Tuesday, May 21, 2013

Statistics 101: Central Limit Theorem Simulation in Excel


Understanding Central Limit Theorem is of EXTREME importance in statistics. This concept usually sets the boundary line between people who understand statistics and people who don’t.

What is confusing about this topic is usually terminology – mean of sampling distribution of the mean, standard deviation of the sampling distribution of the mean – you get the point. What I wanted to do is to make this thing easier to comprehend by using Will Hopkins approach by using simulations. Will Hopkins created great and must read series of spreadsheet designed to help out lost souls (like me) comprehend basic statistic concepts. You can find those spreadsheets and description HERE.

What was lacking in Hopkins spreadsheets was simulation of central limit theorem. Thus I have created this spreadsheet using his famous template. I suggest watching lectures by Khan Academy and then playing with the spreadsheet. Modifying various numbers and seeing how that affects the output might give you deeper understanding of this fundamental concept behind all inferential statistics. I believe I still lack some of the nuts and bolts to get, so I might listen to my own advice and check the Khan videos. 






 If you find any errors please be free to notify me.
Click HERE to download the Excel spreadsheet. 


2 comments:

  1. Great work as normal Mladen. I just have a quick question. How did you add the link to automatically download the file. This would really help me out for my own blog.

    Thanks

    John

    ReplyDelete
  2. If you use Dropbox John, put the file in Shared folder. Once you did this right-click and choose copy the link or something like that.

    Thanks for the feedback John. BTW, check the graph on tab 3 - I was referring to something like this couple of day ago. The problem in Excel is that you cannot make this horizontal with band lines which might represent SWC (at least I don't know how to do it :) ). Need to learn ggplot in R :)

    ReplyDelete