Tangents  
 Created: 28 Dec 2005  Copyright © 2005-2006 by owner.
Standard citation procedures apply.
Modified: 26 Oct 2013 
     


Scientific Method

Primer / Refresher

Just about anyone who has taken an adult-level course in one of the natural sciences has encountered something called "scientific method."  However, it is common for people who do not enter the sciences as a profession to forget this fundamentally important, if less than thrilling, chapter.  Thus, when we encounter evangelists deriding a scientific idea as "only a theory," or parents demanding that "alternative theories" be taught in their children's science classes, the topic of scientific method inevitably comes up—or should.  "Do you understand the rigorous challenges and demands of scientific method, which an idea must withstand in order to qualify as a scientific theory?"  (Those who use the phrase "only a theory" evidently do not understand the term as it is used in science.)  "Have your alternative theories satisfied the demanding criteria of scientific method?"  (Unless the alternative theories have survived being independently scrutinized and methodically tested by those professionally trained in natural science, they have not satisfied these criteria, and thus do not qualify as scientific theory.)

 The Purpose of Scientific Method   Science is the methodical investigation of nature, and scientific method is the accepted process for conducting that investigation, with the fundamental understanding that science cannot create or shape reality, but merely attempts to describe and explain it.  Of course, these descriptions and explanations are sometimes found to be less than perfectly accurate (and on rare occasions, utterly false).  Since science accepts that it cannot change reality to conform to its ideas, it must change its ideas to make them conform to reality, as nearly as current methods of observation allow.  Scientific method is a rigorous and thorough discipline for systematically determining, by careful observation and cogent reason, whether or not an idea is in accord with natural reality.  Only after an idea has satisfied all criteria of the scientific method can it be classified as a scientific theory.  Scientific method thereby ensures that the term "scientific theory" is reserved for ideas that have withstood prolonged and intense scrutiny under the unrelenting light of all available pertinent evidence.  Scientific method thus assures us that the scientific theories of today are solidly supported by substantial factual evidence, cogent reason, and exhaustive testing, and have so far survived all attempts to discredit them.

 The Basic Work Cycle   Scientific method demands adherence to principles of meticulous observation, careful reasoning, factual verification of ideas, honest reporting of findings, and critical peer review.  While application of scientific method is in most cases very demanding, its process is relatively easy to understand in principle.  In its simplest form, scientific method employs a basic loop of activity: observation, explanation, prediction, and back to observation, in a progressive cycle.

  • Observation of a phenomenon gives rise to speculation about an explanation for it.

  • The tentative explanation is used to formulate a testable prediction.

  • The prediction is revealed to be correct or incorrect by subsequent observation.

A correct prediction causes the cycle to be repeated, in order to explore other aspects and limits of the phenomenon, refining the explanations and predictions in the process.  However, if a prediction turns out to be incorrect, the cycle branches.

  • If a discrepancy can be accounted for, the explanation may be revised and a new prediction tested.

  • If a discrepancy cannot be reconciled, the explanation may be rejected or replaced.

If a faulty explanation is revised or replaced, the new version is then run through the cycle in like fashion.  However, if we run out of explanations or predictions, then we set the matter aside as unresolved, until either additional evidence comes to light or new ideas emerge.

Important considerations in scientific experimentation include:

  • Repetition: Does the same phenomenon occur repeatedly and consistently?

  • Universality: Does the same phenomenon occur under all conditions?

  • Verification: Do accumulating results support or contradict the prediction?

  • Exception: In cases where the phenomenon changes or does not occur at all, can those variations and exceptions be rationally explained?

 The Overall Process   Beyond the basic observation-explanation-prediction cycle, scientific method entails a number of interconnected stages: initial development, independent investigation, demonstration of value, and conflict resolution.  The overall sequence can be outlined as follows:

  • Initial Development

    • Observation: A phenomenon is observed to exhibit a pattern or suggest a relationship.

    • Speculation: An idea is proposed to explain the pattern or to define the apparent relationship.

    • Hypothesis: Speculation is formulated as a falsifiable hypotheses, a statement that can be shown to be false if there is any instance in which it is observed not to hold true under specified conditions.

    • Experimentation: A strategy is devised and executed to test whether or not the hypothesis consistently holds true, with either controlled experimentation or methodical observation, along with collecting, recording, and correlating of pertinent data—including reports of any unexpected results.

  • Independent Investigation

    • Publication: Results of successful experiments, including any discrepancies noted, are published in a scientific journal in order to elicit questions, comments, and independent testing.

    • Peer review: Interested scientists scrutinize the published method and results, raising questions and attempting to spot any errors or omissions.

    • Independent testing: They also conduct further experiments or observations of their own devising, in an attempt to show whether or not the hypothesis holds true under all realistically conceivable conditions.

  • Demonstration of Value

    • Explanatory value: A theory is a general explanation for specific observations.  It must actually explain in a way that makes the phenomenon more understandable in terms of nature; it cannot simply invoke mysterious powers which are themselves unexplainable.

    • Predictive value: A theory predicts the behavior of a phenomenon, its causes, its results, and its aspects.  A theory  must be reliably predictive. furnishing a solid basis for further inquiry; otherwise it is not of much value to science.

  • Conflict Resolution

    • Competition: If there are competing theories regarding the same phenomenon, then they must be compared, to determine whether any of them affords a significant advantage in explanatory or predictive power, or whether any of them appears to be more consistently in accord with the available body of evidence.

    • Conflict: If there is conflict between new and established theories, then further examination and testing of both is called for in order to resolve discrepancies, to refute one of the theories, or to identify the separate conditions and limits under which each theory holds true.

At any stage of the process, if significant discrepancies are discovered and cannot be credibly accounted for, then the hypothesis is effectively "falsified," and must be either rejected, or else revised and started through the process once again.  In addition, especially in the case of slow processes or rare events, it might well take years or decades (indeed, sometimes centuries or longer) to acquire the necessary data to confirm a hypothesis, so scientific method can be frustratingly time-consuming.  (That is where accurate measurement and record-keeping become important.  Old but reliable data, considered in ways not previously imagined, sometimes lead to significant new perspectives, insights, and discoveries.)

 Limitations   Is scientific method perfect?  No, just as our knowledge, senses, and tools of measurement are not perfect, to say nothing of our often biased interests.  Faulty hypotheses sometimes do become accepted as theory.  However, scientific method is an ongoing process.  It continually reexamines existing theories in light of new evidence, and odds are overwhelming that any error will come to light eventually.  The rigorous process ensures that surviving scientific theories are very durable.  If a theory is shown to be flawed, it is often the case that it need not be discarded, but only qualified or modified to bring it back into compliance with advancing knowledge.  A case in point is Newton's theory of gravitation, whose calculations were regarded as "natural law" for three centuries, until superseded by those of Einstein, which better explain the workings of gravity and predict its behavior under extraordinary conditions, as confirmed by actual observation.  Notwithstanding, Newton's simpler calculations continue to provide adequate accuracy for most practical purposes, even to calculating multi-year, multi-billion-kilometer space vehicle trajectories.

 Assembling the Puzzle   Occasionally there arises a theory that ties several other theories together, bringing a grand order and cohesion to what previously seemed disconnected bits and pieces.  In the 17th century, Isaac Newton's laws of motion constituted the first such unifying advance for modern physics.  In the 19th century, Charles Darwin's theory of biological evolution did much the same for biology, making sense of the underlying connectedness and relationships among a bewildering diversity of species, as well as enhancing understanding of genetics, geology, and paleontology.  In the 20th century, Alfred Wegener's theory of plate tectonics performed the same service for geology, at once explaining a variety of phenomena, from the complementary coastlines of Africa and South America to mirror-image magnetic patterns on opposite sides of seafloor trenches, from earthquake fault lines to chains of volcanoes, and even the fossil record of identical biological species on long-separated landmasses.  When such a theory reveals the previously hidden relationships among centuries of diverse disciplines and locks them into place, there can be little doubt that the theory is thereby confirmed in its essentials, even if some of its details have yet to be ironed out.

 Misperceptions   A common myth (a remnant of smug overconfidence of the 19th century, that science had essentially answered all questions worth asking) is that science is supposed to prove ideas true or false.  This is not the case.  Science deals in probabilities, not certainties (though some probabilities approach certainty closely enough that the minuscule margin of doubt can be disregarded for practical purposes).  Science draws conclusions only on the basis of currently available evidence within the current constraints of observation and measurement, with the full understanding that any conclusion might subsequently be invalidated by evidence yet to be discovered.  So the most science can say about a theory is that all evidence discovered so far supports it, and none contradicts it.  Confirmation of a single wrong prediction or contradictory fact can cause even the most widely accepted  theory to be challenged and rejected.  Thus the existence of an extensive body of coherent theory that has withstood decades or even centuries of intense scrutiny attests to the rareness with which such rejection occurs, and to the effectiveness of scientific method in weeding out ideas that are not in accord with the reality of nature.

 Perspective   In dealing with nature, there is always a risk of error or misinterpretation; where humans are involved, it simply cannot be avoided.  Indeed, in science a margin of error is always expected and is routinely taken into account, and the whole point of scientific method is to identify and reject the inevitable misinterpretations.  But no human process is perfect, so science makes no claim that its findings are absolutely certain..  On the other hand, science is one discipline that tends to be self-correcting, through continual reexamination and retesting of existing theories in light of new evidence.  Consequently, despite its shortcomings, scientific method has demonstrated itself the most reliable and verifiable way yet to learn about nature with a high degree of confidence, even if we must accept that reliability, verifiability, and confidence do not equate to certainty.

Thus, if someone claims to know that an idea is "an absolutely scientifically proven fact," we can be sure only that he or she knows nothing of the sort, and understands little if anything about science itself.  On the other hand, if we know that an idea is a scientific theory, we can be confident that, though the idea is not absolutely certain, it is better supported by abundant evidence and cogent reason than are most ideas in human experience—a far cry from the mere speculation that typically passes for "theory" in non-scientific pursuits.

=SAJ=