--- In FairfieldLife@yahoogroups.com, "authfriend" <jst...@...> wrote:
>
> --- In FairfieldLife@yahoogroups.com, "drpetersutphen" <drpetersutphen@>
> wrote:
> >
> > I missed this post.
>
> And apparently didn't read it when you did see it.
>
> Every one of the commenters Brian quotes--and Brian
> himself--pointed out that the study couldn't be taken
> as conclusive because the sample size was too small
> to be statistically significant. They all said further
> study would be needed.
>
> > This research means nothing because it is
> > methodologically flawed. How were the measures taken?
> > What was the "control period".
>
> Well, actually you don't know whether it was
> "methodologically flawed," other than the small
> sample size, because Brian didn't say how the
> measures were taken (although he did explain the
> "control period"; apparently you didn't read that
> part).
>
> You have to *know what the methodology was* before
> you can say whether it was flawed.
>
> > Anyone with even a little training in doing this type
> > of research will see huge holes in it.
>
> Again, other than the small sample size, they won't
> see "huge holes in it" from this post because the
> post doesn't give any of the methodological details.
>
You can tell from the post that the study is methodologically flawed beyond any
"sample size" problem. For example, according to the post people knew when
the so called control period ran and what was the period people were
meditating--the research wasn't blinded. For goodness sakes, they used the
police recreation club! Everybody had to know the meditators were in town. If
they had reported bang up positive results I would question them because of the
defective design. The community may have put more police on the streets when
the meditators were in town. Or people could have been on good behavior because
company was around. I would be especially concerned about confounding
variables because the post quoted by you and Peter said: "I compiled this list
for a proposal to the Bermuda Police Commissioner when I was living there. He
was so impressed with the research he offered in-kind support from the Bermuda
Police for a demonstration of the Maharishi Effect." Correlation research is
problematic anyway. They just added to the problems with their poor design.
And that is just one glaring problem. There are others as well.
Another example, which is raised in the linked materials, is that the crime
rate in the area is to small to yield statistically significant results in a
short time period. If that is the case they should have known that going in
and thus the design was flawed for that reason alone.
Any number of things could have occurred to confound the results. No
conclusions can be drawn about anything, not even as a pilot study worthy of
further research. It doesn't even rise to the level of being inconclusive.
Well, I am off again after a quick check-in.
