What Is Data?
Data begins with reality, but it is not reality itself. It is a recorded representation of people, events, measurements, systems, behaviors, or processes.
Data is recorded representation
Something happens in the real world. A person visits a clinic, a machine records a temperature, a customer purchases a product, or a sensor detects motion. When information about that event is recorded, it becomes data.
The record may be a number, word, image, sound, date, category, location, signal, or document. The form can change, but the central idea remains the same: data is information captured about something.
Core idea
Data is a representation of reality created through observation, measurement, classification, or recording.
Because data is a representation, it is always incomplete. A row in a spreadsheet may represent a person, but it does not contain the full person. A medical image may represent anatomy and biological activity, but it does not contain every fact about the patient.
One reality can produce many kinds of data
The same real-world event can be represented differently depending on what is observed and how it is recorded.
Healthcare
A patient visit may produce age, symptoms, diagnosis codes, laboratory values, imaging studies, medications, and clinical notes.
Business
A purchase may produce product name, quantity, price, store, timestamp, payment method, and customer account information.
Transportation
A trip may produce origin, destination, speed, travel time, distance, fuel use, traffic conditions, and route information.
How data is created
An event or condition exists
Something occurs, changes, or is present in the real world.
Someone decides what matters
A person or system selects which characteristics should be observed or measured.
A method records the observation
The information may be entered manually, measured by an instrument, generated by software, or collected by a sensor.
The record becomes part of a dataset
Individual observations are organized so they can be reviewed, compared, analyzed, and communicated.
Data is not automatically truth
Data can be useful evidence, but recording something does not guarantee that it is complete, accurate, unbiased, or properly understood.
Measurements can be wrong. Categories can be poorly defined. Records can be missing. Sensors can fail. People can enter incorrect values. Important context may never be captured.
Responsible interpretation
A data scientist should ask not only, “What does this dataset say?” but also, “How was it created, what does it leave out, and how much confidence should we place in it?”
Practice: Identify the representation
Choose one ordinary event, such as a medical visit, grocery purchase, work shift, school class, or car trip. Then answer:
- What happened in reality?
- What information could be recorded?
- Which variables would appear in the dataset?
- What important details might not be recorded?
- How could the record be inaccurate or misunderstood?
Create your first dataset
Questions to deepen your understanding
- Can two datasets describe the same event differently and both still be useful?
- Who decides what gets measured and what remains invisible?
- What is lost when a complex human experience is reduced to a number or category?
- How could incomplete data lead to a confident but incorrect conclusion?
Why this matters in real work
In healthcare, business, research, government, and engineering, decisions are often made from recorded information. Good analysis therefore begins before statistics or programming. It begins by understanding what the data represents and how faithfully it reflects the real situation.
This lesson forms the foundation for everything that follows. Before cleaning, visualizing, modeling, or predicting, we must understand the relationship between reality and the records created from it.