Skip to content
Omnikalc

Scheffé contrast test calculator

Enter the four group means, the contrast coefficients, the group size and the error mean square from an ANOVA to get the contrast estimate, its standard error, the Scheffé F and the adjusted p-value.

Updated on
Your inputs

Value out of range

Value out of range

Value out of range

Value out of range

Value out of range

Value out of range

Value out of range

Value out of range

Value out of range

Value out of range

Results

Contrast estimate

-6.000

Scheffé F (contrast squared over (k − 1) variances)

1.800

Scheffé adjusted p-value

0.1577

Standard error of the contrast
2.582
t ratio of the contrast
-2.324
Degrees of freedom (3 and error)
56
Unadjusted p-value (t test)
0.0238
Sum of the coefficients (must be 0)
0.00

The Scheffé contrast test calculator gives the estimate and standard error of a contrast among four group means, the Scheffé F and the adjusted p-value that holds for any contrast chosen after an ANOVA, along with the unadjusted t test for comparison.

How the calculation works

  • Enter the four group means, the observations per group and the error mean square of the ANOVA.
  • Enter the coefficients of the contrast; they must add to zero.
  • Read the estimate, the Scheffé F and the adjusted p-value.

Example

With means of 24, 25, 27 and 28, groups of 15 and an error mean square of 25, the contrast estimates -6.000 with a standard error of 2.582. The Scheffé F is 1.800 with 56 error degrees of freedom, so the adjusted p-value is 0.1577 against an unadjusted 0.0238. The coefficients add to 0.00.

Why it matters

The calculator helps students, researchers and analysts who found a significant ANOVA and want to know which groups differ, especially for complex comparisons such as two treatments against two controls, without inflating the false positive rate.

Good to know

Use it with equal group sizes and the error mean square of a one-way ANOVA with four groups. Choose the coefficients so that they add to zero, and keep the same scale: 1, 1, −1, −1 estimates the difference of the means of the two pairs multiplied by 2. For unequal group sizes the standard error changes, and for pairwise comparisons only Tukey’s HSD is more powerful. The test assumes normal errors and equal variances.

The ANOVA calculator gives the F ratio and the error mean square, the omega squared calculator reports the effect size, the Bonferroni correction calculator adjusts planned comparisons and the Welch ANOVA calculator handles unequal variances.

Frequently asked questions

What is the Scheffé test?

The Scheffé test is a post-hoc procedure that tests any contrast of the group means in an ANOVA while controlling the familywise error rate for all possible contrasts, even those chosen after seeing the data. A contrast is a weighted sum of means whose weights add to zero, like the comparison of two groups against the other two.

How is the statistic computed?

The contrast is ψ = Σ cᵢ × meanᵢ, with a standard error √(MSE × Σ cᵢ² / n) for groups of n observations. The Scheffé F is ψ² / ((k − 1) × SE²), compared with an F distribution with k − 1 and N − k degrees of freedom. A contrast is significant if the F is greater than the critical value.

Why is it conservative?

Because it protects against all possible contrasts, including the complex ones, the Scheffé test has less power than tests made for pairwise comparisons, such as Tukey’s HSD. It is the best choice when you test complex contrasts or contrasts suggested by the data, and not only the planned pairs.

What does a zero sum of coefficients mean?

The coefficients must add to zero for the contrast to compare means and not their level. The comparison of groups 1 and 2 with groups 3 and 4 uses 1, 1, −1 and −1, and a comparison of groups 1 and 4 uses 1, 0, 0 and −1. The calculator shows the sum so that you can check your coefficients.

Was this calculator helpful?

Thanks for your feedback!

Results are estimates provided for information purposes. They do not replace professional advice.