One-Way ANOVA Calculator

Enter each group as a list of numbers, with a semicolon between groups.

Each group is a list of numbers; put a semicolon between groups. At least 2 groups of 2 or more.

Try

Result

You entered3 groups, N=9\text{3 groups, } N = 9

F=3, p=0.125, do not reject H0 at the 5% level\displaystyle F = 3,\ p = 0.125,\ \text{do not reject H0 at the 5\% level}

ANOVA table

SourceSSdfMSFBetween6233Within661Total128\displaystyle \begin{array}{l|rrrr} \text{Source} & SS & df & MS & F \\ \hline \text{Between} & 6 & 2 & 3 & 3 \\ \text{Within} & 6 & 6 & 1 & \\ \text{Total} & 12 & 8 & & \end{array}

Sum of squares between

6\displaystyle 6

Sum of squares within

6\displaystyle 6
Show the work
  1. Group means and grand mean

    There are k=3k = 3 groups with N=9N = 9 values in all. The grand mean xˉ\bar{x} is the mean of every value together.

    xˉ1=2, xˉ2=3, xˉ3=4, xˉ=3\bar{x}_{1} = 2,\ \bar{x}_{2} = 3,\ \bar{x}_{3} = 4,\ \bar{x} = 3
  2. Sum of squares between the groups

    This measures how far each group mean sits from the grand mean, weighted by the size nin_i of the group.

    SSB=∑ni(xˉi−xˉ)2=3(2−3)2+3(3−3)2+3(4−3)2=3+0+3=6\begin{aligned} SS_B &= \sum n_i \left(\bar{x}_i - \bar{x}\right)^2 \\ &= 3\left(2 - 3\right)^2 + 3\left(3 - 3\right)^2 + 3\left(4 - 3\right)^2 \\ &= 3 + 0 + 3 \\ &= 6 \end{aligned}
  3. Sum of squares within the groups

    This measures how far each value sits from the mean of its own group, which is the spread that is not explained by the groups. Each term below is one group's total.

    SSW=∑(x−xˉi)2=2+2+2=6\begin{aligned} SS_W &= \sum \left(x - \bar{x}_i\right)^2 \\ &= 2 + 2 + 2 \\ &= 6 \end{aligned}
  4. Mean squares and the F statistic

    Divide each sum of squares by its degrees of freedom, k−1=2k - 1 = 2 and N−k=6N - k = 6, then compare them.

    F=SSB/(k−1)SSW/(N−k)=6/26/6=31=3\begin{aligned} F &= \dfrac{SS_B / (k - 1)}{SS_W / (N - k)} \\ &= \frac{6 / 2}{6 / 6} \\ &= \frac{3}{1} \\ &= 3 \end{aligned}
  5. The p-value from the F distribution

    The p-value is the probability that an F statistic with 22 and 66 degrees of freedom is at least as large as 33 when all the group means are equal.

    p=P(F2,6≥3)≈0.125p = P\left(F_{2,6} \ge 3\right) \approx 0.125
  6. Decision at the 5% level

    Since p=0.125p = 0.125 is not below 0.050.05, there is not enough evidence against the null hypothesis H0H_0 that the means are equal.

  7. Result

    F=3F = 3, p=0.125p = 0.125: do not reject H0 at the 5% level.

Worked examples