A researcher wishes to know if a small city's education level affects the violent crime rate in the small city. He uses 6 small midwestern cities as a sample. The two variables he will study are:

• percentage of 16- to 19-year olds not in school and not high school graduates for each city
• the number of reported violent crimes per 1000 residents for each city

The collected data is

% not in school or graduated	24.2	19.0	18.2	14.9	19.0	17.5
violent crime per 1000 residents	13.0	4.4	9.3	1.3	0.8	3.6

What is the correlation coefficient? Round to three decimal places if necessary.