Spearman correlation

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Spearman correlation

iheb bougmiza
Dear list
  Can we calculate the spearman correlation between 4 binary variables coded in 0 and 1 rather than the chi 2 test ? Wich is better ?
  Thanks


M. Iheb BOUGMIZA

Adresse: 90, Rue de l'Ourcq, Bat B. 75019 Paris
Tél : 06 98 82 36 53
  Mailto: [hidden email]


---------------------------------
 Yahoo! Mail réinvente le mail ! Découvrez le nouveau Yahoo! Mail et son interface révolutionnaire.
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Spearman correlation

Swank, Paul R
Spearmna is not a good idea with dichotomous data. It is a rank oreder statistics and with 0,1 vars you have too many ties. Phi would be better. or Kendall's tau.

Paul R. Swank, Ph.D.
Professor, Developmental Pediatrics
Director of Research, Center for Improving the Readiness of Children for Learning and Education (C.I.R.C.L.E.)
Medical School
UT Health Science Center at Houston

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of iheb bougmiza
Sent: Thursday, June 22, 2006 9:31 AM
To: [hidden email]
Subject: Spearman correlation

Dear list
  Can we calculate the spearman correlation between 4 binary variables coded in 0 and 1 rather than the chi 2 test ? Wich is better ?
  Thanks


M. Iheb BOUGMIZA

Adresse: 90, Rue de l'Ourcq, Bat B. 75019 Paris Tél : 06 98 82 36 53
  Mailto: [hidden email]


---------------------------------
 Yahoo! Mail réinvente le mail ! Découvrez le nouveau Yahoo! Mail et son interface révolutionnaire.
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Spearman correlation

Hector Maletta
In reply to this post by iheb bougmiza
Rather than the Spearmann (ordinal) correlation, it is more appropriate to
apply the Pearson (linear) correlation, which is itself closely related to
chi square. The Pearson correlation for two binary variables is equivalent
to the phi coefficient, which is itself defined as chi square divided as the
square root of (X2/n) where X2 stands for chi square, and n is sample size.
Notice also that the percentage difference for one variable between values
of the other variable equals the corresponding regression coefficients. For
instance, if X=sex (0=male, 1=female) and Y=some Yes-No question, and the
Yes percentage is 50% for men and 80% for women, the regression equation
would be Yes=0.50 + 0.30(Sex).
Hector

-----Mensaje original-----
De: SPSSX(r) Discussion [mailto:[hidden email]] En nombre de iheb
bougmiza
Enviado el: Thursday, June 22, 2006 11:31 AM
Para: [hidden email]
Asunto: Spearman correlation

Dear list
  Can we calculate the spearman correlation between 4 binary variables coded
in 0 and 1 rather than the chi 2 test ? Wich is better ?
  Thanks


M. Iheb BOUGMIZA

Adresse: 90, Rue de l'Ourcq, Bat B. 75019 Paris
Tél : 06 98 82 36 53
  Mailto: [hidden email]


---------------------------------
 Yahoo! Mail réinvente le mail ! Découvrez le nouveau Yahoo! Mail et son
interface révolutionnaire.
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Spearman correlation

Spousta Jan
In reply to this post by iheb bougmiza
HI,

For binary variables, Spearmann and Pearson correlation are always equal, so it is up to you what to use.

Greetings

Jan

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Hector Maletta
Sent: Thursday, June 22, 2006 5:56 PM
To: [hidden email]
Subject: Re: Spearman correlation

Rather than the Spearmann (ordinal) correlation, it is more appropriate to apply the Pearson (linear) correlation, which is itself closely related to chi square. The Pearson correlation for two binary variables is equivalent to the phi coefficient, which is itself defined as chi square divided as the square root of (X2/n) where X2 stands for chi square, and n is sample size.
Notice also that the percentage difference for one variable between values of the other variable equals the corresponding regression coefficients. For instance, if X=sex (0=male, 1=female) and Y=some Yes-No question, and the Yes percentage is 50% for men and 80% for women, the regression equation would be Yes=0.50 + 0.30(Sex).
Hector

-----Mensaje original-----
De: SPSSX(r) Discussion [mailto:[hidden email]] En nombre de iheb bougmiza Enviado el: Thursday, June 22, 2006 11:31 AM
Para: [hidden email]
Asunto: Spearman correlation

Dear list
  Can we calculate the spearman correlation between 4 binary variables coded in 0 and 1 rather than the chi 2 test ? Wich is better ?
  Thanks


M. Iheb BOUGMIZA

Adresse: 90, Rue de l'Ourcq, Bat B. 75019 Paris Tél : 06 98 82 36 53
  Mailto: [hidden email]


---------------------------------
 Yahoo! Mail réinvente le mail ! Découvrez le nouveau Yahoo! Mail et son interface révolutionnaire.
Loading...