Spearmans Rank Help

Printable View

March 16th 2011, 10:22 AM
Teague

Spearmans Rank Help

Hey, I think I'm posting in the right section of the forum. I'm not really sure what Spearman Rank Correlation test falls under. I'm having trouble understanding it fully. I have two sets of data.

Data 1
6.3
5.3
6.0
1.7
1.7
3.2
6.3
6.0
4.9

Data 2
5.7
6.3
5.8
1.7
1.7
2.2
6.3
6.3
6.0

Obviously using these two sets of data I get the probability of 0.683333. I think thats right anyway :(

I was just wondering how you find the 0.05 value and 0.01 value?

Any help is much appreciated :) xx
March 17th 2011, 10:26 AM
Sambit

Given n pairs of observations, $(x_i,y_i)$ , the $x_i$ values are assigned a rank value and, separately, the values are $y_i$ assigned a rank. For each pair $(x_i,y_i)$ , the corresponding difference $d_i$ between the $x_i$ and $y_i$ ranks is found. The value $R$ is: $R=\sum_{i=1}^n d_i^2$

The test statistic is then: $Z=\frac{6R-n(n^2-1)}{n(n+1)\sqrt{n+1}}$ which is approximately Normally distributed.
You can now perform the test using $\tau_{0.05}$ and $\tau_{0.01}$ points of a Standard Normal distribution.
March 18th 2011, 05:07 AM
Teague

Another question?

Can i just ask why i see loads of different critical value tables out there on the net offering loads of different values each time I view a different one? Cheers x
March 18th 2011, 08:00 AM
Sambit

Critical values of any distribution depend on the Level of the test. Which level do you want to consider?