How to Calculate Critical Values by Hand

Critical values are thresholds used in hypothesis testing to decide whether to reject the null hypothesis. While our calculator gives instant results, understanding how to calculate critical values by hand builds statistical intuition. This guide walks you through the manual process for the Z, t, Chi-Square, and F distributions. If you need a refresher on what critical values are, check out our What Is a Critical Value article.

What You’ll Need

• Your significance level (α) — typically 0.05, 0.01, or 0.10
• Knowledge of the test type: one-tailed (left or right) or two-tailed
• Degrees of freedom (df) for t, Chi-Square, and F distributions
• A standard normal (Z) table
• A t-distribution table
• A Chi-Square distribution table
• An F-distribution table
• Pencil and paper (or a quiet workspace)

Step-by-Step Process

1. Identify the distribution and test type. Is your test using Z, t, Chi-Square, or F? Determine if it’s a left-tailed, right-tailed, or two-tailed test.
2. Establish the significance level (α). α is the probability of rejecting the null hypothesis when it is true. Common values: 0.05 for 95% confidence, 0.01 for 99% confidence.
3. Adjust α for two-tailed tests. For a two-tailed test, split α into two equal tails: α/2 on each side. For one-tailed tests, use α directly.
4. Find the critical value using the appropriate formula or table. Each distribution has its own method:
   • Z-distribution: Look up the cumulative probability (1−α for right tail, α for left tail, or 1−α/2 for two-tailed) in the Z-table.
   • t-distribution: With known df, locate the t-value at the desired tail probability from a t-table.
   • Chi-Square distribution: With known df, find the value at the cumulative probability from a Chi-Square table.
   • F-distribution: With df1 and df2, locate the F-value at the right-tail probability from an F-table.
5. Record the critical value. This is the threshold your test statistic must exceed to reject H₀.
6. Verify with a calculation (optional). Use inverse-CDF formulas if you have the appropriate tools.
7. Compare your test statistic. If the absolute value of the test statistic exceeds the critical value, reject H₀.

Worked Example 1: Z-Test (One-Tailed, Right-Tailed)

Scenario: A company claims its battery lasts more than 100 hours. Sample size n=36, population standard deviation σ=10 hours, α=0.05. Test the claim (right-tailed).

1. Distribution: Z (n ≥ 30, σ known).
2. α = 0.05, one-tailed right.
3. No α adjustment needed.
4. Look up the Z-table for cumulative probability 0.95 (since right tail α=0.05 means 95% below). The Z-value with 0.95 cumulative probability is about 1.645.
5. Critical value Zα = 1.645.
6. Calculate test statistic: Z = (X̄ - μ) / (σ/√n). If X̄=103, then Z = (103-100)/(10/√36) = 3/(10/6)=1.8.
7. Since 1.8 > 1.645, reject H₀.

Worked Example 2: t-Test (Two-Tailed)

Scenario: A diet program claims average weight loss of 5 kg. A sample of 11 participants shows mean loss of 4.2 kg with sample standard deviation 1.5 kg. α=0.05, test the claim (two-tailed).

1. Distribution: t (n=11 < 30, σ unknown).
2. α = 0.05, two-tailed. So use α/2 = 0.025 in each tail.
3. Degrees of freedom: df = n-1 = 10.
4. Look up t-table for df=10, cumulative probability 0.975 (since 1−α/2 = 0.975). The t-value is approximately 2.228.
5. Critical values: ±2.228.
6. Test statistic: t = (4.2-5)/(1.5/√11) = -0.8/(1.5/3.317) = -0.8/0.452 ≈ -1.770.
7. Since |−1.770| = 1.770 < 2.228, fail to reject H₀.

Common Pitfalls

• Using the wrong tail: Always confirm if your test is one-tailed (left or right) or two-tailed. The critical value differs drastically.
• Forgetting to adjust α for two-tailed tests: Use α/2 in each tail when looking up critical values.
• Confusing degrees of freedom: For t-test, df = n-1; for Chi-Square, df depends on the number of categories minus constraints; for F, df1 and df2 from numerator and denominator.
• Mistaking the direction: In a right-tailed test, the critical value is positive; left-tailed is negative; two-tailed has both positive and negative.
• Using sample standard deviation σ when population σ is unknown: Use t-distribution, not Z.

Final Thoughts

Mastering manual calculation reinforces your understanding of hypothesis testing. For complex or large data, use our Critical Value Calculator. For more details on formulas and interpretation, see our Critical Value Formula & Calculation and How to Interpret Critical Values guides. If you need values for different distributions, visit Critical Values for Z, t, Chi-Square, F.