Module – 3
Types & Sources of data, collection of data, measurement & scaling
techniques, Techniques of data analysis
Types & Sources of data
• Data is a collection of facts — values or measurements — that can be
processed or analyzed to derive meaningful insights.
Data can come from two main sources:
• Primary Data – collected firsthand for a specific purpose.
• Secondary Data – collected previously by someone else for another
purpose.
Primary data
•Information collected directly from the source.
Example: conducting a survey asking people about
their favorite ice cream flavor.
•The responses you receive are primary data.
•This data is original, specific, and valuable for drawing
conclusions and making decisions.
•✅ Advantages: Firsthand, specific, and accurate.
•⚠️Disadvantages: Time-consuming and costly.
Secondary Data
•Data already collected by someone else for a different purpose.
•Sources: Books, articles, government reports, research studies.
•Saves time and effort, but requires verifying its reliability
and relevance.
•Example: Using government reports on ice cream sales growth
over the decade to strengthen your analysis.
Types of Data – Qualitative & Quantitative
•Data can be categorized into:
[Link] Data – descriptive, expressed in words.
[Link] Data – measurable, expressed in numbers.
•Quantitative data answers questions like how much or how many.
•Qualitative data answers questions like what kind or why
Quantitative Data
•Objective, measurable, numeric.
•Deals with quantities, e.g., number of students, height, or income.
•Subtypes:
•Discrete data: Countable values (e.g., number of goals scored).
•Continuous data: Measurable values with decimals (e.g., weight, height).
•Example:
•Number of goals in a football match — discrete (1, 2, 3…).
•Weight of players — continuous (e.g., 70.5 kg).
Qualitative Data
•Subjective and descriptive, expressed in words, not numbers.
•Describes qualities or characteristics.
•Subtypes:
•Nominal: Categories without order (e.g., gender, color).
•Ordinal: Ordered categories (e.g., rating scales, satisfaction levels).
•Example: Dog’s breed, color, and energy level.
Example – The Dog Case Study 🐕
Qualitative attributes:
•Breed: German Shepherd
•Color: Black and Brown
•Energy: Full of energy
Quantitative attributes:
•Age: 5 years (Discrete)
•Number of Puppies: 3 (Discrete)
•Weight: 23.5 kg (Continuous)
•Height: 24 inches (Continuous)
➡️The dog has both qualitative and quantitative characteristics.
Levels of Measurement
• Statistics has 4 levels of measurement (also called scales of
measurement):
• Nominal
• Ordinal
• Interval
• Ratio
• Each level increases in precision and helps determine which
statistical tests can be used.
1st Level – Nominal Scale
•The simplest level of measurement.
•Describes variables without any natural order or ranking.
•Variables are named or labeled only.
• Cannot be ordered or measured.
•Can be qualitative or quantitative.
• Examples:
• Preferred smartphone brand: Apple, Samsung, Nokia
• Gender: Male / Female
• Location: City / Town
• Type of pet: Dog / Cat / Fish
• Type of house: Apartment / Villa / Cottage
2nd Level – Ordinal Scale
• “Ordinal” → relates to order or ranking.
• Has a natural order but no fixed interval between values.
• The order matters, but the difference between ranks does not.
• ⚠️No true zero point.
• Examples:
• Temperature preference: Cold → Hot → Hottest
• Satisfaction scale: Strongly Agree → Agree → Neutral → Disagree →
Strongly Disagree
3rd Level – Interval Scale
• Represents equal intervals between measurements.
• Contains all properties of the ordinal scale.
• Allows calculation of differences between values.
• ❌ No absolute zero.
• Examples:
• Temperature in Celsius
• Difference between 20°C and 30°C = 10°C
• Difference between 60°C and 70°C = 10°C
• Time and Calendar Years also follow the interval scale.
4th Level – Ratio Scale
• The highest and most informative level of measurement.
• Has:
• Order
• Equal intervals
• Absolute zero (true zero point)
• Examples:
• Height, Weight, Length, Age, Income
• A person 6 feet tall is twice as tall as one 3 feet tall.
• 0 weight = absence of weight → true zero exists.
Techniques of Data Analysis
• Descriptive Analysis: Summarizes the main characteristics of a dataset (e.g.,
mean, median, mode, standard deviation, frequencies).
• Engineering use: Summarizing test results, visualizing data distributions (histograms,
scatter plots), initial data exploration.
• Inferential Analysis: Draws conclusions and makes predictions or inferences
about a population based on a sample of data.
• Engineering use: Hypothesis testing (t-tests, z-tests, ANOVA) to compare material strengths
or process efficiency, confidence intervals for reliability estimates.
• Regression Analysis: Examines the relationship between a dependent variable
and one or more independent variables to understand how changes in one affect
the other.
• Engineering use: Predicting material fatigue based on applied load and temperature,
forecasting project costs or timelines.
Techniques of Data Analysis
• Time Series Analysis: Analyzes data points collected over a sequence of
time to identify trends, cycles, or seasonal patterns.
• Engineering use: Monitoring sensor data from equipment over time to predict failure
(predictive maintenance), analyzing power consumption trends.
• Qualitative Analysis: Interpreting non-numerical data to identify themes,
patterns, and underlying meanings.
• Engineering use: Analyzing interview feedback, categorizing types of equipment
failures, interpreting open-ended survey responses.
• Machine Learning/Data Mining: Using algorithms to discover patterns in
large datasets, build predictive models, and automate decision-making.
• Engineering use: Optimizing manufacturing processes, quality control automation,
training systems to detect anomalies in data streams.