MATLAB: How to calculate the 95% confidence interval for spearman correlation values

ciconfidence intervalcorrelationspearman

I am trying to get the 95% CI of the spearman correlation of 2 vectors, but I can't figure out how obtain that with the function

corr(x1,x2,'Type','Spearman', 'tail', 'both');

Does anyone know any built-in fuction to calculate that?

I don't think there are formulas for that but you can estimate it with bootstrapping:

mycorr = @(x1,x2) corr(x1,x2,'Type','Spearman', 'tail', 'both')
nIterations = 10000;
[lower, upper] = bootci(nIterations,{mycorr,x1,x2})

The 'corr' function will compute p-values, but not confidence intervals.

And 'corrcoef' can compute confidence intervals, but only for Pearson correlation.

As an alternative to programming up the calculations directly,

you may want to consider using 'bootci' to evaluate confidence intervals.

Use de corrcoef function. In your case, it would be:

[R,P,RLO,RUP]= corrcoef(a, b, 'alpha', 0.05)

Where RLO and RUP give you the lower and upper limit of the confidence interval at the 95% level.