a study of studies finds "problems with experimental and statistical methods" mean findings may be somewhat questionable. issues contributing to this include small sample sizes, biases and selective reporting.
news/blog links - kinja - technorati - daypop - blogdex - boing boing - fark - metafilter - memeorandum - watching america - lucianne - instapundit - best of the web - oh, that liberal media - kaus files - daily kos - talking points memo - wonkette - scott rosenberg - mozilla - bugmenot - avg anti-virus - ad-aware |