Your data analysis solution
Name is required.
Email address is required.
Invalid email address
Answer is required.
Exceeding max length of 5KB

How to handle non-normality of data AFTER transformation

By Norman | Sep 01, 2017 09:02PM CEST

I have a data set that I have transformed by sq root and log10 transformations and the data set does still not follow a normal distribution, and the ANOVA standardized residuals are still not following a normal distribution by Shapiro -wilks and other tests.
Any suggestions what to do to normalize this ?
Accept the result and note this in the results section ?

Best Answer
By Sébastien | Sep 04, 2017 09:30AM CEST | XLSTAT Agent

Hi Norman,

Have you tried considering the removal of outliers (if this is justified, of course)?
How many observations do you have? Sometimes when you have too many, normality tests become way too powerful, and a graphical verification of the assumption (QQ plot) can be enough, without a confirmation using the normality test.
We hope you’re enjoying your XLSTAT trial. Let us know if you need any other information :)

Jean Paul


Up

0

Down

By Sébastien | Sep 04, 2017 09:30AM CEST | XLSTAT Agent

Hi Norman,

Have you tried considering the removal of outliers (if this is justified, of course)?
How many observations do you have? Sometimes when you have too many, normality tests become way too powerful, and a graphical verification of the assumption (QQ plot) can be enough, without a confirmation using the normality test.
We hope you’re enjoying your XLSTAT trial. Let us know if you need any other information :)

Jean Paul

Post Your Public Answer

Contact Us

1c26995d494fb3061dd0ae8571ffc0a4@xlstat.desk-mail.com
https://cdn.desk.com/
false
desk
Loading
seconds ago
a minute ago
minutes ago
an hour ago
hours ago
a day ago
days ago
about
false
Invalid characters found
/customer/portal/articles/autocomplete
9283