The word "repeatable" in this context is usually stated as "replicable." When you conduct research and get a particular result, if your research is properly designed and is to have any applicability whatsoever, others should be able to perform the same study and get the same or similar results. For example, if I were to try a particular medication on a group of patients, let's say blood pressure medication, and my results showed that the medication lowered people's blood pressure 30%, this might be a medication worth prescribing to people. But if other researchers who did the same study showed that the blood pressure medication had little or no effect, we would say that the results were not replicable. If other researchers cannot obtain the same or similar results again, this means that the results of the study do not form a proper or reasonable basis upon which to start giving people with hypertension this particular medication.
It is important to bear in mind, though, that all the conditions of the study must be the same, since if any variable changes, replicability is not the problem. If the original study used people aged 45-50, and another study used people from 50-65, different results might have been based upon the change in that variable. So, in order for a study to be replicable (or repeatable), all of the conditions must be the same. Replicability is an extremely important aspect of research. If a finding is made, others should be able to make the same finding, and if they do not, whatever the finding is, it should not be relied upon to make any sort of medical, scientific, or policy decisions.