Yes, and what should you do if you get such a result? Or any result for that matter? You should *check* to see if your equipment is malfunctioning. You should *check* and *rethink* to see if your experiment is badly designed. You should *check* that you have your instruments calibrated. You should, in fact, consider as many possibilities as possible and make sure none of them show your results to be wrong. And, in fact, one way to tell the difference between a good scientist and a bad one is whether they do exactly this.
There was actually a case very similar to this not so long ago. The people at CERN found that, by their measurements, neutrinos were going faster than light.
How should they react? If the results were true, it would be ground-breaking and revolutionary. It would require a revision of special relativity. That would be a HUGE deal.
So, what did they do? They went over everything they could think of that would be a problem: was their equipment malfunctioning? Did they neglect to consider the effects from general relativity? Was there a design flaw that they hadn't considered? Was a calibration off? When they were unable to come up any explanation for why they might get a false result, they published their data.
And this is *exactly* what they should have done. Maybe, just maybe, they found a case where neutrinos do move faster than light. But, being human, they presented the case and asked for input.
Eventually, it was found that an extra loop in a wire caused a delay in a signal that negated the result. When the same experiment was done with the correction, no superluminous neutrinos were found.
So, yes, there are always possible alternative explanations. If you come up with one, express it and then *test* to see if that explanation works in practice. If it does, it might well negate your conclusions. Otherwise, you stick with the results of the measurements. Does it seem like an infinite regress? It might. To you. In practice, that regress terminates fairly quickly in any given situation.
You seem to be of the opinion that experiments are done once and for all. But it is *exactly* for the reasons you mention that this is not the case. It shows why *independent* studies and peer review are required to do any actual science.