Tag Archives: scientific method

“The sun’ll come out, tomorrow, betcha bottom dollar…”

“The sun’ll come out, tomorrow”

Whenever people end a Facebook post – or a tweet – with “do your own research,” there’s a chance that they’ve done no such thing themselves. Or to be more specific, they’re unlikely to have gone through a set of procedures that are what folks who actually do research for a living would go through. There are hundreds of books out there that explain, in excruciating detail, what “research” is but I’d bet money on the probability that the number of folks who have read ONE book on the research process is significantly lower than the number of people who believe the world is secretly governed by lizard people [1], that aliens regularly visit the earth [2], and that if there is no evidence for X, that in itself is proof that X exists [3].

Research is essentially a tool used by scientists to test the veracity of hypotheses. Or in words of one syllable, a way to find out if what you think is true is, in fact, true. But let’s go down the rabbit hole a little more here and try to avoid the words “true” and “truth” because like “commonsense,” so many people claim to know what it is but often turn out to be as subjective in their thinking as everyone else. For many people, “truth” is simply what they believe to be true and may have zero bearing on any actual, observable, demonstrable, measurable quality. The majority of people reading this will believe in some supernatural deity (or deities) with absolute conviction that this is “true,” yet given the opportunity to provide any supporting evidence that is objective and replicable, they will fail stunningly. Faith is not Truth, and Truth is not Belief. Anyone can have faith in something or someone, and anyone can believe in any ridiculous things they like. As I’ve said for years, the First Amendment of the US constitution pretty much guarantees your right to be a complete idiot, and it’s a testament to its power that so many people are making the most of it.

Deciding that something is “true” turns out to be a devilishly difficult and intellectually irritating thing. Even when Little Orphan Annie sings, “The sun’ll come out, tomorrow, betcha bottom dollar that tomorrow, they’ll be sun,” the scientist in you should be thinking, “Oh yeah, how do you know that’s true?” Annie is definitely on my list of “irritating imps” from literature, with the number one spot reserved for Dickens’ Tiny Tim, but she can’t be blamed for at least attempting to make a statement that has all the appearance of being unassailably truthful.

To be more accurate, we’d first have to make sure we all shared the same definitions of the hypotheses she’s espousing; that “The sun will come out tomorrow.” In relation to “the sun,” it would be churlish to suggest that someone might interpret what she’s referring to as Canis Major or Corona Borealis but if we want to be as accurate as we can, we’d make explicit that “the sun” is the star closest to Earth some 93 million miles away. Then, by “come out,” we mean that it will take a position in the sky above the horizon, regardless of whether it is obscured by clouds, smoke, overhanging branches, or a roof. That way is someone tries argue that, “I was in my cellar and didn’t see the sun so it didn’t ‘come out'” you can happily sneer and smack them on the head with a large dead cod and remind them of the definition. This also infers that we must have some way of measuring the act of “coming out” that is objective, which means it doesn’t really on belief, opinions, points-of-view, or any other internal mental phenomena), and repeatable by others, which means anyone else could do the same measurement. Finally, “tomorrow” also needs to be defined as being a point 24 hours after an initial point. If it’s 48 hours before it appears again, then that’s not tomorrow but “every two days.”

But it doesn’t stop here, does it? The location in which I take the measure may also affect the truth of the statement. It may work for Annie who’s in New York city but what about Pippi Longstocking on her visit to grandma’s house in Svalbard, Norway. Being inside the Arctic circle means she’s singing, “The sun’ll come out, in April, betcha bottom Krone that in April, they’ll be sun.” Clearly there is still a sense that “the sun will come up” at some point but it’s not as clean cut as it appears in the Tomorrow song.

Now, given that we’re happy with defining our terms and restrict our experiments to places closer to the equator, what happens when we find that we see the sun rise on Jan 1 at 7:20 AM, then again on Jan 2 at 7:21 AM. Can we say, “Yes, it’s true, the sun’ll come out tomorrow?” You could try and maybe many people would believe you. But the problem is that this is based on ONE sample and it’s a very poor piece of research that takes one sample and concludes it’s the truth. That’s like watching one video on YouTube of someone checking ballot papers and then concluding that the entire election was a scam (unless it was a vote for your candidate, in which case it was true!) No, you’d want to take another sample 24 hours later on Jan 3. Then Jan 4. Then Jan 5. And keep doing that until… well, when?

This is another feature of research that people can misunderstand; that there is an “end” to research [4]. Their understanding is that “research” and “scientific analysis” provide truth and “an answer” when, in reality, that’s not quite right! Back with the rising sun, if you were to say it’s true that the sun will come up tomorrow, what you’re really saying is “given that we’ve observed a sunrise since records began, and that the laws of physics which apply to the universe and how planets revolve around suns have not changed, there is probability close to one that the sun will come up tomorrow.” Is there a chance that the sun may NOT come up? Could there be an unforeseen cosmic accident (black hole hitting the sun) that may wipe out sun and the Earth during the night? Well, maybe. And that is where all research ends up.

Research is about creating hypotheses – ideas that may or may not be true – and then dreaming up ways to test them. You determine in advance what measurements you consider will support the hypotheses and after testing, see if those measures were achieved. If they were, then you can say, “yeah, the objective measure support my hypothesis and it’s therefore worth considering as ‘useful’ or ‘important.'” The more measures you take and the more support you find, then the more meaningful the results. Better still, if lots of other people do the same measures and get the same results, then your hypothesis becomes tougher, stronger, and develops a bit of a cocky swagger.

But although thousands upon thousands of repeated experiments can support a hypothesis, it only takes ONE discrepancy to shatter it into tiny pieces. This notion is sometimes called the “Black Swan” approach. If you set up the theory that “All swans are white” you can spend years out in the world looking at swans and see that every swan you comes across is white. You can then have hundreds of other bird watchers report back that every swan they’ve seen is white. But as soon as ONE person person provides evidence that they’ve found a black swan, the “All swans are white” is now wrong; it’s downgraded at best to “Not all swans are white” or “Most swan are white” or “All swans in the world, except for in the village of Little Turdington that are black, are white.”

Of course, this also introduces the issue of determining what constitutes valid data. The ornithologist from Little Turdington may have taken a photo of a white swan but, armed with a copy of PaintShop Pro 2021, changed the color to black in an attempt to discredit the original hypothesis. Real researchers wouldn’t take that one photo and accept it as proof but more likely send out lots of other researchers to LIttle Turdington to see if they, too, can observe and measure black swans. As more evidence of black swans appears, then the truth, or veracity, of the original claim is diminished.

Science progresses step by step by step. One hypothesis leads to another and depends on what has gone before. In the development of a vaccine for COVID-19, one of the claims by vaccine-skeptics is that it was developed too quickly and not tested enough, which completely ignores that fact that work on similar vaccines has been going on for years, and that all the work that has gone before always sets the scene for what comes next. Sir Isaac Newton, whose “Theory of Gravity” is still a theory [5], is credited with saying that, “If I have seen further, it is by standing upon the shoulders of giants,” and that’s pretty much what research is all about.

And if you don’t believe me, do your own research šŸ™‚

Notes

[1] According to Wikipedia, the increasingly deranged ex-sports commentator, David Icke in the UK, believes that “tall, blood-drinking, shape-shifting reptilian humanoids from the Alpha Draconis star system, now hiding in underground bases, are the force behind a worldwide conspiracy against humanity.” He is, sadly, not alone. Proof, should you need it, includes videos of famous people (the Queen, Hillary Clinton, the Pope) all blinking like lizards, so it must be true.

[2] Given that (a) reports of aliens from outer space visiting earth have been around for as long as folks have been writing and (b) not one has yet gone through the “take me to your leader” process, it makes me wonder just how “alien” their thinking must be. Imagine investing unimaginable amounts of time, money, and energy into building interstellar spaceships that can carry you anywhere in the universe, traveling millions and millions of miles -perhaps over immense periods of time – and then landing in some backwoods to pick up a random person so you can stick an anal probe up their arse before flying back home. Now that IS some alien psychology!

[3] The best conspiracy theories are the ones that evolve in such a way that they become completely impervious to any evidence against them, and where all and any statements are “proof” of the theory. Any “proof” of election fraud in the US 2020 election is seen by conspiracy theorists as “clear evidence” for the stolen election; any absence of evidence of fraud is “proof” that the fraud was effective and therefore also evidence for the stolen election! According to one of my favorite philosophers of science, Karl Popper, this type of thinking is referred to a “metaphysical” because it is incapable – by design – of being tested by any objective measurement and is, ergo, non-scientific. Popper was no fan of Psychoanalysis and always deemed it a metaphysical philosophy rather than a science. Religion, as a phenomenon, is also metaphysical in nature in that there are no measurements we can do to test for the existence of a supernatural entity, or entities (it seems unfair to throw Odin, Thor, and Loki into the “myth basket” while allowing Jehovah and Allah special status as “real”).

[4] All research article end with “much more research is needed,” which is (a) probably true and (b) setting the scene for the next grant application. If anyone can cite a paper that concludes with “and, ergo, there is no need to look at this topic anymore because we now have all the answers” I’d love a copy.

[5] Just because something is labeled a “theory” doesn’t imply that is is somehow wrong. Both “gravity” and “evolution” are spoken of as “theories” but if you chose to believe gravity isn’t “real,” I suggest throwing yourself off a high building and seeing how well your disbelief stops you from hitting the ground at 150 miles per hour and leaving you intact enough to stand up and say, “See, I told you it was just a theory.” In this case, you’d be well advised to treat the “theory of gravity” as a practical reality.

“I don’t care what the research says…”

A colleague of mine was asking for some references to support the notion that kids with severe learning difficulties can learn to use high frequency core words (such as want, stop, and get) because they were being told that what these kiddos really use (or need) are words like toy, cookie, and banana. I duly provided a quick sample of peer-reviewed articles and shared the information with other colleagues. And what the hell, I’ll share them with you, dear reader, in the References section at the end of this piece.

Reading the research

Reading the research

But another of my friends also commented that there are still those folks who respond with comment such as, “I don’t care what the research says, I don’t care who these kids are. These are not the kids I’mĀ working with. The kids I’m working with just aren’t going to use these words.”

So what do you do about this? At what point does being “critical of the research” become “ignoring the research because I don’t believe it.”? In the world of Physics, it’s hard to say, “I don’t care what the research says, I’m still going to fly using my arms as wings.” Mathematicians don’t say, “I don’t care what the research says, 1 + 1 does equal 7.” And it’s a brave doctor who would say, “I don’t care what the research says, you go right ahead and smoke 40 cigarettes a day and you’ll be just fine.”

No-one would argue that Speech and Language Pathology as a profession will ever achieve the rigid, statistical certainties of physics and mathematics, but what doesĀ it say about our profession if we openly admit to ignoring “the research” because it doesn’t fit with our individual experience? There are certainly enough practicesĀ  in Speech Pathology that are hotly debated (non-speech oral motor exercises, facilitated communication, sensory integration therapy) and yet still being used. But all of these are open to criticism and lend themselves to experimental testing, whereas an opinion based on personal experience is not. I could tell you that I have used facilitated communication successfully, but that is still personal testimony until I can provide you withĀ  some measurable, testable, and replicable evidence. This is one of the underlying notions of evidence-based practice in action.

However, it’s Ā one thing to talk about using evidence-based practice but another to actual walk the walk. If the evidence suggests that something you are doing is, at best, ineffective (at worst, damaging), how willing are you to change your mind? If 50% of research articles say what you’re doing is wrong, how convinced are you? What about 60%? Or 90%? At what level of evidence do you decide to say, “OK, I was wrong” and make a change?

If there’s anything certain about “certainty” it’s that it’s uncertain! Am I certain that teaching the word getĀ to a child with severe cognitive impairments is, in some sense, more “correct” or “right” than teaching teddy? No, I am not. But what I can do is look at as many published studies of what words kids typically use, at what ages, and with what frequency, and then feel more confident that get is used statistically more often across studies. This doesn’t meanĀ teddy is “wrong,” nor does it preclude someone publishing an article tomorrow that shows the word teddy being learned 10x faster than the word get among 300 3-year-olds with severe learning problems.

But until then, the current evidence based on the research already done is, in fact, all we have. Anything else is speculation and guesswork, and no more accurate than tossing a couple of dice or throwing a dart at a word board.

Being wrong isn’t the problem. Unwillingness to change in the face of evidence is.

References
Banajee, M., DiCarlo, C., & Buras Stricklin, S. (2003). Core Vocabulary Determination for Toddlers. Augmentative and Alternative Communication, 19(2), 67-73.

Dada, S., & Alant, E. (2009). The effect of aided language stimulation on vocabulary acquisition in children with little or no functional speech. Am J Speech Lang Pathol, 18(1), 50-64.

Fried-Oken, M., & More, L. (1992). An initial vocabulary for nonspeaking preschool children based on developmental and environmental language sources. Augmentative and Alternative Communication, 8(1), 41-56.

Marvin, C.A., Beukelman, D.R. and Bilyeu, D. (1994). Vocabulary use patterns in preschool children: effects of context and time sampling. Augmentative and Alternative Communication, 10, 224-236.

Raban, B. (1987). The spoken vocabulary of five-year old children. Reading, England: The Reading and Language Information Centre.

Quackery, Hokum, Baloney: Separating Science from Stupidity

Suppose I told you that somewhere between Earth and Mars there is an invisible teapot that orbits the sun once every 666 days. The teapot is invisible because it is cloaked using technology developed by space aliens, who left it there to monitor our progress. They believe that once we make contact with the teapot, an alarm will sound and they will return to see if we are truly worthy of being galactic citizens.

Teapot in space

The Orbiting Teapot

The question you need to ask is; “Is that true?” and if so, “How do I know it’s true?” This is, of course, the fundamental question for Science. What do we know and how do we know it.

The invisible teapot was created by theĀ philosopher Bertrand Russell back in 1952 and went like this:

If I were to suggest that between the Earth and Mars there is a china teapot revolving about the sun in an elliptical orbit, nobody would be able to disprove my assertion provided I were careful to add that the teapot is too small to be revealed even by our most powerful telescopes. But if I were to go on to say that, since my assertion cannot be disproved, it is intolerable presumption on the part of human reason to doubt it, I should rightly be thought to be talking nonsense.

But this is the 21st century, after all, and we are all sophisticated, intelligent people, and we have a wealth of scientific knowledge and instrumentation to help us test for the presence of the teapot. The “powerful telescopes” of 1952 have been replaced by much more sophisticated technology and we can now see much more of our solar system.

Bertrand Russell

Bertrand Russell

In principle, therefore, we could focus all the world’s telescopes along the elliptic plane and search for the pot. We already know what it looks like, we know it is between Earth and Mars, and we know it is cloaked. The cloaking business may make it trickier but we also know from current research that “cloaking” is little more than deflecting light around a mass. We could spot the teapot by looking to see if there is a teapot-sized region of space that makes stars behind it appear to change position; this is because the mass of the pot will cause light to bend ever so slightly (it why in a solar eclipse we can see stars on the edge of the sun that are actually behindit).

The key thing to note here is that we TEST for the presence of the teapot and refuse to accept it on faith. I may be able to spin the most wonderful story about the pot, about how beautiful and splendiferous it can be, and how much it has changed my life, but if all I have is my personal perceptions and ideas, you would be right to treat what I say as bullshit of the highest order.

The only way for me to prove that I am right is to provide evidence of the pot. If the telescopes suddenly reveal a sea-green piece of revolving pottery, orbited by teacups (hey, there may be more to the teapot than I knew!) then you should start taking me more seriously. When several independent observatories have pictures, and all independently identify its location by numerical coordinates, and spectrograph analyses all show its chemical structure, then I’m pretty much vindicated. And if after a few years NASA’s latest “Pot Probe” reaches that location and scoops up the teapot into its gaping maw, then that’s likely to be as much proof as any reasonable person would require to be able to say, “Yes, there IS a teapot in outer space.”

Testability is a cornerstone of Science. And the thing that has to be tested is a HYPOTHESIS[1], which is defined as;

A proposition or principle put forth or stated (without any reference to its correspondence with fact) merely as a basis for reasoning or argument, or as a premiss from which to draw a conclusion.

The aim of Science is to test a hypothesis, that is, to see if it is true or false. Now in reality, you don’t actually prove something to be “true,” you “support” it. Truth and support are two very different things. If my hypothesis is “All swans are white,” I can test it by sitting by a river bank photographing every swan than lands on the water in front of me. If I have several friends across the world do the same thing, we might find that all the pictures we have turn out to be white swans. Does this mean that “All swans are white” is true? Nope, it just means that there is overwhelming support, based on many observations and measurements by many people, that swans are white. However – and here’s the kicker – if we find just ONE black swan, the hypothesis is dead in the water. Gone. No amount of evidence can make a hypothesis true, but just one observation can make it false.

This is the principle of falsification, promoted and discussed at great length by the great philosopher of Science, Karl Popper, whose Logic of Scientific Discovery[2] isĀ  a classic in the field. For a more relaxed read (and by “relaxed” I mean “requires a little concentration” as opposed to “Oh, my frickin’ head’s about to explode!”) you might prefer Objective Knowledge: An Evolutionary Approach[3], published in 1972.

Karl Popper

Karl Popper

It’s also explained eloquently by another Carl, Carl Sagan, in his 1996 book The Demon-Haunted World: Science as a Candle in the Dark. I can’t recommend this book enough to students who are eager to learn about the scientific method in an enjoyable and entertaining fashion. It is, perhaps, his best and most lucid book, and it beats me why this isn’t recommended as a high-school text or at least an undergraduate offering to all students. Many people have a woeful understanding of what science and the scientific method are all about and this one book explains it so well.

One particularly practical offering is Chapter 12: The Fine Art of Baloney Detection, where Sagan offers a number of ways to check whether a proposition or hypothesis is valid. Here’s the list for Baloney Detection:

  • Wherever possibleĀ you need to findĀ independent confirmation of the facts. One person or test does not a hypothesis prove!
  • Encourage and engage in debate on the evidence by knowledgeable proponents of all points of view.
  • Don’t fall for arguments from authority alone; I may have a PhD in Astrophysics but that doesn’t mean there IS a teapot.
  • Be prepared to try multiple hypotheses.
  • Try not to get overly attached to a hypothesis just because it’s yours, THis is harder than you might think.
  • Measure, measure, measure. Objective numbers always trump personal beliefs, no matter how many folks share that belief.
  • If there is a chain of argument every link in the chain must work.
  • Sharpen up Occam’s Razor[3] – if there are two hypothesis that could explain the data equally well, choose the simpler.
  • Check to seeĀ if the hypothesis can, at least in principle, be falsified: Is itĀ testable?Ā If it isn’tĀ testable, it isn’t science![4]
  • Can other people replicate the experiment and get the same result?
  • Conduct control experiments, especially “double-blind” experiments where the person taking measurements is not aware of the test and control subjects.
  • Check for confounding factors; make sure you separate as many of the variablesĀ as you can.

This is why evidence-based practice is so important. It separates the speculative from the scientific. The current rush to buy iDevicesĀ as a blanket solution for those individuals who need an AAC device is a good example of where hypotheses precede evidence. When someone turns up at the clinic doors with a kid, an iPad, and a recommendation from a video on YouTube that “this is the answer,” what do you say? There are many purported “evidential” video clips on the Internet that are well-meaninged attempts by parents to show how their kids have “improved” by using technology, but with no pre-testing and no measure of what “improvement” is, it’s impossible to call this evidence.

In their desire to help people with communication problems, it’s sometimes easier to believe in orbiting teapots than measure performance.

Notes
[1] The word hypothesis comes directly from the Greek į½‘Ļ€ĻŒĪøĪµĻƒĪ¹Ļ‚ and means “placing under.” į½‘Ļ€ĻŒ is “under” and you see this in words such as hypodermic (under the skin), hypothalamus (under the thalamus), and hypochondria (under the breast-bone). The ĪøĪ­ĻƒĪ¹Ļ‚ part orignal referred to the action of placing a foot or hand down to beat time in poetry or music, and it became, by extension, the putting down of a proposition or idea.

[2] Popper, K.R. (1935) LogikĀ der Forschung (The Logic of Research) , Vienna: Springer; trans. The Logic of Scientific Discovery, London: Hutchinson, 1959.

[3] Popper, K.R. (1972) Objective Knowledge, Oxford: Clarendon Press. If you just want to focus on just one chapter, try Chapter 6: Of Clouds and Clocks, which can be read somewhat independently of the book as a whole, and is less dense than some of the earlier chapters. Popper isn’t the easiest of folks to read and in truth, I still have a hard time with much of his stuff on probability because of the math and logic involved, but he’s well worth the effort.

[4] “PluralitasĀ non est ponendaĀ sine neccesitate” or “plurality should not be posited without necessity.” This is attributed to William of Ockam (1285-1349), an English Franciscan Monk and philosopher, who used this premise in much of his work and thinking, although the notion was actually a common principle in medieval thought. The actual phrase, Occam’s Razor, appeared first in 1852 and was used byĀ the astronomer andĀ physicist, William Rowen Hamilton. No mention of his looking for a teapot…

[5] The difference between Science and Pseudoscience often comes down to this rule of Testability. An idea that is inherently untestable is called metaphysical or speculation. You may well believe passionately that there are fairies at the bottom of your garden but unless you can subject them to testing, they are no more real than my orbiting teapot.