 1,200,000
H004 2 950,000
Time Series Data
• Data collected on a single subject over multiple time periods.
• Focuses on how value changes over time.
• Each observation is tied to a timestamp.
• Can be used for trend, seasonality, and forecasting analysis.
Time Series Data
• Time series can be broken down into 4 components:
• Trend: the general direction or pattern of change of the data, whether it be
decreasing, increasing, or stable over a period of time. For example, inflation and
price are generally trending up (increasing)
• Seasonality: the repeating pattern that occur at predictable intervals, usually tied to
seasons or calendar. Example: tourism increase in holiday season, electricity use
increase in the summer.
• Cyclic component: long-term, non-seasonal changes in a time series that occur in
irregular but recurring cycles. Example: Economy cycle (Expansion → Peak →
Recession → Trough), real estate pricing cycle (price booms followed by downturns).
• Irregular remainder (white noise): the random, unpredictable fluctuations in a time
series that cannot be explained by the trend, seasonality, or cyclic patterns. Example:
natural disasters, unexpected economic shock created by new policies.
Time Series Data

Component Description Example

Long-term movement or direction Gradual increase in average
Trend (T)
in the data over time temperature over decades
Repeating short-term pattern at
Seasonality (S) regular intervals (e.g., monthly, Higher tourism in holiday season
yearly)
Long-term up-and-down
Business cycles: boom → recession
Cyclic Component (C) movements not of fixed period
→ recovery
(often economic cycles)
Unpredictable, short-term
Sudden drop in stock price due to
Irregular/Random (I) fluctuations due to random or
new policies
unusual events
Time Series Data
• Electricity bill of a household in 2024

Month Electricity Bill (VND)

Jan 2024 950,000
Feb 2024 870,000
Mar 2024 1,050,000
Apr 2024 1,200,000
May 2024 1,100,000
Panel Data
• Can also be called Longitudinal data.
• Data that observes multiple subjects over multiple time periods.
• Combines the elements of both cross-sectional and time series data.
• Enable analysis for changes within each subjects as well as differences
between subjects.
Panel Data
• Data on 3 households, the number of motorbikes in each household
and their monthly electricity bill in June and July 2025
Monthly Electricity Bill
Household ID Month Number of Motorbikes
(VND)
H001 Jan 2025 2 950,000
H001 Feb 2025 2 870,000
H002 Jan 2025 0 1,000,000
H002 Feb 2025 1 880,000
H003 Jan 2025 3 1,200,000
H003 Feb 2025 2 1,250,000
Aspect Cross-Sectional Data Time Series Data Panel Data (Longitudinal)

Data collected at a single Data collected from a single Data collected from
Definition point in time from multiple unit over multiple time multiple units over
units periods multiple time periods

Time Component No Yes Yes

Many units (e.g., people,

Observation Units One unit over time Many units over time
firms) at one time

Changes within and

Differences between Trends and patterns over
Focus between subjects over
subjects time
time

Market research, opinion Economic, behavioral, and

Use Cases Forecasting, trend analysis
polls, census policy studies

Multi-dimensional: units ×
Data Structure One-time snapshot Time-based sequence
time
 Tables and charts for appropriate data
representation by type
• Below is dataset of 10 students, with 5 variables: gender, satisfaction
level, IQ score, weekly study hours, and grade.
Satisfaction Level
Weekly Study
(Ordinal: 1=Very Grade (GPA, Ratio,
ID Name Gender (Nominal) IQ Score (Interval) Hours (Ratio,
Dissatisfied, Continuous)
Continuous)
5=Very Satisfied)
1 Alice Female 4 112 12.5 3.4
2 Ben Male 3 105 8.0 2.8
3 Gina Female 5 120 18.0 3.9
4 Dan Male 2 101 6.0 2.5
5 Cara Female 5 118 15.0 3.8
6 Ivy Female 2 102 7.0 2.7
7 Hugo Male 3 107 9.5 3.0
8 Eva Female 4 109 10.0 3.2
9 Jack Male 4 115 14.0 3.6
10 Finn Female 1 98 5.5 2.4
Tables and charts for appropriate data
representation by type
• For Qualitative (Categorical) data:
• Nominal data:
• Frequency table: presents the number of occurrences (counts) of each category within a
dataset, helps summarize large amounts of categorical data into a readable format.

Gender Frequency
Female 6
Male 4

Frequency Table
Tables and charts for appropriate data
representation by type
• For Qualitative (Categorical) data:
• Nominal data:
• Bar chart: represents categorical data with rectangular bars. The length of each bar
corresponds to the frequency or proportion of that category.
• Pie chart: shows the proportion of each category as slices of a circle, useful for displaying
part-to-whole relationships.
Tables and charts for appropriate data
representation by type
• For Qualitative (Categorical) data:
• Ordinal data:
• Ordered Frequency table: frequency table where categories are arranged based on their
inherent order. It helps in understanding trends or patterns across ranked groups.
• Cumulative Frequency Table: shows the running total of frequencies up to each category.
It's useful for understanding how values accumulate across ordered categories.
Satisfaction Level Frequency Cumulative Frequency
1 1 1
2 2 3
3 2 5
4 3 8
5 2 10
Tables and charts for appropriate data
representation by type
• For Qualitative (Categorical) data:
• Ordinal data:
• Ordered Bar Chart: arranges bars in a logical sequence based on category order. This
enhances the readability of ordinal data trends.
• Pareto Chart: bar chart sorted in descending order of frequency, often paired with a
cumulative line. It helps identify the most significant categories contributing to a dataset.
Tables and charts for appropriate data
representation by type
• For Quantitative data:
• Discrete data:
• Frequency Table: table for discrete data lists each unique value and its corresponding
count. It’s suitable when the number of values is small and easily distinguishable.
Weekly Hours Worked Frequency
20 1
28 1
32 1
36 1
40 1
42 1
45 1
50 1
Total 8
Tables and charts for appropriate data
representation by type
• For Quantitative data:
• Discrete data:
• Grouped Frequency Table: when discrete data includes a wide range of values, similar
values are grouped into intervals to simplify analysis. This is especially helpful for large
datasets.

Weekly Study Hours Range Frequency

0–5 0
6–10 6
11–15 3
16–20 1
Tables and charts for appropriate data
representation by type
• For Quantitative data:
• Discrete data:
• Bar Chart: work well for discrete data with a limited number of values. Each bar
represents a specific value and its frequency.
• Histogram (with gaps): a histogram with visible gaps between bars is sometimes used for
discrete data to emphasize the discrete nature of the values.
Tables and charts for appropriate data
representation by type
• For Quantitative data:
• Continuous data:
• Grouped Frequency Table with Class Intervals: data is grouped into intervals (called
"classes") to summarize distributions. Each interval represents a range of values, and
frequencies show how many data points fall into each range.

Class Interval Frequency

5-9 1
10-14 2
15-19 3
20-24 2
25-29 1
30-34 1
Tables and charts for appropriate data
representation by type
• For Quantitative data:
• Continuous data:
• Relative Frequency Table: shows the proportion or percentage of data points in each
category or class interval, rather than raw counts, useful for comparing distributions
across different sample sizes.

GPA Range Relative Frequency

0.0–2.5 0.2
2.6–3.0 0.3
3.1–3.5 0.2
3.6–4.0 0.3
Tables and charts for appropriate data
representation by type
• For Quantitative data:
• Continuous data:
• Histogram: displays the frequency distribution of continuous data using adjacent bars.
Unlike bar charts, the bars touch to represent the continuous nature of the data.
• Box Plot (Box-and-Whisker Plot): visually summarizes the distribution of continuous data
using five key values: minimum, lower quartile, median, upper quartile, and maximum. It
also highlights potential outliers.
• Line Chart: connects data points with lines, making it ideal for tracking trends over time
in continuous variables.
Tables and charts for appropriate data
representation by type
Tables and charts for appropriate data
representation by type
• In some cases, we may be interested in more than 1 variable, the data
that look at more than 1 variable is called Bivariate data (for 2
variables) or Multivariate data (for multiple variables). This type data
is useful for investigation of the relationships, correlations, or
comparisons among multiple variables.
Tables and charts for appropriate data
representation by type
• Contingency Table: summarizes the relationship between two
categorical variables. It allows analysis of possible associations or
dependencies.

Gender 0.0–2.5 2.6–3.0 3.1–3.5 3.6–4.0

Female 2 2 2 0
Male 0 1 2 1

Contingency Table (Gender × GPA range)

Tables and charts for appropriate data
representation by type
• Summary Statistics Table: includes descriptive statistics such as mean,
median, mode, and standard deviation for numerical variables, often
broken down by categories (e.g., average income by education level).
Statistic IQ Score Weekly Study Hours GPA
Count 10.0 10.0 10.0
Mean 106.4 10.3 3.02
Std 6.84 3.91 0.46
Min 96.0 5.0 2.3
25% 101.25 7.0 2.75
50% 105.0 10.5 3.05
75% 111.25 13.0 3.35
Max 118.0 17.0 3.8
Tables and charts for appropriate data
representation by type
• Scatter Plot: displays the relationship between two numerical
variables. Each point represents one observation, making it useful for
spotting trends, patterns, or correlations.
Tables and charts for appropriate data
representation by type
• Stacked Bar Chart: shows how individual parts contribute to the
whole across categories. It’s useful for comparing distributions within
a group while showing total values.
• Grouped Bar Chart: compares categories across multiple groups (e.g.,
test scores by gender and grade level). Each group of bars represents
a different subgroup.
Tables and charts for appropriate data
representation by type
Descriptive statistics
• Descriptive statistics are used to summarize and present data so that
patterns and key features become easier to understand.
• Methods involve numerical measures such as mean, median, mode,
range, and standard deviation.
• Descriptive statistics describe the dataset at hand without making
predictions or generalizations about a larger population.
• Descriptive statistics can be divided into 4 subgroups:
• Measures of Central Tendency
• Measures of Dispersion
• Measures of Position
• Shape Descriptors
Descriptive statistics
• Measures of Central Tendency: describe the “center” or the typical
value of the dataset
• Mean: sum of all values divided by the number of observations.
• Represents the balance point of the distribution of the data
• Sensitive to outliers (extreme values)
𝑛
෌𝑖=1 𝑥𝑖
• Formula: 𝑥ҧ =
𝑛
• Where 𝑥𝑖 = each data point, and 𝑛 is the number of observations
• Median: the middle value when data is arranged in ascending order. If there is
an even number of observations, it is the average of the two middle values
Descriptive statistics
• Measures of Central Tendency: describe the “center” or the typical
value of the dataset
• Median: the middle value when data is arranged in ascending order. If there is
an even number of observations, it is the average of the two middle values.
• Less affected by outliers.
• Formula:
• If 𝑛 is odd: 𝑀𝑒𝑑𝑖𝑎𝑛 = 𝑥 𝑛+1
2
𝑥 𝑛 +𝑥 𝑛+1
2
• If n is even:𝑀𝑒𝑑𝑖𝑎𝑛 = 2
2
• Mode: the value of category that appears most frequently in the dataset, a
dataset can have no mode, one mode (unimodal), or multiple modes
(bimodal/multimodal).
Descriptive statistics
• Measures of Dispersion (Variability): describe how spread out the
data is.
• Range: the difference between the largest and the smallest value in a dataset.
Highly sensitive to outliers.
• Formula: 𝑅𝑎𝑛𝑔𝑒 = 𝑥𝑚𝑎𝑥 − 𝑥𝑚𝑖𝑛
• Variance: measures the average squared deviation of each value from the
mean. Higher variance indicates that data point are more spread out from the
mean.
𝑛
෌𝑖=1 𝑥𝑖 −𝑥ҧ 2
• Formula: 𝑠 2 =
𝑛−1
Descriptive statistics
• Measures of Dispersion (Variability): describe how spread out the
data is.
• Standard deviation: the square root of variance, measure the spread of the
data from the mean in the same units as the data. Clearer sense of deviation
from the mean.
𝑛
෌𝑖=1 𝑥𝑖 −𝑥ҧ 2
• Formula: s =
𝑛−1
• Interquartile range (IQR): the difference between the third quartile (Q3) and
the first quartile (Q1). Measures the spread of the middle 50% of the data,
resistant to outliers.
• Formula: IQR = 𝑄3 − 𝑄1
Descriptive statistics
• Measures of Position:
• Percentiles: indicates the value below which a given percentage of
observations fall.
𝑘 𝑛+1
• Formula: 𝑃𝑘 =
100
• Where 𝑃 is the percentile rank position and 𝑘 is the desired percentile
• For example: a ranked dataset of 100 value from 1 to 100 {1,2,3,…,100} will have the 90th
percentile at position 90 which also has the value of 90.
• Quartiles: similar to percentile but instead divide the data set into 4 parts
with Q1 being the 25th percentile, Q2 is the 50th percentile (median), and Q3
is the 75th percentile.
Descriptive statistics
• Measures of Position:
• Z-score (standard score): expresses how many standard deviations a value is
from the mean. Positive scores are above the mean, and negative scores are
below.
𝑥−𝑥ҧ
• Formula: 𝑧 =
𝑠
Descriptive statistics
• Shape descriptors (distribution shape):
• Skewness: measures the degree of asymmetry in the distribution of the data.
Positive skew means the tail is on the right while negative skew means the tail
is on the left.
𝑛
෌i=1 𝑥𝑖 −𝑥ҧ 3
• Formula: Skewness =
𝑛−1 𝑠 3
Descriptive statistics
• Shape descriptors (distribution shape):
• Kurtosis: measures the “tailedness” of the data distribution, indicating
whether data are heavy-tailed of light-tailed compare to a normal
distribution. Higher kurtosis means that the data has more extreme values.
𝑛
෌i=1 𝑥𝑖 −𝑥ҧ 4
• Formula: Kurtosis =
𝑛−1 𝑠 4
• If the kurtosis value is > 3, the data is Leptokurtic, meaning data is more clustered
around the mean. If the kurtosis value is < 3, the data is Platykurtic, meaning data is less
clustered around the mean. If the kurtosis value is = 3, the data is Mesokurtic, meaning
the distribution follows the normal distribution curve.
Applications of Data Analysis
• Qualitative analysis:
• Examines non-numerical data to understand concepts, opinions, behaviors, or
experiences.
• Focuses on studying themes and patterns of the subject.
• Example: analyzing customers’ feedback to identify where improvements are
needed.
• Quantitative analysis:
• Examines numerical data to measure quantities, compare groups, and identify
relationships using statistical methods.
• Focuses on the numerical and statistical interpretation.
• Example: Calculating students’ average GPA and testing if there is a
correlation between GPA and weekly study hours.
Real-world application
• Customer segmentation: identify group of customers with similar
buying habits.
• Sales forecast: predict demand to plan production accordingly.
• Evaluation of marketing campaign: measure the return on investment
(ROI) of the campaign.
• Risk assessment: evaluate the risk that a loan applicant carry (credit
score).
• Clinical trials: evaluate the effectiveness of new drugs/prevention
capabilities of new vaccines.
Real-world application
• Curriculum effectiveness: measure and compare the impact different
teaching methods has on students’ performance.
• Census analysis: investigate population trend.
• Example:
• During COVID-19, Vietnam’s Ministry of Health used case data and mobility
tracking to forecast outbreak hotspots and allocate resources effectively.
• Loyalty card programs used by supermarkets to create personalized
promotion to customers based on their purchasing habits.
• Tracking on monthly spending to create informed budgeting choices.
Ethical concerns
• Privacy and confidentiality: data collected must have the owner’s consent
and must not be identifiable.
• Accuracy and integrity: data must be up to date and complete. Absolutely
do not “cherry-pick” – use data that supports a specific agenda.
• Bias and fairness: data must try to reflect the population instead of trying
to reinforce bias/stereotypes.
• Example:
• U.S National Security Agency mass surveillance program where they collected data
from citizens without consent.
• Data trading of your phone number, name, personal information to
scammers/marketers which results in scam/spam calls.
=> All entities whether private or governments can misuse data.
End of Chapter 1
Good luck and happy learning!