When sensible people speak of «the scientific method», what they really mean is something quite grand, covering a range of ideas including:
Section 2 takes up the discussion of how science is really done.
Let’s take a brief digression to talk about the terminology: Voltaire famously remarked that the Holy Roman Empire was neither holy, nor Roman, nor an empire. It’s an idiomatic expression. In English, there are hundreds of such expressions, as discussed in reference 1. An idiomatic expression must never be taken literally; otherwise, it crosses the boundary from idiomatic to idiotic.
It is important to realize that «the scientific method» is an idiomatic expression, and can be quite misleading. For starters, it ought to be plural not singular, because there is not any single method for doing science, just as there is not any cut-and-dried method for writing a novel. In fact, scientists use a great many methods. This point is discussed in reference 2 and elsewhere. Alas, misconceptions about this are appallingly common among non-scientists.
Furthermore, many of the “methods” used in science – and in research in particular – are not nearly as methodical as non-scientists seem to think. In fact, although some scientific activities are highly methodical, many others are not. One is reminded of Pólya’s dictum: Try something; if that doesn’t work, try something else. Exploration, heuristics, and serendipity have always been an imporant part of science, and always will be.
It is easy to get deceived about this. Part of the problem is that scientific publications are expected to be systematic and methodical – even if most of the work that preceded the result was not. As discusseed in section 3.9, it is an infamously common mistake to confuse the way scientific results are initially obtained with the way they are explained.
People who have not done any research, nor even seen it done, sometimes equate «the scientific method» with a step-by-step hypothesis-testing approach, perhaps along the lines spelled out on the poster discussed in section 4. This is a travesty. It makes research seem hundreds of times easier than it really is. Therefore it is an insult to every researcher – past, present, and future. Similarly, it is a disservice to students who may be thinking of becoming scientists, since it gives them a false impression of what they are getting into. In addition, overemphasis on one particular method makes it impossible to understand the history of science.
Some of these misconceptions are discussed in more detail in section 3 and section 4.
Here are some of the principles that guide how science is done:
Note that even when a useful prediction is possible, you still need to do some work to distinguish good predictions from bad predictions. For example, if the payoff is 8-to-1 you will make money if you predict sum=7 every time, but you will lose money if you make a dumb prediction such as snake-eyes every time.
In particular, here is a list of terms that could be used to describe scientific results. Scientists and other thoughtful experts use these words almost interchangeably, without bothering to assign narrow definitions. (Sometimes there are slight variations in connotation, but the signal-to-noise ratio is very poor.) Therefore it is a colossal waste of time when non-experts pretend to assign precise meanings to these terms.
Famous examples include:
Most of these examples could be reworded with no change in meaning: Maxwell’s laws, Newton’s algorithm, Parseval’s identity, et cetera.
Sometimes a change in wording would change the meaning of a phrase ... but still, neither meaning is systematic. For example, the Maxwell relations are distinct from the Maxwell equations. However, this is due to historical accident and idiomatic interpretation, not to any systematic difference in meaning of the individual words. The same can be said of the the distinction between Laplace’s law and the Laplace equation.
|As mentioned in item 11 and section 3.4, any scientific law, theory, principle, etc. comes with provisos and with limits to its validity. It may take several sentences, several paragraphs, or several pages to explain these provisos and limits. This commonly requires conveying thousands upon thousands of bits of information.||Changing a single word – such as renaming a “law” to a “theory” or vice versa – conveys only a few bits of information. This is nowhere near being sufficient to describe our degree of confidence in the idea, let alone describe the provisos or the limits of validity.|
We therefore conclude that the idea of a “law” gradually developing into a “theory” is provably absurd. Trying to judge the validity or importance of an idea based on whether it is called a “law” or a “theory” must be considered an extreme form of judging a book by its cover. As such, it is the opposite of science and the enemy of critical thinking.
|One usage refers to a coherent system of evidence, principles, and methods, offering a comprehensive understanding of a broad topic. (This is much grander than any single rule or fact.)||The other usage refers to a hypothesis, conjecture, or mere speculation.|
This ambiguity can cause all sorts of problems if you’re not careful. See section 3.11 for the details on this.
In many cases a model can be called a theory, and a theory can be called a model. The two words do not mean exactly the same thing, but since both definitions are so broad and fuzzy, it is not worth fussing over the details.
This idiomatic expression strikes me as unnecessary and abusive. If you want to talk about strategy, call it “strategy” (not «the scientific method»). If you want to talk about the nature of science, call it “the nature of science” (not «the scientific method»). For example, the NSTA has a nice discussion of the nature of science (reference 5).
Every increase in computer power increases the importance of calculation, computation, and numerical modeling.
Often calculation and computation go hand-in-hand with visualization (item 21).
Nowadays computers are heavily used to help run the experiments, help solve the equations, and help visualize everything.
For additional discussion of “thinking skills” per se – including how to learn, and how to teach thinking skills – see reference 7.
See also reference 8, reference 9, reference 6, and reference 10 for sensible discussions of what science is, and how scientists do science.
It is important to know the difference between science and pseudo-science. Good discussions of this topic include reference 11 and reference 12.
Let’s take a moment to apply the scientific method to itself. That is, let’s objectively consider the question of whether the scientific method actually works.
Alas, according to all available data, it does not always work very well. According to some measures, more people in the US believe in astrology than believe in evolution, as discussed in reference 13. Scientific evidence works reasonably well for convincing scientists, but for the other 99.999% of the population, not so much. Depending on how it is presented, additional evidence might make people less inclined to believe the right answer, as discussed in reference 14.
To be effective in the real world, a scientist needs two skill sets: one for obtaining the results, and another for persuading non-scientists to believe the results.
There is an important distinction between deceptive and absurd. An idea that makes wrong predictions every time is absurd, and is not dangerous, because nobody will pay any attention to it. The most dangerous ideas are the ones that are often correct or nearly correct, but then betray you at some critical moment.
Pernicious fallacies are pernicious precisely because they are not absurd. They work OK some of the time, especially in simple “textbook” situations … but alas they do not work reliably.
You need not worry about the “most erroneous” errors. You should worry about the most deceptive and most destructive errors.
You should avoid using fallacious arguments, and you should object loudly if somebody tries to use them on you. Common examples of unscientific thinking include:
As mentioned in item 3 and item 11, most rules have limitations on the accuracy and/or their range of validity. You should neither over-react nor under-react to these limitations.
Consider the contrast: Equation 1 is very different from equation 2:
|x = y provided a, b, and c (1)|
|x = y (2)|
which means x = y in all generality.
It is a common mistake to mislearn, misremember, or misunderstand the provisos, and thereby to overestimate the range of validity of such a rule.
There are several ways such mistakes can come about. I’ve seen cases where the textbook soft-pedals the provisos “in the interest of simplicity” (at the expense of correctness). I’ve seen even more cases where the text and the teacher emphasize the restrictions in equation 1, yet some students gloss over the provisos and therefore learn the wrong thing, namely equation 2.
Another possibility is that we don’t fully know the provisos. A good example concerns the Wiedemann-Franz law. There are good theoretical reasons to expect it to be true, and experiments have shown it to be reliably true over a very wide range of conditions. That was the whole story until the discovery of superconductivity. The Wiedemann-Franz law does not apply to superconductors, and you will get spectacularly wrong predictions if you try to apply it to superconductors. My point is that before the discovery of superconductivity – which was a complete surprise – there was no way anyone could have had the slightest idea that there was any such limitation to the Wiedemann-Franz law.
As mentioned in item 14, offering non-specific and/or non-constructive criticism doesn’t help anybody.
It is important to keep track of the limitations of each model, and to communicate the limitations. If you see some folks at risk of error because they are disregarding the limitations, it is helpful to remind them of the limitations. Sometimes it is worth trying to find improved ways of expressing the limitations.
If a model stands in need of improvement, the best thing you can do is to improve it. Devise a rule that has more accuracy and fewer limitations. (You may find this is more easily said than done.) Communicate the new rule to the community, and explain why it is better.
If you can’t devise a better rule on your own, you might hire a scientist to do it for you. (Again, you might find that devising accurate, robust models is more easily said than done.)
There’s a rule that says “don’t borrow trouble”. Conversely, you shouldn’t spread trouble around, either. Let me explain:
Suppose a rule is good enough to solve Joe’s problem, but is too limited to solve Moe’s problems. Then it’s not constructive for Moe to complain about what Joe is doing. It’s none of Moe’s business. If Moe accuses Joe of using a “wrong” rule, the accusation is false; the rule is good for Joe’s purposes, and Moe should not project his problems onto Joe. Conversely, if Joe notices that the rule is too limited to handle Moe’s problem, that is no reason for Joe to distrust the rule within its proper limitations.
This is worth mentioning, because some people want “the truth” and think “the truth” must be exact and unlimited. Conversely they think anything that has limitations must be worthless. (This is an extreme form of over-reacting to the limitations of a model.) It is a very serious, very common problem. See section 3.6 for more on this.
If Joe and Moe choose to work together to devise a new, grander model that has fewer limitations, so that it can handle both their problems, that is great – but it is their choice, not their obligation, and should not be an impediment to using the old model to solve Joe’s problems.
Sometimes we are faced with black-and-white choices, as indicated in figure 2.
More often, though, the choices form a one-dimensional continuum: not just black and white, but all shades of gray in between, as indicated in figure 3.
It is an all-too-common mistake to see things in black-and-white when really there is a continuum. This well-known fallacy has been called by many names, including false dichotomy, black-and-white fallacy, four-legs-good two-legs-bad, Manichaean fallacy, et cetera.
To say the same thing again, it is all too common for people to assume that everything that is not black is completely white, everything that is not white is completely black, everything that is not perfect is worthless, everything that is not completely true is completely false, their friends are always good and their enemies are always evil, et cetera.
A related but more-subtle fallacy is to assume that all things that are not perfect are equally imperfect. In contrast, the fact is that point B in figure 3 is much blacker than point A, even though neither one is perfectly black nor perfectly white.
Understanding this is a crucial part of scientific thinking, because as mentioned in item 3, scientists are continually dealing with rules that are inexact or otherwise imperfect. The point is that we must make judgments about which rules are better or worse for this-or-that application. We cannot just say they are all imperfect and leave it at that. They are definitely not equally imperfect.
Actually, sophisticated thinking requires even more than shades of gray. Often things must be evaluated in multiple dimensions, evaluated according to multiple criteria at once, as indicated in figure 4. Option A is better for some purposes, and option B is better for other purposes.
See reference 13 for more about the distinction between truth and knowledge.
As discussed in reference 4, there are two kinds of statements: assertions and hypotheses. Unlike an ordinary assertion, a hypothesis is stated without regard to whether it is true. It might be known true, known false, probable, improbable, or whatever.
A hypothesis is not a prediction or even a guess. If I toss a coin, there are two hypotheses as to the outcome: it might come up heads, and it might come up tails. I do not need to make a prediction before doing the experiment.
Any experiment worthy of the name involves at least two possible outcomes. If the outcome of an activity is completely predictable, there is only one hypothesis that needs to be considered ... but then the activity is not an experiment. It might be a demonstration or an exercise, but it’s not an experiment.
I don’t want to argue about the meaning of words. If in your mind the word “hypothesis” means a prediction or a guess, then we need to find another word that is free of any such meaning. Sometimes it helps to talk about “scenarios” or “outcomes”. Perhaps the best terminology is “things that could happen”; everybody understands that, even at the grade-school level.
No matter what word you use, the point is that science is not a guessing game. As mentioned in item 15, section 3.8, and section 10.3, the bedrock principle is:
This is required for safety if nothing else. This is the rule in daily life, in science, in farming, in small business, in big business, and even in the Boy Scouts.
It is common for professional scientists to hold a planning meeting, to come up with a list of things that could happen. Everybody contributes to the list:
The same team-based approach works well in a classroom situation.
In any case, you need a list. To say the same thing the other way: Focusing on a single hypothesis, to the neglect of other plausible hypotheses, would be unsafe as well as illogical.
Again: a hypothesis is not a prediction or even a guess; it is just something to be considered. One of the principal results of a scientific inquiry is the ability to make useful predictions – but this is a result, not a prerequisite. It is the ending point, not the starting point.
|After doing the experiment for the first time, you should be able to make predictions about what will happen in subsequent experiments of the same type.||Before doing the first experiment, there is no way to predict the outcome with any certainty. The best you can do is to consider all the plausible things that could happen.|
More generally, there is no point in doing experiments where the outcome is known in advance. The point of doing experiments is to learn something we didn’t already know. The experiment must have at least two possible things that could happen, or it isn’t an experiment at all.
In many cases, after a scientific investigation is complete or nearly complete, it can be summarized in terms of hypothesis testing. That is, using 20/20 hindsight we can make a list of hypotheses and say which are consistent with the results and which are ruled out by the results. One should not imagine, however, that all scientific work is motivated by hypotheses or organized in advance in terms of hypotheses. Some is, and but a great deal isn’t.
Science – especially exploration and research – usually involves a multi-stage iterative process, where the results of early stages are used to guide the later stages. The early stages are not well described in terms of hypothesis testing, unless we abuse the terminology by including ultra-vague hypotheses such as “I hypothesize that if we explore the jungle we might find something interesting”.
Typical example: When Bardeen, Brattain, and Shockley did their famous work, they started from the vague conjecture that a semiconductor amplifier device could be built. This turned out to be true, but it was neither novel nor specific. The general idea had been patented decades earlier by Lilienfield. Indeed a glance at the following table would have led almost anyone to a vague hypothesis about semiconductor triodes:
|vacuum-tube diode (known)||vacuum-tube triode (known)|
|semiconductor diode (known)||???|
The problem was, all non-vague early hypotheses about this topic turned out to be false. It is easy to speculate about semiconductor amplifiers, but hard to make one that actually works. The devil is in the details. Bardeen, Brattain, and Shockley had to do a tremendous amount of work. Experiments led to new theories, which led to new experiments ... and so on, iteratively. Many iterations were required before they figured out the details and built a transistor that worked.
Example: When Kamerlingh Onnes began his famous experiments, he was not entertaining any hypotheses involving superconductivity. He was wondering what the y-intercept would be on the graph of resistivity versus temperature; it had never occurred to him (or anyone else) that the graph might have an x-intercept instead.
Example: When Jansky began his famous experiments, he was not entertaining any hypotheses about radio astronomy. He spent over a year taking data before he discovered that part of the signal had a period of one sidereal day. At this instant – and not before – the correct hypothesis came to mind: that part of the signal was emanating from somewhere far outside the solar system. The point is that a very great deal of scientific and engineering activity preceded the historic hypothesis.
Looking back with 20/20 hindsight we can analyze and summarize Jansky’s work in terms of hypotheses ruled out or not ruled out ... but hindsight is not a useful method to the researcher who is doing the original work.
Example: On the day when Fleming discovered penicillin, he was not entertaining any hypotheses about penicillin, antibiotics, or anything remotely similar. The key observation was the result of a lucky accident. Of course, after the discovery, he considered various hypotheses that might explain the observations, but the point remains: the hypotheses came after the observations, and did not guide the initial discovery.
Counterexample: At the opposite extreme, in a typical forensic DNA-testing laboratory, two very specific hypotheses are being entertained: Either sample A is consistent with sample B, or it isn’t. This may be “scientific”, but it isn’t research.
Example: When the BATSE team discovered TGFs (terrestrial gamma flashes), they weren’t looking for them. They were not “testing the hypothesis” that TGFs exist. The spacecraft was intended to look for cosmic gamma-ray sources. Then they noticed, hey wait a minute, some of the flashes are coming from the wrong direction.
Example: Theoretical calculation (as in item 22) does not usually proceed by means of hypothesis testing. If you are asked to multiply 17 by 29, I suppose you “could” do it by testing a series of hypotheses:
However, I don’t recommend that approach. Reliable and efficient long-multiplication algorithms are available.
Theoretical physics involves a great deal of calculation. Overall, it is not well described as hypothesis testing.
Example: The same goes for experiments. Even very simple experiments are often not well described by hypothesis testing. If you are asked to count the number of beans in a given jar, you could contrive all sorts of hypotheses, just as we did in the previous example:
However, none of those hypotheses would do you much good. At some point, if you want an accurate result, you have to count the beans.
Bad Example: Christopher Columbus started out with the expectation that he could sail to India. He did not do a good job of considering all the plausible hypotheses. Instead, he picked one hypothesis and more-or-less assumed it was true. The problem is that when his expectation was not fulfilled, he tried to fudge the data to conform to his expectation, rather than reporting what he actually discovered. This is flagrantly unscientific behavior.
It is extremely common to set out expecting to discover one thing and instead to discover another (e.g. Fleming, Jansky, Kamerlingh Onnes, and innumerable others). This is so common that it even has a name: serendipity. Serendipity is not blind luck. It is the opposite of blind luck; it is the kind of luck that you earn by being smart, keeping your eyes open, and visiting places where interesting stuff is likely to occur.
As the proverb says: If the only tool you have is a hammer, everything begins to look like a nail. Now, I have nothing against hammers, and I have nothing against hypothesis testing. But the fact remains that in many circumstances, hypothesis testing is not the right tool for the job. Scientists know how to use many different tools.
It is common for people who don’t understand science to radically overemphasize the hypothesis-testing model, and to underestimate the number of iterative stages required before a good set of hypotheses can be formulated. It is a common but ghastly mistake to think that a good set of hypotheses can be written down in advance, and then simply tested.
Overemphasizing hypothesis-testing tends to overstate the importance of deduction and to understate the importance of induction, exploration, and serendipity.
If you are going to think in terms of hypotheses at all, you should do your best to consider all the plausible hypotheses, i.e. all the plausible things that could happen.
For example, in a simple coin-tossing experiment, if you hypothesize that the coin will come up heads, you should at least (!) consider the hypothesis that it will come up tails. As mentioned in section 4, if an “experiment” has only one possible outcome, it’s not an experiment; it’s just some kind of construction project or demonstration. Even if you are repeating a conventional experiment and expect the conventional result to be confirmed, you are at least implicitly considering the possibility that it won’t be confirmed; otherwise the whole exercise is pointless.
During the planning stage, you don’t need to consider in detail every imaginable scenario, but you should do your best to consider all the plausible scenarios. Even beginners should devise a flexible plan, flexible enough to cover all the things that are reasonably likely to happen.
First of all, you need to do this for safety. Secondly, you ought to do this because it greatly increases the chance that the experiment will be worthwhile.
After the planning stage, you should keep your eyes open, looking for anomalies, i.e. looking for any kind of unexpected results. If the preliminary experiments don’t go as planned, that’s useful information. You can use that by going back to the planning stage, formulating a better plan, and starting over from there. This iterative process is very important.
There seems to be widespread misunderstanding of the role of planning, prediction, and iteration. Science is not a guessing game. One of the great advantages of being a scientist is that you don’t have to guess! Instead, you make a list of all the plausible things that could happen and proceed from there. This minimizes the amount of guesswork, and minimizes the amount of unwelcome surprises.
Consider a typical high-school science fair. The presentations can be placed into three categories:
At a typical science fair, there are appallingly few (if any) presentations that fall into category III. This is no accident; the students were taught to do things this way, as you can see in figure 5 in section 4.
What makes it worse is that generally the projects in category II are severely downgraded relative category I. This is a travesty of science, because it rewards guessing the answer in advance – or doing “experiments” where the result is a foregone conclusion – as opposed to doing an experiment that actually accomplishes something and tells you something you didn’t already know.
The students in category II would have been very much better off if they had transferred to category III, that is, if they had considered a larger set of hypotheses. This is the easiest thing in the world to do, since you can rewrite the list of hypotheses after doing the experiment.
When I am judging a science fair, I tend to downgrade everyone in categories I and II equally, for failing to consider all the plausible hypotheses. Science is not a TV show where “the” myth is either busted or confirmed.
I do not give any preference to category I, because science is not a guessing game, and especially because I don’t want to encourage doing experiments where the result is a foregone conclusion.
This reminds me of how science was done in the old Soviet Union. Scientists were rewarded for “successful” projects, so they always did the work in advance, and then submitted proposals to “discover” things they already knew about.
The thing that is extra-sad about category II is that in the real world, scientists are free to change their hypotheses. They routinely change them again and again, iteratively, as they learn more.
As pointed out in reference 6 and reference 10, there is a huge difference between how a scientific result is initially obtained and how it is explained. It is an infamously common mistake to confuse these two things.
|When explaining a scientific result, the explanation should be linear and logical.||When initially obtaining a scientific result, the process is almost always messy and nonlinear.|
|A text or a scientific paper should say what we know, and give the best evidence for it. This is all you need in order to apply the result to practical problems. See reference 18.||The real history involves a great deal of backtracking out of blind alleys. The details are interesting if you want to learn how science is done, but uninteresting and irrelevant if all you want is the result.|
After a goodly amount of data has been collected, it makes sense to make a list of hypotheses, and to decide which are consistent with the data and which are not. This a posteriori list of hypotheses need not bear any relationship to whatever set of hypotheses (if any) you had before seeing the data.
Oftentimes (albeit not necessarily), a good way to structure a scientific paper is to discuss the hypotheses that are consistent with the data, and to discuss hypotheses that were a priori plausible but are not a posteriori consistent with the data.
Let’s be clear: If you see somebody touting a cut-and-dried step-by-step approach to doing science, based on hypothesis testing, it means they have read about scientific results but never actually done any science, at least not any research or exploration.
Consider the contrast:
|When non-scientists try to guess how science is done, all-too-often they imagine simple laboratory experiments, where the scientist tightly controls all the relevant variables, makes changes, and observes the results. This represents some sort of ideal case.||In real life, tightly-controlled laboratory experiments might be physically impossible, prohibitively expensive, and/or unethical.|
On the other hand, sometimes some aspects of the observations can be checked by experiment. For example, stellar spectra can be compared with laboratory spectra. However, the point remains: the primary subject matter (including stars, galaxies, etc.) remains beyond the reach of real experimentation.
As mentioned in item 7, the word “theory” can be used in two radically different ways.
|One usage refers to a coherent system of evidence, principles, and methods, offering a comprehensive understanding of a broad topic. (This is much grander than any single rule or fact.)||The other usage refers to a hypothesis, conjecture, or mere speculation.|
Remarkably, both usages are correct, and the ambiguity can be traced back more than 2000 years. Both meanings are used by scientists and non-scientists alike.
It is important to be aware of this, because the ambiguity is routinely used as a a sucker-punch, used by persons who are attacking science in general and evolution in particular. It is best to avoid the word “theory” entirely when debating such persons. Don’t be a sucker.
Here’s a constructive suggestion: When a word is ambiguous, we can always coin new words. In particular, we can replace the word “theory” with kyriotheory and hypotheory. These words are complementary to the words thesis and hypothesis (respectively), in the following sense:
|A thesis is a statement you put forward and offer to defend with evidence and logic.||A hypothesis is a statement you put forward without offering to defend it. Indeed, in a proof by contradiction, you put forward a hypothesis with the intention of disproving it, not supporting it.|
|A kyriotheory is something you see that is supported by comprehensive evidence and logic.||A hypotheory is something you see that lacks support. It could have come from a hypothesis, a conjecture, or a mere speculation.|
Note that the words “thesis” and “theory” do not come from the same root. They are complementary, in the sense that showing and seeing are complementary; that is, they describe the same process from opposite points of view.
The prefix “hypo-” comes from the Greek ύπο- meaning literally “below”, hence lesser in stature or (figuratively) lesser in importance. (The corresponding Latin prefix is “sub-”, which is perhaps more familiar.)
The prefix “kyrio-” comes from the Greek κύριο- meaning powerful, authoritative, masterful, or canonical. The English word kyriolexy refers to using words according to their literal, canonical meaning, as opposed to figurative or obscure meanings. Similarly, the word “Kyrie” (meaning Lord or Master) shows up in liturgical and/or musical contexts.
Here’s an example showing why these words are useful:
- * If somebody says “Oh, that’s just a theory” they must be talking about a hypotheory.
- * The theory of evolution is not a hypotheory. It is a kyriotheory.
- * If somebody says “Evolution is just a theory” they don’t know what they’re talking about. It’s not just a theory, it’s a kyriotheory. It’s a comprehensive body of facts and logic.
Here’s another suggestion: You can always say what you mean using plain English:
|If you mean comprehensive understanding, don’t say “theory” — say comprehensive understanding.||If you mean conjecture or speculation, don’t say “theory” — say conjecture or speculation.|
In particular, rather than referring to Darwin’s “theory” of evolution, it would be better to speak of our modern comprehensive understanding of evolution. (Darwin made an epochal contribution to our understanding, but things didn’t stop there. Nowadays Darwin’s evidence and reasoning are only a few strands in a very large tapestry.)
The poster shown figure 5 codifies a large number of misconceptions. Such posters – and the misconceptions they represent – are horrifyingly common.
As discussed in section 1, real scientists use many methods. There is not any cut-and-dried method for doing science, just as there is not any cut-and-dried method for writing a novel.
Here are some of the problems with the poster in figure 5.
As one example among many, theoretical physics and pure mathematics are perfectly respectable sciences, just not experimental sciences. They do not even remotely conform to the experiment-based “method” described on the poster. It is quite offensive to theorists to imply that their work is not science.
If the so-called experiment has only one possible thing that could happen, then it’s not an experiment, and it’s certainly not research; it’s just some kind of construction project or demonstration. See section 3.8.
This point is quite fundamental: In the early stages of a scientific inquiry, it is not reasonable (let alone necessary) to make a specific prediction. By way of analogy, if you were on a jury, you would want to wait until after you heard the evidence before deciding what the verdict would be. The same idea applies to science. As a specific example, in connection with the celebrated “twelve coins” puzzle (reference 20), before the experiment is done there is no way to make a meaningful prediction about the specific outcome. Any specific prediction would be merely a wild guess, the sort of thing you get from a crystal ball. It would be unscientific as well as pointless.
After the experiment has been done, the situation is different: We can predict that anybody else who measures the same coins will get the same answer – but even then, we should at least consider the possibility that the prediction could go awry. (Furthermore, of course, if they measure different coins, all bets are off.) Let’s keep things straight: Science makes predictions based on the evidence. Making a wild guess before obtaining the evidence is the opposite of science ... and the opposite of common sense.
|When Lewis and Clark started out, they were not testing the specific hypothesis that they would find bighorn sheep in the Rocky Mountains. Nobody on their team had ever heard of bighorn sheep or heard of the Rocky Mountains.||Instead, they had broad, flexible plans. They were hoping to find a wide range of flora and fauna. They weren’t hoping to find humongous jagged mountains athwart their path, and they were not even sufficiently prepared for the possibility. They had to improvise and get help along the way.|
I emphasize this point because it is so very harmful to require a young scientist to guess what the outcome of the experiment will be, and even more harmful to judge the work according to whether the preconceived notions were correct. It destroys curiosity, originality, and creativity. See section 3.8.
Somebody tried to tell me that a feedback arrow, to symbolize iteration, would have been out of place on this poster, because the idea is too complex. I pointed out that in another classroom down the hall from where I spotted this poster, younger students were being shown a poster of the life cycle of an insect, and in a classroom farther down the hall, yet-younger students were being shown a poster of the hydrologic cycle. Does this mean that as the students progress through the school, they become dumber and dumber, to the point where they can no longer understand cycles?
I have to wonder, did the folks who created the Scientific Method poster in figure 5 choose this font specially because they wanted to offend scientists in general and experimentalists in particular? ... Or did they choose it because the only scientists they knew anything about were the villains in monster movies?
In the interests of fairness, it must be pointed out that not everything on the Scientific Methods poster is terrible. In particular, the second item is OK. Scientific activities should have an identifiable purpose. Sometimes the purpose is a bit vague, for example when exploring unknown territory in hopes of discovering “something” ... but still there is a purpose, and the effort is expected to pay off on average.
There is a proverb to the effect that “you can’t beat something with nothing”. That is to say, even though the poster is worse than nothing, it’s hard to ask people to rip down the poster shown in figure 5 unless you can offer some sort of replacement. A possible replacement is discussed in section 5. Other possible replacements are provided by reference 21.
With occasional exceptions, science is a highly social activity. It involves huge amounts of collaboration and communication in various forms. This is not the sort of thing that most people think of when you mention “the scientific method” but it is a very big part of how science is really done. Writing a paper at the end of the project is part of what we mean by communication, but nowhere near all of it.
The process is also highly nonlinear. That is to say, it is iterative, with lots of loops within loops.
In some cases, it proceeds roughly as shown in figure 7, and as described in the following story.
The first time, he shoots it down immediately, pointing out that one step in your outline violates some famous theorem that you should have known about. So you give up on that idea. Eventually you come up with a new idea, and start over at step 1.
Eventually you come up with an idea that withstands preliminary scrutiny. Your buddy helps by filling in some of the missing steps.
Down the hall there is a senior guy who is very knowledgeable and famously skeptical. If you can explain it to him, and he doesn’t find anything wrong with the argument, you’re on the right track.
Also you practice explaining it to non-experts, including people from other disciplines and/or students. This tells you how much background information you must provide to a non-expert to make the story comprehensible.
In an industrial research situation, at some point you explain the idea to your department head. It’s his job to round up the resources – including additional team members – you need to finish the project. In the academic world there are guys who play a similar role. That is, they know what everybody else is doing, and can fix you up with a potential collaborator whom you might not otherwise have met.
At this point, you have a team. It’s a loose-knit informal team. Each person might belong to many different teams at the same time, working part-time for each of them.
The plan should be as flexible and open-ended as possible, within reason. This may involve grouping possible outcomes into general categories, and dealing with things category-by-category rather than detail-by-detail.
You do not need a crystal ball or a Ouija board to predict “the” outcome, precisely because the plan covers all the plausible outcomes.
You start by giving a practice talk in front of a tiny audience, perhaps just one expert and one non-expert. At the end of the talk, they critique your talk, slide by slide ... and also critique the talk overall. You redo the slides to incorporate their suggestions. You give another practice talk and iterate until it converges to something that is clear and correct. That is, clear enough for the non-experts to follow, and correct enough to keep the experts happy.
At this point you give the talk to larger audiences.
Giving talks is also part of the process of recruiting collaborators. Some people will start by replicating and verifying the work. Others will build on the work, adding new layers of ramifications.
Eventually it gets published in a journal, where everybody in the field can read it. You can also put the paper on the web, where yet more people can read it.
In some cases, good way to organize the report might be to list a set of interesting hypotheses that are consistent with the results, plus another set of interesting hypotheses that are ruled out. However, this is not the only format, or even the usual format.
If there is a list of hypotheses, it probably bears little or no resemblance to the list of scenarios you considered during the planning stage. The report should present what you actually learned from the data – not what you thought you would learn. Science is not a guessing game. You get credit for what you actually did, regardless of what you initially thought you were going to do. The purpose of the plan is to provide for safety and to provide for efficient use of resources. The plan is absolutely not a binding contract. It does not constrain the outcome of the work or the format of the report.
The plan should be flexible, whereas the report should be specific. Keep in mind the example of Lewis and Clark: They reported, quite specifically, finding bighorn sheep in the Rocky Mountains. There was nothing like that in the initial plan, nothing nearly so specific, for good reason. It was, however, within the general category of things of things they had planned for.
It takes judgment to decide which hypotheses, if any, to discuss in the report. You need to limit it to interesting hypotheses. Note that there are always an infinite number of uninteresting hypotheses. For example, the hypothesis that 2+2=13 is almost always ruled out, but it is not worth discussing.
See section 3.9 for more discussion of how results are explained, and how this differs from how they are obtained.
In science as in daily life, it is necessary to make approximations, as mentioned in item 3. For example, when you buy shoes, you don’t buy a pair that is exactly the right size; you buy a pair that is close enough to the right size.
Elementary arithmetic is exact, in the sense that 2 plus 2 equals 4 exactly. In contrast, physics, chemistry, biology, etc. are not exact sciences; they are natural sciences. For example, Newton’s law of universal gravitation:
|FI = G|
is one of the greatest triumphs in the history of human thought … but we know it is not exact. It is a very good approximation when the gravitational field is not too strong and not changing too quickly. It is also misleading, because FI is not the only contribution to the weight of ordinary terrestrial objects; there are significant correction terms from other sources including the rotation of the earth, as discussed in reference 22.
It is a common mistake to treat all approximations as equally good, or equally bad.
To say the same thing another way, when you are in a situation that requires making an approximation, that does not give you a license to make a bad approximation. It’s your job to figure out what’s good and what’s bad.
It is not always easy to distinguish good approximations from bad approximations. It requires knowledge, skill, and judgment.
Science rarely offers certainty. Often it offers near certainty, but not absolute certainty. (This is in contrast to religion, which sometimes offers absolute certainty, and to things like elementary arithmetic, which offers absolute certainty over a limited range.)
One of the surest ways to be recognized as a non-scientist is to pretend to be certain when you’re not.
The world is full of uncertainty. It always has been, and always will be. You should not blame science for “causing” this uncertainty, and you should not expect science to eliminate this uncertainty. Instead, science tells us good ways to live in an uncertain world.
Techniques for quantifying uncertainty are discussed in reference 3.
As mentioned in item 19, it is impossible for anyone to do anything without making assumptions.
Remember that a major purpose of scientific methods is to make useful predictions and to avoid mistakes. False assumptions are a common source of serious mistakes.
At this point, non-experts commonly say “don’t make assumptions” or perhaps “check all your assumptions”. Alas, that’s not helpful. After all, most assumptions are true and useful ... otherwise people wouldn’t assume them. The trick is to filter out the tiny minority of assumptions that turn out to be false. This is far easier said than done. There are too many assumptions, and it is impractical to even list them all, let alone check them all.
The real question is, which assumptions should be checked under what conditions? There is no easy answer to this question.
Assumptions can be classified, approximately, as explicit assumptions and implicit assumptions. Explicit assumptions are the ones you know you are making. They are usually not the main problem; you can make a list of the explicit assumptions and then check them one by one.
The big trouble comes from implicit assumptions that aren’t quite true. This includes things that “everybody knows” to be true, but are not in fact true, as discussed in reference 16. They also include rules that have become invalid because you have mistaken the provisos, as discussed in section 3.4.
Skilled scientists can question assumptions somewhat more quickly and more methodically than other folks, because they have had more experience doing it. But it’s never easy. All of us must rack our brains to figure out which assumptions have let us down.
It always looks relatively easy in retrospect. Once somebody has identified the assumption that needed repair, it is easy for everybody else to hop onto the bandwagon.
One sometimes-helpful suggestion is this: If you find a contradiction, inconsistency, or paradox in what you “know”, that is a good reason to start questioning assumptions. Start by questioning the assumptions that are most closely connected to the contradiction.
Some scientists keep lists of paradoxes. If an item stays on the list for a long time, it means there is a problem that is not easily solved, and the solution is likely to be a turning point in the history of science. Examples from the past include the Gibbs paradox, the black-body paradox, various paradoxes associated with the luminiferous ether, the Olbers paradox, et cetera.
An important component of science, especially of scientific research, involves exploring new territory. Commonly assumptions that were valid in the old territory break down in the new territory. Indeed when researchers choose where to explore, they often seek out situations where assumptions can be expected to break down, since that will reveal new information. For more on this, see section 8 and reference 23.
|In ordinary applications, when you want to rely on the model, you should stay safely within the limitations of the model.||In research mode, where the model is the object of research, you are testing the model, not relying on it. Then it makes sense to patrol along the boundaries, to see if the limits need to be tightened or loosened. It also sometimes makes sense to go far beyond the limits, in hopes of making a surprising discovery.|
As discussed in section 3.7, we should not overemphasize hypothesis testing. On the other hand, we should not overemphasize serendipity, either.
All too often, people tend to draw boundaries where no real boundaries exist, and tend to focus on the extremes when reality lies in the middle, far from any extreme. The following table shows some of the wrong ways to look at the situation:
To repeat: reality lies in the middle, usually far from any extreme. The table is typeset on a red background with lots of question marks, to warn you that it is unwise to make the distinctions on each row, and unwise to equate things in each column.
|At one extreme, it would be a mistake to think that research is precisely guided by pre-existing hypotheses.||At the opposite extreme, it would be a mistake to think that research is conducted at random, with no idea what to look for or where to look.|
Many things that are touted as “big discoveries” were actually developed, step by step, combining many small discoveries and many small inventions.
By way of example, when people found out about the magnetrons used for radar in World War II, they assumed it had been a sudden discovery. Actually it had been the subject of years of intense research and development, but the work had been kept secret.
As another example, Pierre and Marie Curie announced in 1898 the “discovery” of radium, but then they had to slave away in a decrepit shed for several years before they could isolate a pure sample and determine just how radioactive it was.
Thomas Edison said that inventing was 1% inspiration and 99% perspiration. I don’t want to argue about the exact percentages; the crucial point is that both are necessary. Inspiration alone won’t get the job done. Perspiration alone won’t get the job done.
I am reminded of the rock star who said it took him 15 years to become an overnight sensation.
Scientists understand probability and statistics. By definition, you can’t make a particular fortunate accident happen on demand, but you can work in an area where valuable discoveries are likely to be made from time to time. Everybody must accept some risk. For example, any sensible farmer knows there is some risk that a freak storm will destroy his entire crop. The key idea is that successful crops are sufficiently common – and the crop is sufficiently valuable – that the farmer makes money on average.
Scientists do not accept all risks, nor do they decline all risks. They accept risks that are likely to pay off well on average.
See reference 23 for more on this.
As mentioned in item 2, a major purpose of scientific methods is to make useful predictions and to avoid mistakes. The known scientific methods are a collection of guidelines that have been found to work reasonably well.
One of the most important steps in avoiding mistakes is to always keep in mind that mistakes are possible. This is so important that this whole section is devoted to emphasizing it and re-expressing it in assorted ways.
James Randi said you should take care not to fool yourself, keeping in mind that “the easiest person to fool is yourself”.
Another word for this is modesty. Being aware of your own fallibility is modest. Pretending you are infallible is immodest.
It is OK to a limited extent to be an advocate for your favorite idea, but you must not get carried away. When you collect data in support of an idea, you must also look just as diligently for data that conflicts with that idea. See section 10.3.
A related form of modesty, which is also crucial for avoiding mistakes, is to not overstate your results. Scientists use certain figures of speech that are designed to avoid overstatement. Among other things, this includes recognizing the distinction between data and the interpretation that you wish to place upon the data. As an illustration, imagine some children go on a field trip to the dairy. Upon their return, they write a childish report that says “cows are brown” – or, worse, “all cows are brown”. A more modest, scientific approach would be to say “the cows we observed were all predominantly brown”. A statement about the observed cows sticks closely to the data, while a generalization about all cows requires a leap beyond the data.
As mentioned in item 11 and section 3.4, practically all scientific results have some limits to their validity, and you must clearly understand and clearly communicate these limits.
Here is a very incomplete sketch of some of the issues that arise when taking data. (This stands in contrast to the rest of this document, which mostly emphasizes the analysis phase.)
Consider the famous Twelve Coins Puzzle as discussed in reference 20. Suppose you find a casino that is willing to pay you $350 for identifying the odd coin, but makes you pay $100 for each weighing. If you weigh the right combinations of coins, you can do the job in three weighings, so you make money every time. In contrast, if you follow a sub-optimal strategy that requires four or more weighings, you will lose money on average.
This scenario is reasonably analogous to many real-world situations. Commonly there’s a significant price for making a measurement, and you want to maximize the amount of information you get for this price.
I mention this because all too often, people claim that a principle of scientific experimentation is to “change only one variable at a time”. It’s easy to see that such a claim is not reliably true. The Twelve Coins Puzzle (reference 20) is one counterexample among many. If each weighing differs from the previous weighing by only one coin, you cannot come anywhere close to an optimal solution.
As discussed in reference 24, the suggestion to “change only one variable at a time” might nevertheless be good advice in some special situations. That’s because the cost of making a measurement is not always the dominant cost in the overall information-gathering process. For example, imagine a situation where gathering the raw data is very cheap, while just plain thinking about it is expensive. Then you might want to follow a strategy, such as changing only one variable at a time, that makes the data easy to interpret, even though you had to do a large number of experiments (much larger than theoretically necessary).
When an experiment involves human collecting or selecting the data, and especially when there is a task involving human subjects, elaborate strict measures must be taken to maintain the integrity of the results.
The problems that can arise are numerous, varied, and often subtle. Some of the variations have names of their own, including:
Techniques that can be used to defend against such problems include placebos and blinding (especially double blinding), plus the use of thoroughly randomized control groups.
For example, if an observer is interviewing a subject, it is all too common for the interviewer to telegraph the desired answer. Double blinding means that neither the interviewer nor the subject knows what answer is desired, so this particular problem cannot arise.
There have been notorious cases where forensic labs “helped” prosecutors by reporting false DNA matches, false bullet matches, et cetera. It is very difficult to prevent this. At the very least there must be strict blinding. There must also be careful calibration against samples from a suitable control group. Last but not least, there must be scrupulously independent testing, to detect departures from protocol in general, and false matches in particular.
Even when there is no human subject, for instance if a human observer is looking at an inanimate dial, there are many ways that bias can creep in. Sometimes blinding helps, but there may be serious practical disadvantages to this. Nowadays often the simplest thing to do is to digitize the readings and stream them into a file on a computer ... either instead of or (preferably) in addition to having a human observer.
It is OK to a limited extent to be an advocate for your favorite idea, but you must not get carried away. When you collect data in support of an idea, you must also look just as diligently for data that conflicts with that idea. Then you must weigh all the data fairly, and disclose all the data when you discuss your idea. (If you don’t do this it is called “selecting” the data, which is considered a form of scientific misconduct.)
The same applies to theories: It does not suffice to show that your favorite theory does a good job of fitting the data. You should diligently search for other theories that do a comparably good job of fitting the data.
This is what sets science apart from debating and lawyering, where advocacy is carried to an extreme, and it is considered acceptable to skip or make light of data that tends to support the “opposing side”.
Before discussing the principles, let’s look at a couple of example scenarios.
Scenario #1: Suppose you are tossing a coin, perhaps to generate random numbers for use elsewhere, or perhaps to test whether the coin is fair. As mentioned in item 15, section 2, section 3.7, and especially section 3.8, you should consider all the plausible outcomes. One way to do this is as follows:
In this scenario, outcome #3 is a “catch-all” hypothesis. Perhaps the coin misses the table and lands on the floor. Perhaps a magpie steals the coin and flies away with it before you can determine whether it is heads-up or tails-up. The nice thing is about this hypothesis is that we can write it down without specifying all the details. We do however need to know one crucial detail, as you can see from the following contrast:
|Suppose the failure of the coin to land on the table is not correlated with whether it would have landed heads-up or tails-up. In this case, we can simply veto the aberrant events. We toss another coin and continue.||Suppose the failure is correlated with what the outcome would have been. Perhaps a particularly malicious magpie is more likely to steal the coin if it is heads-up. Perhaps one side of the coin is slippery, making it more likely to slide off the table. In such a case, it is not acceptable to simply veto the aberrant events, because this would introduce a bias into the results.|
So, the key point is, we need to understand outcome #3 well enough to make sure it does not introduce any bias. We hope we can determine this without understanding every detail.
In science, depending on context, the phrase “selecting the data” can have very nasty connotations. For example, if we hear that “so-and-so was accused of selecting the data” it means that he allegedly manipulated the data so to introduce a bias.
|Sometimes this is done for bad motives, with the intention of supporting some preconceived notion.||Sometimes this is done for seemingly-good motives. Sometimes people think they can “improve” the data by rejecting data that is “implausible” or “out of range”.|
It must be emphasized that this is considered misconduct, no matter what the motives. Don’t do it! Even in routine situations, it introduces a bias into the results. Perhaps more importantly, it deprives you of the chance to make a non-routine discovery, by figuring out why the data is out of range.
Here’s another scenario involving a catch-all hypothesis: Suppose we want to compute the average aspect ratio of eggs, i.e. the ratio of length to width. There are millions of eggs available, and we can’t afford to measure them all, so we choose a subset and measure that. We must make sure that the sample is chosen in such a way that no bias is introduced.
The experiment can have various outcomes. We can categorize them as follows. For each egg:
For example, perhaps the egg falls to the floor and gets smashed before the measurements are complete. Again, we must be careful. There are two possibilities:
|Suppose the breakage is uncorrelated with the aspect ratio. In such a case we can veto the broken egg and measure another instead.||Suppose that high-aspect-ratio eggs are more likely to be dropped. This could seriously degrade the aspect-ratio measurement. If there is any significant chance of this, we need to account for it, or redesign the experiment to prevent such problems, and start over.|
So, here are some principles:
Recall that a measurement generally has both a nominal value and an uncertainty (“error bars”) as discussed in reference 3. Vetoing out-of-range data is particularly likely to distort the error bars, which is unacceptable, even if the nominal value is not greatly affected.
In advanced experiments, it is sometimes necessary to have a trigger or veto or triage mechanism, i.e. some rule that selects which data will be kept and which will be discarded. Doing this right is very, very tricky ... so tricky that usually it is simpler, cheaper, and all-around better to just keep all the data. Also: when you write up your results, you should describe the trigger criterion in detail, so that readers can judge for themselves whether the trigger introduced any significant bias.
To repeat: The phrase “selecting the data” may not sound nasty the first time you hear it, but the connotations are very nasty indeed. There are various things that could be going on:
The burden of proof is on you, to show that whatever you are doing is legitimate, i.e. that it does not bias the conclusions.
Keep in mind the common-sense principle that you should never put yourself in a position where the first mistake is fatal.
In the real world, scientists and engineers do simulations and dry runs. They build pilot plants before committing to full-scale operation, so that most of their mistakes will be small mistakes.
Therefore you should arrange your experiment so that you can take some data, then do some analysis, and then come back and take some more data. This allows feedback from the analysis phase back to the data-taking phase, so that you can improve the data-taking if necessary.
This is all the more important for students who are, after all, students not experts, and can be expected to make mistakes. Yet we want students to get good results in the end.
If at all possible, arrange it so that analysis (at least some sort of preliminary analysis) happens in real time, so that if anything funny happens during the experiment, the experimenter knows about it, for reasons explained in section 10.5.
Let me tell a little story. Once upon a time in a mythical place called La Jolla there was a fellow named John who really liked numbers. The more digits the better. He wanted unbiased numbers so dearly that he would have his grad student, Richard, face the instrument with his eyes closed; when John said “Now!” Richard would momentarily open his eyes and observe the number. John would write it in the lab book. After doing this for many days and weeks, they had lots of lab books filled with lots of numbers.
They were looking for some sign of a superfluid transition in liquid 3He at low temperatures. John had already made a string of important discoveries in low-temperature physics, and finding the superfluid would move him from pre-eminence to immortality.
Aside: The nice thing about theoretical predictions is that there are so many to choose from. Predictions of Tc started at about 300 millikelvin. When that was disproved experimentally, the predictions moved to lower temperatures: 100 mK was disproved; 50 mK was disproved; 30 mK was disproved; 10 mK was disproved. In fact John and Richard had checked experimentally down to less than 2.5 mK without seeing anything. So the theorists gave up and lowered their prediction to something in the microKelvin range.
Once upon the same time, in a mythical place called Ithaca, there were a couple of guys named Bob and Doug. They liked numbers OK, but they also liked graphs. They wanted to see the data. And they didn’t want to wait until the experiment was over to analyze the data and see what it meant; they wanted to see the data in real time. This is in the days before computers, so if you wanted a graph you had to suffer for it, fooling with Leeds chart recorders, i.e. mechanical contraptions with paper that always jammed and pens that always clogged.
One Thursday in late November, Doug was watching the chart as the cell cooled through 2.5 mK. There was a glitch on the T versus t trace. Doug circled it and wrote “Glitch!!” on the chart. He warmed back up and saw it again on the way up. And then again on the way back down. He called Bob. Bob put down his plate of turkey and zoomed into the lab. The two of them stayed up all night walking back and forth through the “glitch”.
I’ve seen that chart. It doesn’t look like what I would call a glitch. It’s more of a corner. A small-angle corner, just a barely-visible departure from the underlying linear trend. Eyes are good at spotting corners in otherwise-straight lines.
Doug and Bob assumed, based on the aforementioned experimental and theoretical evidence, that this wasn’t the superfluid transition. They figured it was something going on in the nearby solid 3He. But they eventually figured out that it was indeed the superfluid. Right there at 2.5 mK.
Everybody assumes that if John and Richard had been plotting their data on a strip-chart recorder, they would in fact have discovered the superfluid. But they didn’t.
Now, imagine what it was like working in Bob’s lab after that. With strictness bordering on fanaticism, strip-chart recorders were attached to all the significant variables ... and even some of the not-very-significant variables.
Every so often, a new baby grad student, his fingers stained N different colors from trying to unclog the chart pens, would ask whether we really needed all those chart recorders. Somebody would explain by saying, Once upon a time in a mythical place called La Jolla, .......
Lab notebooks are not supposed to be perfect. If there are no mistakes in the lab book, the lab book is presumably a fraud. You are allowed to mark bad data as bad, but you are not allowed to obliterate it or eradicate it. See reference 25 for an example of a well-kept lab book containing a correction.
You are not allowed to “clean up” the books. You are certainly not allowed to keep two sets of books (a dirty one to record raw data, and a clean one to show off).
There are many reasons for keeping good records. First and foremost, you and your collaborators need the information on a day-to-day basis. During the analysis phase, it is all-too common to find that the data cannot be analyzed, because even though the nominal result of the measurement was recorded, the conditions of the measurement were not adequately recorded. There’s no value in knowing the ordinate if you don’t know the abscissa.
As mentioned in section 10.3, record all the data. If possible, hook up a computer to stream all the data into a file. Record the abscissas as well as the ordinates.
The rule is to keep good records. It is traditional to keep data in a so-called lab book, aka laboratory notebook, which is one way of keeping records, but not the only acceptable way. Good electronic records are an acceptable alternative, and are in some ways better. For example, if a witness signs and dates a page in a lab book, a bad guy might be able to add something to the page later. Sometimes this is detectable, but sometimes not. (Don’t try it.) In contrast, if an electronic document is cryptographically signed, there is no way to alter any part of the document without invalidating the signature.
There are various ways of obtaining unforgeable date-stamps on electronic documents; see reference 26 for an example. If you don’t want to bother with that, one option is to just email the document (or a hash thereof) to your lawyer, with a cover letter saying that you are not asking him to take any action, just to file away the document, so that if there is ever any dispute he can authenticate the date. Note that if the document is huge and/or highly sensitive, you don’t need to send the document itself; it suffices to send a cryptologic HMAC (hashed message authentication code).
For things like blueprints, circuit diagrams and computer programs, electronic records have huge advantages. The point is that such things get revised again and again, and if you just have “the” document, you have no idea who contributed what when. In contrast, a modern revision control system such as git will conveniently keep track of the current version and all previous versions. It also keeps track of who submitted what changes, and when. Submissions can be digitally signed.
If you are collaborating with people at other locations, electronic documents have tremendous advantages.
If you are stuck with an ordinary lab book, and you want to revise or annotate a page after it has been signed, one good way is to attach a sticky note to the page. The note should refer the reader to a later page where your latest&greatest thoughts can be found. Using a sticky note guarantees you can’t be accused of altering the page after it was signed.
In addition to helping you and your collaborators directly, good records have other uses. They are particularly important in connection with patents. Questions about inventorship and priority come up a lot. I’ve seen it first hand, many times. Once I came out on the short end of the stick, which didn’t bother me, because the other guy had records that convinced everybody (including me) that he invented the thing a month before I did, fair and square. More commonly though, one claimant has copious records documenting the gradual evolution of the invention, while the other claimant is just a pretender, with nothing but a bold assertion covering the final, perfect result. Sometimes disputes arise after an application has been filed or after a patent has been granted, but in a large company or large university, disputes can arise intramurally, before the application is filed, when the lawyers try to figure who should be named as inventor(s) on the application.
You should ask your patent attorneys what they want your records to look like ... and then ask them what they will settle for. It seems some lawyers would “prefer” to see every detail recorded in a lab book, then signed and notarized in triplicate ... but they will settle for a lot less formality.