The Pearson r correlation statistic requires that data in each variable must be (See Brown for a discussion of other Pearson r requirements.)
Consider the evaluation data in the "evals.sav" file where students respond the following items:
9. The instructor seemed well prepared for class.
A. Strongly agree B. Agree C. Neutral D. Disagree E. Strongly disagree
These responses are subsequently converted to numbers, where A=1, B=2, C=3, D=4, and E=5. Such data do not meet the requirements of the Pearson r because they are not interval. To perform a correlation on data that are not interval, perform the Spearman rho statistic, which requires that data in each variable must be at least ordinal.
In other words, if the data are ordinal, or if they are interval but not normally distributed, then employ the Spearman rho statistic. If your data are both interval and normally distributed, do not use Spearman rho as it is a weak statistic that should be employed only if absolutely necessary.