Comment by Merrill

Rather than trying to draw policy conclusions from an epidemiological study like this, wouldn't it be more accurate to give drivers measured amounts of THC, put them in a driving simulator, and measure their performance?