I am running basic linear regressions in IBM SPSS Statsitics Viewer, but am only getting psotive R values. Specifically, I am regressing annual growth rates of fish with temperatures seen over there periods of growth. I am getting what would clearly be a strong negative correlation (instances where as temperature increased, growth decreased), with a low p value and high r square, yet the r value remains positive. Even the slope of the line is negative in these instances. How is my output showing a negative slope but positive correlation coefficient? Any advice on this issue would be much appreciated, as only getting positive correlation values when there are negative trends is tiring.