In earlier postings I shared some thoughts on the assumptions and limitations of science. This posting focuses on the power and strength of science as a way of learning and knowing about the natural world.
In my earlier postings you read that one limitation of science is that science we can address only questions that are objective AND empirical, and another limitation is that there is no way we can be absolutely sure that a scientific explanation is correct. These are limitations, but they are also the basis for some of the great power and strength of scientific inquiry.
Strength #1: Independent confirmation or refutation via empirical evidence.
The scientific approach to answering objective questions about the natural world always includes the collection and analysis of empirical evidence. (Objective questions are those that have a definitive answer, and something that is empirical can be investigated through observations via our physical senses or technology that extends those senses.)
Because scientific explanations are based on empirical observations, anyone with access to the right kind of equipment (if needed) can replicate an experiment or collect their own observations independently, and independently test their own evidence to see whether a scientific conclusion is confirmed or refuted. If claimed evidence cannot be replicated, then the conclusion is put into significant doubt and others will carry out their own studies to add to the body of observations which eventually becomes so compelling that the original explanation is either accepted, modified to explain all available evidence, or rejected in favor of a different, but stronger explanation. Plus, this process of discovery and dissemination of scientific explanations includes independent, critical peer-review before it can be accepted for publication in a professional journal. Then, once published, the information is read and assessed by the larger scientific community that in most cases carries out independent tests that allow for independent confirmation or refutation before it is accepted by the larger scientific community as a viable explanation.
This approach minimizes researcher bias and the chances that poor methodology or faulty or poorly supported conclusions will make their way into the accepted body of scientific knowledge.
Point #2: The self-correcting nature of science.
Statistical analysis of empirical data, and consideration of new evidence as it becomes available are a routine part of most scientific studies. Result of these tests provide a statistical level of confidence we have in relation to hypotheses we test. These levels of confidence are based on mutually accepted levels of confidence that are based on statistical critical values that have been calculated by statisticians for data sest of definitive sizes and for each kind of statistical test that exists.
These critical values allow us to identify the likelihood or level of confidence we have in accepting a scientific explanation as valid. The scientific community typically requires a researcher to be at least 95% confident that a particular explanation (hypothesis) should not be rejected before it can be considered a viable possibility.
When a scientist carries out a research project they most often employ two preliminary explanations - hypotheses. One represents the researcher's best prediction of what the outcome or eventual explanation will be. This is called the Alternative or Research Hypothesis. The other hypothesis is called the Null Hypothesis. This hypothesis is a statement that says that the correct explanation is anything other than the Research Hypothesis. The Null Hypothesis is the one that a researcher tests and must decide whether to reject or fail to reject based on the analysis of empirical evidence. The decision about what to do about the Null is determined by the amount of possible error that is associated with the outcome of the statistical tests. What this means in practice is that a researcher must be more than 95% confident that the Null Hypothesis is NOT correct before they reject it. The other 5% represents the amount of error that exists in relation to that decision.
Actually, there are two types of error associated with this kind of decision-making. One type is the possibility of accepting an explanation when it should have been rejected, and the other type is the chance of wrongly rejecting an explanation when it should have been accepted.
So if the outcome of a statistical test shows a p-value (probability value of the null hypothesis being correct, or level of error in decision-making) is greater than 0.05 or 5% the researcher is compelled to fail to reject (i.e., accept) the null hypothesis. If the p-value is smaller than 0.05 or >5% the researcher is compelled to reject the null hypothesis. Only if this happens can the researcher consider the research hypothesis as a viable possible explanation. it does not, however, mean that the research hypothesis is correct. It means only that it has not been rejected as a possible explanation.
Since this process of eliminating possible explanations has been going on systematically for around 300 years now, many, many weak or incorrect scientific explanations have been corrected or rejected in favor of better ones. What this also means is that whenever this process is applied, there is ALWAYS a margin of error, slim though it may be, associated with every decision. What this also means is that as hypotheses are tested and rejected or not, we get progressively closer to describing truths about the natural world and how it works.
About now I hope you are asking yourself "Is it possible to discover absolute truth through science?"
I strongly contend that not only can science discover absolute truth, but that it does so on a regular basis. The problem though is that while absolute truth can be discovered, it is impossible to be absolutely confident that what science has been discovered is the absolute truth.
Because science cannot be absolutely confident in its discovery of truth, scientists continually test explanations with new sets of data collected in new ways or with new types of technology. Hypotheses that that bear up under repeated testing become theories. Theories that have withstood the test of time and many repeated tests for validity are considered strong theories. Those that do not hold up under this type of scrutiny are either modified to explain all previously existing pertinent data and new data, or they are rejected in favor of new explanations that are able to explain all pertinent data.
This approach to testing, re-testing, and improving, or revising and replacing explanations is referred to as "The Self Correcting Nature of Science" which is one of the greatest strengths of the scientific approach to discovery of truth. Dr. John Moore expressed the power of this aspect of science when he wrote, "Great art is eternal; great science tends to be replaced by greater science" in his book, Science as a Way of Knowing: The Foundations of Modern Biology."
In conclusion, science relies on independent review in order to minimize the effects of personal bias and to maximize the quality of scientific explanations. It also includes a systematic process for eliminating weak or incorrect scientific explanations in favor of more complete or better-supported explanations. These two strengths make the scientific approach an extremely powerful way to discover truth about the natural world and how it works.