When and how to use the Central Limit Theorem?

472 Views Asked by At

I'm currently trying to solve an exercise in statistics.

I have got many hundred observations for some measured fluctations.

In a normal Q-Q plot, I have plotted all the observations and a line has been drawed. This is all done by software.

As I can see, the plotted observations do not really differ from a straight line, maybe only a little bit at the beginning of the line and at the end of line. Therefore, as I understand, I can assume that the data (the observations) follow a normal distribution.

However, in the exercise, I'm told to make use of the central limit theorem. But as I understand, this theorem should only be used if I can't assume a normal distribution (if the plotted observations differed from a straight line).

Did I misunderstand something?

1

There are 1 best solutions below

2
On BEST ANSWER

In simple words, the Central Limit Theorem says that the sum of a sufficiently large number of weakly dependent quantities has a distribution close to normal, regardless of their initial distribution.

So

But as I understand, this theorem should only be used if I can't assume a normal distribution (if the plotted observations differed from a straight line).

you don't need any assumptions on prior distributions, and it is really the main key point of CLT.

UPDATE (Thanks to comment by Karl): You don't need any special assumptions on prior distribution while it has well-defined finite mean $\mu$ and variance $\sigma^2$. There exist examples of such distributions which don't satisfy this requirement.