Anderson Darling Normality - Sense Check Please?

**sweep** · 06-22-2011, 03:42 AM

Folks,

Could someone "stats savvy" please road-test the attached code for me?

This function should return a 1x4 array of S, AD, AD*, and p for an AD normality test.

Useage is =ADNormailty(Range)

any thanks,

Dave

**whmcneese** · 06-28-2011, 12:00 PM

Hello Dave,

I compared the results of your function for five data sets using the function contained in the SPC for Excel software (I am the developer of that). Results are identical so it looks like a good function to me.

Bill

**sweep** · 06-29-2011, 11:30 AM

Thanks Bill.

I have had similar agreement - I'll implement it.

**flying007** · 09-20-2011, 04:18 AM

Dear, hello !
Thank you for your share!
And i have some trouble in this code, So can you help me?
I used the formula select the 10 data :
57
67
49
64
73
91
66
87
74
58
but the result output only a value: -102.384, what's wrong happen ?
===
I'm sorry for my careless post .

**zbor** · 09-20-2011, 04:23 AM

wellcome to the forum. .. Please check rules:

2. Don't post a question in the thread of another member -- start your own. If you feel it's particularly relevant, provide a link to the other thread.

**freakok** · 04-25-2012, 06:10 AM

Thank you, thank you very much!

**ddter** · 06-30-2013, 07:41 AM

First of all, thanks for this great function.

One thing however, this function those not take in account the possibility for empty cells.
Therefore I have adapted the code - see attachment

ADNormality.txt

thanks

**Stefan1234** · 03-29-2016, 06:39 AM

In the below mentioned website, the Anderson Darling test is explained.
Can anybody tell me why there is a correction in the calculation of the Z-value (see column L of sheet 2 in the embedded excel-sheet of the below mentioned website).

The P value is not calculated as i/n. But corrected and is now calculated as (i-0,3)/(n+0.4)
Can anybody give some substantation of the used 0.3 and 0.4. The N is maximised at n=200.

https://www.spcforexcel.com/knowledg...-for-normality

**Stefan1234** · 03-29-2016, 06:41 AM

In the below mentioned website, the Anderson Darling test is explained.
Can anybody tell me why there is a correction in the calculation of the Z-value (see column L of sheet 2 in the embedded excel-sheet of the below mentioned website).

The P value is not calculated as i/n. But corrected and is now calculated as (i-0,3)/(n+0.4)
Can anybody give some substantation of the used 0.3 and 0.4. The N is maximised at n=200.

https://www.spcforexcel.com/knowledg...-for-normality

**Stefan1234** · 03-29-2016, 06:41 AM

In the below mentioned website, the Anderson Darling test is explained.
Can anybody tell me why there is a correction in the calculation of the Z-value (see column L of sheet 2 in the embedded excel-sheet of the below mentioned website).

The P value is not calculated as i/n. But corrected and is now calculated as (i-0,3)/(n+0.4)
Can anybody give some substantation of the used 0.3 and 0.4. The N is maximised at n=200.

https://www.spcforexcel.com/knowledg...-for-normality

**zbor** · 03-29-2016, 07:03 AM

Unfortunately your post does not comply with Rule 2 of our Forum RULES. Do not post a question in the thread of another member -- start your own thread.

If you feel an existing thread is particularly relevant to your need, provide a link to the other thread in your new thread.

Old threads are often only monitored by the original participants. New threads not only open you up to all possible participants again, they typically get faster response, too.