# Spearmans rank multiple ranks

• Nov 24th 2006, 09:34 AM
dtwazere
Spearmans rank multiple ranks
I have my data ready for ranking, my only quuestion is when it comes
to ranking a number that has more than one what will i rank it.

E.g I have one number 1 in my data so this gets ranked as one

but i have 5 number 2's in my data so would this be ranked as 2.2
since there is 5 of them and 2.2 would be the average (1/5=.2)

or would it 2.5 for each of them
• Nov 24th 2006, 10:25 AM
earboth
Quote:

Originally Posted by dtwazere
I have my data ready for ranking, my only quuestion is when it comes
to ranking a number that has more than one what will i rank it.
...

Hello,

you have:
rank value
1 → 1
2 → 2
3 → 2
4 → 2
5 → 2
6 → 2
7 → 3.6
... → ...
then you calculate the rank as the average of all similar ranks. In this case all 2's will get the rank $\displaystyle r=\frac{2+3+4+5+6}{5}=4$

Now you have:
1 → 1
4 → 2
4 → 2
4 → 2
4 → 2
4 → 2
7 → 3.6
... → ...

EB
• Nov 24th 2006, 11:42 AM
dtwazere
Hi, thanks for the response, this has caused me more confusion, would it be possible for you to rank my data for me as i am really stuck at this point! I can do the rest of the equation myself.

I would really appreciate this
• Nov 24th 2006, 09:26 PM
earboth
Quote:

Originally Posted by dtwazere
Hi, thanks for the response, this has caused me more confusion, would it be possible for you to rank my data for me as i am really stuck at this point! I can do the rest of the equation myself.
I would really appreciate this

Hello,

Quote:

...would it be possible for you to rank my data for me as i am really stuck at this point!
of course.

I've attached your modified file so you can see what I've done with your data:

1. I copied the time data into the column t' in sorted order.
2. In the column r' are listed all possible ranks in increasing order.
3. In the column r'' are listed the ranks so that equal data get equal ranks. (Notice: In German the comma is the same as the decimal point in English)
4. Now you have a one to one function between data and rank and you can accomplish your original list.

Remark: Of course it looks a little bit screwy to have a rank 11.5. But that doesn't matter for your calculations.

EB
• Nov 25th 2006, 01:39 AM
earboth
Quote:

Originally Posted by dtwazere
Hi, thanks for the response, this has caused me more confusion, would it be possible for you to rank my data for me as i am really stuck at this point! I can do the rest of the equation myself.

Hello,

in the attachment you'll find the completed table.

As you may know you can calculate the ranks by using "average(range with absolute adresses!)". For the one-to-one function you can use a command something like "vertical reference" (I'm sorry I only know the commands in German which are not very helpful for you, I belöieve)

EB
• Nov 25th 2006, 02:31 AM
dtwazere
Thank you all for your help. I will now do my spearmans ranking.

:D You guys are cool :cool:

Would it be possible for you to check my spearmans rank if i upload it as an attachment to see if i got it correct?

If it is not too much trouble.

Thanks, Daniel
• Nov 25th 2006, 05:41 AM
earboth
Quote:

Originally Posted by dtwazere
Thank you all for your help. I will now do my spearmans ranking.
:D You guys are cool :cool:
Would it be possible for you to check my spearmans rank if i upload it as an attachment to see if i got it correct?
If it is not too much trouble.
Thanks, Daniel

Hello, Daniel,

obviously you used $\displaystyle r_S=1-\frac{6\cdot \sum\left(D^2\right)}{n(n^2-1)}$. So everything is fine and correct.
Congratulations!

By the way: If I were you I would write somewhere on the spread-sheet which formula you have used. Otherwise you'll not understand your own homework two weeks later.

EB
• Nov 25th 2006, 11:43 AM
dtwazere
thanks a lot for all your help, i will attempt to do this for my other set of data. If i get confused ill post it here.

Thanks again!:cool: