# Here Are Basic Steps Of Anova Compute Variance In Each Group Sample If Not Given

Here are basic steps of ANOVA:

Compute variance in each group/sample, if not given.

Compute Mean of all variations from each sample. Consider all sample variances as a new set of data,let’s call it set of variances, and compute mean for this numbers.That will be Mean Square Error (MSE) in Lane terminology.

Compute Variance for the set of numbers that represent each sample variance. For Mean of this set use MSE calculated above. Multiply result by number of samples/group.That will be Mean Square Between (MSB) in Lane terminology.

compute F-value as MSB divided by MSE.

Define degree of freedom df1 and df2.Let’s say you have k groups/samples and each sample has n observations/elements. df1 = (number of samples) – 1 = k – 1df2 = (number of samples) × (each sample size – 1) = k(n – 1)

Use Excel Statistics function F.DIST.RT(x, df1, df2) to determine P-value. It will give you the area of Right Tail (RT) in ANOVA F-Distribution.

compare P-value to the given significance level:

if p-value is less than given significance level (typically 0.05)then reject Null Hypothesis;if p-value is greater than given significance level (typically 0.05)then do not reject Null Hypothesis.

Group 1

Group 2Group 3Group 466746965687161625967696572827368467661

64

## Leave a Reply

Want to join the discussion?Feel free to contribute!