we calculate Chi-square by getting the sum(Observed - Expected)^2/(Expected)



You get the expected in each cell by taking row total times column total divided by overall total
Expected Low Drunk Driving and Low Under Age is (22)(20)/50 = 8.8
Expected High Drunk Driving and Low Under Age is (28)(20)/50 = 11.2
Expected Low Drunk Driving and High Under Age is (22)(30)/50 = 13.2
Expected High Drunk Driving and High Under Age is (28)(30)/50 = 16.8

so Chi-square statistic is (8-8.8)^2/8.8 + (14-13.2)^2/13.2 + (12 -11.2)^2/11.2 + (16- 16.8)^2/16.8 = 0.216

Now we get the p-value, with df of (row-1)(column - 1). There are 2 rows and 2 columns in the contingency table, so df = 1. I use this site.... http://www.socscistatistics.com/pvalues/chidistribution.aspx
The p-value is 0.642105. Since the p-value > .05