Avoiding Errors in Reasoning: An Introduction to Logical Fallacies

ABSTRACT

OBJECTIVE. The objective of this article is to provide all students of radiology, from

novices to experts, a framework for understanding common errors in reasoning.

CONCLUSION. By familiarizing ourselves with common fallacies, we can take steps to

reduce the probability that we will commit such errors or fail to recognize and respond
appropriately to them when they occur.

Introduction

It is vital that physicians and other health professionals be adept at clear thinking and
logical reasoning. In medical education, we tend to focus much of our attention on
epidemiology and statistical inference, but in fact these disciplines are but components
of a larger process of medical reasoning [2]. The process of reasoning is present, at
least implicitly, in every radiology report that uses empiric observations and established
principles of reasoning to arrive at a diagnostic conclusion.

No matter how good our observations, faulty reasoning will invariably lead us from the
truth, with adverse consequences for patients, the profession, and our society.
Conversely, high-quality reasoning by physicians will permit the profession of medicine
to flourish. The journal Medical Decision Making, founded in large part through the
efforts of radiologist Lee Lusted, is dedicated to such a purpose [3]. It is vital that all
students of radiology, from novices to experts, develop an understanding of the features
according to which arguments may be judged valid or invalid.

One particularly fruitful way of better understanding medical reasoning is to consider the
various ways in which defective arguments can lead us astray [4]. This is the purpose of
this article. Although the number of possible errors is infinite, we are fortunate that most
invalid arguments generally fall into one of several dozen stereotypical patterns [5].
These patterns are often termed fallacies, from the Latin fallere, to deceive. By briefly
examining the most notable fallacies, we can equip ourselves more readily to recognize
and respond to faulty arguments when we encounter them.

Caveat
s

Before considering the fallacies, we need to be clear that validity and truth are not
synonymous. We can arrive at true conclusions through faulty reasoning, and we can
construct internally valid arguments that lead to false conclusions. For example, we may
arrive at the correct diagnosis of a lesion on a chest radiograph, even though the
process of reasoning we used is faulty. Similarly, when we start with a false premise,
although every step in our reasoning process is valid, we may arrive at a false
conclusion.
Moreover, logical principles are not the only basis on which we make choices.
Nonlogical considerations, such as honor and loyalty, may sometimes sway our
decision making in ways that logic alone would not dictate. To convince people of
something, we may have to appeal to them in ways that transcend mere logic. Similarly,
not all examples of poor reasoning are honest errors, and, in some cases, malicious
deception may be involved.

Ad
Hominem

In Latin, ad hominem means against the person. Ad hominem arguments attempt to

undermine a position by attacking the person taking it. For example, a radiologist might
attack an equipment vendor's claims regarding a new product's capabilities by pointing
out the vendor's financial interest in selling the product. However, the ad hominem
argument does not refute the position itself because it never deals with it directly. The
vendor's claims for the product might be absolutely valid.

Just as we cannot determine the quality of a book merely by inspecting its cover, so too
we cannot state with certainty that a claim is invalid merely because we object to the
person making it. It is natural that we take into account the source of every statement,
but the message and the messenger are logically distinct. We should neither shoot the
messenger because we do not like the message, nor ignore the message because we
do not like the messenger. No matter how intensely we may object to the person
advancing an argument, we have a responsibility to evaluate the argument itself on its
own merits.

Hypocris
y

A variant of the ad hominem argument is the so-called “hypocrisy fallacy.” In this case,
we judge someone's statement as false because it is inconsistent with something else
the person said or did. Yet mere inconsistency between two positions does not establish
that either one is wrong. It merely suggests that both are unlikely to be true. Apparent
inconsistency may be explained by the fact that new information has come to light or
that decision makers have changed their minds.

We may, for example, legitimately change our minds because of new technological or
economic developments. Or perhaps a radiologist recommends examination A for one
patient and examination B for another because the two patients actually differ from one
another in some subtle but significant way. It is when positions change based purely on
pecuniary or political self-interest that the allegation of hypocrisy may come into play.

Appeal to Authority

A related logical fallacy is the appeal to authority. In it, we take a recommendation at

face value merely because the source is a recognized authority. Radiology residents
frequently encounter such reasoning on call when a house officer in another specialty
insists that an examination be performed simply because the attending physician said
so. Authority and expertise do not necessarily imply validity. The history of medicine is
littered with useless and, in some cases, frankly harmful practices, such as bloodletting,
that gained favor based strictly on the authority of an eminent physician or scientist.

While it is reasonable to evaluate the trust-worthiness of a recommendation's source,

mere authority is no substitute for critical investigation. This principle underlies the
current enthusiasm for evidence-based medicine [6]. We need experts to help guide our
practice, but true scholars and statesmen do not expect us to adopt their
recommendations strictly on the basis of their authority. In fact, the best among them
expect us to regard their claims with thoughtful skepticism. In science and medicine, the
best practitioners are their own harshest critics, always trying out alternative
explanations and seeking to uncover hidden biases that might have led them astray.

Expertise itself deserves careful examination. In most cases, expertise is attainable in

only a relatively small domain. Just because people have achieved mastery in one
domain does not automatically qualify them as experts in another. When doubts arise
about the legitimacy of someone's claims to expertise, we can critically examine the
boundaries of that person's expertise, attempt to ensure that no unrecognized biases
are involved, and inquire whether other experts in the field would reach similar
conclusions.

Appeal to Convention

Another dangerous fallacy is the appeal to convention. Just because the majority of
people, or even everyone, says or does something, we cannot be certain that it is the
correct thing to say or do. Many physicians once believed that it was reasonable to use
external beam ionizing radiation to shrink the enlarged thymuses of children with stridor.
The procedure not only turned out to be of no benefit, it actually increased the risk of
other diseases, such as cancer of the thyroid gland, and was largely abandoned by
1960 [7]. Our most important discoveries often occur when thoughtful people begin
asking probing questions.

We need not continue to tolerate inefficiency, ineffectiveness, rude behavior, or even

duplicity merely because the precedent for such behavior has been set or because
others are doing the same thing. For example, just because some researchers are
manipulating their data does not make it right for us to manipulate our data. Moreover,
rigid adherence to convention slows the rate at which we adapt and improve what we
do. If we are not careful, appeals to convention can leave us clinging to little more than
the lowest common denominator of behavior in our field. If we want to identify the most
effective approach to a problem, we cannot simply take a poll.

Flattery
One of the most powerful but ultimately vacuous techniques of persuasion is flattery,
sometimes referred to as apple polishing. All of us enjoy being admired and praised, but
sometimes our appetite for flattery can leave us vulnerable to invalid arguments. For
example, someone may tell us how brilliant they think we are. Naturally impressed with
this person's astute judgment, we may then treat less critically some of the other things
he or she says. People in positions of authority are especially likely targets of flattery.
Examples include faculty members whose students seek good grades and supervisors
whose employees seek retention, promotion, and other rewards.

Emotional Appeal

Closely related to flattery are other forms of emotional appeal such as appeals to fear,
compassion, and the desire to belong. Appeals to fear encourage us to accept a
proposition merely because it frightens us. Yet how likely is the adverse outcome, and
what can we do to prevent it? When we allow fear to cloud our judgment, we set
ourselves up for a violent collision with reality.

Appeals to compassion are attractive, but just because an action would manifest
genuine sympathy does not mean it is the right thing to do. Sometimes doing what the
evidence warrants requires us to set aside our personal feelings. Likewise, we may be
duped into doing something unjustified out of a desire to be part of a group. To give in to
that temptation is to unnecessarily open ourselves up to the criticisms of an unbiased
observer.

Begging the Question

Known in Latin as petitio principii, begging the question occurs when the premises of an
argument assume the truth of its conclusion. An example would be this argument: “Our
group should do everything possible to increase revenue because generating revenue
is what we are in business to do.” It is not unreasonable to assume that a particular
premise is true, but when we do so we should not pretend that we are advancing an
argument or arriving at a new conclusion.

Whenever someone who has been asked to defend a position responds by merely
restating the position, they are begging the question. It also occurs when a position
springs from wishful thinking or is grounded in ignorance. If a vendor simply asserts that
his or her product is best but has little or no knowledge of the competitors' products,
then the question is being begged. The reasoning is more likely to be sound, however, if
the vendor can offer supportive experience and data.

Biased Sample

Biased samples occur when we draw conclusions based on methods of empiric

investigation that are skewed in some way. For example, a radiology chair might argue
that a particular compensation system should be adopted based on the fact that six
other chairs recommended the same system. In fact, it is possible that the chair tends to
speak with friends who share similar views. If a less biased survey were conducted, a
different conclusion might emerge.

A famous example of biased sampling involved the 1948 U.S. presidential contest
between Harry Truman and Thomas Dewey. Based on the results of telephone surveys
conducted by the most respected polling organizations, Dewey was favored over
Truman by more than 10 percentage points. A famous photograph from the Chicago
Daily Tribune shows president-elect Truman holding up an early edition with the
headline “Dewey Defeats Truman,” yet Truman won. What happened? In 1948, only
relatively affluent people owned telephones. By conducting their sample via telephone,
the pollsters had biased the result.

False Dilemma

A particularly insidious fallacy is the false dilemma. This occurs when someone seeking
to influence a decision inappropriately narrows the range of alternatives, often down to
only two. For example, someone might say that a radiology department must either
immediately purchase a state-of-the-art piece of equipment or close its doors forever.
Although it may in fact be the case that the purchase is warranted, other options, such
as a 3- to 6-month delay, may also be reasonable.

An ultimatum is often a particularly dramatic form of false dilemma. People may

threaten that if they do not get their way they will quit, or seek to have someone fired, or
do everything in their power to make life miserable for those standing in their way. It
may take courage to resist such threats. When it comes to sound reasoning, however,
the general level of complexity in human affairs makes it unlikely in any situation that
there are two and only two alternatives, one of which is a little short of doom. In benign
hands, the false dilemma may merely represent an oversimplification, but in the most
dire situations it may reflect deceit that quickly erodes a reputation for probity.

Compromise

The converse of the fallacy of the false dilemma is the fallacy of compromise. We
cannot assume that an intermediate position between two extremes is necessarily the
correct one. There is no question that when conflicts arise compromise is often
desirable. Yet the midpoint is not always the correct balance, and in some cases may
be more unjust than either of the opposed positions. A revealing tale concerning
compromise is told of King Solomon.

A woman's infant died. During the night, she took another woman's infant as her own,
placing her dead infant by the second woman's side. When the second woman awoke,
she knew what had happened. When they presented their dispute to the King, he
ordered that the live infant be cut in two, with half given to each woman. When only one
of the women objected and said she would sooner forfeit her half, he knew that she was
the real mother. In this case, the refusal to accept compromise showed who was
advancing the legitimate claim.
Radiologists should not assume that compromise is always the appropriate course of
action, or that when compromise is called for, the parties should always meet each
other “half way.” For example, a group of cardiologists might argue that they should
perform all of a hospital's vascular angiographies, while a group of radiologists might
argue that cardiologists should perform none. Having the cardiologists assume
responsibility for half the cases is not necessarily the appropriate resolution.
Radiologists who assume that compromise is always the best route need only recall
what happened some years ago with obstetrical sonography, in which most radiology
departments no longer participate.

Slippery Slope

The slippery slope argument attempts to prevent small changes by arguing that they will
inevitably lead to large changes that are widely recognized as undesirable. For
example, a hospital administrator might argue that outpatient imaging clinics are
unacceptable because if every physician began seeing all their patients in outpatient
clinics the hospital would be forced to close. Taking one small step does not necessarily
entail going all the way to the extreme. It is often possible for people and organizations
to make incremental changes without completely overthrowing the established order.
Incrementalism is actually a reasonable approach in many real-life situations.

Straw Man

The straw man is involved when someone misrepresents another's position to make it
easy to overcome. For example, a residency program director might encounter
resistance to proposed reductions in resident work hours. Opponents to such a
reduction might argue that residents will no longer gain enough experience to become
clinically proficient. “If we are graduating incompetent physicians,” the opponent might
argue, “we will be defaulting on our responsibility as a residency program. Therefore,
we should leave the duty hours as they are.” However, it may not be accurate to imply
that any reduction whatsoever leads inexorably to incompetence. Perhaps the training
experience could be altered in such a way that residents suffer no loss of proficiency or
even become more proficient, despite the fact that they are logging fewer hours in the
hospital.

Conclusio
n

In view of the complexity of contemporary radiology and health care, a completely fail-
safe strategy by which we can protect ourselves from errors of reasoning is not
possible. However, by familiarizing ourselves with common fallacies, we can take steps
to reduce the probability that we will commit such errors or fail to recognize them when
they occur. In addition to recognizing and avoiding them, by understanding them better
we can also prepare ourselves to respond effectively when we see them being
committed. To err may be human, but taking the appropriate steps to prevent and
counter-act such errors is divine.