Thanks to Jay Wang and Ivan Zezula's ranking formulae (see "Rank Awful"
postings) I have produced an Excel spreadsheet to calculate both correlation
coefficients for Pogo's data.

> 1. I have two sets of scores which are related to each other.
>     SET A 1100,1040,950,1090,1050,970,900,1110,940,880
>     SET B 108,128,113,103,112,119,96,103,112,92.
>     My question is do I use Pearson's or Spearman's.

Depends on your question! If you are expecting for a linear relationship
between A and B then use Pearson's r, if you are looking for a monotonic
relationship then use Spearman's rho. If you just want to describe the
relationship between the variables then produce a scatterplot; it looks like
there is a non-monotonic relationship (see the scatterplot in the
spreadsheet). In which case neither correlation coefficient would be
appropriate.

Dr Graham D. Smith
Psychology Division
School of Behavioural Studies
University College Northampton
Boughton Green Road
Northampton
NN2 7AL

Tel (01604) 735500 Ext 2393
Email [EMAIL PROTECTED]

Reply via email to