Accuracy compared to what?

[NigelW]

I'm not sure if this is relativism creeping in or what, but I do have a thought I would like to have challenged. This may show my lack of education in science, but I am looking for some help.

 

Theories cannot be accurate compared to empirical reality and definitions/theories are not empirical reality. Accuracy can only be arrived at through comparison against other theories.

 

"Your theory is inaccurate..." is less accurate compared to "...your theory compared to my theory is more accurate..." which is more accurate.

 

It is more accurate because there a comparison.

 

Certainty as well...

 

You cannot be certain of a theory compared to empirical reality. You can only be certain that theory is more or less accurate compared to another theory.

 

You can be certain that the statement "...your theory that cars have jet engines is inaccurate..." is less accurate than "...your theory that cars are driven by jet engines is inaccurate compared to my theory that the internal combustion engine is mainly used..." because I am comparing to another theory.

 

(Theories that are accurate, per UPB in a nutshell, must be empirically tested.)

[TheAuger]

Theories can only be proven false based on other empirical, verifiable observations of reality, not other theories. Theories cant be proven to be "true", only "true enough" (based on certain data).

 

Certainly theories/philosophy could give a shit about our preferences for jet engines or internal combustion engines, which is a matter of aesthetics/taste.

[Pepin]

I think the best approach to responding to this is just to give my own answer as opposed to responding to your argument. I apologize if this is annoying.

 

The accuracy of a theory is based on the comparison of predicted measurement against the actual measurement. If a theory predicts the strength of a magnetic field generated by an inductor with a specific current and voltage running through it, the accuracy of the theory is the predicted value against the measured value.

 

In psychology, the measurement may be a behavior such as lying to avoid consequence. If the theory suggests that the behavior ought to be quite high for particular reasons, an experiment would be conducted to measure the actual number of liars against the predicted amount in conjunction with the tenants of the theory. Though the measurements may confirm the prediction, further examination and testing are often needed to rule out alternative theories which would indicate the result.

 

In terms of comparing theories, it is possible to say one theory is less accurate that another when the measurements of one correlates more with empirical data than another. Newtonian physics is less accurate than special and general relativity because it matches the data less than the competing theory.

 

This does not imply that Newtonian physics is false, rather that the predictions it makes are not very good in particular circumstances, such as close to light speeds.

 

Theories are often models constructed from empirical data. The model is tested by comparing new measurements from unknown untested conditions, against the prediction the current model generates. If we construct a model of magnetic field strength based on testing of inductors with different sizes with the same voltage and current, we predict the magnetic strength based by changing a variable, say voltage, while keeping everything else constant, and compare the prediction of the model to the measured strength.

[NigelW]

I am in full agreement with you, Pepin.

 

I did a quick google search. I think the definition I was looking for was Meta-analysis, anyways...

 

If I compare a theory that I make today against one I made two weeks ago and recorded, then I can know for certain whether or not I was more or less accurate.

 

If I say I am certain that my theory is 100% accurate we can compare that to other past theories. Then the 100% goes out the window.

[dsayers]

Theories cannot be accurate compared to empirical reality and definitions/theories are not empirical reality. Accuracy can only be arrived at through comparison against other theories.

 

Theories can be more/less accurate relative to other theories. However, any given theory's absolute accuracy is derived from the real world; the consistency of which is the basis for logic.

[Pepin]

If I say I am certain that my theory is 100% accurate we can compare that to other past theories. Then the 100% goes out the window.

 

If a theory is fully correlated with current empirical data, then such a theory can be considered 100% accurate within a given margin of error. As an example, if we derive a mathematical formula from empirical data that describes the motion of springs, and this formula describes all of the currently available data, then the model can be considered 100% accurate.

 

In such an instance, another theory cannot be more accurate as there is already a 1:1 correlation of theory and the data. To think of this mathematically, if we have a line plot in which every point can be described by the function f(x) = x, then no other function will describe the line plot more accurately as it is already 100% accurate.

 

In terms of empirical data, a function which describes data will be subject to a margin of error, which means that any prediction a theory makes which is currently 100% accurate will be within the margin or error. Whether the theory holds up a few magnitudes within the margin of error is currently unknown.

 

In the case of the spring, if physicists believe there is more to the story, they may hypothesize and come up with new formula is created which may intend to describe the motion of a spring more accurately than the previous theory. The emphasis is on the "may", as the hypothesis is predictive.

 

In order for this new model to be validated, further empirical testing must be done, which might simply involve increasing the precision of the measurements. If the previous prevailing theory describes the new data, then this is further evidence to support that theory. If the theory diverges from the data, then the theory can be considered accurate on scales above this, but not at these scale. If the hypothesis matches the results, then it can be considered to be more accurate than the previous theory as it not only describes all of the previous data, but all of the current empirical data. If the hypothesis does not match, then physicists will model the results and attempt to construct a new formula and hypothesis.

 

I hope this helps provide a better way of looking at the accuracy of theories. Theories become more or less accurate when compared to new empirical data. A theory cannot become more accurate on its own, or in comparison to another, without empirical testing. And just to provide a caveat, science in the real world often is very messy and does not follow the above form.

[Psychophant]

The degree of inaccuracy of a theory is determined by the lack of knowledge - variable, flaw in logic.

[NigelW]

Theories can be more/less accurate relative to other theories. However, any given theory's absolute accuracy is derived from the real world; the consistency of which is the basis for logic.

 

If a theory, whose definition implies empirical testing, is valid then by definition is must be accurate. But accurate compared to what? Another theory. Absolute accuracy does not exist. I think you are conflating validity and accuracy because the real world is the basis for logic, you are correct, but logic is the basis for validity.

 

I hope this helps provide a better way of looking at the accuracy of theories. Theories become more or less accurate when compared to new empirical data. A theory cannot become more accurate on its own, or in comparison to another, without empirical testing. And just to provide a caveat, science in the real world often is very messy and does not follow the above form.

 

The definition of a theory is that is has already been tested. So when I say that accuracy can only be achieved through comparing theories that implies empirical testing. If a theory is a car, a Lamborghini is faster than a Kia Rio.

 

The definition of a hypothesis is a proposed explanation for a phenomenon. A working hypothesis is a provisionally accepted hypothesis proposed for further research.

 

 

The degree of inaccuracy of a theory is determined by the lack of knowledge - variable, flaw in logic.

 

I am talking about accuracy, not validity. If there is a flaw in logic, then a theory is no longer valid and can not be accurate.

[WasatchMan]

I would recommend watching this video because I find it relevant to this discussion.  A lot of people are confused by what science is and I think this video might help with developing a better perspective on it. 

 

The accuracy, and more precisely "precision", of a theory is determined empirically by testing the limits of it and finding where the theory deviates from reality.

 

 

[dsayers]

If a theory, whose definition implies empirical testing, is valid then by definition is must be accurate. But accurate compared to what? Another theory. Absolute accuracy does not exist.

 

Compared to the real world. I had already stated this, so it is unclear as to why you'd anticipate a different answer without any attempt to refute/correct. Theories are attempts to accurately describe the real world, so it is axiomatic that the real world is the gauge by which to test their accuracy.

 

Also, you just said, "It is absolutely accurate that absolute accuracy does not exist." This is a performative contradiction; the very act of making the claim is demonstration that you accept the claim to be false.

[NigelW]

Also, you just said, "It is absolutely accurate that absolute accuracy does not exist." This is a performative contradiction; the very act of making the claim is demonstration that you accept the claim to be false.

 

 

Accuracy is not a thing that exists, it is a comparison. I can be absolutely accurate in saying that you are incorrect, though. You are saying that I am certain and that I am contradicting myself, but you did not ask what I was comparing. If I say that objective truth does not exist, I would be correct. If I said that objective accuracy does not exist, I would be correct. This is because they are concepts.

 

I made the argument that you were conflating accuracy with validity, which you have not refuted. I am not comparing a theory against empirical reality because that is testing the validity, not the accuracy.

 

I apologize if I may be coming across as obtuse, but it is a valid theory to say that valid theories are measured against other valid theories to measure accuracy because the only way you know that your inaccurate is by comparing apples to apples.

[NigelW]

Theories can only be proven false based on other empirical, verifiable observations of reality, not other theories. Theories cant be proven to be "true", only "true enough" (based on certain data).Certainly theories/philosophy could give a shit about our preferences for jet engines or internal combustion engines, which is a matter of aesthetics/taste.

 

I am not talking about true or false.

Theories can be more/less accurate relative to other theories. However, any given theory's absolute accuracy is derived from the real world; the consistency of which is the basis for logic.

 

What in reality can we compare theories to in order to establish accuracy?

[dsayers]

I accept your correction that concepts do not exist. You initial claim of "absolute accuracy does not exist" was put forth as if it refuted my position. That is to say that contextually, I interpreted that claim as "absolute accuracy is not valid," as claiming that absolute accuracy does not exist would have no bearing on the discussion.

 

Theories are attempts to describe the real world. Therefore their accuracy must be compared to the real world. For example, let us take your theory that absolute accuracy does not exist. We understand that this theory is absolutely accurate because to exist is to be comprised of matter and energy. This is something that cannot be expressed with a grade, hence the ABSOLUTE accuracy of that theory.

[WasatchMan]

If a theory, whose definition implies empirical testing, is valid then by definition is must be accurate. But accurate compared to what? Another theory. Absolute accuracy does not exist. I think you are conflating validity and accuracy because the real world is the basis for logic, you are correct, but logic is the basis for validity.

 

What in reality can we compare theories to in order to establish accuracy?

 

I am not sure I follow this.  The accuracy of a theory is determined by its predictive capability. Its predictive capability is determined by setting up experiments specifically meant to probe how well the theory is able to predict the outcome of the experiment.  Theories are only compared to one another based on which one has a better predictive capability, and is therefore more accurate.  In other words, the answer to the question "accurate compared to what?" is "reality", where reality is the standard of "absolute accuracy".

[NigelW]

Theories are attempts to describe the real world. Therefore their accuracy must be compared to the real world. For example, let us take your theory that absolute accuracy does not exist. We understand that this theory is absolutely accurate because to exist is to be comprised of matter and energy. This is something that cannot be expressed with a grade, hence the ABSOLUTE accuracy of that theory.

 

I've defined validity as conformity to the rules of logic. I've also defined logic as being derived from reality. Lastly I have defined logic as being the basis for Validity. If you test the theory and it is proven, it is valid. It is also accurate compared to a theory that has failed. There is a more accurate way of measuring the accuracy of a theory. You've agreed that it is possible to compare theories against one another. I am arguing that theories, expressed through language, can be compared and that it is more accurate than comparing theories to the real world. Which would be consistent with absolute accuracy, which is absolute accuracy in knowing a theory to be less accurate.

 

 

I am not sure I follow this.  The accuracy of a theory is determined by its predictive capability. Its predictive capability is determined by setting up experiments specifically meant to probe how well the theory is able to predict the outcome of the experiment.  Theories are only compared to one another based on which one has a better predictive capability, and is therefore more accurate.  In other words, the answer to the question "accurate compared to what?" is "reality", where reality is the standard of "absolute accuracy".

 

You are making a comparison to something in reality. I am saying it is more accurate to say that we compare theories against other theories. If you disagree with something else in particular I would appreciate it being pointed out.

[WasatchMan]

You are making a comparison to something in reality. I am saying it is more accurate to say that we compare theories against other theories. If you disagree with something else in particular I would appreciate it being pointed out.

 

How do you compare one theory against another? Reality.  The whole comparison is based on the ability of one theories predictive capability vs. anothers. To say one theory is more accurate than another is exactly like saying one theory better predicts phenomena in reality than another.  There is no other standard if you believe in objective reality.

 

[edit]: I think an example may be needed here.  The accuracy of Newtonian gravity was not determined by comparing it to Relativity, experimental evidence discovered the accuracy of Newtonian gravity to be limited well before Relativity was ever imagined.  Relativity was determined to be more accurate because it more closely matches the results of experimental evidence. Furthermore, the accuracy of Relativity is known to not be absolute through experiment, however there is not as of yet a more accurate theory of gravity than Relativity to compare it to. Reality is the only external measure to gauge accuracy. [/edit]

[dsayers]

I am arguing that theories, expressed through language, can be compared and that it is more accurate than comparing theories to the real world.

 

And I am refuting this claim by pointing out that the accuracy of a theory can only be derived from its comparison to the real world. Meaning that in order to compare one theory to another, you must first compare each one to the real world. 

 

Don't get me wrong. I'm enjoying the mental exercise. For example, above I claimed that theories can be relatively accurate compared to one another but absolutely accurate compared to the real world. However, this is a contradiction unless a theory could also be relatively accurate compared to reality. I'm still trying to think of whether or not I accept the potential that a theory could be accurate on a gradient compared to the real world.

[NigelW]

And I am refuting this claim by pointing out that the accuracy of a theory can only be derived from its comparison to the real world. Meaning that in order to compare one theory to another, you must first compare each one to the real world. 

 

Defining validity and accuracy would be helpful. I don't accept this as a refutation.

 

 

Don't get me wrong. I'm enjoying the mental exercise. For example, above I claimed that theories can be relatively accurate compared to one another but absolutely accurate compared to the real world. However, this is a contradiction unless a theory could also be relatively accurate compared to reality. I'm still trying to think of whether or not I accept the potential that a theory could be accurate on a gradient compared to the real world.

 

I don't follow you. Definitions, please.

[NigelW]

My theory is:

 

Theory Vs. Reality = Validity

 

Theory Vs. Theory = Accuracy

 

Both require validity by definition and validity requires empirical testing.

[WasatchMan]

My theory is:

 

Theory Vs. Reality = Validity

 

Theory Vs. Theory = Accuracy

 

Both require validity by definition and validity requires empirical testing.

 

This is not true.  Any intro to physics class will use an arrow hitting a bulls eye as a way to explain accuracy.  Accuracy is how close the arrow is to hitting the bulls eye, not how close you are to the bulls eye compared to someone else.  The shot of the arrow analogizes the prediction of the theory, the bulls eye represents actual experimental results.

[NigelW]

This is not true.

 

The statements to follow or mine?

[WasatchMan]

The statements to follow or mine?

 

Search "accuracy" in google, then click images.  Tell me what you see...

[NigelW]

So, what do we call hitting the target at all? Oh wait... valid?

 

And what do we call an arrow closer to the bulls-eye... more accurate?

 

They are both valid, someone is just a better shot... which is what I am saying.