[Previous entry: "London police gun down innocent man"] [Main Index] [Next entry: "Gordon P. reviews two movies"]

07/25/2005 Archived Entry: "Gordon on my last FOX/ifeminists column"

Gordon P. writes, Just read your column from last week (like I said, I've been busy :-( I like it very much; it's good to see someone clearly explain why conclusions from "hard science" and "soft science" may not be equally valid! Thought you might be interested in another reason why "soft science" research is often so bad.... Click on 'more' to continue.

Often, proposals are written by researchers who have not bothered to do more than a cursory literature search first, and these studies are then approved by panels who are _also_ not familiar with the most recent research, and whose members may not realize that the proposed study is redundant.

According to a recent report by Dean Ferguson of the Ottawa Health Institute published in the journal _Clinical Trials_(v.2, p. 218, 2005-Jun) [reported in a New Scientist article, sorry, subscription required], it is not unusual for research doctors to carry out redundant "Randomized Clinical Trials" on drugs and therapies that have already been "proven" in earlier clinical trials. The driving force for these redundant studies does not appear to be either "Publish or Perish" pressure or financial incentives from drug companies (albeit both _do_ appear to play a role), but merely that the researchers _have not bothered to do more than a cursory literature search_ before submitting the proposal --- and the grant-approval committees are likewise not up on the most current research, so redundant studies get funded on drugs and therapies that have already been "proven" in earlier trials.

According to the article, as many as tens of thousands of patients may have died because they received placebos instead of drugs that had previously been "proven" to be therapeutic. (NOTE, I keep putting "proven" in scare-quotes because as noted in Robert Matthews' "The Great Health Hoax," many of the studies that have "proven" a therapy are themselves statistically flawed,