Data Science Academy · Level 1 · Module 2

Types and Structures of Data

Data can represent categories, quantities, text, images, measurements, documents, signals, or events. Understanding its type and structure determines how it should be stored, analyzed, and interpreted.

Lesson overview

What you will learn and create

This lesson introduces the major ways data is classified and organized.

Distinguish

Qualitative and quantitative data

Recognize whether a variable describes a category or measures a numerical quantity.

Recognize

Data structure

Identify structured, semi-structured, and unstructured information.

Understand

Measurement scales

Explain how nominal, ordinal, interval, and ratio variables differ.

Build

A data classification guide

Classify variables from a real-world dataset by type, structure, and measurement scale.

Learn

Variables carry different kinds of meaning

A variable is a characteristic recorded for an observation. Age, diagnosis, temperature, product category, travel time, and customer comments are all variables, but they do not represent information in the same way.

The type of a variable affects which summaries, visualizations, and analytical methods are appropriate. A category such as blood type should not be treated like a numerical measurement such as body temperature.

Core idea

Before analyzing a variable, determine what it means—not merely how it appears in a file.

Meaning

Qualitative and quantitative data

Qualitative data

Qualitative data represents qualities, names, groups, labels, or categories.

  • Blood type
  • Department name
  • Product category
  • Diagnosis
  • Customer feedback

Quantitative data

Quantitative data represents amounts, counts, measurements, or numerical values.

  • Age
  • Weight
  • Temperature
  • Number of purchases
  • Travel time
Storage and organization

Structured, semi-structured, and unstructured data

Structured data

Information organized into clearly defined rows, columns, fields, and data types. Examples include spreadsheets, relational databases, and transaction tables.

Semi-structured data

Information that does not follow a traditional table but contains labels, keys, tags, or markers that provide organization. Examples include JSON, XML, emails, and system logs.

Unstructured data

Information without a fixed row-and-column format. Examples include clinical notes, documents, photographs, medical images, audio recordings, and videos.

Measurement

Four common measurement scales

Measurement scales describe the meaning and relationships contained within recorded values.

Scale Meaning Example Important property
Nominal Names or categories Blood type No natural order
Ordinal Ordered categories Pain: mild, moderate, severe Order exists, distance is uncertain
Interval Equal numerical intervals Temperature in Celsius Zero does not mean absence
Ratio Equal intervals with a true zero Weight or duration Ratios are meaningful

Practice: Classify each variable

For each variable below, identify whether it is qualitative or quantitative and determine its likely measurement scale.

  1. Patient identification number
  2. Age in years
  3. Satisfaction rating: poor, fair, good, excellent
  4. Appointment date
  5. Diagnosis category
  6. Travel distance in miles
  7. Written clinical note
Build

Create a data classification guide

Use the small dataset you created in Module 1 or choose another everyday dataset. Create a table with these columns:

Variable Meaning Qualitative or quantitative Measurement scale Structure or format
Date When the event occurred Quantitative or temporal Interval Structured
Event type Category of event Qualitative Nominal Structured
Duration Length of the event Quantitative Ratio Structured

Add at least five variables. For each one, explain why you selected that classification.

Reflect

Questions to deepen your understanding

  • Can a number function as a label rather than a measurement?
  • Could the same variable be stored differently in two systems?
  • What analytical mistakes could occur when an ordinal variable is treated as a precise numerical measurement?
  • Why is unstructured data often more difficult to analyze?
Apply

Why data type matters in real work

Healthcare systems contain numerical measurements, diagnosis categories, medical images, clinical notes, timestamps, and signals. Business systems may contain transaction tables, customer comments, emails, product images, and activity logs.

Each type requires different methods for storage, preparation, visualization, and analysis. Correctly identifying data type is therefore one of the first responsibilities in any data science project.