Basic Primer on Science and the Scientific Process

I've already run into some pretty screwed up notions about science. I'm not saying Christians are the only ones who have them - there might be any number of people out there who don't understand the basics of the scientific process and science in general - but, so far, the ones I've found least likely to even listen to what the situation really is happen to be religious "nuts." Note that "nuts" doesn't describe all religious people, not by even a large margin. I think most church-goers are perfectly reasonable people who understand what's important in their lives, use their various religious teachings to help themselves be better people and manage to do so without forgoing independent thought.

But I get very frustrated trying to speak with people for whom facts, data and logic mean nothing. And they are the ones most vociferous that scientists want to be gods and the ones with the very least understanding of what science is.

What is science and the scientific process? Explaining observed conditions in a systematic way.

What that means is that first and last, there's data. No data, no science. Data could be observed conditions (weather, climate, astronomy, living things in the wild). Data could be the results experimentation, like noting what happens when something is fertilized or dropped from a height and dragged along a rough surface.

Why data first? Because the first step in science is trying to formulate explanations for what data you have. Each potential explanation (called hypothesis) must fit the available data. That does not mean that every data point is explained by the hypothesis, but it does mean that none of the data negates or makes the hypothesis impossible. Data that doesn't fit is not ignored but set aside to either look for other explanations (hopefully ones that mesh with the other working hypotheses) or more data that might link it all together.

The next step is to take a working hypotheses and try to validate it, show that it's the right explanation. Again, this is done through data, either more observed information or experimentation. If increasing amounts of data stop supporting the hypothesis (i.e. the data that doesn't fit starts to be significant) or if a single hard fact negates it, the hypothesis is discounted and other hypotheses are pursued.

When a sufficient quantity and quality of verifiable data supports a hypothesis, it becomes a theory. Some theories have such an overwhelming preponderance of data (and no other competing theory) that scientists stop looking for alternate explanations but seek to refine the existing theory, treating it as effectively fact, even if every aspect of it isn't yet understood. Examples of this include plate tectonics and evolution. Yet, even the strongest theories can conceivably be derailed by a single incontrovertible fact. Usually, however, well regarded and broadly accepted theories are not derailed because, to become that way, they have been picked apart, and experimented with and challenged with exceeding rigor.

What's a fact? It's an observation or an experimental result that is (a) verifiable and (b) repeatable. If Dr. Strangelove says he had a huge scientific breakthrough but no one else saw or measured it and/or no one else (preferably outside his lab) can reproduce the result, it's not a fact. Anything Dr. Strangelove postulates based on this spurious result is going to be treated very skeptically by scientists until other independently verified data is found that supports it.

A fact is not something someone wrote down in history, per se, unless there is objective evidence to support that assertion. For example, if someone in some period of time claimed there was an eclipse and astronomy calculation indicated an eclipse would have indeed happened about that time, it would be considered factual. If someone said there was huge volcanic eruption at some point in time and some particular location, but there is no evidence of it in the archaeology or geology of the area, other explanations (including an error on the part of the recorder) would be looked for. That doesn't mean it would be considered nonsense unless there were strong counterindications, but without corroborative evidence, it would be considered speculative rather than factual.

That doesn't necessarily mean that the recorder was lying. They could be describing a previously unknown phenomena in terms they understand, like Aristotle assuming previously dried up ponds could spontaneously create life when refilled with water or that heavier objects were subject to greater gravitational acceleration than lighter ones. His measurement and observational tools were insufficient to allow him to get the full story. As tools become more precise or complete and more data is gathered, science moves forward, sometimes adapting original theories to the new data, sometimes discarding it when it fails to live up to the new data.

What is NOT science?

Starting with a premise and trying to prove it's so, discarding anything that doesn't agree with the beginning premise or negates it. This is not science and scientists don't do it. Even if some scientists along the way fall into the trap of ignoring data that doesn't agree, other scientists will find out and correct the situation. That is bad science and scientists are well aware of the trust they have been given and won't knowingly condone such actions.

I have to add the irony (or more accurately the hypocrisy) that the folks mostly likely to level this accusation are usually the same ones trying to promote a "scientific theory" based on an assertion made in some religious text or another that either has no scientific basis (other than the text) or is directly negated by facts. Science comes from facts, not despite them. As soon as you've decided the answer before you've looked at the data, you are no longer doing science. Period.

Which doesn't mean you can't believe whatever you want, just think of it as miraculous or magical or whatever makes you happy. Just don't try to warp science to make it fit.

Science is not secret. There are some scientific breakthroughs and technological advances that take place behind closed doors, either for patent protection or national security (i.e. military developers), but that is far and away a small portion of science. Science is one of those places where nothing is accepted by the scientific community until you show you work (in this case your data) and survive that data and your conclusions being challenged. There are two reasons for this: one is that science often leads to the greater good for people - understanding weather patterns or geological indications. The other is that something that can't take the rigor of scientific review is probably not true.

Note that that very openness is often a significant source of stress for scientists who, if they have a controversial finding, spend all of their time defending what are often trivial aspects or odd data points by people who have no real understanding of the science behind it.

Disproved vs. Unexplained. I spoke briefly about data that may not fit a theory not being discounted, but also not precluding a theory's veracity. In general, that data would be unexplained, like the platypus, for example, which doesn't fit neatly into the usual patterns when discussing evolution. To the best of my knowledge, no one has a definitive explanation for that, so it's unexplained, but doesn't negate the patterns of natural selection and adaptation directly observed and evaluated via genetic review.

Disproving a notion/hypothesis/theory is a completely different beast. The hard ones is when an assertion is made that is not readily proved nor disproved. Even if it seems unlikely (as the original assertion that vaccines caused autism, for example), in order to disprove the notion, a great deal of testing and observation had to take place, showing that the use of the accused element in the vaccine did not have the effect suggested as well as considerable data showing the same level of autism when using no vaccines or vaccines without it. Unfortunately, before the baseless accusation (which was part of the problem) was thoroughly disproved, so many people had already bought into it that the notion is still prevalent. And, as data was gathered, a large number of children were put at risk, were unnecessarily ill and even died as a results.

The other much easier way to disprove a theory is hard data already in hand that negates it absolutely. For example, if one's "theory" was that the earth was ten thousand years old, this could be immediately disproved since there are hard paleontological and geological and archeological data from far far before that. There are, in fact, human records that go back much further than that. Hence, disproved.

Scientists are not gods (Lord knows I don't want to be one) or all knowing. You don't have to tell us, we know that. That's WHY we insist on the rigor we do, why we check each others' work, why the data takes precedence over the most impressive reputation. We know mistakes on our part can affect many other people. We don't like being wrong, but good scientists would rather be wrong and caught than think they are right when they're not.

I can't speak for all scientists, but I don't want to take anyone's belief system away. Have at. Enjoy. Just don't call it science and don't teach it like it is in schools.

Cross posted on Gather.

Read more...

RS Classic: Dissing Science

Another blast from the past.

Truth often suffers more by the heat of its defenders than from the arguments of its opposers.
–William Penn

Recently, I found myself hot under the collar. Now, don’t get me wrong, I have a temper. But this was all about science and I work very hard to keep my cool and professionalism when it comes to science – passion and science = bad science.

But I was beginning to understand why so many scientists and naysayers are getting into arguments just short of fistfights. It’s hard for even the best scientist to keep his or her temper down when they are being accused of lying and laziness.

This discussion is not going to be about religion. I truly believe in freedom of religion including anyone’s belief in Thor or the monster that hides under the bed. I have not been granted omniscience (aside from believing in fairies myself) and have no right to tell you what to believe in as long as (a) you don’t try to force it down my throat and (b) don’t try to pretend it’s science.

Science is a fascinating beastie. It’s all about finding out what and how reality is and, if possible, figuring out how to manipulate it for the betterment of mankind. Fortunately and unfortunately, reality is pretty slippery. It provides some facts pretty easily, some with a lot of work and some, well, it still hasn’t provided a clear answer. Fortunately, it’s what we don’t know that makes it fun. Unfortunately, what we don’t know (or, worse, think we know) can be very dangerous. For instance, long ago they tried blood transfusions under the correct assumption that blood can help when blood’s been lost. Unfortunately, their lack of knowledge about blood factors meant that some transfusions didn’t go too well so that it was an act of desperation until blood typing was discovered early in the 20th century.

It is because of this, the thrill and excitement (and the risk) of what we know and don’t know that scientists voluntarily subject themselves to strict rules and processes of rigor. Rules include repeatability (preferably independently), adequate controls, and theories that can be disabled with a single immutable fact (which is not the same as a single data point). All the data must be accounted for or explained; one cannot pick and choose the data. And, to be really recognized, it needs to withstand something called peer review.

Peer review? You expect them to police themselves? Yes, for two reasons. First, one cannot evaluate the science of a proposal, conclusion or bit of research without an in depth understanding of science and scientific processes. Generally, this effectively limits you to scientists or “former” scientists (if such things exist). Secondly, scientists want to be right. Being wrong, spectacularly, is not how any scientist wants to be remembered.

Yeah, but don’t we look after our own? Actually, not so much and there’s good reason. There is no benefit for a scientist to give a free pass on the work of another scientist. If the work is bogus or sloppy, the reviewers would be impugned (rightly) along with the researcher by their failure to look at it critically. Wherein politicians (and some other, but not all, professions) can generally make mistake after mistake and survive, in the scientific world, a single instance of dishonesty and/or sloppy work can ruin a career.

If someone does research and writes conclusions, a reviewer is obligated to try to shoot holes in it, find the problems, look for errors, whether the reviewer agrees with it or not, or we do a disservice not only to the rest of the world, but to the researcher. If it’s wrong, we don’t want to hang our hats on it (and we save the researcher some embarrassment). If it’s right, we want it to be as bulletproof as possible.

And people in the same field do not all agree (HAHAHAHAHA!) - far from it. It is, in fact, the hemming and hawing on details that have let so many less familiar with the process think that so many scientists disagree on global warming, when what they are really doing is trying to understand not IF there will be changes, but how much, how fast and what we can do to minimize it. On that, I feel comfortable saying consensus has not been reached. But it doesn’t mean the science is invalid.

Don’t get me wrong. Scientists get excited by findings, breakthroughs, new possibilities, verifications, etc. If we didn’t, we wouldn’t be here. And, yeah, anyone who does work they’re proud of likes recognition. That’s why we peer review and take our time accepting new ideas. If an idea can take being tested independently and repeatedly, if it can stand the test of time and many different people trying to poke holes in it, it stands a much better chance of being valid that if it came from one excited individual.

That’s one reason I get so hot under the collar at being accused of laxity or dishonesty. I’m part of those very checks and balances. Does nothing get by us? Sadly, yes, mistakes are made here and there. But they usually get caught eventually and the numbers (percentage-wise), I feel, are low. Truth told, I don’t know of any other profession that examines itself so critically, puts in so many impediments and challenges to ensure that integrity, that objectivity that makes science what it is. Because people do use that information in a life and death way, we have to be responsible with what we say.

Your doctor doesn’t have the same review process when he gives a prescription (though the basis for the treatments he'll suggest should have gone through the same rigorous process). Your mechanic doesn't even have that. Your lawyer can do all kinds of stupid things (at hundreds of dollars an hour). And there’s not much you can do to preclude mistakes being made by those you trust.

In most cases, there is little you can do except fire someone (after the fact, mostly likely) or in cases of egregious error, sue. In most cases, you just have to live with it. Think of how much better government might be if positions and arguments in government were vetted with the same rigor as science. (I find it ironic that people are more likely to be believe others who have a vested interest in persuading you to a point of view - bankers, big business, politicians – than those that have none.)

Believe me, folks, we don’t want to tell you bad news. There’s no glory or riches in telling people what they don’t want to hear. If we tell you, it’s because that’s what the evidence is telling us and we don’t want our unwillingness to put it on the line to cause anyone to be hurt unnecessarily. At the least, we have no interest in looking stupid.

In science, we really try to weed out the mistakes before you ever see our results. We’re human, but we want the truth out there more than we want our name on it. We’re not perfect, but we’re trying to do the right thing.

Cut us a little slack, k?

You know, some things never change.

Read more...