I'm very new to the statistic methods but at the beginning of my wonderful math-journey I found some problems.

Two days ago I got to know the ANOVA method. I'm kind of familiar with it today, but there's one issue that I can not understand.

I know how to calculate F, and df1 so as df2. But how about F critical value? How can I calculate this? (I calculated F, df1, df2, have n, k and alpha given.)

  • $\begingroup$ You do not really calculate it yourself, because it is a quantile of the F-distribution. So you read it off a table (or use a calculator) given your $\alpha$, df1, and df2. $\endgroup$ – hplieninger Dec 2 '13 at 13:43
  • $\begingroup$ what if I would like to implement it by myself? (I would!) so then, how to calculate it, do I need a big math knowledge? $\endgroup$ – mazix Dec 2 '13 at 14:08
  • 1
    $\begingroup$ There is no closed-form expression for quantiles of the F-distribution. I don't know of any method not involving numeric integration. $\endgroup$ – Michael M Dec 2 '13 at 14:59
  • 1
    $\begingroup$ Yes. As for the normal distribution, probabilities and quantiles can't be found by plugging values in some formula. $\endgroup$ – Michael M Dec 2 '13 at 21:01
  • 1
    $\begingroup$ You can use a statistical package like 'R' (function qf) or relay on tabulated calues as suggested by @hayöpe. An example of such table is given in google.ch/… $\endgroup$ – Michael M Dec 2 '13 at 21:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.