FX uses the scientific method better then JAMA?

Brian W. Tague taguebw at wfu.edu
Fri Dec 8 17:46:02 EST 1995

<U27111 at uicvm.uic.edu> wrote:
>The breakfast time show on the cable channel FX today showed
>something which just about every mother everywhere knows... but
>apparently the editors (and peer-reviewers) of JAMA does not?
>That sugar does indeed cause hyper-activity in children!

>To test this premise... the breakfast time show decided to run
>their own experiment (using the scientific method).  They had two
>children who had not yet eaten breakfast... one was feed a high
>protein breakfast of sausage and eggs, and the other one a high
>sugar breakfast of cereal and donuts.

>There was a definite (obvious) difference in these two children's

>So...  as the doctor in the program pointed out... many would
>criticize this short study because it wasn't done as a blind study
>(we knew which child was which and there may have been bias).
>But my god... the difference was so pronounced - I really don't see
>how that would matter.  It certainly shows something is going on...
>and it's not what JAMA reports!

>How could the people at JAMA publish such an obvious false premise?
>What does this tell us about the process of peer review and
>And isn't it sad when a cable channel breakfast show can use the
>scientific method better then the publishers of JAMA?
>Just some food for thought,

First off: the lack of a double blind analysis in the TV show is a 
serious problem, and not just with the bias WE might have observing and 
reporting on the children's activity. It has been shown in numerous 
studies in situations like this that if we expect, in this case, one of 
the children to act differently then our expectation will change their 
behavior regardless of the treatment.

Second: I doubt very much that the JAMA study had a sample size of one. 
That makes this "experiment" useless.

The TV show "experiment" you describe is pseudoscience at its worst and 
has absolutely no significance.

My 2 electrons
Brian W. Tague

More information about the Bioforum mailing list