Section No. 1 | 2 | 3 | 4 | 5

By DAVID NG

“You see something interesting…”

This little phrase is often the start of the scientific process. In that it all begins when someone, possibly you, has noticed something intriguing. This doesn’t mean that it has to be interesting to everyone – just as long as it’s interesting to someone. In fact, sometimes, the science will stop right there. In other words, the act of just “observing” might be good enough – think about how everyone would feel if you were the first to discover a certain kind of creature.

Still, most conventional views of science would assume that you’ve seen something curious enough to merit the question “why?” And it is in that inspired act of asking a question, where arguably the most important part of the scientific method takes form.

We are, of course, referring to the notion of the hypothesis: which according to the Oxford Dictionary is defined as:

“A supposition or proposed explanation made on the basis of limited evidence as a starting point for further investigation”

For us, in less eloquent terms, we say that this is the part where you try very hard to “think of a reason why.” Furthermore, when you do this, you inadvertently set the scene for the next stage of the method by defining how a person might “figure out ways to check your reason why.” To a scientist, this last phrase is a colloquial way of talking about experiments.

For fun, let’s explore these concepts by using an example. Here, we’ll focus on some interesting observations that were noted in China during the early 1980’s. Essentially, what folks observed was that there was a discernable decline in Chinese stork numbers [1]. As well, there was also a drop in fertility rates [2]. In other words, storks in China were disappearing and the Chinese appeared to be having less babies.

But why?

At one level, we might suppose that the two are not at all related.  It could simply be a correlation and nothing more. But for the purposes of our discussion, let us suppose that we are trying to surmise whether the two are ultimately connected – whether there was truly a causative element involved.

Here, some of the hypotheses might assume that there is a continuum involved, in that one of the observations is actually directly responsible for and logically leads to the other observation. For instance, if we play into the stork/baby mythology, where storks do indeed deliver babies, perhaps we can say that the decline in stork numbers was in turn causing the baby effect. Others, however, might ponder whether there is a more central reason for the two trends. In this case, we might talk about a hypothesis that suggests one prominent thing at play that is simultaneously responsible for both outcomes.

Here, we can try to distinguish the two scenarios by looking at the evidence more closely. Does the stork decline happen before or after the baby decline? What exactly are the numbers associated with the declines? Important, because even with the drop taken into account, the actual numbers of storks might still be more than enough to cover the number of babies born. In any event, as you can see, a hypothesis can be quite nuanced and is really only a small step in a much longer path.

For amusement’s sake, let’s take this idea of nuance even further and look at the three potential hypotheses presented below [3]. Here, they all focus on a core reason (environment, economics, or aliens) that could explain our observations.

Looking at these flowcharts more carefully, you can see that when accessed logically, they all work. Even with the somewhat interesting inclusion of aliens, the fact remains that all three could be considered acceptable, worthy even. However, this is very different from a hypothesis being valid. Validity, which aims to make sure that what you say is indeed true, or at least true under every logical interpretation, is a much higher bar to meet. It is something that needs to be earned through the critical examination of evidence.

Now this is an especially important word, so we’ll once again invite the gravitas of the Oxford Dictionary to provide a definition:

“Ground for belief; testimony or facts tending to prove or disprove any conclusion.”

But such grounds can take several different forms, constituting strong or weak evidence.  If we focus on the alien claim in particular, evidence might look a little like this:

1. We found an alien! And we have proof!

Here, we have important evidence from the point of view of addressing one critical question: are aliens real? It is crucial because it could be said that this detail is a major stumbling block in the alien hypothesis. However, proof of the existence of aliens isn’t in of itself strong evidence to support the hypothesis. This is because it doesn’t address any of the specific ideas and mechanisms put forth to explain our stork and baby narrative. Ideally, you would want to see data that demonstrated the involvement of our said alien with either storks or babies – actually, you would like to see both.

2. We found an alien eating a stork! We also found an alien with a baby on a leash! And we have proof!

This type of evidence is better, but it is still technically weak. This is because just having this data isn’t necessarily conclusive. What if the stork you see is, in actual fact, American? What if the pet baby is not Chinese? What if it is Chinese, but not in fact, from China? What if it is a result of alien cloning techniques? As you can see, the scientific mind will take what might otherwise appear convincing, and deconstruct it skeptically. A scientific mind will continually probe, and continually look for flaws in the evidence.

3. We found an alien eating a stork, and we have biochemical proof that the stork is from China! We found an alien with a pet baby, and we even saw the alien take the baby from a family in China! And we have proof!

Now, we’re getting closer, but now the issue is in the matter of whether this evidence represents an impactful occurrence. In other words, this particular data is really only good for showing the loss of one stork and of one baby. Obviously, this can hardly validate the observation that whole populations have dropped, which means that better evidence would also provide a better sense of the numbers involved. This particular stork and this particular baby could represent a simple coincident.  Furthermore, what if this data was flawed for other reasons? Perhaps, if we had decided to observe the baby a little longer, we would have noticed that the baby was in fact being kept for food! This doesn’t change the fact that aliens may still be responsible for the drop in numbers, but it does nevertheless alter the sentiment of the current hypothesis significantly.

Of course, this type of review can go on and on. And the thing is: it does. What we have here is this continual cycle of coming up with hypotheses, coming up with ways to address the hypotheses, coming up with evidence, and then reevaluating everything over and over and over again.

Hopefully, you can see why this can very easily become an arduously slow process, although that’s not to say that it is always slow. More importantly, you should be able to appreciate how the process can lead to varying outcomes.  It could lead to revisiting old ideas.  It could naturally result in conflicting views.  It could even cause your explanation (and also possibly the scientific discipline) to change directions dramatically. Imagine, if you will, that the real answer to our stork and baby scenario was a little bit about everything – a little mythology, a little environment, a little economics, and even a little bit about aliens. If this were the case, you could probably appreciate how difficult that complete story might be to tease out.

Here’s another thing to note:  if you think about this process carefully, you will soon realize that the continual acquiring of scientific evidence never actually proves anything to be one hundred percent certain. It can only modify or support an existing hypothesis, although by supporting it relentlessly, a hypothesis can get stronger and stronger and perhaps one day rise to the rank of a scientific theory or scientific law. But even there, there is no certainty that there won’t be something that comes along to discredit that idea in a single stroke. This is Karl Popper’s take on the philosophy of science: that at the end of the day, you cannot prove something to be true: you can really only prove something to be false.  This might take a moment to ponder, but if you do so with our alien example, you’ll note that this description does fit.

On the whole, our little alien discussion hopefully provides a window into how the scientific method works. But if we are honest with ourselves, we should also admit to glossing over something very important. Specifically, it’s in the parts where we have very nonchalantly uttered the phrase, “And we have proof!”

This bit, we will spend some more time on in the next section, as it considers how we distinguish strong evidence from weak evidence. Which is all the more daunting these days, since it’s quite likely you might not even understand the technical details of the evidence. Indeed, it might even be completely alien to you.

Notes:
1. Ma, Ming; Dai, Cai. “The fate of the White Stork (Ciconia ciconia asiatica) in Xinjiang, China”. Abstract Volume. 23rd International Ornithological Congress, Beijing, August 11–17, 2002. p. 352.

2. S Horiuchi. “Stagnation in the decline of the world population growth rate during the 1980s.” Science 7 August 1992: Vol. 257 no. 5071 pp. 761-765

3. Pollution and economic scenarios via personal communication with Hadi Dowlatabadi.

(3rd draft)