As I understood it, the study was suggesting that certain demographics are not well represented in the field; that there is data overlooked. I did not take that to mean it was validating purely subjective methods.
Unfortunately, the study now seems to be behind a paywall, so I cannot check. If there's a relevant area you could quote to show that's the case, I'd appreciate it.
Crucially for feminist glaciology, feminist political ecology argues for the integration of alternative ways of knowing, beyond diverse women?s knowledges to include ?more broadly ? the unsettling of Eurocentric knowledges, the questioning of dominant assumptions, and the diversification of modes and methods of knowledge production through the incorporation of everyday lived experiences, storytelling, narrative, and visual methods.
Alternative ways of knowing is referencing following non-objective methods. Questioning dominant assumptions would be fine (and indeed a cornerstone of science) were the methods of supplanting these assumptions objective.
St. Germain, LeGuin, Khan, and many others ? from Roni Horn (2009) to Pauline Couture (2005) ? approach glaciers from distant and varied disciplinary and artistic spaces compared with glaciologists or even anthropologists studying human-glacier interactions. Such alternative representations of glaciers are rarely incorporated or even acknowledged within greater discourses of glaciology and global environmental change research. Yet their voices should not simply be disregarded, overshadowed by Western science, or, worse, relegated from policy contexts where, in fact, the human experience with ice matters greatly. These alternative representations from the visual and literary arts do more than simply offer cross-disciplinary perspectives on the cryosphere. Instead, they reveal entirely different approaches, interactions, relationships, perceptions, values, emotions, knowledges, and ways of knowing and interacting with dynamic environments. They decenter the natural sciences, disrupt masculinity, deconstruct embedded power structures, depart fromhomogenous and masculinist narratives about glaciers, and empower and incorporate different ways of seeing, interacting, and representing glaciers ? all key goals of feminist glaciology.
Sorry for the long quote but I think the context is important here.
This is the best part. Can you decode the message?
The author is suggesting that the facts derived from the natural sciences should take a backseat to the "knowledges" and "different ways of knowing" from arts and literature because only the latter follow a feminist narrative.
Analysis [http://www.livescience.com/20896-science-scientific-method.html] is a part of the process. The source of the data is something which must be analysed.
Plainly, you cannot simply say "X is not a method. Hence, it is not considered within the methodology context". That's simplistic sophistry. A dozen things which are not themselves methods need nonetheless be considered during prediction, analysis, reproduction, etc.
Yes, and analysis requires setting up in the methodology part because it requires a method. You need to figure out how to analyse your data, what method to use. Prediction requires a method (prediction useing method A need not give the same result as prediction using method B), and to be honest I've never heard of anyone considering reproduction during the methodology part of the process - it follows naturally from documenting your entire process and providing all the conditions under which you performed your experiments. But if that's in there I will accept it as truth.
Analysis is right there, as an essential step in the process. So is conclusion. Both frequently include interpretation of data, rather than objective reportage.
Nobody is talking about random data. Only interpretation. Professional interpretation; rigorous interpretation; nonetheless, frequently non-objective.
Yes, and analysis must be objective. You must provide the grounds for your results and the process through which you arrived there. If you try to publish your manuscript without showing how you got the numbers (which is around 66% of the entire manuscript) chances are your peer review will just be three lines of "No, just no". Numbers all the way, and a well-defined process. That sounds very objective to me. You can of course choose to skip all that, but then it's an issue with the scientist, not the process.
The Conclusion need not be objective since, as you say, it must include some form of interpretation or extrapolation. The Discussion should not be objective, since then it's not discussing and the numbers are already covered in the Results.
Blatantly. You would, however, very frequently see terms like, "this would suggest [...]" or "this indicates [...]".
The examples you use-- "0.7 true", "0.45 representative"-- are (presumably) here to indicate the absurdity of presenting non-mathematical concepts in mathematical terms. But that was rather the same point as I was making when I said you cannot reduce validity to a binary state.
Exactly, but you are missing a component; the premise is the number. "The model had an AUC of 0.88, suggesting that the principle is valid for the selected subpopulation. Expected noise from previous quality audits was 0.09".
Note that there is no level of validity discussed, ever. Let's try this another way; If you can find me a study where a hypothesis' validity or veracity is discussed as a non-binary value I will concede the argument.
I've read a lot of academic crap, and I've seen grant proposals get turned down for a tenth of the money, that sounded a tenth as lunatic. By analogy, this is like saying, "We want to study rocketry from a homeopathic perspective. Traditionally it has been believed that only reaction mass can propel a rocket, but intuitively some people have long held that a little sugar water and meditation is just as effective. Historically these claims have been dismissed by the rocketry elite, and we see no reason (as laypeople) why that should be."
Crucially for feminist glaciology, feminist political ecology argues for the integration of alternative ways of knowing, beyond diverse women?s knowledges to include ?more broadly ? the unsettling of Eurocentric knowledges, the questioning of dominant assumptions, and the diversification of modes and methods of knowledge production through the incorporation of everyday lived experiences, storytelling, narrative, and visual methods.
Alternative ways of knowing is referencing following non-objective methods. Questioning dominant assumptions would be fine (and indeed a cornerstone of science) were the methods of supplanting these assumptions objective.
St. Germain, LeGuin, Khan, and many others ? from Roni Horn (2009) to Pauline Couture (2005) ? approach glaciers from distant and varied disciplinary and artistic spaces compared with glaciologists or even anthropologists studying human-glacier interactions. Such alternative representations of glaciers are rarely incorporated or even acknowledged within greater discourses of glaciology and global environmental change research. Yet their voices should not simply be disregarded, overshadowed by Western science, or, worse, relegated from policy contexts where, in fact, the human experience with ice matters greatly. These alternative representations from the visual and literary arts do more than simply offer cross-disciplinary perspectives on the cryosphere. Instead, they reveal entirely different approaches, interactions, relationships, perceptions, values, emotions, knowledges, and ways of knowing and interacting with dynamic environments. They decenter the natural sciences, disrupt masculinity, deconstruct embedded power structures, depart fromhomogenous and masculinist narratives about glaciers, and empower and incorporate different ways of seeing, interacting, and representing glaciers ? all key goals of feminist glaciology.
Well, I'll be. It does seem ridiculous to incorporate "storytelling" into scientific research. Interesting, perhaps, in anthropology or cultural studies; not so much in glaciology.
Fallow said:
Yes, and analysis must be objective. You must provide the grounds for your results and the process through which you arrived there. If you try to publish your manuscript without showing how you got the numbers (which is around 66% of the entire manuscript) chances are your peer review will just be three lines of "No, just no". Numbers all the way, and a well-defined process. That sounds very objective to me. You can of course choose to skip all that, but then it's an issue with the scientist, not the process.
The Conclusion need not be objective since, as you say, it must include some form of interpretation or extrapolation. The Discussion should not be objective, since then it's not discussing and the numbers are already covered in the Results.
So must analysis. If you're being objective, then you go no further than the data itself; and that is not analysis, by definition.
What you describe above-- providing the grounds for results, the process you used-- that's merely explaining the method. You've got to provide interpretation of the results as well: to argue what they represent, the extent, the validity. All subjective.
Fallow said:
Exactly, but you are missing a component; the premise is the number. "The model had an AUC of 0.88, suggesting that the principle is valid for the selected subpopulation. Expected noise from previous quality audits was 0.09".
Note that there is no level of validity discussed, ever. Let's try this another way; If you can find me a study where a hypothesis' validity or veracity is discussed as a non-binary value I will concede the argument.
The premise is objective; the interpretation, that it's valid for the selected population, is not. It's a binary state, but still, a call was made.
Now, that call can be based on rigorous scientific agreement on what point it is considered valid. There can be an accepted numerical point in the field. But that point is not some inherent characteristic of science or mathematics.
As for validity "expressed as a non-binary value"; obviously, there is no scale, which is (I get the impression) what you're imagining. Whenever a study concludes that its results could indicate X, that's a non-binary statement on the validity of the results. Take a quick look here [http://classic.sciencemag.org/content/308/5721/541.full?sid=c71e714a-f525-4ddd-9a2d-28b8cd9e3d64];
AAAS said:
"However, it appears that in recent times this large mixed population of floating and tidewater glaciers has responded synchronously to a climatic forcing, indicating, at least statistically, that such glaciers should be predictable on decadal time scales. Because we know that the loss of floating ice shelves can cause acceleration of inland glaciers, these observations suggest that the cumulative loss of ice at the fronts of these glaciers may be leading to an increased drainage of the Antarctic Peninsula that is more widespread than previously thought."
Emphasis mine. These are not definitive, committal, binary statements. They refer to likelihoods and possibilities. That's what I was referring to; no imaginary numerical scale of validity somewhere.
Analysis is exactly that. You take your results, you interpret them according to a very specific framework (preferrably one you have designed in the methodology part of your study), and based on the result of your interpretation, you draw logical conclusions. All of this is "no further than the data itself". Extrapolating "what could this imply" is not done in the analysis.
What you describe above-- providing the grounds for results, the process you used-- that's merely explaining the method. You've got to provide interpretation of the results as well: to argue what they represent, the extent, the validity. All subjective.
Arguing what the numbers represent is done in the conclusions, the extent more frequently in the discussion, and the validity is not argued. Perhaps what you mean is to argue for the strengths and weaknesses of the data? That is quite common.
The premise is objective; the interpretation, that it's valid for the selected population, is not. It's a binary state, but still, a call was made.
The premise is objective, the interpretation follows a framework and must also be objective (if it actually is in the end depends on the scientist). I don't understand what you means by "the interpretation, that it's valid for the selected population ... ". Validity and interpretation are different things and validity is never interpreted. Validity doesn't pertain to whether you are right or wrong in your hypothesis, it means only that your process, whatever it may be and whatever it may test, is giving you correct answers. In the other direction, the interpretation is always valid. Whether it gives an answer to the hypothesis is another matter entirely.
Now, that call can be based on rigorous scientific agreement on what point it is considered valid. There can be an accepted numerical point in the field. But that point is not some inherent characteristic of science or mathematics.
I'm not sure you are talking about the scientific (logic) version of valid [http://www.iep.utm.edu/val-snd/] anymore.
As for validity "expressed as a non-binary value"; obviously, there is no scale, which is (I get the impression) what you're imagining. Whenever a study concludes that its results could indicate X, that's a non-binary statement on the validity of the results. Take a quick look here [http://classic.sciencemag.org/content/308/5721/541.full?sid=c71e714a-f525-4ddd-9a2d-28b8cd9e3d64];
No, when using terms like "suggests", "indicates", "implies" etc you are saying that you cannot assert the validity of the argument, and thus make no claim to. You are saying "Hey, this is pretty likely and makes sense but we haven't run the tests and so cannot claim this is the case". It's the difference between "smoking is associated with lung cancer" and "smoking is a cause of lung cancer". And that is a massive difference requiring a lot of evidence (and is the reason it took so long to officially assert the relationship and smack the tobacco companies with a stick). Hell, writing medical articles it's not enough to say "A suggests an increase of B to be causal of C", you need to be even more careful and say "A may potentially suggest an increase of B to be causal of C" because even the thought of a claim should be avoided.
I'm gonna leave the article you paraphrased here because I think looking at it now would give you a different perspective of it.
AAAS said:
"However, it appears that in recent times this large mixed population of floating and tidewater glaciers has responded synchronously to a climatic forcing, indicating, at least statistically, that such glaciers should be predictable on decadal time scales. Because we know that the loss of floating ice shelves can cause acceleration of inland glaciers, these observations suggest that the cumulative loss of ice at the fronts of these glaciers may be leading to an increased drainage of the Antarctic Peninsula that is more widespread than previously thought."
Analysis is exactly that. You take your results, you interpret them according to a very specific framework (preferrably one you have designed in the methodology part of your study), and based on the result of your interpretation, you draw logical conclusions. All of this is "no further than the data itself". Extrapolating "what could this imply" is not done in the analysis.
"Interpret", "draw logical conclusions", none of this is objective. If it goes no further than the data itself, then you're simply presenting your findings, and saying not a thing about them.
Fallow said:
The premise is objective, the interpretation follows a framework and must also be objective (if it actually is in the end depends on the scientist). I don't understand what you means by "the interpretation, that it's valid for the selected population ... ". Validity and interpretation are different things and validity is never interpreted. Validity doesn't pertain to whether you are right or wrong in your hypothesis, it means only that your process, whatever it may be and whatever it may test, is giving you correct answers. In the other direction, the interpretation is always valid. Whether it gives an answer to the hypothesis is another matter entirely.
Validity pertains directly to whether the study can be taken to be accurate. That's it's very essence.
The individual in your example has concluded that it's valid for the selected population. To state that necessitated him making a call. It was not written in the data itself, there was no demonstrable and precise point at which it became valid. He made a call.
Fallow said:
I'm gonna leave the article you paraphrased here because I think looking at it now would give you a different perspective of it.
AAAS said:
"However, it appears that in recent times this large mixed population of floating and tidewater glaciers has responded synchronously to a climatic forcing, indicating, at least statistically, that such glaciers should be predictable on decadal time scales. Because we know that the loss of floating ice shelves can cause acceleration of inland glaciers, these observations suggest that the cumulative loss of ice at the fronts of these glaciers may be leading to an increased drainage of the Antarctic Peninsula that is more widespread than previously thought."
The article does as it always did; presents several possibilities indicated, to some extent, by the data. Which is the very core of my point. You're not denying that it does so, as indicated by how you've left my emphasis standing-- and as such I'm at a loss as to what alternative perspective you expect me to have come to.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
