Spearmans Rank Correlation Coefficient in Coursework Watch
Anything above 0.75 can be said to be strong positive, (age of car, loss in value) when 1 is perfect positive.
<= -0.75 strong negative (age of car vs Value)
-1 Perfect Positive.
Otherwise weak pos/neg coming closer to zero, until about +ve and -ve .25 when there is little to no correlation.
This is best illustrated using a scatter graph first, describing what it looks like, THEN using Spearman to put a quantity to your observations.
I used standard deviation in an earlier part of my coursework. How easy is it to compare SD and SRCC?
Don't go down this road without knowing what you are talking about though.
eg: Age Price Rank x Rank y d*2
2 5000 1 2 1
5 4050 3 1 4
3 6000 2 3 1
D*2 is the difference between rank x and rank y squared.
Rs= 1- 6 x sumd*2
Calculate the Rs value ,as in here, 1-6x6
3x(3*2-1) = 0.5
Use table to find critical value. (n=3 and significance level eg: 5%)
State Ho and H1 and show that the test stat (here 0.5) is either greater than the c.v. or less than. This will show that you should accept H0 or H1, which ever is this case and then you can talk about this shows hows there is rank correltion or isn't what ever you stated in your hypothesis'.
Hope this helps, I know how to do it, it just isn't easier to explain over thenet, (without a diagram etc.)
EDIT: I mean how to calculate the thingy not a table
Im doing the colour test.
i am doing y7 girls word y7 girls colour
y10 boy and girl word and colour and so on
Im not sure which ones to do the spearmans rank on.
Would river width by depth in spearmans rank be a 2 or 1 tailed test ???