🚀 go-pugleaf

In article <1992Aug6.211342.7966@icd.ab.com> adnan@icd.ab.com (Adnan C. Yaqub) writes:
>
>>>>>> On Thu, 6 Aug 1992 19:02:14 GMT, jgreely@morganucodon.cis.ohio-state.edu (J Greely) said:
>J> An interesting study, which seems to have been designed to reinforce
>J> the experimenter's bias.  Why this particular "test"?  How did he
>J> choose his samples?  Were they given information on normal sentencing
>J> guidelines for such a crime, or gauged for their previous knowledge of
>J> common sentences for rape?  How were the subjects selected (general
>J> population, volunteers, judges, parents, etc)?
>
>These are valid questions.  However, before you pass judgment, read
>the study.  You seem to assume, because of the experiment devised,
>that the researcher is biased.  Suppose you wanted to prove that
>pornography does or does not have harmful side effects.  What
>experiments would you choose?

Ah, scientific method.  It helps to understand it before using it.  It
is a wonderful tool.  From your statement above - "Suppose you wanted
to prove" reveals that it would not matter what experiments were
selected or created.  Because the key word which tends to invalidate
any conclusions drawn from such experiments is "wanted."  There should
be no subjective desire to prove anything - only a desire to find out
what is really going on.

If I wanted to prove that pornography was harmful, I could devise just
about any experiment that involves some use or form of pornography, but
my bias to prove will make me create tests to that hypothesis that will
always be slanted in the direction of proving my hypothesis.  I will
always search for the data that confirms what I want to prove and
declare it valid, but I will tend to declare data that does not support
what I want to prove as invalid.  I will write subjective questions.  I
will create subjective tests.

The only objective way to attempt to answer such a complex social
question utilizing scientific method would be to create experiments that
assume many possible conclusions.  For example,

Assumption #1:  Pornography has harmful effects on society

Hypothesis #1-1:  Pornography can induce people to commit sexual
related crimes after exposure to said pornography.

Hypothesis #1-2:  Pornography can induce people to sleep excessively.

Hypothesis #1-3:  Pornography causes cancer.

Assumption #2:  Pornography does not have harmful effects on society

Hypothesis #2-1:  There is no cause and effect relationship correlation 
between use of pornography and the commission of sexual crimes after
exposure to pornography.

Those are just a few of the assumptions and just two out of the infinite 
number of hypotheses.

Of course, many hypotheses are irrelevant (1-2, 1-3).  Our brains are 
very good at filtering them out.  Our brains are also very good at
stacking the deck in the direction that we "want" and making our science
appear to be valid.

Valid tests to indicate a cause and effect relationship between
pornography use and criminal sexual activity should examine only the
effect of the one variable - pornography - and should be measured
against a control group that does not have to deal with that one
variable.  Also, another group should have to deal with a placebo or
fake pornography.  The number of samples should be statistically
significant.

All summaries of tests should objectively discuss all outside factors
or variables that might have been involved and could not be
controlled.  A low-level of control of other variables dictates that the
experimenter not draw any definitive conclusions, because (s)he can not
be sure what effect these other variables had on the experiment.

That's why data that points to a conclusion about social behavior that
was indicated by only one variable is usually disregarded.  The
experimenters refuse to acknowledge that other variables might be
tainting their experiments.  If the're so careless as to make such a
glaring error, then they could also get causality and effect reversed.
When your logic is flawed, anything is possible.

One of the reasons people like J. Greely can just jump in and say, "I
doubt that this experiment really points to conclusion X" is because he
understands that 1) the experimenters had an objective - prove something,
and that 2) they would not be able to isolate this single variable, yet
claim that it was the only contributing factor to the conclusions they
drew.  It is not science, it is politics.

These experiments cited do nothing to consider the affects of: 1) low
self-esteem of the criminal, 2) causes of that low self-esteem (acne,
low intelligence, genetic disfigurement or just bad looks, abusive
childhood, abusive adulthood, lack of communication skills, lack of
achievement, lack of any rewards in life, etc. etc. etc.), 3) effects
of puritanical guilt over the sex act, 4) effects of financial status
of criminal (Hey, if they had the money, they might have just paid for
it - rather than "take"), 5) effects of gravitational pull - the
variables just go on and on and on.

That is why it is virtually impossible to create a scientific
experiment that really tests a theory about society.  There are too
many variables.  The scientific method tends to break down with complex
problems.  With three or more variables, chaos is quickly reached in
the data and nothing can be predicted accurately.

So, how do we make determinations about the value or lack of value of
pornography in society?  Well, intuitively.  We are expert systems.  We
can make decisions based on a preponderance of the evidence.  The
evidence is not in on the cause and effect relationship between violent
sexual crime and pornography use.  There is some evidence that points
to a relationship of sexual criminals also being pornography users -
namely that many convicted sex offenders admitted to being pornography
users.  There is also evidence that many convicted sex offenders
admitted to being milk drinkers.  There is little evidence that
suggests that pornography users also being sex offenders.  The vast
majority of pornography users are NOT sex offenders.

So, what can we say about the evidence that we have so far?  Not a
whole lot really.  We can say, "We don't know if there is a strong
cause and effect relationship between pornography use and the commission
of psycho-sexually deviant acts (as defined by criminal law).  There is
evidence that people who commit sexual offenses tend to also use
pornography.  There is little valid (objective), conclusive scientific
data that points to pornography use being an impetus to the commission
of criminal sex acts."  That's about as precise as we can get.

Many people who have used pornography and have felt little desire to
commit criminal sex acts as a result of that experience tend to think
those other variables have more to do with psycho-sexually deviant
behavior than does pornography use.  They think that people who might
have been sexually abused themselves might tend to replay that behavior
- thinking that it is not such abnormal behavior.  They believe that
low self-esteem due to many factors can drive people to commit
desperate acts.  These variables seem like likely candidates for a
strong cause and effect relationship to the commission of criminal sex
acts.

The "Ban Pornography" purists tend to think that ridding the world of
pornography would have a very good effect on society.  

Like I said, we decide these things intuitively.  These folks have been
told that pornography and many things about nudity and the sex act are
bad.  So they are very uncomfortable with both the availability of
pornography and the wide-spread criminal sexual behavior.  In their
minds, they are putting 2 and 2 together and getting 4.

There is little objective scientific evidence, though, that this is the
case.  In the cultures which have enacted laws to ban pornography, the
effects were sometimes the opposite of what was expected, the
commission of criminal sex acts actually increased.  But different
studies have indicated different conclusions.  

Objective people recognize that sexual crime is a serious problem and
complex problem that involves many, many variables - not just one.
Just because it is a complex problem, does not mean that it can't be
fixed.  We need to examine all the variables that appear to be relevant
and determine, as best we can, their effects, then try to regulate
those variables, or eliminate them.  We're in the process of doing just
that.  This discussion is leading us in the direction of figuring out
what to do with this problem.  What it lacks in legwork, it makes up
for in intellectual exchange.

This discussion will not find the answers.  We need a preponderance of
the evidence, and we aren't providing any evidence.  What we are now
doing is discussing why the ground rules of such an investigation are
so important.  We can only come to agreement in our conclusions when we
all believe that the information that points to that conclusion is
valid.  You won't get a majority of folks to agree on resolutions to the
problem until we agree what is causing the problem.  If a majority is
"pre-convinced," then logical, reasoning members of society will point
out the problems with that bias.  Biased conclusions would create
attempts to solve the problems that would not work at all - in fact, 
they might make the problem worse.

Meese Commission Study.  Don't you remember, "Ed Meese is a Pig!!!"?
More than half the folks in this country believed `Ed Meese was a Pig.'
So, were these same people going to believe the results of his study?
No way.  And when scientists reviewed his study, they found just what
they thought they might - a lack of objective science, illogical
reasoning, puritanical religious zeal, bias in collecting and reporting
data, bias in selecting appropriate and relevant hypotheses to test,
irrelevant suggestions for solving the problem, etc.

That study did not clarify this issue or attempt to find the answers.
It was politics.  

Aaron L. Hoffmeyer
TR@CBNEA.ATT.COM