Basic Scientific Method Question

An inference is something which is assumed based on apparent evidence. For example, if i turn the corner, and find you standing above the bleeding body of a man, and you have a smoking gun in your hand, i might reasonably infer that you have just shot the man. Of course, someone else could have shot him, and thrown down the smoking gun, which you picked up just before i came around the corner--so inferences are only tentative and preliminary explanations of what one observes. Other explanations might just as plausible.

An hypothesis is a systematic statement which seeks to explain a circumstance, and which covers all of the observable data. So, an hypothesis would state that the man bleeding on the pavement was shot (not proven), that the smoking gun is the gun with which he was shot (not proven), and that absent other evidence, the most plausible inference is that you shot him, but that this is not proven.

A prediction is a statement which can be derived from a theory (not necessarily from an inference or an hypothesis) which will describe what data will be seen in specific circumstances. I can't really apply it to the "smoking gun" scenario i have provided, other that to make a series of statements about how people perceive things, and then predict that anyone coming around the corner to find a man bleeding on the sidewalk and you standing there with a smoking gun in your hand is going to assume that you shot him.

The ability to predict is essential to a theory. In science, theory is not a vague idea or a notion of cause and effect, it is as close to a statement of truth as science ever comes. A theory will predict certain observable cause and effect situations, and if it fails to accurately predict, the theory must either be modified to account for the data, or be abandoned until a theory can be propounded which accounts for the data. All theories have been derived from hypotheses. The hypothesis becomes a theory when it reliably predicts new data, and when it cannot be falsified (proven not to apply in all cases).

I'm not a scientist, nor claim to be one. Others may appear who can explain this better for you.

Don't see how they could; that's a really clear explanation, Set.

Thanks, Boss. I wasn't sure as i wrote it that it were.

Theres not much difference. The first two fit in a hierarchy
An inference is an assertion based upon a small data set

A hypothesis is an assertion in which (so far) all the observable data supports

A Theory is an explanation for a phenomenon or set of related phenomena in which all the observable data supports and no data refutes.

A prediction can be made using the systematics and facts of either a hypothesis or a theory.

We can revise a hypothesis or a theory. We can discard a hypothesis entirely . A theory is more robust in its support basis. Remember, NO data refutes a theory and a hypothesis is just less firmly grounded than a theory


Stuff like Atomic Theory has been tested and used to make predictions and design maxchines . STring "Theory" on the other hand, may just as well be called "String Hypothesis or a string inference" because we dont yet know the entire basis of facts and phenomena this even governs, and the actual supportive data is small.

WOW! We are so lucky to have several giant brains on a2k.

I seldom reply, but I love to read your posts.

One can also understand a scientific hypothesis as a tentative explanation based on observation that can be confirmed (or not) by additional data. As more and more data are explained by the hypothesis, or as the hypothesis is modified to account for new data (in some cases, can "predict" an event), its confirmation becomes stronger.