76 Unit 2 Critical thinking: the basics
fallacy. There are not two different fallacies
there: just two different ways of describing the
same fallacy, one more general than the other.
One could say that there is a correlation
between the people dining at the restaurant
and the people reporting sick. Let’s suppose
the figures for people dining at the Bayside (B)
and reporting symptoms were as shown in the
following diagram:Ate at B Reported sick0444There is a correlation – 4 out of the 4 who
reported sick had all eaten at the restaurant.
But it is a weak correlation: 44 ate there
without reporting sick, and although none
who did not eat there reported sick, we have
no information about those who may have
been sick but did not report it. We have a
plausible hypothesis. But to infer the Bayside’s
guilt from the data alone would be fallacious.
Arguments or inferences that assume causal
connections from correlations alone are
generally flawed.Recognising and avoiding flaws
There are many other classic fallacies and
common reasoning errors besides those you
have seen in this chapter. Some have names
such as ‘slippery slope’ or ‘restricting the
options’ or argumentum ad hominem. Many of
these will feature in Unit 4, and you will learn
to recognise them, so that you can reject
unsound arguments and avoid making
similar errors in your own reasoning.
It is a good idea to keep a diary or notebook
of common flaws that you come across.
(There is a suggestion in the end-of-chapter
assignments on how to organise this.)most likely to jump to a conclusion that may
be false.
[6] is a good example. We are told that a
number of people ate at a certain restaurant
and reported sick the next day, with suspected
food poisoning; then that the restaurant
closed. It is natural enough to assume that
eating in the restaurant caused the people to
be ill. People often justify such assumptions
by saying that there is no other explanation;
or that it is all too unlikely to be a
coincidence. But on reflection there often are
other possible explanations; and coincidences
do happen.Cause and correlation
The post hoc fallacy is itself an example of a
more general reasoning error known variously
as the ‘false cause’ or ‘mistaken cause’ or
‘cause–correlation fallacy’; or more
descriptively as confusing correlation with
cause. A correlation is any observed
connection between two claims or two facts,
particularly between two sets of data or trends.
For instance, if there were an observed upward
trend in violent crime in a city, at a time when
sales of violent computer games were on the
increase, it would be right to say there was
some correlation between the two trends.
It would also be tempting to conclude that
the games were at least a factor in causing the
actual violence to increase. Many people
make this inference, and not unreasonably,
since a significant number of computer games
have violent content. It is perfectly justified to
claim that if such games did turn out to be a
cause of violent crime it would be no surprise,
and it would help to explain the trend in a
convincing manner. But the plausibility of an
explanation does not make it true. It can be
posited as a reasonable hypothesis (see
Chapter 2.1), but not safely inferred.
The inferences from [6], and the reasoning
in [7], also exhibit the cause–correlation