Hi folks, I have some data where I have to investigate with a suitable QQ-plot wheter or not the measurement erros are normally distributed. The distributed data are form the residuals of a lineair regression, that can be considered as aproximations of the actuel measurement errors. This is the graph: http://r.789695.n4.nabble.com/file/n4633819/QQplot.jpg How can I recognize, with reference to the graph, if the errors are normally distributed or not. And with that conclusion in mind, is there some relationship between those variables? Thank You Noor -- View this message in context: http://r.789695.n4.nabble.com/QQplot-normally-distributed-tp4633819.html Sent from the R help mailing list archive at Nabble.com.
Hi, Following a straight line indicates less evidence towards non-normality. But QQ-Plot is an exploratory tool. You can confirm your ideas obtained from the QQ-Plot via noramlity tests such as Shapiro-Wilk test. See shapiro.test under stats package and nortest package. Ozgur -- View this message in context: http://r.789695.n4.nabble.com/QQplot-normally-distributed-tp4633819p4633820.html Sent from the R help mailing list archive at Nabble.com.
Hi, But what are the functions of the outliers on the left and right? Does they influence the normal distribution? -- View this message in context: http://r.789695.n4.nabble.com/QQplot-normally-distributed-tp4633819p4633823.html Sent from the R help mailing list archive at Nabble.com.
Hi, Try boxplot for outliers. To decide whether they influence significantly, try confirmatory normality tests. Ozgur -- View this message in context: http://r.789695.n4.nabble.com/QQplot-normally-distributed-tp4633819p4633830.html Sent from the R help mailing list archive at Nabble.com.
Or uou can try library(car) ?qqPlot use that with argument simulate=TRUE, which will give a simulated envelope around the curve for comparison. Kjetil On Tue, Jun 19, 2012 at 9:30 AM, ?zg?r Asar <oasar at metu.edu.tr> wrote:> Hi, > > Try boxplot for outliers. > > To decide whether they influence significantly, try confirmatory normality > tests. > > Ozgur > > -- > View this message in context: http://r.789695.n4.nabble.com/QQplot-normally-distributed-tp4633819p4633830.html > Sent from the R help mailing list archive at Nabble.com. > > ______________________________________________ > R-help at 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.
Dear Kjetil, Simulated point-wise confidence envelopes are available from qqPlot() only for studentized residuals from linear and generalized linear models. For an independent sample of observations, the confidence envelopes produced by qqPlot() are based on the standard errors of the order statistics for the reference distribution. Best, John ------------------------------------------------ John Fox Sen. William McMaster Prof. of Social Statistics Department of Sociology McMaster University Hamilton, Ontario, Canada http://socserv.mcmaster.ca/jfox/ On Tue, 19 Jun 2012 15:21:07 -0400 Kjetil Halvorsen <kjetilbrinchmannhalvorsen at gmail.com> wrote:> Or uou can try > library(car) > > ?qqPlot > > use that with argument simulate=TRUE, which will give a simulated > envelope around the curve for > comparison. > > Kjetil > > On Tue, Jun 19, 2012 at 9:30 AM, ?zg?r Asar <oasar at metu.edu.tr> wrote: > > Hi, > > > > Try boxplot for outliers. > > > > To decide whether they influence significantly, try confirmatory normality > > tests. > > > > Ozgur > > > > -- > > View this message in context: http://r.789695.n4.nabble.com/QQplot-normally-distributed-tp4633819p4633830.html > > Sent from the R help mailing list archive at Nabble.com. > > > > ______________________________________________ > > R-help at 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 at 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.
Hi, So in my example, I can say that the data comes from a moderate normal distribution because the points more at the right lay straight to a straight line, then the points at the left. Please a confirmation here. But what is the information above (that the data is from a normal distribution) say about the relationship between two variables? -- View this message in context: http://r.789695.n4.nabble.com/QQplot-normally-distributed-tp4633819p4633918.html Sent from the R help mailing list archive at Nabble.com.
>Hi,>So in my example, I can say that the data comes from a moderate normaldistribution because the points more at the >right lay straight to a straight line, then the points at the left. Please a confirmation here.>But what is the information above (that the data is from a normaldistribution) say about the relationship between two >variables? Dear Noor, Which two variables? Ozgur -- View this message in context: http://r.789695.n4.nabble.com/QQplot-normally-distributed-tp4633819p4633920.html Sent from the R help mailing list archive at Nabble.com.
Hi
See the file http://www.mijnbestand.nl/Bestand-6ZPTBYDLBZQI.txt here .
That file contains the results of a study on breathing resistance in
children with asthma and children with cystic fibrosis to investigate wheter
there is a relationship between breathing resistance and length in each of
the two groups. Consider the cystic fibrosis group.
See the
http://www.mijnbestand.nl/Bestand-G68PRQKUB4QX.txthttp://www.mijnbestand.nl/Bestand-G68PRQKUB4QX.txt
R-code for what I have now
I already have respectivly: a) QQplot with best fit line of resistance
against length, b) investigation linear relationship with a t-test, linear
regression line (see QQplot at part A), c) testing with a different test
statistic, and than the two question I find hard:
To perform the tests in parts b and e, it was assumed that the measurement
errors are normally
distributed. Test with a suitable QQ-plot whether or not this assumption
is justified.
Recall that the residuals of a linear regression can be considered as
approximations of the actual
measurement errors.
And: What is, based on all your results, your final conclusion about the
relationship between breathing
resistance and length?
I think it has to do something with the Central Limit Theorem?
--
View this message in context:
http://r.789695.n4.nabble.com/QQplot-normally-distributed-tp4633819p4633927.html
Sent from the R help mailing list archive at Nabble.com.