The study *daylia* refers to is very interesting in its full text. I used a hierarchical regression analysis in my dissertation on a similarly small sample: their's 74, mine 109 subjects. That small a sample is too small to generalize to a larger population but it is an acceptable size in behavioral research. The amounts of pot they smoked was self-reported so there is no way to control for THC quantity but they averaged it at 5-5% THC to have a stable measure. e
Neither sample (mine about anxiety disorders, their's about pot smokers)was randomly selected which means that the statistical test has considerably less validity. The self selection quality of the sample is interesting because the IQ levels of the subjects (WISC test) appears to be slightly above average to begin with...It was longitudinal which gives it much more validity. The standard deviation on this test is 15. I just skimmed it so I haven't really given it a good workover but the most significant variable in the study is probably the fact that the subjects were.......








CANADIANS!!!!!!   


Now what could we conjecture about this!!!!!! LOL
love, harpy