The physicists explained that, in experimental tests, there are three loopholes that allow observed violations of local realism to still be explained by local realistic theories. These three loopholes can involve locality (if there is not a large enough distance separating the two objects at the time of measurement), the freedom to choose any measurement settings (so measurement settings may be influenced by hidden variables, or vice versa), and fair sampling (a small fraction of observed objects may not accurately represent all objects due to detection inefficiencies).
Previous experiments have closed the first loophole, which was done by ensuring a large spatial separation between the two objects (in this case, two quantum mechanically entangled photons) so that measurements of the objects could not be influenced by each other. Special relativity then ensures that the objects cannot influence each other, since no physical signals can travel faster than the speed of light. In these experiments, classically unexplainable correlations were still observed between the objects, indicating a violation of local realism. (The fair sampling loophole was closed in another earlier experiment using ions, where large detection efficiencies can be reached.)