Thus, 60% of the variation in scores is explained by the variation in hours worked. The square root of this estimate yields the correlation coefficient between exam scores and hours worked and this is computed as 0.775 (the negative sign needs re-introduction here). Note from table 1 that the covariance between the two variables is –133.5 ÷ 7 = –19.071. If we divide this by the product of the standard deviations (4.749×5.182), we get the simple correlation coefficient (–19.071) ÷ (4.749×5.182) = –0.775.