On 03/09/2008 10:33 AM, Williams, Robin wrote:
Hi,
I am looking for a normality test in R to see if a vector of data I have
can be assumed to be normally distributed and hence used in a linear
regression.
Raw data that is suitable for standard linear regression is normally
distributed, but the mean varies from observation to observation. The
necessary assumption is that the errors are normally distributed with
zero mean, but the data itself also includes the non-random parts of the
model. The effect of the varying means is that the data will generally
*not* appear to come from a normal distribution if you just throw it all
into a vector and look at it.
So let's assume you're working with residuals from a linear fit. The
residuals should be normally distributed with mean zero, but their
variances won't be equal. It may be that in a large dataset this will
be enough to get a false declaration of non-normality even with
perfectly normal errors. In a small dataset you'll rarely have enough
power to detect non-normality.
So overall, don't use something like shapiro.test for what you have in
mind. Any recent regression text should give advice on model
diagnostics that will do a better job.
help.search("normality test")
suggests the Shapiro test, ?shapiro.test.
Now maybe I am interpreting things incorrectly (as is usually the case),
am I right in assuming that this is a composite test for normality, and
hence a high p-value would suggest that the sample is normally
distributed?
A low p-value (e.g. p < 0.05) could suggest there is evidence of
non-normality, but p > 0.05 just shows a lack of evidence. In the case
where the data is truly normally distributed, you'd expect p to be
uniformly distributed between 0 and 1. (I have an article in the
current American Statistician suggesting ways to teach p-values to
emphasize this; unfortunately, it seems to be a surprise to a lot of
people.)
Duncan Murdoch
As a test I did
shapiro.test(rnorm(4500))
a few times, and achieved very different p-values, so I cannot be sure.
I had assumed that a random sample of 4500 would have a very high
p-value on all occasions but it appears not, this is interesting.
Are there any other tests that people would recommend over this one in
the base packages? I assume not as help.search did not suggest any.
So am I right about a high p-value suggesting normality?
Many thanks for any help.
Robin Williams
Met Office summer intern - Health Forecasting
[EMAIL PROTECTED]
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
