By Victor J. Stenger
Stenger: Evidence Doesn't Support
In his lengthy response in the last issue of Research
News, Dr. Weaver commendably admits my central point, which is that
the reports of a positive connection between religious behavior and health
are not sufficiently significant to draw the frequently heard conclusion
that the existence of the supernatural has empirical support. As he puts
it, "Even if there is a beneficial relationship between religion and health,
one cannot therefore prove from those findings that there is a God or Divine
involvement in events.
Unfortunately, this is not the way the bulk of the studies that Dr. Weaver
cites are perceived by the public. Many of the authors of these reports
appear on TV, are interviewed by other media and write popular books in
which they make claims that far exceed the significance of their results.
Even if nothing supernatural is being claimed, the implication that
significant health benefits result specifically from religious behavior
is not supported by the evidence. As I mentioned in my original essay,
and as pointed out by Sloan and others, epidemiological studies are extremely
difficult to analyze because of confounding factors. Thus, if it is found
that churchgoers are healthier than those who stay home on Sunday morning,
one must allow for those who are too ill to get out of bed. I mentioned
one such paper that was withdrawn after it was found that the immobility
of the ill was not taken into account.
In most cases, the published studies have been reported in medical
journals. Unfortunately, the publication standards of medical journals
are quite low compared to other science fields such as physics, which makes
them inappropriate venues for extraordinary claims. The low standard is
presumably necessary to assure that possible helpful therapies are not
kept from needy patients for too long. In many cases their efficacy, is
far from proven and quote studies commonly disputed by later studies.
Unlike physicians, however, physicists and other non-medical scientists
are not in the business of saving lives but rather that of investigating
extraordinary phenomena. Those who claim efficacy for therapies that go
beyond established medical science should be bound by the stricter standards
of those fields that deal with extraordinary claims. I am not asking that
they be held to a higher standard, just the same standard that my colleagues
and I were required to meet in 1998 when we published the claim that the
neutrino has mass.
Scientific journals require a minimum standard of statistical significance
for the publication of a claimed empirical effect. The “probability level”
quoted in many scientific papers is what statisticians call the “p-value.”
For example, suppose an effect is reported with a p-value of 0.05, the
typical threshold of medical journals.
Technically, this means that in a long sequence of identical experiments
we would expect to observe an effect as great or greater produced by statistical
fluctuations in five percent of the cases. This not the same as the “likelihood”
or “probability” that the results were due to chance, although authors
often wrongly state it in this way. In fact, it is always possible to get
any observed effect by chance. You simply have to repeat the experiment
To illustrate this point, let me use as an example one of the highly
publicized experiments involving remote intercessory prayer. In these studies
at least the investigators can attempt to control for confounding factors
in a way not possible in epidemiological studies.
Cardiologist Randolph Byrd has claimed that coronary patients benefit
from blind, distant intercessory prayer. But his p-value is only 0.05,
apparently sufficient for publication in the regional Southern Medical
Journal. However, Byrd’s observations would be expected from statistical
fluctuations on average every twenty experiments. And, since null results
are rarely reported, one can easily imagine 19 other unpublished experiments
showing no effect with their data sitting in file drawers.
Another study along the same line as Byrd’s has been published in a
major international medical journal, Archives of Internal Medicine,
with nine co-authors. There, positive results are reported at a p-value
0.04, but for different criteria than Byrd’s. In fact, they fail to confirm
Byrd’s specific results.
No respectable physics journal would publish a result with a p-value
of a few percents. If it did, many papers would contain false claims that
were only statistical artifacts, wreaking havoc with the whole research
enterprise. In fact, the publication standard in physics is typically a
p-value of 0.0001, that is, only one in 10,000 similar experiments would
be expected to produce the reported effect or a greater one as a statistical
This was the case for our discovery of neutrino mass and all of the
other great discoveries in physics over the past 50 years. Furthermore,
I can cite from experience many investigations that reported effects at
levels of significance far higher than any of these religious studies that
failed to be confirmed by independent experiments.
To simply say that many studies have appeared in peer-reviewed journals
does not imply that they have demonstrated anything extraordinary that
cannot be explained by known material processes. I am sure going to church
lowers your blood pressure for an hour or so. The experiments referred
to by Dr. Weaver have been seriously misinterpreted by those who seek scientific
confirmation for their religious beliefs. If anything, the lack of significance
of such a huge body of work can be taken as empirical evidence for the
absence of a supernatural power that plays an active role in human affairs.
Victor J. Stenger is professor emeritus of physics
at the University of Hawaii and visiting fellow in philosophy at the University