Contributors: KatieC, Emily, Ming-Ray, and Dani.
Attempts to replicate several famous scientific studies result in a declining effect, meaning that the extent of an observed effect gradually decreases when the test is repeated. For example, a study claimed that birds preferred mates with more symmetrical appearances, supported by data that eight out of eight females chose mores symmetrical mates. The next year nine out of twelve confirmed the theory, and with each year, fewer and fewer birds exhibited the inclinations of the first birds, until it was no longer a correlation.
Some explain it as a "regression to the mean" - that the first result was an outlier and as time progressed, the results neared the true outcome of the experiment; others suggest the possibility that the act of observing changes the nature of reality - such as a thermometer will only ever measure the temperature of a substance AND the thermometer itself, and never ; and the vast numbers of unknown variables at play that could have affected the study without the researcher’s knowledge. This does not appear to apply to constants such as the law of gravity, and Newton’s Laws.
How reliable is science? (How do we know there won’t be a declining effect on everything we are “sure” about?) How reliable is the inductive method?
Most, if not all the conclusions that we call “knowledge”, have been reached through inductive reasoning - the assumption that because some principle has held true in the past, it will continue to hold true in the future. In fact, the whole scientific process is centered around the examination of one variable - the independent - and is based upon another assumption, that it is the only variable affecting the outcome of the experiment. The experiment is repeated adnauseam before it can be “induced” that it will always happen. Ironically, by definition, “assumption” is “a thing that is accepted as true or as certain to happen” based upon past evidence - induction.
The decline effect raises the question ofwhat is certain and what is uncertain in science. If a universally accepted theory or law could gradually decline and someday become so insignificant that the theory is disregarded entirely, it implies that science could be heading towards nothing now. A great finding might only turn into nothing as more trials are carried out. With the decline effect in consideration, the only way for us to be sure if a theory will show the decline effect or not is to try it out infinitely, but that is impossible. People will never be able to carry out a trail infinite times, and will therefore never know if the theory stands in the infinite tries. Then, people might claim that there is no point of studying science anymore if everything just might disappear someday. The act of observing changes the nature of the observed, and so the natural state of the world can never be observed.
The implications of this article are that inductive reasoning is not reliable because the universe is too arbitrary to be organized, and thus all scientific pursuits are a waste of time. However it would be pragmatic to keep in mind that the research that this article is based on, was ALSO based on induction - because many theories have demonstrated the decline effect, there is the tentative conclusion that ALL theories will eventually demonstrate the decline effect. To discount induction based on this article, would be to discount the article itself, a pointlessly paradoxical cycle.
Another example that can be traced back to the same mysterious decline effect can be seen by Jonathan Schooler’s study on memory. Schooler tried to prove that people who had to describe their memories were less able to remember them than people who didn’t have to describe their memories. The results of his study showed that people who had been asked to describe the face of a robber in great detail after watching a video were less good at recognizing the face than those that did not engage in any description at all. His first experiments were positive, and thus supported his theory about “verbal overshadowing”. However, over the next five or six years when he tried to replicate the study, the effect declined 30%, and then another 30% when he tried to replicate it a second time. Schooler’s study demonstrates the decline effect. Again, this may or not be due to the limitations of the scientific method. Some even suggest that the decline effect itself is not an erosion of truth, but instead, a failure of theorizing (Anderson).
No comments:
Post a Comment