How science works: Let’s say you want to see if there’s a correlation between medicine A and disorder B. There are three options—A is correlated with more B, A is correlated with less B, or there is no correlation. If A is correlated with more B, then the majority of studies will show a higher rate of B in people given A. Some studies, by chance, will not, unless for some reason every person given medicine A gets disorder B and no one else ever gets it. If A is correlated with less B, the reverse is true—the majority of studies will show a lower rate of B in people given A, but some studies (again, by chance) will not. Some might show no correlation, some might indicate that A is correlated with more B, but the overwhelming majority of studies will show a negative correlation. If A is not correlated with B, in any way, then most studies will show no correlation, some (by chance) will show a positive correlation, and some (by chance) will show a negative correlation. If there was never a single study that implied some correlation, that would be highly unusual. Note that I say these things happen by chance, but they can also be affected by poor study design and researcher bias, and if you’re looking at studies reported by the media (rather than the actual journals) they are most definitely affected by the media reporting things it wants to report, and misinterpreting many studies (deliberately or otherwise). In any case, the overwhelming majority of studies suggests no correlation between vaccines and autism. Any one particular study disagreeing with that is essentially irrelevant in light of the thousands that don’t.