Data are "derived" from observations. Alternatively said, observations become data when we record them. Otherwise, they remain as observations. By example, the only reason we know what Galileo observed during his observations of the moon is because he recorded his observations in pictures, making them data.
The next step in the chain is to "derive" a result. This involves analysis or interpretation.
In summary, the chain is
The one word you seek is result.
When you want to further qualify the level of effort taken in making observations, you can use words such as focused, directed, or even scientific as opposed to random or undirected. The phrase "unscientific observations" is not necessarily a "good" way to express the opposite of scientific. So, data obtained from directed observations have perhaps a higher standard than data obtained from random observations.
Suppose instead that you need a qualifying metric that defines the trust that you have in a result that is obtained by the analysis of data from (scientific) experiments or observations. Consider these criteria:
- verifiable - able to be found by others
- certifiable - attestable
- precise - reproducible
- accurate - indistinguishable from truth
- measurable - subject to quantification through observation
- robust - impervious to unpredictable outcomes
- empirical - derived through observation
- first-principled - derived through reasoning
You may make variations. For example, a verifiable and robust result is one that others can readily find and that has no ambiguities. A robustly verifiable result is one where the method that one uses to find the result has no ambiguities.
A result obtained from experimental observations is essentially empirical by nature. First principled reasoning may be used to define the accuracy of the result. Replicate, certifiable experiments are needed to establish the degree of precision on the result.