Normal Probability Plot Question
Six Sigma – iSixSigma › Forums › Old Forums › General › Normal Probability Plot Question
 This topic has 4 replies, 4 voices, and was last updated 16 years, 6 months ago by mZ.

AuthorPosts

November 10, 2005 at 9:33 pm #41371
I ran a normal prob plot on some data today. It’s continuous data and it is several weeks worth of data points.
I have a pvalue of 0.032 which i believe indicates that the data is normal.
I am new however and I hate seeing an output with values on it that i dont understand. What does the Asquared value of 0.757 represent? The thing is, its directly above the pvalue and if i show somebody this chart and they ask me what asquared means..and i dont have a good explanation then the credibilty of the entire plot could be in question.
Also, I dont understand what the probablilty scale is telling me. The line is running up and to the right. I know that the dots are N and that they tell a story but what is the line telling me?
I’m very new at this and i just want to really understand what the tools are telling me.
thanks
geekus0November 10, 2005 at 9:35 pm #129580AndersonDarling…..forgot to mention.
0November 10, 2005 at 10:05 pm #129583Pvalue = 0.05
Asquared is the test statistic for the AndersonDarling Normality test. It is a measure of how closely a dataset follows the normal distribution. The null hypothesis for this test is that the data is normal. So if you get an Asquared that is fairly large, then you will get a small pvalue and thus reject the null hypothesis. Small Asquared values imply large pvalues, thus you cannot reject the null hypothesis.
There is a ton of information on this site, and little investigation will go a long ways, just enter a keyword search and do some reading . The answers your looking for are here
CT0November 10, 2005 at 10:06 pm #129584
jimmie65Participant@jimmie65 Include @jimmie65 in your post and this person will
be notified via email.Try the help menu in Minitab. Open the dialog box for probabilty plot and clock the Help button, then search through that entry. You’ll probably find some interpretation information under Example.
If it’s not there, type your term in the search box.0November 10, 2005 at 10:33 pm #129585CT:
normal = Pvalue >= 0.05? Is this right?
If the test is against a theoretical Normal distribution, as the pvalue getting larger, it means your data is getting closer to a normal distribution. When the Asq is close to 1, then you can say your data is distriuted normally.
My two cents.
mz0 
AuthorPosts
The forum ‘General’ is closed to new topics and replies.