I’m interested in the Bayesian approach and subjectivity of scientists. I had recently read that scientists used to develop their theories and synthesize theoretical frameworks introducing beliefs, making subjective inferences and sometimes hadn't even experimented with the phenomenon they were talking about. Those cases include Alexander von Humboldt, Albert Einstein, Charles Darwin and Sigmund Freud, and others. In those terms, it’s necessary to be sure when and why the conclusions are objective, if they are.
Our perception is that, following the scientific method, our conclusions are objective but our interpretations are subjective. It is difficult to only make statements relating to phenomena with which we have directly experimented. How else could we build upon the current body of knowledge? Evidence-based inference is still subjective as it can be interpreted multiple ways. At least in health and medical research how can we inform change/intervention without such subjective inference. I agree that it is necessary to understand the uncertainty of the interpretations from our research results. Quantifying this uncertainty will be difficult. G
Even in the most formal research areas like mathematics, objectivity (or thruth) of a theorem that was formally correctly proved still depends on an assumption that the underlying axioms (that cannot be proved) hold. Theories are not facts. Theories are developed based on all researchers' knowledge and intuition. They can be tested, but rarely proved.
It is important to clearly and transparently present the assumptions on which research is based, describe the research design and data analysis methods, and present your results in a way that will allow others to perhaps interpret them differently. If a researcher makes assumptions that are not likely to hold, than conclusions will not be objective, disregarding the scientific methods she applies.
Unfortunately, assumptions are often not clearly stated in research papers. Some researchers apply statistical methods even though assumptions for their use are clearly not met. If you are not sure about the assumptions underlying the research, it lowers your confidence in research conclusions. I agree with Gavin, quantifying this confidence is quiet a challenge.
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis