Statistics does not attempt to describe reality perfectly. Instead, it builds simplified representations of reality called statistical models. A model is a structured way to explain variability, uncertainty, and patterns in data.
Every statistical conclusion rests on a model, whether acknowledged or not. Unstated models are often the most dangerous ones.
A statistical model is an assumption-driven framework that connects data to an underlying process.
A model typically specifies:
Models do not claim truth. They claim usefulness within stated assumptions.
When a lecturer predicts student performance using study hours, attendance, and prior grades, a model is being used.
When an economist predicts inflation using historical trends, a model is being used.
When a medical researcher estimates treatment effectiveness, a model is being used.
In each case, assumptions shape conclusions.
Deterministic thinking assumes the same input always produces the same output.