PDA

View Full Version : How can we exempt bias from Astronomy?



klam22
2020-Apr-24, 10:38 AM
In Astronomy, as well as other sciences, bias can play a big part in coming to a scientific conclusion. For example, if a researcher wanted to find something, itʻs most likely theyʻre going to find it because bias is playing a big role in driving their conclusion in a certain direction. A good example is the scientific papers on the Messier 73 objects. Two papers were published months apart on the M73 objects, one claiming these stars had no relationship with each other and the other claiming it was an open cluster. The two papers used the exact same data and came to two complete opposite conclusions. This is a perfect example of confirmation bias at play where bias can drive the way one interprets data.
How can the Astronomy community do their best in getting rid of this bias when researching?

PetersCreek
2020-Apr-24, 02:23 PM
Welcome to the CosmoQuest forums, klam22. How did you reach the conclusion that confirmation bias was responisble for their disagreement? Can you cite the papers?

Roger E. Moore
2020-Apr-24, 04:03 PM
Agree, need to see the papers. Can you post links, please? Also clarify the M73 issue.

klam22
2020-Apr-27, 09:23 AM
https://arxiv.org/abs/astro-ph/0003372
https://arxiv.org/abs/astro-ph/0001238

Or you guys can check out this Youtube video: https://www.youtube.com/watch?v=R4g2lfsCjJM&feature=youtu.be

Thanks for your responses!

PetersCreek
2020-Apr-27, 04:21 PM
https://arxiv.org/abs/astro-ph/0003372
https://arxiv.org/abs/astro-ph/0001238

Or you guys can check out this Youtube video: https://www.youtube.com/watch?v=R4g2lfsCjJM&feature=youtu.be

Thanks for your responses!


Thanks for the citations. Back to the first of my questions, though: how did you reach the conclusion that confirmation bias was responisble for their disagreement?

Strange
2020-Apr-28, 07:56 AM
In Astronomy, as well as other sciences, bias can play a big part in coming to a scientific conclusion. For example, if a researcher wanted to find something, itʻs most likely theyʻre going to find it because bias is playing a big role in driving their conclusion in a certain direction.

While I agree that individual researcher's bias, and the model they prefer, may affect how they interpret results. Scientists are human after all. So I don't think you can eliminate bias.

(I don't agree with the suggestion that they can find whatever they want; if it isn't there, they can't find it.)

Interpreting data in two (or more) different ways is very common in science. It is usually resolved by gathering more evidence that resolves the question. Or increases in our understanding that allows one interpretation to be rules out.

There was a good article on exactly this by Ethan Siegel today: https://www.forbes.com/sites/startswithabang/2020/04/28/only-evidence-not-compelling-arguments-can-settle-scientific-debates/#5e2ddc6b194c


So, you've arrived at a crossroads: you think the world works in a certain way, and someone else disagrees with you and thinks the world works in a different way. You've both got your reasons as to why you're convinced that your way is right and the other person is wrong, but for some reason, you cannot come to an agreement with one another.
...
The only solution that's scientifically valid is to obtain the critical evidence: a lesson we all need to be reminded of.

He uses the example of the debate between Shapley and Curtis about whether nebulae were photo-stars or separate galaxies ("island universes"). Some of the interpretations were based on inaccurate data, an incomplete understanding of the physics around stars, etc. Once more evidence and better theories developed then the debate could be resolved. A similar thing happened with cosmology: when I was young, the steady state and big bang models could both be argued for based on the evidence available at the time.

Copernicus
2020-Sep-03, 03:58 AM
In Astronomy, as well as other sciences, bias can play a big part in coming to a scientific conclusion. For example, if a researcher wanted to find something, itʻs most likely theyʻre going to find it because bias is playing a big role in driving their conclusion in a certain direction. A good example is the scientific papers on the Messier 73 objects. Two papers were published months apart on the M73 objects, one claiming these stars had no relationship with each other and the other claiming it was an open cluster. The two papers used the exact same data and came to two complete opposite conclusions. This is a perfect example of confirmation bias at play where bias can drive the way one interprets data.
How can the Astronomy community do their best in getting rid of this bias when researching?

Only evolution is unbiased, but frequently wrong for many iterations until it is right.

Kay Burton
2020-Sep-18, 08:06 AM
In Astronomy, as well as other sciences, bias can play a big part in coming to a scientific conclusion. For example, if a researcher wanted to find something, itʻs most likely theyʻre going to find it because bias is playing a big role in driving their conclusion in a certain direction. A good example is the scientific papers on the Messier 73 objects. Two papers were published months apart on the M73 objects, one claiming these stars had no relationship with each other and the other claiming it was an open cluster. The two papers used the exact same data and came to two complete opposite conclusions. This is a perfect example of confirmation bias at play where bias can drive the way one interprets data.
How can the Astronomy community do their best in getting rid of this bias when researching?

As for me, astronomy is a fairly young science. Yes, studies of the starry sky began hundreds of years ago, but real technical possibilities for research have appeared quite recently. It seems to me that this is what causes different research results. Astronomers only accumulate knowledge. Therefore, it seems to me, we need to wait a few more years until the laws of outer space are studied in more detail.

agassiz
2020-Oct-17, 02:39 AM
Only evolution is unbiased, but frequently wrong for many iterations until it is right.

Evolution isn't a science in the way Astronomy is, and anyway, evolution is at least biased towards survival.


If the conclusion is that astronomy is at least biased somewhat, what are the particular biases that astronomy is susceptible to?

21st Century Schizoid Man
2020-Oct-17, 04:36 AM
You can't.

If someone has developed some theory, become famous for it, built a career on it, etc., how are they likely to respond to evidence that emerges, suggesting the theory may be wrong?

The only think we can do is try to deal with bias when it comes up, and having a diversity of researchers, not all working together and not all funded by the same entity, is probably the best we're going to do. Individuals will still exhibit bias, but they will remain, individuals.

Copernicus
2020-Oct-31, 01:06 AM
Evolution isn't a science in the way Astronomy is, and anyway, evolution is at least biased towards survival.


If the conclusion is that astronomy is at least biased somewhat, what are the particular biases that astronomy is susceptible to?

I'm saying humans are biased and fallible, but over time, these bias disappear one death and one discovery at a time. Then science evolves to be better.

Noclevername
2020-Dec-06, 08:47 AM
Only evolution is unbiased, but frequently wrong for many iterations until it is right.

Sexual selection could be called a bias of evolution. Changes unrelated to the practical art of survival become entrenched in the reproductive process.

Noclevername
2020-Dec-06, 09:04 AM
The OP question can be paraphrased as, how do we eliminate cognitive bias from the scientific process?

We can't. We are human beings, not Vulcans. Total objectivity is an ideal to be strived for, not a fully reachable goal.

We can try to recognize and minimize or compensate for the subjective factors that color our models and results, but we will always have perception filters built in because our brains work that way. As many others have stated, the best way to counter bias is more and more sources of consistently accurate, repeatable data from multiple sources.

That, and the attitude that objectivity and dispassionate analysis of factual evidence are innately good things (a view that is not universal).

21st Century Schizoid Man
2020-Dec-06, 11:52 AM
We are human beings, not Vulcans.

Speak for yourself!

Noclevername
2020-Dec-06, 01:35 PM
Speak for yourself!

Live long and prosper.

Jens
2020-Dec-07, 03:59 AM
I'm really repeating what others have said, but really, you can't get rid of bias.

I would also mentions that the scientific method was developed partly as a way to try to overcome confirmation basis. And so yes, experiments, and the existence of different researchers, helps to overcome it.

Swift
2020-Dec-07, 04:58 PM
https://arxiv.org/abs/astro-ph/0003372
https://arxiv.org/abs/astro-ph/0001238

Or you guys can check out this Youtube video: https://www.youtube.com/watch?v=R4g2lfsCjJM&feature=youtu.be

Thanks for your responses!
Thanks for the citations. Back to the first of my questions, though: how did you reach the conclusion that confirmation bias was responisble for their disagreement?
I go back to Peterscreek's question to the OP, how do we know that the differences in conclusions in these two papers are from confirmation bias? I've only looked at the abstracts, but I'm not convinced the two groups are looking at the exact same data set, nor have the analyzed their data in the same way. The differences may be experimental methods, not confirmation bias.

Is concluding the differences are from confirmation bias caused by the confirmation bias of those coming to that conclusion? ;)

Copernicus
2020-Dec-23, 07:42 AM
And we haven't even begun to separate out, out of the millions or billions of scientific publications, are the ones that we pay attention to. Much less know, which ones are really accurate.