After
taking a long flight from Australia to Pakistan, my mother-in-law was suffering
from sharp otalgia, or ear ache. She consulted three doctors and each suggested
a different treatment. One of them prescribed oral antibiotics, the other
advised chewing gums to release ear pressure and using over-the-counter ear
drops, and the third one explained in detail why a surgery was inevitable. My
mother-in-law, petrified at the word ‘surgery’, decided to take antibiotics and
ear-drops to alleviate her discomfort. It did work out, but the proposition of
a surgery still disturbs her.
Doctors
differing in diagnosis for the same disease is not something new. This
unpleasing inconsistency in professional judgment is a case of noise; an
often-overlooked shortcoming of the human behavior that is described in detail
in the book “Noise: A Flaw in Human Judgement.” This book is authored by the
celebrated Nobel laureate Daniel Kahneman (writer of “Thinking, Fast and Slow”),
former McKinsey partner and Professor at HEC Paris Olivier Sibony, and esteemed
legal scholar and co-author of “Nudge” Cass R. Sunstein. The book discusses in length how to detect and
measure the presence of ‘noise’ and the remedial ways to reduce it.
The
book begins with a distinction between noise and bias. While bias is systematic
error; presenting a regular deviation, noise refers to a random scatter. In
simpler words, noise is attributed as the inconsistency in decisions, made by
different people or even the same person repeatedly, in decisions or
circumstances where uniformity is expected. With the help of scenarios, studies,
and statistics, the authors stress the presence of noise that plagues many
fields and situations, such as medicine, law, education, and performance
reviews. An error in judgement in such fields may induce people to lose faith
in their institutions.
With
every moment of every day, the human mind makes judgements. When individuals make
noisy decisions, they err randomly, but when they make biased judgements, their
decisions show a similar pattern with every observation. Decisions made by
different individuals vary, but that variation – or noise – is also present among
groups of individuals. Kahneman, Sibony and Sunstein share their insights on
why humans are vulnerable to noise when making decisions, how to identify
noise, and how it could be avoided in order to make accurate judgments that
will help shape a fairer world.
The
authors proceed by explaining the different types of noise – pattern noise, and
occasion noise. They assert that noise in systems is partly because different
professionals follow different practices. In the field of medicine, for
instance, there is startling evidence that different medical practitioners
agree only two-thirds of the time to the same treatment of a disease. This is
known as “pattern noise”: a case where professionals differ with each other on the
same issue. An example of pattern noise in judicial system, as penned by the
authors, notes, “Two men, neither
of whom had a criminal record, were convicted for cashing counterfeit checks in
the amounts of $58.40 and $35.20, respectively. The first man was sentenced to
fifteen years, the second to 30 days.” Judges should be interchangeable, but if
the verdicts for similar cases vary by such huge margins, the judicial system
becomes flawed. Such variations add more error in the system than cancel
each other out.
The
authors suggest that systems are full of noise because specialists who work in
the same field commit contradictory practices. They name it “occasion noise” –
disparity in judgments caused by mood, time constraints, weariness, or weather
conditions – factors that should not affect decisions but, in fact, do. For
instance, teachers give better grades if they recently received a salary raise.
Noise
is inherent whenever judgments are made. But what could be done to minimize
this noise? Should these noisy systems be replaced with machines? Or, how about
only considering a group’s average judgement and disregarding the individual
one? The authors are extremely reluctant to agree on these proposed solutions. They
are not willing to give up on human judgements, but rather recommend people to
adopt better “decision hygiene”. Decision hygiene refers to procedures that
help reduce noise. One such procedure could be to have multiple independent
judgements, and taking an average of it. That is to say that if there are five
members in a group, then the average of individual judgments should be
preferred to a collective group judgement.
Since
systems are distressed with noise, the authors suggest conducting a “noise
audit” – a method to identify noise and improve the quality of decisions. If
organizations perform regular noise audits of their team, it will allow them to
find the causes of these variabilities and will enable them to implement
processes to alleviate the level of noise in decision making structures. Just
as organizations and policy-makers are employing “nudges”, if they employ
“decision hygiene” and “noise audits”, they will be in a better position to
make fairer decisions that would save lives, time, money, and resources.
Noise: A Flaw in Human Judgement
is a lengthy, but a well-written and captivating book. It shows us the many
instances where we become prey to bad decisions. Every leader, academic and
decision-maker must specifically read the parts where the effects of noise are
emphasized and how to reduce them. Also, people who are interested in learning more
about cognitive biases will find the book a value addition to their knowledge. Although
the book is assembled with stories, experiments, and evidences, some of its
chapters are too scholarly to be understood by many. Especially readers who are
not well-versed in statistics will find the book a bit too technical and
boring.
By Daniel Kahneman,
Olivier Sibony, Cass R. Sunstein
Little, Brown Spark.
454 pp.
