UPDATE: This is no longer the problem – the real problem is outlined here.

No use trying to hide it: I need to look everything up when doing simple statistics.

Perhaps you can help me with a simple question?

Here goes:

Let’s say you are convinced that Mondays, Tuesdays and Wednesdays differ as to the amount of rainfall. To test this, you measure the amount of rainfall (in millimeters) on these three days over a period of six weeks.

You get (averages):

Monday: 10mm

Tuesday: 30mm

Wednesday: 15mm

I suppose the null hypothesis must be that the rainfall did not differ.

How do you express the strength of the difference in these three cases? That is: How certain are you that the measured difference is non-coincidental?

Help would be appreciated!

Comments are closed.