Central Limit Theorem

#1
Hello,

I need help with this. Calculate the weekly return for BIT and construct a histogram. Does the data on return rates appear
normally distributed?

The data I have used is from a sample with a sample size of 335. So, even though it looks right-skewed, can CLT be applied? Weekly Return Rate of BIT.png Weekly Return Rate of BIT.png

Weekly Return Rate of BIT.png
 

obh

Active Member
#2
Hi Gkhidhir,

The CLT is related to the distribution of the sum of distributions (or average)
So what do you try to do with the data?

Since there are more than 30 observations and the data seems only slightly skewed you can assume the sum/average distribute normally.

You can check the normality and skewed level using the:http://www.statskingdom.com/320ShapiroWilk.html
But this is going back to the question of the use, for example, I wouldn't run a normality check if I want to compare averages of such a big slightly skewed data.
 

obh

Active Member
#4
Okay back to your originally question

The data I have used is from a sample with a sample size of 335. So, even though it looks right-skewed, can CLT be applied?

z-scores do you find evidence of outliers
I agreed, without any test, (no need in this case) it looks right-skewed and not normally distributed. (just for the fun, you can still use the link I send you, I'm sure it will approve it)

yes, CLT can be applied for sum/average.

Try to calculate the outliers (I'm sure you will find some outliers)