Course contentsShow
← Data Science
Lesson 1442 of 2,14536. Causal Inference BasicsFree lesson

Limitations and Practical Constraints

When RCTs are infeasible, expensive, or unethical, and the role of observational alternatives.

Limitations and Practical Constraints

What You'll Learn

You'll understand when randomized controlled trials (RCTs) become impractical or impossible, and why observational methods become necessary alternatives.

Why RCTs Aren't Always the Answer

While RCTs are the gold standard for establishing causality, they face significant real-world limitations:

Feasibility Constraints

Some experiments simply cannot be randomized. You can't randomly assign people to smoke cigarettes for 20 years to study lung cancer risk—participants won't comply, and recruitment would fail. Similarly, you can't randomly assign companies to different economic policies or cities to climate conditions.

Cost and Time

RCTs are expensive and slow. A pharmaceutical trial costs millions and takes years. For business decisions or policy questions requiring quick answers, waiting for a multi-year RCT isn't realistic. Sometimes the opportunity window closes before results arrive.

Ethical Barriers

Many causal questions involve interventions we cannot ethically impose. We can't deny potentially life-saving treatments to a control group, expose children to harmful substances, or randomly assign people to poverty to study its effects.

External Validity Trade-offs

Even when feasible, highly controlled RCT environments may not reflect real-world conditions. A drug tested in carefully selected trial participants might behave differently in the general population.

The Observational Alternative

When RCTs fail these constraints, researchers turn to observational methods—analyzing data from natural experiments, administrative records, or existing datasets. These require careful control for confounders and stronger assumptions, but they allow us to study important causal questions that RCTs cannot address.

Key Takeaway: RCTs provide the strongest causal evidence but are often impractical due to cost, ethics, feasibility, or time constraints—making well-designed observational studies essential tools in the causal inference toolkit.