Problems With the Scientific Method
- Thread starter TruthSeeker
- Start date
- Status
You'll be sorry you asked:
False = P(!True)
So if something is false, there is a probability that it could be true?
So what exactly is science dealing with?
Sure! You seem to think there is such a thing as absolute truth or falsity. Interesting.So if something is false, there is a probability that it could be true?
Likelyhoods, probabilities, refinement, shake well, repeat.
Hi Supes, long time no see.
If I may, I think I can see where both you and Sam are coming from. (Or maybe not)
Sam, yes or no?
Supes, yes or no?
If I may, I think I can see where both you and Sam are coming from. (Or maybe not)
It's spherical or not. The degree of sphericity is subject to variation. When is sphere not a sphere.
Black holes exist or they do not. Our knowledge of this truth is uncertain.
Again, it does or it does not, our certainty of this is less than 100%.
Sam, yes or no?
Supes, yes or no?
So science deals in assumptions, probabilities and inferences, not truth or falsehood. Which is what I said.:shrug:
Not quite. You are missing the fact that science needs to come to at least provisional conclusions about a given thing, otherwise there would never be a growing foundation of results to build on. Those conclusions are TRUE or FALSE based on probabilities and inferences.
Science must deal in the level of confidence or certainty of things, just like everyone else.
you're still talking about assumptions leading from inferences and inferences based on those assumptions.:bugeye:
This thread was inspired by medicine-related news. For a long time, it was believed that lycopene, a substance found predominantely in tomatoes, helped to fight cancer. However, recent research shows that lycopene actually worsens cancer! At the same time, they now believe that apple peel is very helpful to fight against cancer.
Ok. I'm not a genius, but there is only one truth. Lycopene is either helpful or harmful. So why does one research shows that it's helpful and the other harmful?
Some scenarios:
1) We gained new knowledge before the new research, therefore, the scientific method is limited by our knowledge range.
2) Our methods of research improved, therefore the scientific method is limited by the accuracy our research methods.
3) Our statistical methods are insufficient to produce reliable results.
4) Other.
5) A combination of the above
I would think one of the main problems is that the use of statistical methods are insufficient, on their own, to produce a reliable result. One must create logical explanations if they want to create an accurate theory. For instance, if apple peel seems to really help, then you have to specify the components in the apple that produce such results AND not only test it with statistical methods, but also carefully research how those components interact with our bodies in different circumstances. Of course, the difficulty in this scenario would be in terms of time and money restrictions, as such pedantic practices would be extremely time consuming.
Anyways... any thoughts?
Clinical trials, no matter how painstakingly conducted, still isolate subjects into a controlled study. An isolated study, though informative, cannot field every situation, just as you posit regarding statistical method.
However, I think that one of the more common flaws in clinical research is that they are treating a symptom more than they are a condition and vise versa. In this light, a great many factors can come into play, including incidences where a condition is detected and treated as a symptom, and where a symptom is a detected and treated as a condition.
Inaccurate correlations occur way too often in statistical research, resulting in confusion as well as adverse events in a patient's or subject's health.
No!
I think we just settled the issue of whether science deals in truth or falsity above. A scientist will tell you that it is absolutely true that the earth is spherical.
I suggest we agree that science accepts the truth or falsity of demonstrable binary propositions, and assigns a degree of certainty to everything else.
Yes?
Is that a true no/yes? or a true = P(!F) no/yes?
It's a query regarding the degree of certainty you would assign to the preceding proposition.
:m:I should have gone to bed.
I know what you both mean I think, and I can't find the words.
What isn't a binary proposition?
Electrons exist or they do not.
Our knowledge of their behaviour/ properties is subject to statistical uncertainty.
Erm. Do they exist, or are they only a way we have of explaining things?
I may go to bed. And not sleep.
More coffee, first.
I know what you both mean I think, and I can't find the words.
What isn't a binary proposition?
Electrons exist or they do not.
Our knowledge of their behaviour/ properties is subject to statistical uncertainty.
Erm. Do they exist, or are they only a way we have of explaining things?
I may go to bed. And not sleep.
More coffee, first.
It's a query regarding the degree of certainty you would assign to the preceding proposition.
At what level of significance?
I should have gone to bed.
I know what you both mean I think, and I can't find the words.
What isn't a binary proposition?
Hmmm...
Maybe an example.
Quantum mechanics is probably the most wildly successful theory in human history. It predicts the behavior of everything but gravity. So you could say that QM is a true representation of nature. But 100% true? Could there be a better, more refined truth?
Newtonian mechanics is true. Except in high gravity or at high relative speeds. Then it's superseeded by general or special relativity.
I will stick to my "truth is provisional" idea regarding such non-binary questions.
- Status