28 ■ CHAPTER 02 Evaluating Scientific Claims
SCIENCE
within schools, and a willingness to label the
condition. But after Wakefield suggested that the
MMR vaccine might be causing autism, the press
and other organizations began to report that the
rise in autism rates was caused by the increased
use of vaccines.
Linking rising rates of autism with increased
use of vaccines is a correlation. Correlation
means that two or more aspects of the natural
world behave in an interrelated manner: if one
shows a particular value for one aspect, we can
predict a value for the other aspect. But correla-
tion does not establish causation, in which a
change in one aspect causes a change in another.
Correlations may suggest possible causes for aphenomenon, but they do not establish a cause-
and-effect relationship. For example, there is
also a correlation between organic food sales
and the increase in autism. From 1998 to 2007,
rates of autism increased hand in hand with
organic food sales. They are correlated, but there
is no scientific evidence that eating organic food
causes autism (Figure 2.6).
Another correlation that spurred fears about
vaccines is that the onset of autism symptoms
occurs at about the same age that children
receive vaccinations. Most children receive the
MMR vaccine at about 15 months old, which is
shortly before the first symptoms of autism are
often noticed. Parents of children with autism
saw their children begin to exhibit symptoms
of the illness after their vaccination and there-
fore directly observed a correlation between the
administration of the vaccine and the onset of
an autism spectrum disorder.
Yet, again, correlation does not prove causation.
Only scientific experiments can demonstrate
causation. So, was Wakefield’s claim based on good
science? Did the MMR vaccine cause autism?Real or Pseudo?
Unfor tunately, sometimes a claim that super-
ficial ly looks like science is actually pseu-
doscience. Pseudoscience is characterized by
scientific-sounding statements, beliefs, or prac-
tices that are not actually based on the scientific
method. Asking a few simple questions at each
step of the progression toward a claim purported
to be “scientific” can help you distinguish science
from pseudoscience (Figure 2.7).
Using these criteria, Eaton analyzed Wake-
field’s claim. It was quickly obvious that his study
did not live up to the standards of good scien-
tific research (Figure 2.8). First, the study was
small—only 12 children participated—yet the
conclusions were grand: Wakefield suggested
that all children should stop receiving the MMR
vaccine. Sample size is extremely important in
observational studies, where small samples may
skew data to one extreme. Large sample sizes are
more representative of the population, are less
likely to be affected by outliers, and provide the
power to draw more accurate conclusions.
A second problem with the study was that
it was not made up of a random sample of1997199819992000200120022003200420052006200720082009
YearThe real cause of increasing autism prevalence?0 0
100,000
200,000
300,000
5,000
10,000
15,000
20,000
25,000
Sales ($ millions)Individuals diagnosedAutism
Organic food salesFigure 2.6
Correlation is not causation: organic food and autism
Reddit user Jasonp55 created a tongue-in-cheek demonstration of why it is
important not to assume causation from correlation. Jason used real data on
organic food sales and the prevalence of autism from 1998 to 2007. The two
are highly correlated, but one does not cause the other.
Q1: How much did organic food sales grow during the period covered
in the graph? How much did the incidence of autism grow?Q2: Why might both organic food sales and autism prevalence
have increased during this time period? A Reddit user in the original
discussion thread suggested that both might be affected by
increasing wealth in the United States. How might increased wealth
affect these variables?Q3: In what way has the vaccine-autism debate been confused by
people misinterpreting correlation as causation?