An rct with n=40 is better than the junk non-controlled studies that inexplicably get funded and accepted by journals but probably meaningless for something involving something as subjective as mental health.