Friday, May 18, 2012
Friday, April 27, 2012
Sane facts and plausible results
In the previous post I mentioned "sane facts" and "plausible results"; let me discuss that for a moment. Sane facts means actual facts, nothing special, just actual facts, not made up ones. Jus to give you an example: "when I drop a rock from my hand it will fall towards the ground" see, not too hard is it. "when I drop a rock from my hand it will fall towards the ground with a constant velocity" - False, look carefully and you'll notice it picks up speed; at some point it will hit terminal velocity, but i am nowhere near as tall as I'd have to be for that to have an effect. This you can actually test yourself, drop a rock from a height, and take the time, drop it from twice the height and take the time. If you get twice the time, the rock falls with a constant velocity, but if you get anything lower than twice the time, the rock accelerates. I have now provided you with facts, and I sowed you how you can test it.
Plausible results are a bit trickier to explain but bear with me on this. If you do an experiment, the conclusion you get from the experiment needs to be related to the actual experiment. Let's take a simple experiment, banging two rocks together. Now just to add some scienceyness to it, we are going to measure the temperature increase the banging had on the rocks. This experiment would not reveal the curvature of the earth, so if you conclude that the earth is round(ish) and has a radius of 6367.5 kilometers, I am not going to tell you your conclusion is wrong. I am however going to tell you that the experiment you did never, ever could tell you that. Your result is therefore not a plausible result for the experiment, it doesn't make "that exact result, forever invalid" it just means that that experiment would never reveal that conclusion.
Plausible results are a bit trickier to explain but bear with me on this. If you do an experiment, the conclusion you get from the experiment needs to be related to the actual experiment. Let's take a simple experiment, banging two rocks together. Now just to add some scienceyness to it, we are going to measure the temperature increase the banging had on the rocks. This experiment would not reveal the curvature of the earth, so if you conclude that the earth is round(ish) and has a radius of 6367.5 kilometers, I am not going to tell you your conclusion is wrong. I am however going to tell you that the experiment you did never, ever could tell you that. Your result is therefore not a plausible result for the experiment, it doesn't make "that exact result, forever invalid" it just means that that experiment would never reveal that conclusion.
Why Peer Review?
Why is peer reviewing so important to be able to actually prove that your argument is indeed scientific. Anyone can claim something and then say it's scientific, it doesn't make it scientific.
Peer reviewing simply means that; at least one other scientist, not affiliated with you, with knowledge of the matter in question, has looked at the paper. After looking at your paper he/she has come to the conclusion that your facts seems real and plausible, and that your conclusions add up and actually makes sense considering the facts presented. There are a few key words in that are really important here, "other scientist", "not affiliated", "facts" and "add up". If someone does an inadvertent mistake, they won't usually see it themselves. The same foes for someone affiliated with you, because they may actually let you get away with a less thorough check; or in cases of pseudosciences like "The dangers of Dihydrogen monoxide", the scientist affiliated with you may be a practitioner of the same pseudoscience and willing to present the same false facts.
I am not going to be very demanding, present one paper that has gone trough a scrutiny process by someone with knowledge of the subject in question, and in their official capacity approved it as "not pulled straight out of the authors butt". However I will demand a three party relationship, author, publisher and reviewer need to be from separate institutions, and not directly affiliated; in very short terms, it means that none of the three parties can have an "office right down the hall" from any of the other parties. And for the record, home offices are not a way to circumvent this, use common sense.
Peer reviewing simply means that; at least one other scientist, not affiliated with you, with knowledge of the matter in question, has looked at the paper. After looking at your paper he/she has come to the conclusion that your facts seems real and plausible, and that your conclusions add up and actually makes sense considering the facts presented. There are a few key words in that are really important here, "other scientist", "not affiliated", "facts" and "add up". If someone does an inadvertent mistake, they won't usually see it themselves. The same foes for someone affiliated with you, because they may actually let you get away with a less thorough check; or in cases of pseudosciences like "The dangers of Dihydrogen monoxide", the scientist affiliated with you may be a practitioner of the same pseudoscience and willing to present the same false facts.
I am not going to be very demanding, present one paper that has gone trough a scrutiny process by someone with knowledge of the subject in question, and in their official capacity approved it as "not pulled straight out of the authors butt". However I will demand a three party relationship, author, publisher and reviewer need to be from separate institutions, and not directly affiliated; in very short terms, it means that none of the three parties can have an "office right down the hall" from any of the other parties. And for the record, home offices are not a way to circumvent this, use common sense.
Subscribe to:
Posts (Atom)