Prove me wrong

Embed from Getty Images

Galaxies are big and far away. Two consequences of this are that you cannot rotate them and you cannot walk around them to look at them from a different angle. As part of my degree I did a literature review on a certain class of galaxies known as ‘BL Lacertae Objects’. These are highly luminous and scientists think this may be because they are spewing out a jet of radiation that aligns with our line of sight from Earth. This would mean that there are other galaxies out there that have these jets but that don’t appear to us as BL Lacertae Objects.

In my literature review, I looked into the possibility that BL Lacterae Objects and Fanaroff-Riley Type 1 galaxies are the same, with the latter being the parent population. Scientists try to address this question by counting the number of each, modelling the jets and seeing if this is consistent.

This is science but it is not an experiment. In my recent post on the use of science in education, I wrote about the kinds of experiments I am conducting as part of my PhD. But the scientific method is a disciplined form of inquiry and experiments are just one form that such an inquiry may take.

Pinning down exactly what the scientific method involves can be hard. Some would want to include the process of peer review. I would wish to stress that, in formulating a scientific hypothesis, scientists draw on prior knowledge and look for mechanisms that are plausible, consistent with that prior knowledge and no more complicated than they need to be. However, I would not include that in my definition.

Instead, I think that the core of the scientific method involves the testing of falsifiable hypotheses:

This idea of falsifiability is closely associated with the philosopher Karl Popper and it is the key way of differentiating between science and nonscience. It is particularly useful in providing warning signs while developing a new scientific theory. For instance, Cognitive Load Theory (CLT) has veered into the unfalisfiable with the concept of ‘germane’ load. I would argue that this issue has now been addressed but critics of CLT can justifiably point to it as a serious concern.

The field of education has a habit of generating unfalsifiable ideas. If I were to encourage teachers to ask one question of themselves and of others it would be, “How would we know if this idea was wrong?” If you can answer this question then you have a starting point for evaluating the evidence. If you can’t then we have something that is ‘not even wrong‘. At least disproved hypotheses serve the useful purpose of delineating truth and falsehood. Unfalsifiable assertions serve very little purpose at all and waste too much of our time.

Advertisements

7 Comments on “Prove me wrong”

  1. Luke says:

    I think most philosophers of science would say that Popper was simply wrong about falsifiability being the defining characteristic of science. As both Kuhn and Lakatos pointed out, scientists simply don’t abandon their ideas when they are ‘falsified’. For instance, what would it take you (on the assumption you endorse CLT) to abandon the idea that learning involves a change in long term memory? The Lakatosian answer is that if you’re a good scientist, then there is no single finding or experiment that could get you to abandon this idea: it’s part of your research programme’s hard core.

    • Greg Ashman says:

      I think this is a view of science as sitting in the individual. No, an individual scientist with an investment in a theory may hang on to it despite evidence and, in doing so, she might even play a useful role. However, it is the scientific community as a whole that would decide when to move on.

      • Luke says:

        That’s part of it, but I think it’s more than that. The community doesn’t abandon theories that have been falsified either. Typically the community either simply ignores anomalous observations or accommodates them by adding some kind of auxilary rescue hypothesis that explains them away while leaving important bits of the theory intact. This latter strategy is always possible, even if the rescue hypothesis seems crazy (this was how Neptune was predicted, as a rescue hypothesis designed to save Newton’s theory of gravity from anomalous observations).

        Lakatos directly addresses the claim that falsification is the hallmark of science: “History of science, of course, is full of accounts of how crucial experiments allegedly killed theories. But such accounts are fabricated long after the theory had been abandoned. Had Popper ever asked a Newtonian scientist under what experimental conditions he would abandon Newtonian theory, some Newtonian scientists would have been exactly as nonplussed as are some Marxists.” (Methodology of Scientific Research Programmes, p. 4).

        An analogous challenge to you (or the entire community of cognitive psychologists) might be: what kind of evidence would cause you/them to abandon a cognitive account of learning and adopt a behaviourist account (or a sociocultural account, or whatever)? The correct answer to this question has to be that there’s no single piece of evidence that could reasonably cause this outcome. But that doesn’t mean that cognitive psychology is not a science.

      • harrysblakey says:

        Luke,
        I see a few issues with your line of argument. If Popper had asked a scientist what experiments would cause him to abandon the Newtonian theory a correct answer would have been the Michelson-Morley experiment would bring into question application of the theory in some circumstances. Lakatos argument is based on an imagined Newtonian scientist that only exists in his head. We should be alerted by this lack of experimental rigor on his part.

        But even today Newtonian theories are not abandoned and are taught as part of the foundation of an education in science. This is because the theories are not false but have limitations or need additional theories to explain certain measured results. By the time Popper was writing Newtonian theories had been well tested by experiments that would have falsified them in the conditions tested so an adjustment to the theory was exactly what we would expect based on any new findings.

        To see how much science works by actively testing for falsification you would have to look at new theories being tested currently. To claim it doesn’t work like that you would have to show there are not lots of grad students writing up negative outcomes of experiments as they seek to produce some new and reliable finding. Do you really think that is not happening?

        That we don’t see long standing theories completely abandoned due to new findings of falsity is just survivor bias. These are theories that are not completely false. The new hypotheses that are abandoned completely don’t get much attention.

  2. […] Prolific blogger Greg Ashman wrote a piece recently on the need for educational ideas to be tested by scientific means.  […]

  3. monkrob says:

    Inspired me to pose a hypothesis that needs testing.

    https://assistantprincipalsofficeblog.wordpress.com/2017/04/30/how-do-we-test-this-hypothesis/

    Much of your PISA analysis shows the hypothesis is false.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s