Can you make assumptions in science?

Post Reply New Topic

I asked this in another thread, but as it was off-topic, I decided to make a new thread.

In science, is it alright to make assumptions? Here's the what-if:

Let's say that John's scientific experience is substantial. Can John make assumptions based on that experience, in the process of his thesis, and still call his conclusions valid?

I guess you can make assumptions based upon first principles and previous research but they would still need to be verified in some way (citations etc) Remember the famous quote in a letter by Isaac Newton "If I have seen further it is by standing on the shoulders of giants".

Conjecture is a good starting point for a hypothesis but it must be backed up by evidence. and if the evidence refutes it then the conjecture is wrong. Remember the correct way to view scientific knowledge is that our current understanding of natural laws must be seen as the best approximation we have rather than absolute facts.

I think that depends on what the assumption is.

All phenomena have natural causes

Correlation implies causation

Rates of a phenomena remain steady so future data can be accurately extrapolated by looking at current rates

Now let's look at somebody with substantial scientific experience making all three assumptions in the process of her thesis (which, oddly, she does present as a thesis rather than a hypothesis). Are her conclusions valid?
http://www.medicaldaily.com/autism-rates-increase-2025-glyphosate-herbicide-may-be-responsible-future-half-316388

Dr. Seneff has substantial scientific experience. But here's the catch: it's experience in computer science and artificial intelligence. This does not cross over to biology. She starts with a foundational assumption of science: all phenomena have natural causes. I think this is a safe assumption and it has yet to be disproven.

Then she makes a thesis (there is a toxic substance that is causing autism rates to increase) and proceeds as though this is proven and she is merely illustrating that proof. Not such a good idea. She then goes on to make two assumptions that have frequently been disproven, but maybe not in the field of computers and artificial intelligence.

She assumes that correlation implies causation- that a correlation (over time) of an increase in glycophospate use with an increase of autism diagnoses implies that the glycophosphate use caused the autism.

She also assumes that current rates of increasing autism diagnosis will continue exactly as currently observed and so by 2025 half the children born will be autistic. I am wondering if her substantial scientific experience in computers is what caused her to make this assumption. Nobody in the field of medicine makes such assumptions but in computer science, such an assumption even has a name: Moore's Law
http://en.wikipedia.org/wiki/Moore%27s_law


I have no doubt she is familiar with Moore's Law. I wonder if she's so familiar with it that she (subconsciously?) applied it to medicine and assumed that autism increases would continue at the current rate, as computer hardware changes do. This is one of the problems of assumptions outside your own field. It's safer to stick to assumptions that apply to all of science, such as "phenomena have natural causes" .

Sure, you can make assumptions in science; but without valid empirical evidence to back up those assumptions, you may as well be selling snake-oil.

I suppose this is aimed at me. I couldn't find a question mark in your other post so I didn't know what "what if" you might be on about.

I don't have time to get into it just now, and I'd be more confident of a reasonable discussion if the terms were clearly defined. It's an Alinskyist trick to provoke an argument then render it absurd by changing the premises as the "argument" progresses.

If you define "assumptions" as a possible explanation for observations of phenomena to be tested by experiment to develop a thesis it would be fair enough... but that's not what "assumptions" means.

Be back later.

All science is based on assumptions as we lack absolute knowledge. Thus, Science is about assuming as little as possible. If there are two models that results in the same answer, then the one would take the one with the lower amount of assumptions.

Occam's Razor? A fine research method, indeed!

Among competing hypotheses, the one with the fewest assumptions should be selected. Other, more complicated solutions may ultimately prove correct, but -- in the absence of certainty -- the fewer assumptions that are made, the better.

It also helps to base all of the assumptions on provable principles -- no miracles or supernatural interventions, no unknown forces, no magic and no psychic abilities.

An assumption must also occur only at the beginning of a scientific quest, and not at the end.

Example: "If we assume that a function of X can yield all instances of Y, then all that remains is to determine what the most appropriate function of X should be."

It also helps to be able to state clearly what assumptions you might be making. (Even if they are "standard" in the field.)

Nah, because until he can "show his work" so that others could test his predictions, this would border on "hunch" or "trust me" and would definitely be in the realm of "artist".

I agree, and.. I agree.
In essence, I guess an assumption is a stance taken on which you base the next step, or response, or the path forwards... something like that.

And so far, the response from others seems to all agree that the results of an assumption are only valid once the assumption has been tested. Does experience not count, or does that still make it subjective?

Ockham's Razor is a fine tool for helping to sort out what's a real proposition from a deceptive sales-pitch hidden in a pile of blather or waffle. The Materialism (empiricism) you apparently demand does not, cannot exist without a metaphysical substrate of, at least, life and intellect and will to desire and perceive the metaphysical stuff called "truth" or "reality".
As such, assumptions have no place in science.
[/quote]

Experience, by it's very definition, is almost entirely subjective.

Experience, by it's very definition, is almost entirely subjective.

Wooee! I sure can't work this site. One minute I can't post, then it posts double, then another, and it won't let me delete any because "the post you're trying to delete doesn't exist.

It's a good tool, yes, but not an infallible tool.


And yet we all make them.. and rely on them.. and believe in our assumptions. Do we not?


And even if we acknowledge our subjectivity, we remain convinced we are correct - and too often sans evidence.

Make assumptions, but try to check them.

The classic one is whether your data is normally distributed.