Welcome

Occasionally, I feel the need to transfer thoughts from some corner of my mind to some forgotten corner of the blogosphere. So this is the space where I do that.


The postings here are a good cross-section of my interests. There are quite a few posts on some philosophical thoughts. There are also more professional posts on areas of strategy, IT Management, and data science.


I hope they are enjoyable and thought-provoking to read. Please leave comments and let me know what you think. I would enjoy the opportunity to engage in a conversation on these topics.


Sunday, November 6, 2011

Scientifically Speaking

One of the big challenges faced by scientists is communicating their findings to the public.  Very often, it is the case that they report one thing, but the public hears something completely different.  Scientific language often gets interpreted differently by people who only speak Common English and not Scientific English.

For example, let's take one of the tactics used by Creationists, those people who refuse to believe the fact of Evolution because it contradicts their religious beliefs.  Creationists like to make a point that evolution is "only a theory".  In common usage, a theory is just a guess that someone makes.  We say things like, "I have a theory that the dog ate my sandwich".  It's a guess we make, and we're going to go out and look for evidence of it.  In Scientific Language, a theory is the current best explanation that we have for something that we observe in the universe.   A theory of gravity explains how we think that gravity works.  We all know that gravity exists.  We don't say, "I have a theory that gravity exists".   No, it's an observed fact.  However, scientists have a theory of how gravity works that explains gravity and takes into account relativity and quantum physics.   In Scientific Language, the word we use for a guess is a hypothesis.  A scientist would say, "My hypothesis for the disappearance of my mid-day sustenance is that the canine inhabitant of this domicile has consumed it".

When we talk about the Theory of Evolution, we're talking about our explanation for how evolution has occurred and what mechanisms drive it.  We're not making a guess that evolution could be true.  Evolution is an observed fact of nature.  There is no doubt about it.  However, there is still scientific debate over some details of the explanation.  Such debate is healthy because it is how we learn more.  It doesn't indicate that there is doubt about whether evolution is true or not.

Another example where I see this language disconnect is when talking about the results of some study or experiment.  For example, suppose a study was performed that was aimed at determining if acupuncture was effective for reducing pain.  The scientists would report, "the study showed no evidence that acupuncture was effective at treating pain".   The problem here is that to say "the study showed no evidence" leaves room for the reader to think that there could be evidence, but it just wasn't found by this study.  It's like saying, "there is no evidence that there was a second shooter who really killed President John F. Kennedy".  People who want to believe in a second shooter can easily believe that the investigators simply didn't look in the right place to find the evidence of the second shooter.

Here's where the real problem is.  In fact, scientists know that no study is final proof of something.  There is always more to know.  In addition, experimental studies are carefully designed to test specific hypotheses.  There are very strict conditions specified that are used to determine if the hypothesis is correct or not.  So a given study is only looking at one part of a problem.  There could be other evidence that this study wasn't looking for, and thus wouldn't find.  Every theory is tentative - it could change if more evidence is discovered.  That's the strength of science.

On the other hand, the fact that an experiment was conducted and had negative results is itself evidence.  If we test acupuncture and find that there is no difference in pain whether acupuncture was used or not, then we have evidence that acupuncture did not improve the pain.  We shouldn't say "there is no evidence that acupuncture improved the pain".   We should say, "the evidence of this study shows that acupuncture does not improve pain".

In some cases, many experiments have been done with no positive results. At that point, scientists should be clear and simply say, "given all the evidence that has been gathered, we conclude that acupuncture does not work".   Simply adding the statement about having gathered data gives more weight to the conclusion.   As scientists, we have to be true to our value of being accurate in what we say.  However, if we are going to communicate our findings in a way that people can understand and use to make better decisions in their lives, then we have to speak in a way that is understood by people who don't speak Scientific English.

No comments: