[R] cochran-grubbs tests results

[Greg Snow]

The fact that you are asking what a p-value is indicates that you do not have enough of a background for us to be able to help you.  It is not that we do not want to help, just that anything we could say in an e-mail would likely do more harm than good at this point.

You should either consult with a professional statistician, or take some stats classes to get to the level where we can give meaningful answers to your questions.


For those familiar with the fortunes package, fortune(122) applies here.

-- 
Gregory (Greg) L. Snow Ph.D.
Statistical Data Center
Intermountain Healthcare
greg.snow at imail.org
801.408.8111


> -----Original Message-----
> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-
> project.org] On Behalf Of miguelgonzalez
> Sent: Wednesday, September 15, 2010 4:41 AM
> To: r-help at r-project.org
> Subject: [R] cochran-grubbs tests results
> 
> 
> Hello,
> 
> I'm new in this R world and I don't know much about statistics, but now
> I
> have to analize some data and I've got some first queries yet:
> 
> I have 5 sets of area mesures and each set has 5 repetitions.
> My first step is to check data looking for outliers. I've used the
> outliers
> package. I have to use the cochran test and the grubbs test in case I
> find
> any outlier. The problem is that I don't know how to interpret the
> results.
> Cochran test finds an outlier among variances. Grubbs test finds an
> outlier
> within the set of measures wich corresponds to the variance that is an
> outlier. I repeat the Cochran test discarding that measure. And it
> finds a
> new outlier and so on. This is what I've done first:
> 
> ---------------------------------------------------------------------
> > varianza
> [1] 9.979937e-07 1.131360e-07 4.140418e-06 2.184604e-07 3.662077e-07
> > data1<-c(rep(4,5))
> > cochran.test(varianza, data1, inlying = FALSE)
> 
>         Cochran test for outlying variance
> 
> data:  varianza
> C = 0.7094, df = 4, k = 5, p-value = 0.007645
> alternative hypothesis: Group 3 has outlying variance
> sample estimates:
>            1            2            3            4            5
> 9.979937e-07 1.131360e-07 4.140418e-06 2.184604e-07 3.662077e-07
> -----------------------------------------------------------------------
> ---------
> 
> It would be really helpful if someone could tell me what C, df, K and
> p-value mean.
> Besides, I don't know if alternative hypothesis (in the grubbs test)
> means
> that the highest value is an outlier to be discarded or just the worst
> value, but not necessarily to be discarded.
> 
> -----------------------------------------------------------------------
> -----
> > grubbs.test(serie3, type = 10, opposite = FALSE, two.sided = FALSE)
> 
>         Grubbs test for one outlier
> 
> data:  serie3
> G = 1.3384, U = 0.2039, p-value = 0.2155
> alternative hypothesis: highest value 0.238877332 is an outlier
> -----------------------------------------------------------------------
> ------
> 
> As you see I'm quite lost.
> Thank you in advance and I'll keep trying.
> 
> Gracias
> 
> --
> View this message in context: http://r.789695.n4.nabble.com/cochran-
> grubbs-tests-results-tp2540295p2540295.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.