Understanding Sample Size, Power and Effect Size

Before you touch a calculator, you need the ideas behind the number. This level turns the intimidating language of power, alpha, beta and effect size into plain intuition, so "How many participants do we need?" becomes a question you can reason about rather than fear.

Who this level is for

Anyone preparing a protocol, grant or proposal who freezes at the sample-size section: researchers, clinicians, coordinators, graduate students and program staff at any career stage, across low- and middle-income countries and beyond. No statistical background assumed, and no algebra.

What you'll learn

• Why sample size is a design decision made early, not a formality bolted on at the end
• The real cost of under-powered and over-powered studies (wasted resources, ethics, feasibility)
• How to restate a research question as null and alternative hypotheses
• Type I and Type II errors, made intuitive with a simple jury analogy
• What statistical power (80% or 90%) actually means, and why we rarely go lower
• How to choose an effect size that reflects a clinically meaningful difference, not a random guess
• How precision, margin of error and confidence intervals drive descriptive studies

You finish with: Your null and alternative hypotheses + a justified effect size.