1. Adam Gentry Adam Gentry United States says:

    This study was obviously paid for by special interests or undertaken with the knowledge a positive finding would attract media interest and future funding.

    Notice how no data is revealed.  But even if it had been, I have no doubt everything would be statistically significant and many wildly significant at the p < 0.0001 level.  I see this all the time with large sample sizes.  The researchers never learned that effect size is not independent of sample size.  In the sciences, we had been accustomed to dealing with sample sizes in the 30 - 300 range.  It was difficult to recruit study participants.  Bug in the age of the Internet and Software, databases offer ready access to thousands of data points from participants who do not even know they're supplying data for a study.  When the sample size increases to levels observed here (over 10,000), unless you choose a test of strength of association that is independent of sample size, the difference between the means will become statistically significant no matter how small it is as long as it is not zero.  There was an article published in a Statistics journal that suggested as much as 12% of the research literature (and growing) is contaminated with positive findings that are artifacts of sample size.  I encountered this when I reviewed research published by a group of hospital employees  who were prospective employers.  I couldn't believe it.  They were proud of the article.  Everything they tested -- and I mean EVERYTHING -- was statistically significant at the p < 0.0001 level.  But when you looked at the groups being compared (the stats compared here were percentages rather than means), the differences that were statistically significant included 14% vs. 9% and 36% vs. 31%.  In no universe are these differences large enough to be meaningful to anyone.  

    What's also suspicious is the substances that are all linked to this terribly frightening adverse event are incredibly varied in their pharmacological effects and phenomenology, sharing in common only the fact they are unpopular with politicians, soccer moms, temperance groups, and law enforcement.  

    Did you get the message?  This is your HEART on drugs.

    I am leading a movement to restore for chronic pain patients access to Rx painkillers thar have been involuntary discontinued by their doctors under extraordinary, career-existential pressure from the DEA and State Medical Boards.  That's my COI here.

The opinions expressed here are the views of the writer and do not necessarily reflect the views and opinions of News Medical.
Post a new comment