How to get p-val from correlation coefficient and number of sample?
39 vues (au cours des 30 derniers jours)
Afficher commentaires plus anciens
frewise
le 7 Nov 2011
Réponse apportée : Gregory Pelletier
le 24 Jan 2024
Hi, all!
We know that in MATLAB, [r,p]=corrcoef(A) could return the correlation coefficient and p-val.
But who knows how p-val is calculated, what is the formulas?
Thanks for your help!
Réponse acceptée
Wayne King
le 7 Nov 2011
The p-value is based on a t statistic
t = r* sqrt((n-2)/(1-r^2))
which has n-2 degrees of freedom.
You can find a discussion of the sampling distribution of r both when the null hypothesis is that r=0 and when you are testing for nonzero values of r here:
Plus de réponses (2)
Wayne King
le 7 Nov 2011
??? I told you that the the t above is t-distributed with n-2 degrees of freedoms. Once you know the distribution, the p-value is the probability that you would observe a value at least that extreme in that distribution.
For example, if you had a t-value of 3 and you had 20 samples, then
1-tcdf(3,18)
gives you the probability in the upper tail. Since the t-distribution is symmetric. If you are doing a two-sample hypothesis test, you have to double that probability to get the p-value (you have equal probability in the lower tail)
For example:
rng default
x = randn(18,2);
[r,p] = corrcoef(x);
gives
r = -0.0242
t = r*sqrt(16/(1-r^2));
2*tcdf(t,16)
which is exactly what MATLAB gives in the p matrix.
Gregory Pelletier
le 24 Jan 2024
Here is how matlab calculates the p-value in corrcoef (see p_check below for the manual calculation of the p-value compared with p from corrcoef):
load hospital
X = [hospital.Weight hospital.BloodPressure];
[R, p] = corrcoef(X)
N = size(X,1);
t = sqrt(N-2).*R./sqrt(1-R.^2);
s = tcdf(t,N-2);
p_check = 2 * min(s,1-s)
% R =
% 1.0000e+00 1.5579e-01 2.2269e-01
% 1.5579e-01 1.0000e+00 5.1184e-01
% 2.2269e-01 5.1184e-01 1.0000e+00
% p =
% 1.0000e+00 1.2168e-01 2.5953e-02
% 1.2168e-01 1.0000e+00 5.2460e-08
% 2.5953e-02 5.2460e-08 1.0000e+00
% p_check =
% 0 1.2168e-01 2.5953e-02
% 1.2168e-01 0 5.2460e-08
% 2.5953e-02 5.2460e-08 0
0 commentaires
Voir également
Catégories
En savoir plus sur Hypothesis Tests dans Help Center et File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!