Assuming a single centre study has been conducted with enough sample size, random sampling and high internal validity, is it safe to assume that the results can be generalized?
No, all you have proven is that it is true for that population at that time.
Even if the population was international, then it is still only true at that time.
However, if you have a good theory on WHY your results are true, then it is safer to generalise. Generalising is dangerous. That is why we plonk it in our Discussions section, and not our Results section.
I agree with Lan, to generalize any true result we need a strong theory behind it which could be implacable globally... But I think we should always proposed to generalization of any true result...
If you are thorough and detailed in describing the population and setting in your study, you increase the possibilities to generalize. But you can't do this yourself, it's up to the reader to judge whether your findings can be applied to his/her population/setting.