Studentized Range q Table

What Is a Studentized Range (q) Table?

The studentized range q table provides critical values for Tukey's Honest Significant Difference (HSD) test. After ANOVA reveals significant differences among groups, Tukey's HSD identifies which specific pairs of means differ significantly.

The studentized range statistic is:

$q = \frac{\max(\bar{X}_i) - \min(\bar{X}_i)}{\sqrt{MS_E/n}}$

When to Use This Table

Use the q table after:

1. Conducting ANOVA and finding a significant F-statistic
2. Wanting to make all pairwise comparisons between group means
3. Needing to control the familywise error rate

Tukey's HSD is appropriate when:

• All groups have equal (or nearly equal) sample sizes
• You want to compare every group to every other group

How to Read the Table

1. Determine k (number of groups being compared)
2. Calculate df (error degrees of freedom from ANOVA = N - k)
3. Select your α level (typically 0.05 or 0.01)
4. Find q at the intersection

Example: Comparing 4 treatment groups with 36 total observations:

• k = 4
• df = 36 - 4 = 32 ≈ 30 (use closest value)
• At α = 0.05, q ≈ 3.85

Calculating Tukey's HSD

The Honest Significant Difference is:

$HSD = q \times \sqrt{\frac{MS_E}{n}}$

Where:

• q = studentized range value from table
• MS_E = mean square error from ANOVA
• n = sample size per group

Decision: If |Mean₁ - Mean₂| ≥ HSD, the means are significantly different.

Worked Example

Scenario: Three teaching methods tested on 30 students (10 per group)

• ANOVA shows significant differences (F = 4.82, p < 0.05)
• MS_E = 25
• Means: Method A = 75, Method B = 82, Method C = 78

Calculation:

1. k = 3, df = 30 - 3 = 27
2. From table (α = 0.05): q ≈ 3.51
3. HSD = 3.51 × √(25/10) = 3.51 × 1.58 = 5.55

Comparisons:

• |A - B| = |75 - 82| = 7 > 5.55 ✓ Significant
• |A - C| = |75 - 78| = 3 < 5.55 ✗ Not significant
• |B - C| = |82 - 78| = 4 < 5.55 ✗ Not significant

Unequal Sample Sizes: Tukey-Kramer

When group sizes differ, use the Tukey-Kramer modification:

$HSD_{ij} = q \times \sqrt{\frac{MS_E}{2}\left(\frac{1}{n_i} + \frac{1}{n_j}\right)}$

This produces a different HSD for each pair of groups.

Comparison with Other Post-Hoc Tests

Why "Studentized"?

The term "studentized" refers to dividing by an estimate of standard deviation (like the t-distribution, named after "Student"). The range is standardized using the pooled within-group variability.

Key Properties

• Controls familywise Type I error rate at α
• More powerful than Bonferroni when making all pairwise comparisons
• Conservative (may miss some true differences)
• Assumes normality and homogeneity of variance

Related Tools

• One-Way ANOVA Calculator — Initial test
• Tukey's HSD Calculator — Perform the test
• Bonferroni Correction Calculator — Alternative method

Summary

The studentized range q table is essential for post-hoc analysis following ANOVA. It provides critical values that allow you to calculate Tukey's HSD—the minimum difference between means required for statistical significance. By controlling the familywise error rate, Tukey's test maintains the overall α level across all possible pairwise comparisons, making it the go-to choice when examining which specific groups differ after finding a significant overall effect.