For the bivariate correlations reported in the report, the strength of the relationships between pairs of variables was provided using a scale of magnitudes. Following Cohen (1988), reported magnitudes adopted the notion of a scale of small, moderate, and large sized relationships, qualitative terms that allow interpretation of the strength of a relationship through the concept of effect size. Cohen suggested that for a scale of the proportion of variance accounted for (the square of the correlation coefficient, r²), one might use a value of 0.01 to signify a small effect size, 0.09 for moderate, and 0.25 for large. Some latitude is appropriate in determining the scale of effect sizes within the context of the analysis. The magnitudes reported in this report were based on a scale in which the effect is small if r² is less than 0.05, moderate if r² is at least 0.05 but less than 0.25, and large if r² is 0.25 or greater.