Presented for your consideration without comment:
The interesting thing is the "Use of Marijuana to Self-Medicate BD" study is useless. Monitoring a small number of bipolar people for 6 days will give you completely random results. It's too short a period to remove moods from the equation.