Faithful Still Seeking "Proof" in Prayer Studies
(Top Posts - Distance From Belief
in theism - 081301, updated 121009)

Preface

For those of you not all that satisfied with the faith
route to god(s), curious about the efforts of pseudo-
science to "prove" that talking to god(s) has some
magic power of some kind, the following is offered
for your titillation, countered by doubts/skepticism
regarding said efforts ...

- - -

Can Prayer Heal?
Scientists Suggest Recovery May Be the Hand of God at Work
August 13, 2001
  http://abcnews.go.com/2020/story?id=132674&page=1

--- --- --- --- --- --- --- --- --- --- --- ---

Prayer (from the Skeptic's Dictionary)
  http://www.skepdic.com/prayer.html

- - -

Prayer Test
  http://prohuman.net/disbelief/prayer_test.htm

- - -

Prayer Studies
  http://www.infidels.org/library/modern/gary_posner/godccu.html

- - -

Re: Prayer Test

Prayer Works (like get real already, pseudoscience
rears its ugly head again)...

...and here's the proof
  http://archinte.ama-assn.org/cgi/content/abstract/159/19/2273

---

Comments: If half of this group received NO prayer and the other
half of this group received NO prayer, wouldn't the results be
EXACTLY the same as those that appeared in this article?

If the half of this group that did NOT receive prayer did BETTER
than the half of this group that DID receive prayer, would anyone
propose that prayer HARMS patients? Of course not.

Isn't this type of study absurd being that praying for names and
claiming it impacts live humans is ridiculous on its face?

Isn't this type of study superstition, not science? After all, in a
half and half study of 1,000 people, chances are one random
group is going to do differently than another random group.
Therefore, you've got a 50/50 chance of either group being
better or worse off with no proof whatsoever that any requested
activity on the part of 3rd parties having any impact whatsoever
on the result.

How's this for a test - have a group pray for one half of the
names and have another group pray for the other half of the
names. Based on the flawed logic of this study, whatever group
does better would "prove" that the pray'ers for that group were
better pray'ers, eh? The power of prayer? Absurd. Superstition.
Ridiculous.

Another idea - have a group curse one half of the names and
have another group curse the other half of the names. Based
on the flawed logic of this study, whatever group does worse
would "prove" that the cursers for that group were better
cursers, eh? The power of cursing? Absurd. Superstition.
Ridiculous.

---

Given a 50 percent shot at conducting a study and getting the
result desired, one must wonder how many of these types of
studies have been conducted and buried when the result turned
up tails instead of heads, eh?

---

From a follow-up reply in that thread,
posted on October 29, 1999:

Grizzly wrote in message <3819A61F.4F4E5E2D@infidel.org>...
>Reading the news story brings up one glaring fact. It doesn't say
>how the patients were assigned to groups. Random assignment
>is essential in this study to try and draw conclusions. The news
>article said that the primary investigator (Harris) provided the
>names for about half the patients to praying Christians.
>
>But it doesn't say how he did it. The news article also says that
>these same researchers reviewed about a thousand case histories.
>If they reviewed these charts before assigning people to get the
>"prayer effect", it is possible that the researchers may have given
>more optimistic cases to the "prayer effect" group. Even unknow-
>ingly.

There is a possibility of bias in the assignment in that the assignment
of names was made by the chaplain's secretary (see below for de-
tails) and the secretary was responsible for contacting the prayer
group leaders with names of people to pray for.

>
>But to respond to Dan, there is not a 50/50 chance that one group
>will be better (if standard inferential statistics are applied). If the
>groups are truly randomly assigned, and there is no effect, then
>you would expect to see no statistically significant differences
>between the groups. There may be very small differences due
>to chance, but most researchers would reject these differences
>based on statistical probability.

Per the study, using an analysis method by Byrd, the prayer
group and the non-prayer group scored statistically the same
(the Byrd technique rated patient conditions as good, intermed-
iate, or bad):
http://archinte.ama-assn.org/issues/v159n19/fig_tab/ioi90043_t5.html

(link no longer available, requires purchase of the full text of
the article for access)

The difference showed up when using a more complex rating
system which scored based on specific treatment/results on
a range of 35 separate criteria and scores from 1 to 6:
http://archinte.ama-assn.org/issues/v159n19/fig_tab/ioi90043_t3.html

(link no longer available, requires purchase of the full text of
the article for access)

It naturally follows that as the complexity of the rating criteria
increases, the chances for two random groups having the same
score decreases. Therefore, chances are quite high, with 35
separate criteria, that the scores for two random groups would
not be the same.

>
>Grizzly

At the study details, within the Archive of Internal Medicine
web site, an email address was listed for Dr. Harris. I sent him
the following email with questions regarding the study but the
email was returned as undeliverable. If anyone can contact Dr.
Harris, submit the following questions, and get back to us on
his response, I would be very grateful. The questions regard
doubts/concerns I have as to the methodology used in the study
(other concerns, listed in my previous post to this newsgroup,
are not included in this questionnaire to Dr. Harris).

---

Email I tried to send to Dr. Harris ...

Subject: Prayer Study

Questions:

1) Regarding your statement "Although our findings would be ex-
pected to occur by chance alone only 1 out of 25 times that such an
experiment was conducted, chance still remains a possible explana-
tion of our results.", where does the 1 out of 25 figure come from?

2) Per the difference in MAHI-CCU scores, you consider 6.35 to
7.13 to be statistically significant. What would you consider to be
statistically insignificant? For example, would 6.35 to 6. 79 be
considered statistically insignificant?

3) If you had found that the non-supplementary intercessory prayer
group had scored 6.35 to 7.13 for the prayer group, would you
have published the study? In other words, is your study biased in
that only a positive intercessory prayer result would have been
reported and the opposite would not, for political or professional
or religious reasons, have been reported?

4) The group assignment was made by the chaplain's secretary and
the study was associated with Saint Lukes. Doesn't that prejudice
the study in that bias, subconscious or otherwise, would have been
likely given the religious nature of the environment and of the person
instituting the critical assignment of names to groups, not to mention
said person being responsible for contacting prayer group leaders?

5) Regarding your statement "Thus, there was an unknowable and
uncontrollable (but presumed similar) level of "background" prayer
being offered for patients in both groups.", isn't that presumptive
and scientifically flawed in that the amount of prayer offered out-
side the study was not controlled? In other words, it is impossible
to know how much prayer of a direct and personal nature was
offered for the two groups. You're merely assuming the group that
had the blind prayer got more prayer than the other group.

6) Regarding your statement "After elimination of 6 patients who
were waiting for cardiac transplantation, 1013 were randomized
(Figure 1), 484 (48%) to the prayer group and 529 (52%) to the
usual care group. This difference in sample sizes was most likely
caused by chance (P=.18). After subsequent removal of those
patients who spent less than 24 hours in the CCU, 524 remained
in the usual care group and 466 in the prayer group.", it would
appear that the prayer group consisted of a healthier group of
people than the non-prayer group.

After all, 466/484 (95.9 %) of the prayer group stayed over a day,
whereas 524/529 (99.1 %) of the non-prayer group stayed more
than a day. This appears to show bias from the start, with a healthier
group of patients chosen for the prayer group as compared to the
non-prayer group.

7) You stated 66 % of the non-prayer group was men and 61 %
of the prayer group was men. Therefore, based on the non-prayer
group being 524 and the prayer group being 466, that would =
.66 x 524 = 346 men in the non-prayer group and
.61 x 466 = 284 men in the prayer group.

Wouldn't the fact there were 62 more men in the non-prayer group
bias that group towards significantly higher MAHI-CCU scores,
based on a sexual bias? Conversely, with 178 women in the non-
prayer group and 172 women in the prayer group, the scores for
women would balance out.

8) You state the prayer group average stay was 6.48 days com-
pared to 5.97 for the non-prayer group with the two longest stays
being in the prayer group. Yet, you follow that up with a claim
that this fact is statistically insignificant. Why did you comment on
throwing out the two long stays in your evaluation (after all, they
were being prayed for, too) and what, pray tell, would you con-
sider to be a significant difference in length of stays? 7 days to 6?
8 days to 6? 9 days to 6?

9) Using the same exact criteria listed above, but with one group
being asked to stare at a green card and recite names rather than
pray for the names, do you expect the result would be the same?
Why or why not? Note - your answer here may be indicative of
your personal bias in favor of religion and prayer, so please con-
sider this question carefully before responding.

Thanks for your consideration of and response to these questions.
I await your reply with anticipation and earnestly desire to under-
stand the basis for your conclusion so I can evaluate the data and
conclusions presented in your study.

- - -

Re: Prayer Test (response to a poster who wanted to 'believe'
that the Saint Lukes study was valid)

Saint Lukes. Chaplain's secretary assigning names. 35 criteria
with 6 ratings per criteria for a total of 210 possible assessments
for any individual - result - one group with substantially more
men and roughly the same number of women scored different
than the other group. Same study, same groups, a simple criteria
of good, intermediate or bad for each patient. The result? "no
difference" (ratings by Byrd). The results? Inconclusive, I'm
sure you'll admit, in a quiet moment of reason.

As to psychology and claims prayers are good for you, I'll
keep eating right, exercising, taking medication and vitamins,
and depending on humans, with the power of positive thinking
(a natural trait of mine, on my good days) on a realistic and
hopeful basis, thank you.

Prayer unanswered is probably a lot more frustrating and anxiety
ridden than you'd like to admit, don't you think, maybe? Since
most prayers *are* demonstrably unanswered (with no reason
to doubt the demonstrable lack of an answer), are you really so
sure that prayer is as positive a psychological force as you seem,
a priori, to feel it is? Is your prayer-promotion more the result
of desire for what you want to be true than it is common sense
regarding what actually is true, in this life, on this earth, at this
time? Maybe?

Thanks for your consideration. Can I hear a Hallelujah, there,
glory be to logic and reason? Amen. (-:

- - -

Religion-Health Connection Claims Questioned

Don't Mix Spiritual and Medical Care, Scientists Warn

Date: Posted 2/23/1999
  http://www.sciencedaily.com/releases/1999/02/990223081754.htm

Excerpt: "A team of Columbia University scientists is
warning against incorporating spiritual and religious
interventions into medical practice.

'We believe even in the best of studies the evidence of
a relationship between religion, spirituality and health
is weak and inconsistent,' write Richard Sloan, PhD,
and colleagues in the February 19 issue of The Lancet,
a leading British medical journal.

'Much of the scientific data for claims about religion
and health is highly questionable,' the authors said.
'It provides no empirical justification for the introduction
of religious activities into clinical medicine.'

... even the best of studies that attempt to show health
benefits connected with prayer, church attendance and
other forms of spirituality contain serious methodological
problems.

They reviewed hundreds of studies and identified several
dozen which illustrated flaws that they said characterize
much of the literature. The studies often involved small
numbers of subjects and failed to control for other factors
that could account for the findings, such as age, health
status, and health behaviors. They said that other studies
they examined failed to present the findings fully or failed
to make appropriate statistical adjustments.

... It is just as inappropriate for physicians to counsel
religious practice for the sake of better health as it is for
them to advise an unmarried patient to marry because
the data show that marriage is linked with lower death
rates.

The possibility exists of doing harm if a patient infers
that illness is a result of insufficient faith. 'No one can
object to respectful support for those patients who draw
upon religious faith in times of illness,' write Sloan and
colleagues. 'However, until these ethical issues are re-
solved, suggestions, intentional or otherwise, that reli-
gious activity will promote health or, conversely, that
illness is the result of insufficient faith, are unwarranted.' "

- - -

Study: Religious Struggle May Indicate Greater Risk
of Death Among Elderly

Date: Posted 8/13/2001
  
http://www.sciencedaily.com/releases/2001/08/010813081207.htm

Excerpt: "A study of 595 hospital patients suggests older patients
who are wrestling with religious beliefs during an illness may have
an increased risk of dying, according to researchers from Bowling
Green State University and Duke University Medical Center.

While several previous studies have demonstrated a reduced risk
of death with more frequent church attendance, this is the first study
to look closely at certain negative forms of religiousness as predict-
ors of mortality.

Feelings of "being abandoned or punished by God," "believing the
devil caused their illnesses" or "feeling abandoned by one's faith
community" were identified as key factors in risk of death among
elderly participants, said Dr. Harold G. Koenig, one of the authors
of the study and an associate professor of psychiatry at Duke
University Medical Center.

The results of the study appear in the Aug. 13, 2001 issue of the
Archives of Internal Medicine. "This study reminds us that religion
is a rich, complex process, one that represents a potent resource
for people facing problems and one that can, at times, be a source
of problems in itself," ..."

Comments: Or, put another way, if you get sick and your god/faith
don't come to the rescue, you're on the downside of the naturalistic
fate which is your destiny no matter what ... after all, the faith-
exposed who get sicker are unlikely to have their faith strengthened
and the faith-exposed who get better are likely to have their faith
strengthened. Evidence for the supernatural in both cases? Nada.

- - -