I was reading and wondering what the half-life of market-related facts
Truth decay: The half-life of facts
Much of what we believe to be factual has an expiration date, but the good news is that we can see it coming
IN DENTAL school, my grandfather was taught the number of chromosomes in a human cell. But there was a problem.
Then in 1956, Joe Hin Tjio and Albert
Levan tried a new technique for looking at cells. They counted over and
over until they were certain they could not be wrong. When they
announced their result, other researchers remarked that they had counted
the same, but figured they must have made a mistake. Tjio and Levan had
counted only 46 chromosomes, and they were right.
Science has always been about getting
closer to the truth, and anybody who understands it knows that a
continual transformation of accepted knowledge along the way is how it
works. However, sometimes it can feel random and unsettling. Smoking has
gone from doctor-recommended to deadly. Eating meat used to be good for
you, then bad, then good again; now it's a matter of opinion. The age
at which women are told to get mammograms has risen. We used to think
that Earth was the centre of the universe, and our planet has since been
demoted. I have no idea any longer whether or not red wine is good for
me.
It turns out that there is order
within the shifting noise. The good news emerging from my field of
scientometrics - the quantitative study of science - is that in the
aggregate there are regularities to the changes, and we can even
identify how fast facts will decay over time. This means we don't have
to be all at sea in a world of changing knowledge. Portions of what we
assume to be true will eventually be overturned, but realising that
patterns exist could help to identify the apparent truths that are
poised to expire.
It is obvious that scientific
knowledge is continually updated through new discoveries and the
replication of studies, but until recent years little attention had been
paid to how fast this change occurs. In particular, few had attempted
to quantify how long it would take what we know at any given moment to
become untrue, or replaced with a closer approximation of the truth.
Among the first groups to measure this
churning of knowledge was a team of researchers at Pitié-Salpêtrière
hospital in Paris, France. To get a handle on it,
Thierry Poynard
and his colleagues chose to focus on medical fields in which they
specialised: cirrhosis and hepatitis, two areas related to liver
diseases. They took nearly 500 articles in these fields from over 50
years and gave them to a panel of experts to examine. Each expert was
charged with saying whether the paper was factual, out-of-date or
disproved (
Annals of Internal Medicine, vol 136, p 888).
Through doing this, Poynard and his
colleagues were able to create a simple chart that showed the amount of
factual content that had persisted over the previous decades (
see diagram).
They found something striking: a clear decay in the number of papers
that were still valid. Furthermore, it was possible to get a clear
measurement for the "half-life" of facts in these fields by looking at
where the curve crosses 50 per cent on this chart: 45 years.
Essentially, information can be
likened to a radioactive material. Medical knowledge about cirrhosis or
hepatitis takes about 45 years for half of it to become out of date or
disproved.
The half-life metaphor does not match
up exactly to its radioactive namesake. For one thing, the time frames
being examined make it difficult to see if the decay is truly
exponential. Similarly, depending on the maturity of the field, the
half-life need not be constant. Indeed, it has surely changed as
medicine transitioned from art to science. Nonetheless, half-lives can
be a useful way of thinking about the decay of knowledge.
We can't predict which individual
papers will be overturned, of course, just like we can't tell when
individual radioactive atoms will decay, but we can observe the
aggregate and see that there are rules for how a field changes over
time. The cirrhosis and hepatitis results were nearly identical to an
earlier study that examined the overturning of information in surgery.
Two Australian surgeons found that half of the facts in that field also
become false every 45 years (
The Lancet, vol 350, p 1752)
Unfortunately, convening a panel of
experts to comb through all of science's past conclusions isn't
feasible. So we have to sacrifice precision for our ability to look at
lots of science relatively quickly. One simpler way to do this is by
looking at the lifetime of citations, the coin of the scientific realm
and the metric by which the impact of a research paper is measured.
To understand the decay in the truth
of a paper, we can measure how long it takes for people to stop citing
the average paper in a field. Whether it is no longer interesting, no
longer relevant or has been contradicted by new research, this paper is
no longer a part of the living scientific literature. The amount of time
it takes for others to stop citing half of the literature in a field is
also a half-life of sorts.
Through this we can begin to get rough
estimates of the half-lives of many fields. For example, a study of all
the papers in the
Physical Review journals, a cluster of periodicals of great importance to physicists, found that the half-life in physics is about 10 years (
arxiv.org/abs/physics/0407137).
This is the opposite of what is found
in journal articles, where knowledge at the frontiers of the hard
sciences is overturned more rapidly than in the social sciences. This
may be because immediate repeated experimentation can be more
straightforward in the physical sciences than in the social sciences,
where the data can be noisier.
So drawing strong conclusions about
the differences between the calculated half-lives of disparate fields
should be approached with caution, because of the various influencing
factors and different methods employed to gauge their longevity. Still,
the point remains that specific bodies of knowledge can be viewed as
having different expiry dates. This leads one to wonder whether
reframing knowledge in this way should change how we interpret the facts
that each of us use to navigate the world. After all, identifying that
certain bodies of facts will endure longer than others might influence
how we act on information - and identify what facts we should be most
wary of.
The spectrum of truth
Let's imagine that we were to line up
facts on a spectrum according to how quickly they change. On the far
left, we have the groups of fast-changing facts, which are constantly in
flux. These would be our knowledge about what level the stock markets
closed at yesterday, or predictions for weather around the globe. You
might say that these areas have tiny half-lives: the length of time they
stay true is very brief.
On the far right, we have the very
slow-changing knowledge, featuring the facts that for all practical
purposes are constant. Individually, this might include what we've
learned about the number of continents on the planet, and on a group
level, you might place pretty much everything the ancient Greeks wrote
about geometry.
In between, we would have the groups
of facts that change, but not too quickly. This knowledge might change
over the course of years or decades or a single lifetime. These facts in
the middle are what I refer to as "mesofacts" - and their relatively
slow rate of change may mean that many people fail to acknowledge their
ultimate transience and looming expiry.
Knowledge about nutrition might be
considered to be one of these mesofactual areas. For example, in the US,
we no longer use the four basic food groups - meats, dairy,
cereals/breads, fruits/vegetables. There are now five food groups placed
on a plate. Between these points there was a food pyramid that
underwent multiple overhauls. We have also experienced long-term changes
with regard to whether we should eat fatty foods, carbohydrates and
many other things.
Another mesofactual realm might be how
we are supposed to take care of babies. Each generation has a set of
facts that changes, from whether babies should sleep on their backs or
stomachs to whether pregnant women can safely smoke or drink alcohol.
Mesofacts are all around us, and just
acknowledging their existence can be useful. After all, as studies of
our perception of slow shifts in the natural world have revealed, we are
often blind to gradual change (see "
Too slow to see").
Of course, this doesn't mean that
everything is going to be overturned. If a government-funded study
promotes a certain health practice, for example, we should not
immediately dismiss it as being based on one of many evanescent facts.
That would be foolishly throwing the informational baby out with the
bath water. But we should not be shocked if that advice is later
contravened.
Fortunately, we are getting better at
internalising this particular truth: many medical schools inform their
students that within five years, half of what they have been taught will
be wrong - and the teachers don't know which half it will be.
Thankfully, facts don't expire
arbitrarily. Even though knowledge changes, the astounding thing is that
it changes in a regular manner. If we recognise this, we'll be better
prepared to live in the rapidly changing world around us.
![[IMFresets.jpg]](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi-T5ph9NkNfhPgAcbrfDDKxTmrRZvImWZLum0mcfZVOSYgL7RyISA29H4d_aaowO-J7qxBFFSvSiKopnB8Zue2wcSq8fCtJN_Tw6tpXHCEa9FSoy0S59nUoT8a3FW5oVZOxGZNufaaSJI/s400/IMFresets.jpg)
