# [R] How to interpret Spearman Correlation

Frank Harrell f.harrell at vanderbilt.edu
Wed Nov 2 13:23:19 CET 2011

```What David was getting at is that you interpreted the P-value as one minus
the P-value, not a safe practice.  There is also some question about whether
it would have been better to recommend a good book or course.
Frank

Raphael Saldanha wrote:
>
> Hi David,
>
> This is not private tutoring, just someone trying to help, and I'm sorry
> for my distraction.
>
> On Tue, Nov 1, 2011 at 10:34 PM, David Winsemius <dwinsemius@>wrote:
>
>> Shahab;
>>
>> You would be well advised not to seek private tutoring from someone on
>> the
>> Internet who tells you that a p-value of 0.008736 is "not significant".
>>
>>
>>
>> On Nov 1, 2011, at 8:09 PM, Raphael Saldanha <saldanha.plangeo@>
>> wrote:
>>
>> > Hi Shahab,
>> >
>> > This test shows that there is some positive statistical correlation,
>> BUT
>> > the p-value of the test - this is, the level of significance - shows
>> that
>> > the correlation is not statistically significant at 95% confidence
>> level.
>> > So, the correlation may be equal to zero.
>> >
>> > To understand this concepts in a good way, you need to be secure about
>> > variance and hypothesis test.
>> >
>> > I can help you more if you need. Send me a direct mail (this list is
>> for
>> > with Statistics.
>> >
>> > My e-mail: saldanha.plangeo@
>> >
>> > On Tue, Nov 1, 2011 at 8:58 PM, shahab <shahab.mokari@> wrote:
>> >
>> >> Hi,
>> >>
>> >> I am not really familiar with Correlation foundations, although I read
>> >> a lot. So maybe if someone kindly help me to interpret the following
>> >> results.
>> >> I had the following R commands:
>> >>
>> >> correlation <-cor( vector_CitationProximity , vector_Impact, method =
>> >> "spearman", use="na.or.complete")
>> >> cor_test<-cor.test(vector_CitationProximity, vector_Impact,
>> >> method="spearman")
>> >>
>> >> and the results are:
>> >> "correlation"
>> >> Correlation =  0.04715686
>> >>
>> >> "cor_test"
>> >> Spearman's rank correlation rho
>> >>
>> >> data:  vector_CitationProximity and vector_Impact
>> >> S = 5581032104, p-value = 0.008736
>> >> alternative hypothesis: true rho is not equal to 0
>> >> sample estimates:
>> >>      rho
>> >> 0.04582115
>> >>
>> >>
>> >> So apparently, there is positive correlation between two given
>> >> variables since Correlation =  0.04715686  > 0
>> >> However I couldn't interpret the significance ?' what does "rho" say?
>> >> Is there any simple sample that I can read and try to understand? I am
>> >> do confused in understanding how significance can be interpreted.
>> >>
>> >> Thanks,
>> >>
>> >> /Shahab
>> >>
>> >> ______________________________________________
>> >> R-help@ mailing list
>> >> https://stat.ethz.ch/mailman/listinfo/r-help
>> >> http://www.R-project.org/posting-guide.html
>> >> and provide commented, minimal, self-contained, reproducible code.
>> >>
>> >
>> >
>> >
>> > --
>> > Atenciosamente,
>> >
>> > Raphael Saldanha
>> > saldanha.plangeo@
>> >
>> >    [[alternative HTML version deleted]]
>> >
>> > ______________________________________________
>> > R-help@ mailing list
>> > https://stat.ethz.ch/mailman/listinfo/r-help
>> http://www.R-project.org/posting-guide.html
>> > and provide commented, minimal, self-contained, reproducible code.
>>
>
>
>
> --
> Atenciosamente,
>
> Raphael Saldanha
> saldanha.plangeo@
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help@ mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

-----
Frank Harrell
Department of Biostatistics, Vanderbilt University
--
View this message in context: http://r.789695.n4.nabble.com/How-to-interpret-Spearman-Correlation-tp3965809p3972797.html
Sent from the R help mailing list archive at Nabble.com.

```