Short answer: scientific evidence is information gathered in a disciplined way that helps support, weaken, or refine a claim about the world. It can include measurements, observations, experiments, models, records, images, samples, surveys, and statistical results.
Not all evidence has the same strength. A single anecdote may suggest a question worth studying, but it rarely settles a scientific issue. A carefully designed experiment with clear controls usually carries more weight. A systematic review of many good studies may carry more weight still. The strength of evidence depends on the question being asked and the method used to answer it.
Evidence is not just data
Data become scientific evidence when they are connected to a clear claim and a reliable method. A table of numbers is not automatically meaningful. Readers need to know how the numbers were collected, what was excluded, what was measured, and whether the method could have distorted the result.
For example, a survey can provide useful evidence about attitudes, but it may be weak evidence about actual behaviour if people misremember or answer in socially acceptable ways. A laboratory experiment can show a causal mechanism, but it may not always reflect real-world conditions. An observational study can reveal patterns in large populations, but it may struggle to prove what caused them.
How to judge it
Good scientific evidence is usually transparent, relevant, and open to challenge. The method should be described clearly enough for others to understand it. The evidence should answer the actual question, not a more convenient substitute. The conclusion should be proportionate to the quality of the evidence.
A careful reader should also distinguish between evidence and interpretation. Scientists often agree about measurements but disagree about what they mean. That disagreement is not always a scandal. It can be part of the process by which evidence is tested, criticised, and made stronger.