'But the fact is, that is the body of Lord Chartis!' ...
'My love, you make the classic error. That is not a fact, that is a conclusion drawn from facts.' ...
'You mean, it is not Chartis?' ...
'Oh, no. It is certainly Chartis. I was merely objecting to your choice of words.'
- Steven Brust, Jhegaala
The scientific method consists of several steps from asking the initial question to drawing a final conclusion. Not all of the intermediary steps need rigorously be followed. Without scientific observation, however, there is no scientific method; and without a final conclusion, there is no scientific discovery.
Scientific observation is the collection of data, either directly through the senses or by means of calibrated instruments. Using instruments for measurement reduces the amount of subjectivity inherent to the senses. Reproducibility of results is crucial to the scientific method, so obtaining data that can be checked and confirmed by others is essential. Thus, as Brust so pithily points out, scientific fact is nothing other than objective and verifiable observation.
Scientific observation is used twice during the scientific method: to obtain initial (baseline) knowledge about the item, and then again after the test is conducted, to measure any changes as a result of the test. Although he does not explicitly say it, Brust implies that his speakers have previously seen Lord Chartis while he was alive. This would give them a baseline scientific observation against which to compare the body before them now.
Yet scientific observation is only a collection of facts. To go even one step further, to try in any way to interpret those facts or make an inference from those facts, goes outside the realm of scientific observation. An inference drawn from facts is not itself a fact.
Based on the data obtained through scientific observation, a conclusion can be drawn. This conclusion should be a logical inference based on the observed data and any changes in that data before and after testing. This conclusion is sometimes known as the scientific discovery.
While scientific observation must always be verifiable, a given experiment may reach a different conclusion when conducted at a different time or by different people. What is logical depends entirely on what premises are used. Brust uses at least three unspoken premises in reaching his conclusion: that the speakers' senses are reliable, that a dead body resembles the living man, and that neither the living man nor the dead body are in any way disguised.
Thus scientific discovery always depends both on the accuracy of data that can be obtained and on the validity of the logical premises. In developing his laws of motion and gravitation, Newton took as his premises that space, time, and motion were absolute, unaffected by the presence or absence of objects. Based on our everyday world, there is no reason to think otherwise. Yet when these assumptions were finally called into question and then tested using the scientific method, they turned out to be false: and general relativity was born.