Hello! Thank you so much for your detailed explanation. I have been at this for hours, and I can't figure out why about halfway through the data on my F(Yn+1-i) column turns into the #N/A error message, and no matter what I do it won't change. I have a n=1046. Right at mark 525 it gives me the error sign. Any thoughts?
Hi, Ramdular. Thanks for watching my youtube videos. Sooner or later, I will try to upload Ryan Joiner test in Excel after studying it. Many thanks, Sangwoo.
Hi! I have a question. Since there are ties in the data, shouldn't the rank (i) contain ties as well? And is the Anderson-Darling Test a good normality test for data with ties?
Hi, Kris. I hope you are well and thank you for your good question. In terms of " And is the Anderson-Darling Test a good normality test for data with ties?", please refer to the following site : variation.com/wp-content/distribution_analyzer_help/hs140.htm It says " The Anderson-Darling test is severely affected by ties in the data due to poor precision. In terms of "tied data", as far as I know, the tied data are always an issue for normality test. Please run the normality test by ad.test for the following data(80 samples), then you will get this output. > ad.test(test_data_1$X1) Anderson-Darling normality test data: test_data_1$X1 A = 0.26166, p-value = 0.6973 It means that this data satisfies normality. However, if you generate tied data (let's say 3 times) for this data, you will get this output > ad.test(test_data_3$X1) Anderson-Darling normality test data: test_data_3$X1 A = 0.7993, p-value = 0.03793 It means that this data does not satisfy normality, although it looks like normal distribution. Please check it from histogram) Based on my statistical inference knowledge, I recommend to add small randomly generated number (e.g. rnorm(240) ) with original data in order not to have tied data. (* Assume that Z and X1 are independent. If Z follows normal distribution and Z + X1 follows normal distribution, then X1 follows normal distribution) Then you can have this output. It means that it follows normal distribution. > test_data_3_1 ad.test(test_data_3_1$X1) Anderson-Darling normality test data: test_data_3_1$X1 A = 0.43283, p-value = 0.3011 [Original data, 80 samples] 5.5 5.7 5.8 5.8 5.9 6.0 6.1 6.1 6.3 6.3 6.4 6.4 6.4 6.5 6.5 6.7 6.7 6.7 6.7 6.7 6.8 6.8 6.8 6.8 6.8 6.8 6.8 7.0 7.0 7.0 7.0 7.1 7.1 7.1 7.1 7.1 7.1 7.2 7.2 7.2 7.2 7.2 7.2 7.3 7.3 7.3 7.3 7.3 7.5 7.5 7.5 7.5 7.5 7.5 7.7 7.7 7.7 7.7 7.7 7.7 7.8 7.8 7.8 7.8 7.8 8.0 8.0 8.0 8.1 8.1 8.1 8.3 8.3 8.4 8.4 8.5 8.5 8.6 8.7 8.8
Very good explanation. Thank you. One question. Why wouldn't the value in cell D51 be 1.00, instead of .98....? It is a cumulative function and I thought the final percentage in a cumulative function is always 100% It seems important because when the logarithm of (1-1) is calculated it would be an error. LN(0)= "error"
Hi, Tom. Thanks for your question. As far as I know, the main reason of not having exact 1 is that it is calculated from sample data from normal distribution, which has a mean and variance. Again, if we calculate the CDF from population, it will be converged to 1. Many thanks, Sangwoo.
![fellow fellow - พรุ่งนี้ไม่มีใครรู้ feat. INK WARUNTORN [OFFICIAL MV]](http://i.ytimg.com/vi/QP6JyjYx_W0/mqdefault.jpg)
Thank you very much. The VDO is really clear and explains everything very well.
Hi. Thanks for your positive feedback.
Let's keep in touch.
Hello! Thank you so much for your detailed explanation. I have been at this for hours, and I can't figure out why about halfway through the data on my F(Yn+1-i)
column turns into the #N/A error message, and no matter what I do it won't change. I have a n=1046. Right at mark 525 it gives me the error sign. Any thoughts?
Hi, Luisa. Thanks for watching my TH-cam video. Could you share your working file via my Gmail to have a look?
Very good Explanation Thank you
One request you to make video on how to calculate Ryan Joiner test in Excel sheet
Hi, Ramdular.
Thanks for watching my youtube videos.
Sooner or later, I will try to upload Ryan Joiner test in Excel after studying it.
Many thanks, Sangwoo.
Omg great explained thank you very much sir!
THANKS
Hi! I have a question. Since there are ties in the data, shouldn't the rank (i) contain ties as well? And is the Anderson-Darling Test a good normality test for data with ties?
Hi, Kris. I hope you are well and thank you for your good question.
In terms of " And is the Anderson-Darling Test a good normality test for data with ties?",
please refer to the following site : variation.com/wp-content/distribution_analyzer_help/hs140.htm
It says " The Anderson-Darling test is severely affected by ties in the data due to poor precision.
In terms of "tied data",
as far as I know, the tied data are always an issue for normality test.
Please run the normality test by ad.test for the following data(80 samples), then you will get this output.
> ad.test(test_data_1$X1)
Anderson-Darling normality test
data: test_data_1$X1
A = 0.26166, p-value = 0.6973
It means that this data satisfies normality.
However, if you generate tied data (let's say 3 times) for this data, you will get this output
> ad.test(test_data_3$X1)
Anderson-Darling normality test
data: test_data_3$X1
A = 0.7993, p-value = 0.03793
It means that this data does not satisfy normality, although it looks like normal distribution. Please check it from histogram)
Based on my statistical inference knowledge, I recommend to add small randomly generated number (e.g. rnorm(240) ) with original data in order not to have tied data.
(* Assume that Z and X1 are independent. If Z follows normal distribution and Z + X1 follows normal distribution, then X1 follows normal distribution)
Then you can have this output.
It means that it follows normal distribution.
> test_data_3_1 ad.test(test_data_3_1$X1)
Anderson-Darling normality test
data: test_data_3_1$X1
A = 0.43283, p-value = 0.3011
[Original data, 80 samples]
5.5
5.7
5.8
5.8
5.9
6.0
6.1
6.1
6.3
6.3
6.4
6.4
6.4
6.5
6.5
6.7
6.7
6.7
6.7
6.7
6.8
6.8
6.8
6.8
6.8
6.8
6.8
7.0
7.0
7.0
7.0
7.1
7.1
7.1
7.1
7.1
7.1
7.2
7.2
7.2
7.2
7.2
7.2
7.3
7.3
7.3
7.3
7.3
7.5
7.5
7.5
7.5
7.5
7.5
7.7
7.7
7.7
7.7
7.7
7.7
7.8
7.8
7.8
7.8
7.8
8.0
8.0
8.0
8.1
8.1
8.1
8.3
8.3
8.4
8.4
8.5
8.5
8.6
8.7
8.8
Very good explanation. Thank you.
One question.
Why wouldn't the value in cell D51 be 1.00, instead of .98....? It is a cumulative function and I thought the final percentage in a cumulative function is always 100%
It seems important because when the logarithm of (1-1) is calculated it would be an error. LN(0)= "error"
Hi, Tom. Thanks for your question. As far as I know, the main reason of not having exact 1 is that it is calculated from sample data from normal distribution, which has a mean and variance. Again, if we calculate the CDF from population, it will be converged to 1. Many thanks, Sangwoo.
thank you, very useful.
Thank you!!
What if my AD is 20?
Hi, Muffin Lover.
Could you share more about your issues? I couldn't fully get your issues in terms of "my AD is 20"?
Cheers, Sangwoo.
thank you, very useful.
Thanks for subscribing my youtube channel. Let's keep in touch. :)