The individuals polled did not include a representative sample of
the population. Results are unreliable if they involve persons
whose ages, religious beliefs, gender, race, nationality, etc. are
Loaded questions are asked.
The answers are not completely thought out.
Confusing terminology is used. 1
For example, one 1980 poll asked similar questions, worded in two different
"Do you think there should be an amendment to the Constitution
prohibiting abortions, or shouldn't there be such an amendment?"
Those in favor of an amendment: 29%; opposed 67%. Here,
the pollsters used the term "abortion" which they probably
selected as the most emotionally neutral term that they could
"Do you believe there should be an amendment to the Constitution
protecting the life of the unborn child, or shouldn't there be such an
amendment?" In favor of an amendment: 50%; opposed
34%. Here, the implication is that an abortion kills a child.
Meanwhile, a poll conducted one year later asked a heavily weighted question:
"The decision on whether or not to perform an abortion rests
with the consenting patient, and should be performed by a licensed
physician in conformance with good medical practice." 90%
were in favor of this statement. But the question is deceptive. The
pollsters combined two questions in to one: should a woman be able to
decide to have an abortion, and should it be done by a skilled
Pro-choice groups could focus on the answer to the first question, and claim
that less than 30% of the population is pro-life. They could use the results of
the third question and claim 10%. Pro-life groups could center on the second
question and claim that half the population was pro-life. And both groups would
be correct - at least they could support their assertion with statistics.
Another example of biased questions is seen in a Wirthlin Worldwide
National Quorum poll, conducted on behalf of the pro-life group Faith2Action
in 2002-DEC 4 (N=1001; margin of error = 3.2%)
One question was worded: "In light of recent medical advances such
as in-utero surgery and 3-D ultrasound technology, which reveals the
unborn child's body and facial features in detail, are you in favor of
restoring legal protection for unborn children?" 68% of those sampled
were strongly or somewhat in favor. A person who disagreed with this
question would be in the position of not protecting children.
Another was worded: "Would you favor judicial nominees to the U.S.
Supreme Court who uphold laws that restore protection for unborn children?"
66% of those sampled were strongly or somewhat in favor. The implication
is that some nominees might refuse to uphold existing laws.
One deficiency with most polls is that they require snap judgments from the
subjects, without examining all of the implications of their decision. For example:
The Gallup Poll reports on a yearly basis the percentage of
American adults who oppose abortion access for all reasons.
Results have ranged from 22% in 1975 to 12% in 1995. The value in the
year 2000 was 19%
Yet, a Time-CNN poll in 1992 showed that only 11% of American adults
would withhold an abortion needed to save the life of the woman, and
only 12%would prohibit an abortion if the woman's health is in
As a minimum, we need to know were the exact questions asked, of whom was it
asked, how many individuals were sampled, and under what conditions were the
questions asked. Does the question talk about:
terminating a pregnancy,...or murdering babies?
women's rights,...or killing children?
access to an abortion,...or compulsory parenthood for every pregnant woman?
whether the individual personally approves or disapproves of abortion
for themselves,... or what laws should be in place to prohibit a woman
who wants an abortion from getting one.
Even with a careful set of questions, and a representative sampling of the
population, there remains one overwhelming factor that reduces the effectiveness
of polls. Some individuals or their partners have never had a crisis pregnancy,
or have never had a close friend with an unwanted pregnancy. They are liable to
answer questions on the basis of their religious or political views. But if they
have to face a crisis pregnancy and are forced to think through all the options,
they often change their opinion from pro-life to pro-choice, or vice versa.
Unfortunately, polls tend to weight the opinion of those who have never
thoroughly examined the questions equally with those who have. Polls are exactly
what they claim to be: an expression of the opinion of the public -- whether
informed or not.
A better poll:
The following questionnaire is provided to political
candidates by Project Vote Smart in order to determine their
positions on a range of abortion questions. With few modifications, it could
form the basis of an improved public opinion poll:
Indicate which principles you support (if any) concerning
a) Abortions should always be illegal.
b) Abortions should be illegal when the fetus is viable, with or without life support.
c) Abortions should always be legally available.
d) Abortions should be legal only within the first trimester of pregnancy.
e) Abortions should be legal when the pregnancy resulted from incest or rape.
f) Abortions should be legal when the life of the woman is endangered.
g) Abortions should be limited by waiting periods and notification requirements as decided by each state government.
h) Prohibit the dilation and extraction procedure, also known as "partial birth" abortion.
i) Prohibit public funding of abortions and public funding of organizations that advocate or perform abortions.
j) Support "buffer-zones" by requiring demonstrators to stay at least five feet from abortion clinic doorways
k) Provide funding for family planning programs as a means to decrease the number of abortions.
The ideal poll:
In the author's opinion, the best polling technique would be for the agency to pose a number of
situations to their subjects -- describing a scenario that a pregnant woman
might find herself in, and asking if the woman should be allowed to have an
abortion if she wishes one.
For example, a poll could be taken in which 50% of the subjects were asked
the following questions. The other 50% would be asked the same questions, with "baby"
substituted for "fetus:"
A woman became pregnant due to a contraceptive failure. She is a
university student and does not want to interrupt her education.
A 16 year old teenager in high school conceived four weeks ago, was
abandoned by her boyfriend, and believes that her parents will throw her
out of the house if they find out.
A 16 year old teenager in high school is pregnant and has the full
support of her family who offer to help her raise the child. But she
feels that a baby would seriously interfere with her educational plans.
A woman became pregnant as the result of a rape.
A woman became pregnant as a result of incest.
A rare situation has come up in which continuing the pregnancy would
result in the death of the fetus and the woman.
An amniocentesis procedure shows that the fetus suffers from the
Down's Syndrome genetic defect.
A rare situation has come up in which the woman would suffer very
serious and perhaps permanent health problems if she cannot have an
And, most important, the identical survey would be conducted on a yearly
basis so that long-term trends could be determined.