John Bell explored this question, basically in dialog with Einstein. Both were seeking to defend realism. Einstein was also seeking to defend locality.
In their work they defined realism and locality in these ways:
- Realism: there is a universe that we can characterize without referencing an observer
- Locality: A light speed cone limits possible interactions between objects in our universe
The answer by Philomath cites the definition of realism from the EPR paper. "Without disturbing the system" is the key phrase that if it cannot be satisfied, includes an observer in reality. Another key phrase is "we can predict with certainty (i.e. with probability equal to unity) the value of a physical quantity". There is consensus among physicists that based on the Heisenberg Uncertainty Principle, we can never specify any pair of physical quantities explicitly within the limits of HUP, so the Uncertainty Principle limits "realism" to only a close approximation of our universe. Additionally, under the Copenhagen theory of QM, there is not a defined value to physical quantities prior to a measurement. So Copenhagen can never be fully realist per HUP, and is not realist at all based on quantum uncertainty before measurement.
Einstein was not just defending realism and relativity, he was also trying to defend determinism. Einstein proposed a Hidden Variables Theory as an alternative QM theory to Copenhagen, which could satisfy both realism and locality, and allow our universe to be deterministic, but this model was tested and refuted.
Bell then showed how no alternative local hidden variable theories could ever match QM. He considered non-locality to be intrinsic to all QM theories. Bell's rationale has been experimentally supported by the discoveries of entanglement between particles that are outside a light-cone for interaction.
As an advocate for realism, and also for determinism, he defended the alternate quantum theory of Bohmian Mechanics. Bohmian Mechanics accepts non-locality, and extends it to not just a few esoteric cases of entangled particles, and instead assumes the entire universe is intrinsically entangled. Realism, rather than a stochastic universe, is necessary for determinism to be possible. Under BM, every quantum event is influenced by every other particle in the universe, no matter how far away, AND every particle has a defined location and other properties (contradicting the HUP).
Bohmian Mechanics is a competing quantum theory to QM, and is in principle testable against QM. Here is a blog summarizing the testing to date, which is trending against BM. https://settheory.net/Bohm In addition to the tests trending against BM, the intrinsic difficulty in performing calculations and making predictions under BM has made it a not very useable theory, and this has further limited its popularity among theoretical physicists. I also suspect that having to make assumptions about how much universe there is beyond our furthest observations, and its uniformity, may also detract from its attraction. A survey of theoretical physicists in 2016 found about 2% were advocates of Bohmian Mechanics.
There are other "interpretations" of QM, some of which are actually competing theories, and others of which may just be interpretations, as possible test cases between them and QM have not always been fleshed out. Einstein was particularly vigorous in clarifying that his "hidden variables interpretation" would make different predictions, and was therefore a competing theory. Other speculative theorizers have not been as vigorous, but over time, it has become clear that at least some of the other "interpretations" are actually competing theories, as with BM. Most of these testable differences are very difficult to test, and BM appears to have been the most evaluated different theory, other than HV.
Other than BM, the other interpretations of quanta are all stochastic, and accept HUP, and therefore are non-realist per Einstein and Bells' definition. They are also non-local.