A question about Spearman's Rank and Mean Values

#1
Hello,

I'm not entirely sure I've put this in the correct forum...

This is possibly a daft question, but I have used SPSS to calculated Spearman's correlation for a series of data sets and I was wondering if I can average the correlation values to give an average r result? :confused:

I have a list of people and their actual age and I am comparing it to estimated ages. Two different people estimated their age, twice, eg:

Spearman's r:
Observer 1a = 0.981
Observer 2b = 0.980
Observer 2a = 0.810
Observer 2b = 0.832

Can I average the correlation values for Observer 1, and then Observer 2, and then average both Observer 1 and 2 together? To get:

Observer 1 average: 0.9805
Observer 2 average: 0.821
Overall Average: 0.90075

Thank you in advance!
 

hlsmith

Less is more. Stay pure. Stay poor.
#2
Not completely following your description. Is it possibly to do a 3-way correlation. Can you link the datasets?
 
#3
Basically, I'm looking at the correlation between estimated age and actual age and I have a Spearman's r value.

But to examine inter- and intra-observer variation, two observers have estimated age, and they've estimated age twice each.

I therefore have four Spearman's r values. I'm just wondering if I can average those Spearman's r values as they are? (Note: the sample sizes are different for each set because the observers decided that they couldn't estimate age for certain individuals)