Hi!
I know that you advise to speak a good english, but thats not my mother tongue, so I hope you apologize any mistakes from me.
I was trying to see check how random Excel is generating numbers, because I read that Excel is not a real random generator. Very recently someone showed me a paper speaking about Excel & randomness, but I still want to test it. So, my problem here is not really to discover about Excel itself but more about the maths and how can I find it by myself.
So, to test the randomness of Excel, I generated lots of numbers so I can check if there is any "anomaly" from what should be expected by chance. Here is what I did:
- I have lots of cells, near 695505, each one generating a number from 1 to 4.
- I counted how many numbers 1 were generated, and numbers 2, 3 and 4. For example, from the near 700000 random numbers, 25,07% are numbers 1 , 24,94% are the number 2 , 24,98% are number 3 and 25,01% are the number 4 . I also calculated the SD (standard deviation): 0,047 and then how many standard deviations each result was from the mean (the mean is always exactly 25%): for number 1, -1,47 SDs away from the mean; 1,26 for number 2; 0,44 for number 3 and -0,23 for number 4.
What should I do next, and how? How can I read and interpret the data so I can check if there is any anomaly? I mean, we should expect by chance that all the numbers were 25% (exactly25,0000...) from the total. Should I calculate the magnitude size effect? An how would I do that in this case?
Thanks for any insights about this
I know that you advise to speak a good english, but thats not my mother tongue, so I hope you apologize any mistakes from me.
I was trying to see check how random Excel is generating numbers, because I read that Excel is not a real random generator. Very recently someone showed me a paper speaking about Excel & randomness, but I still want to test it. So, my problem here is not really to discover about Excel itself but more about the maths and how can I find it by myself.
So, to test the randomness of Excel, I generated lots of numbers so I can check if there is any "anomaly" from what should be expected by chance. Here is what I did:
- I have lots of cells, near 695505, each one generating a number from 1 to 4.
- I counted how many numbers 1 were generated, and numbers 2, 3 and 4. For example, from the near 700000 random numbers, 25,07% are numbers 1 , 24,94% are the number 2 , 24,98% are number 3 and 25,01% are the number 4 . I also calculated the SD (standard deviation): 0,047 and then how many standard deviations each result was from the mean (the mean is always exactly 25%): for number 1, -1,47 SDs away from the mean; 1,26 for number 2; 0,44 for number 3 and -0,23 for number 4.
What should I do next, and how? How can I read and interpret the data so I can check if there is any anomaly? I mean, we should expect by chance that all the numbers were 25% (exactly25,0000...) from the total. Should I calculate the magnitude size effect? An how would I do that in this case?
Thanks for any insights about this