Course contentsShow
← Data Science
Lesson 326 of 2,1459. Hypothesis Testing FoundationsFree lesson

Critical Values

Finding the boundary values that separate rejection and non-rejection regions based on α.

Critical Values

You'll learn how critical values mark the exact cutoff points that tell you when to reject the null hypothesis.

What Are Critical Values?

A critical value is the specific number on the sampling distribution that separates the rejection region from the non-rejection region. Think of it as a fence line: if your test statistic falls on one side, you reject H₀; if it falls on the other side, you don't.

Remember when you set your significance level (α = 0.05, for example)? The critical value is where that α "lives" on the distribution. It's the boundary value that marks off exactly 5% (or whatever α you chose) in the tail(s) of the distribution.

How It Works

Imagine a courtroom where the judge says, "Evidence must be this strong to convict." The critical value is that threshold of strength.

  • For a two-sided test at α = 0.05, you have two critical values (one positive, one negative) that each mark off 2.5% in the tails
  • For a one-sided test at α = 0.05, you have one critical value marking off 5% in just one tail

If your calculated test statistic exceeds the critical value (falls in the rejection region), you reject H₀. If it doesn't reach the critical value, you fail to reject H₀.

Finding Critical Values

Critical values come from statistical tables or software and depend on:

  • Your chosen α level
  • The distribution you're using (Z, t, chi-square, F)
  • Whether your test is one-sided or two-sided
  • Degrees of freedom (for t, chi-square, and F distributions)

Key Takeaway: Critical values are the boundary numbers that define rejection regions—they translate your chosen significance level (α) into concrete decision thresholds on the sampling distribution.