We used a simple Pearson product-moment correlation test to verify the seemingly tight relationship between the
two trends. [FN50] This test measures the *68 linear relationship between two variables on a coefficient scale
between -1 and +1. The closer the correlation coefficient is to +1, the more statistically significant the predictive relationship
between the two variables. Since we are particularly interested in the correlation of the time-series trends
before and after the national effort to increase the harshness of the juvenile justice system, we calculated three separate
correlation coefficients: (1) the correlation coefficient between the homicide responsibility rate for both juveniles
and young adults, (2) the correlation coefficient of the homicide responsibility rates before the national wave of juvenile
justice laws, and (3) the correlation coefficient of the homicide rate during and after the national wave of juvenile
justice legislation. Table 1 shows the results of these three correlation calculations.