In our post-Enlightenment paradigm, you’d be forgiven for thinking everyone would know what science is and how it works, but the vast majority of us don’t seem very clear on it at all.

That’s probably got a lot to do with the way science is taught in school (not exactly riveting), and the way the media portrays scientific discoveries (hyperbole and then some)… and, of course, the way a growing sub-sector of society is responding to the fallibility of the scientific establishment (with pseudo-skeptical paranoia).

Science is misunderstood by proponents and opponents alike, though in very different ways.

Well-meaning but misguided proponents of science often hail new discoveries as reported by the media, hyping conclusions derived from studies and blowing them out of all proportion, often to the embarrassment of the scientists concerned. Although science does lead to new discoveries, this is a slow and methodical process of precision work, and not the glamorous ground-breaking field it is often portrayed as.

Proponents often also confuse science with technology, perhaps because the two are somewhat related. An important distinction, however, is that where science is concerned with explanations of natural-world phenomena, technology is concerned with man-made machines, gadgets and other contraptions that are designed to pad the lives of humans, making them more comfortable or convenient, or at least increase profit margins. So where technology makes use of scientific findings, it is only scientific insofar as it tests its models to make sure they work – at least well enough to be marketed; it doesn’t make discoveries about how the world actually works.

Opponents of science often claim to be skeptics, although this is in contrast to the skepticism exercised by scientists themselves in their methodical integrity to the process of discovery. Anti-science pseudo-skeptics may make claims that science itself is corrupt (some scientists and some organisations undoubtedly are), that many discoveries are being suppressed or covered up (it’s possible, but we’ll probably never know for sure), or that science has a creepy agenda that involves perpetrating harm (what the…?). However, these same pseudo-skeptics don’t tend to apply this same skepticism to the quacks and frauds they turn to for alternatives.

Every day, somewhere on the internet, it is possible to bear witness to claims that scientists are corrupt while no skepticism is applied to the commodified preachings and products of an ex-professional who has been de-registered for fraud, or an amateur quack out to make a buck.

Luckily for SHIFT readers, the SHIFTy editorial team all have an educational and/or professional background in science, and as we’re a non-commercial non-profit organization run by volunteers, it’s reasonably safe to assume we’re not out to sell any false promises, so let’s take a look at what science is really all about….

So, what is science then?

Science is an outgrowth of a branch of philosophy known as empiricism. Empiricism holds that knowledge can only come from sensory experience of a phenomenon – which basically means you have to observe it in one way or another in order to know it’s real.

This empirical observation of how the world works contrasts with other ways of making sense of the world, such as a priori reasoning – which basically draws logical conclusions, but doesn’t base them on evidence; or intuition – just ‘knowing’ or ‘sensing’ something, which can be accurate, but does not make use of evidence to support assertions; or revelation – often a ‘supernatural’ or drug-induced experience – again, unsubstantiated by observable evidence.

Unlike most other ways of making sense of the world, scientific knowledge is tentative, and subject to revision. Whereas our a priori conclusions, intuitions and revelations tend to be fairly rigid in the face of new information, science builds upon a base of previously existing knowledge through inquiry, resulting in increased sophistication of understanding of phenomena.

Scientific knowledge is accumulated through a process known as the scientific method of inquiry – usually shortened to just the ‘scientific method’. The scientific method entails a process of formulating hypotheses, deriving predictions as logical consequences, and carrying out experiments based on those predictions to test the hypothesis.

Formulation of a question

The first stage of the process of inquiry involves coming up with a question you want answered. This question can refer to the explanation of a specific observation, for example “why is the Earth warming?” But it can also be more open-ended than this, for example, “how can humans live more sustainably on Earth?” The way a question is formulated strongly affects the final outcome of the investigation, as it impacts all stages of the process.

Determining a suitable question to ask may prove more complex than it may seem at face value, however. Even underlying our questions are a range of assumptions that we are often unaware of, let alone willing to challenge, and these affect the design of an experiment and the interpretation of scientific results. Coming up with a suitable question requires us to be aware of our own assumptions, and be prepared to challenge them. It also requires the integrity to avoid confirmation bias, in which what is observed is simply what was expected.

Hypothesis

A hypothesis works like a tentative preliminary answer to the initial question. It can be very specific, for example “the Earth is warming due to the impact of increased carbon dioxide emissions on the greenhouse effect.” But it could also be broad, for example, “humans can live more sustainably on Earth by reducing our collective ecological footprint.”

A commonly heard but often misunderstood term is null hypothesis. A null hypothesis isn’t an absence of a hypothesis – it’s the suggestion that the initial hypothesis is false; for example, “increased carbon dioxide emissions have no impact on the greenhouse effect that keeps the Earth warm.” Generally speaking, scientists are out to show that the null hypothesis is false.

A scientific hypothesis must also be falsifiable. This means scientists must be able to identify a possible outcome that conflicts with the predictions – i.e. it’s possible to be proven wrong. If there is no way a hypothesis can be falsified then it is essentially meaningless, as experimentation will demonstrate nothing if there is not more than one possible outcome.

Prediction

Prediction is simply a statement of the logical consequences of the hypothesis. It is possible for a prediction to be correct by sheer coincidence, so the strength of a prediction lies in the extent to which this is likely. The less likely it is that the prediction is correct by pure coincidence, the more convincing it is if it is fulfilled.

A complication arises if two or more hypotheses make the same prediction. If this is the case, then it is impossible to know which of the two hypotheses is supported by the evidence. Therefore, it is necessary to distinguish the hypothesis from likely alternatives by making predictions specific to their hypothesis.

Testing

Testing is the essential component that sets science apart from other fields of inquiry. It is not enough to simply observe anecdotally whether one’s hypothesis seems to be supported by evidence – it must be tested under certain conditions. Experimentation is the method favoured by scientists, as it sets clear parameters for observing whether the real world behaves as predicted by the hypothesis.

To complicate matters, experimental support for the predictions doesn’t mean that the hypothesis is true for certain – future experiments may reveal otherwise. Even if predictions are supported by large numbers of experiments by many scientific teams, this does not automatically render their hypotheses correct. The more rigorous the experiment, and the smaller the margin for error, the better; and those that take the risk of confirming a null hypothesis are the most sound.

Controlled experiments are the means by which scientists ensure their results are not compromised by biases or flukes. Randomised controlled trials, double-blind experiments, and comparisons with placebos or other alternatives are all ways in which scientists seek to reduce the margin for error.

A major issue that emerges from testing is that results of tests are not often generalisable. As experiments test highly specific phenomena in order to ensure test results can either support or disprove their hypotheses, their conclusions relate only to those specific phenomena. This means that evidence for broader ideas or discoveries is accumulated very gradually – so beware those pop-science articles in the news that claim a single experiment has revealed something ground-breaking! Scientists tend to be far more tentative about extrapolating from their conclusions than the media are in reporting them.

Analysis & interpretation

Once data has been gathered through the testing process, then things can begin to take shape. The analysis stage is the stage at which scientists determine what their results are actually showing. From this they can decide what steps are necessary to take next – be it further testing, or the elimination of a hypothesis, or the formulation of a new hypothesis, for example.

Scientists compare the predictions of their hypothesis to those of the null hypothesis in order to determine which explains their data better. If the evidence has falsified the hypothesis, then a new hypothesis needs to be formulated – so it’s back to the drawing board. The experiment may, alternatively, support the hypothesis, but the evidence for this may not be strong enough for high confidence, so other predictions must be tested.

If a hypothesis is strongly supported by the experimental evidence, then it becomes possible to build upon this base by formulating new questions that will provide further insight.

Replication

Replication of an experiment’s findings is necessary for acceptance of any given hypothesis by the scientific community. If an experiment cannot be repeated with the same results – or if it cannot be repeated at all – then this suggests the findings of the original experiment were the result of mistakes or flukes, and not actually supportive of the hypothesis.

Scientists often strive to replicate their own findings for the sake of providing a robust evidence base, but the gold standard for replication is when it is achieved by other scientists. The archiving of experimental data is required as policy by some of the more rigorous journals so that other researchers can test the data and methods, and build upon previous research. This also acts as a safeguard against fraudulent results, as fraudulent scientists – such as Andrew Wakefield, original proponent of the vaccines cause autism myth – are easily caught out.

(A brief note on the high-profile Wakefield fraud: contrary to common misrepresentation, Wakefield only said his research showed the MMR vaccine caused autism in his prelude to spruiking his own patented alternative vaccine triad; he did not condemn vaccines in general.)

Review

Peer review is the process by which experimental findings are evaluated by experts in the field – usually specialists in the specific area under investigation. In order for scientific findings to be published they need to run the gauntlet of peer review, which can be a very unforgiving process, with as many as 80% of studies in some areas failing to make it into print. Contrary to popular belief, peer review does not certify that the results of an experiment are correct; rather it determines the extent to which the experiment itself was carried out with integrity so that its findings were not compromised.

Theory: final destination or just the best explanation… for now?

It is only after a great deal of work has gone into testing and replicating results that an overall theory may begin to emerge, and this can take many years, even decades, to happen. Again contrary to popular misconception, a theory isn’t some half-baked speculation about how the world works; it is the most successful explanation of a phenomenon that has withstood rigorous attempts at falsification – it is arrived at only after repeated failures to disprove a hypothesis over many years or decades.

This is why the theories of evolution and gravity are generally unchallenged in the scientific community – they’re not just ideas for how things might work; they’ve been tested to the point of near-certainty, and they’ve stood the test of time without robust challenge from new discoveries.

Once science arrives at a theory a knowledge base can be built on firm foundations, enabling us to further discover and better understand the natural world in all its spectacular complexity.

But theories can be wrong, or at least incomplete, and so science remains open to new discoveries, and theories remain open to modification or overhaul. In reality the baby rarely gets chucked out with the bathwater, and small modifications to theory are usually sufficient to make sense of the new information.

Every now and then, though, new information comes to light that challenges the very bedrock of our understandings about how the world works, and it’s the job of science to unravel the mysteries and help us make sense of them. It truly is a never-ending process of discovery!