ANALSIS OF VARIANCE- ANOVA - [INTRODUCTION]
Example: Three
rehabilitation methods are being compared.
The goal is to reduce time off work due to (similar) injuries.
Physiotherapy vs chiropractic vs non-steroidal drugs:
Cannot use t-tests three times (danger of 3 x repeating the error of claiming stat sig when not valid)
|
Between groups SS:
Tph2 + Tch2 + Tns2 - T2 Nph Nch Nns N
= 3172 + 4102 + 3632 - 1092 5 5 5 15
= 864.933
|
Now divide each SS by the df to obtain MS:
And then with MS(betw)/MS(within), you have the F ratio...
At this stage, don't worry too much about the correct alternative hypothesis.
Just describe the result in terms of the original data. (See conclusion above)