If a group of scientists have access to one data, from the data they can draw conclusions either through mathematics or just thought experiments.
Those thought experiments is different for any scientist, no one thinks the same especially when the topic is difficult.
For example when talking about parallel universes, scientists have come up with the weirdest examples of a multiverse. Some thinking of a brane universe, while others say that its a landscape universe, quilted universe. All of their 'evidence' seems correct but they have opposite meanings.
A weird analogy is 'religion'. All the religions seem to have 'evidences' (hardly) that attract people towards it, they all make sense but that doesn't mean that their evidence is right.
----
Now if they're trying to break down the data using maths, there could be a great uncertainty and measurement error that if done enough could change the whole idea behind the data.
Interesting question, I can babble for days for this but lets keep it as that