Z-test for Correlation

From GM-RKB
Jump to navigation Jump to search

A Z-test for Correlation is a parametric correlational hypothesis test of continuous variables that is based on a z-test statistic.

  • Context:
    • It can be described and solved by the following procedure:
Test Requirements:
Hypotheses to be tested:
Test Method and Sample Data Analysis:
[math]\displaystyle{ Z_{r,n-1} = \frac{\sqrt{n-3}}{2} [ln (\frac{1+r}{1−r})−ln(\frac{1+\rho_0}{1−\rho_0})] }[/math], with [math]\displaystyle{ r }[/math] being the sample correlation coefficient.
Results and Interpretation:


References

2017

[math]\displaystyle{ W=\frac{1}{2}ln\frac{1+R}{1−R} }[/math]
follows an approximate normal distribution with mean [math]\displaystyle{ E(W)=\frac{1}{2}ln\frac{1+\rho}{1−\rho} }[/math] and variance [math]\displaystyle{ Var(W)=\frac{1}{n−3} }[/math].
The theorem, therefore, allows us to test the general null hypothesis H0:ρ=ρ0H0:ρ=ρ0 against any of the possible alternative hypotheses comparing the test statistic:
[math]\displaystyle{ Z=\frac{\frac{1}{2}ln\frac{1+R}{1−R}−\frac{1}{2}ln\frac{1+\rho_0}{1−ρ0}}{\sqrt{\frac{1}{n−3}}} }[/math]

to a standard normal N(0,1) distribution.