devils advocate here: assuming the data was 100% reliable, we could use that data to determine if there are patterns that lead certain individuals down that path. of course if that data is not reliable we made see patterns where there are none. I prefer not to speculate on the reliability of the data, just giving my perspective